diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18188 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.8303,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4e-05,
+      "loss": 2.8293,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-05,
+      "loss": 2.1855,
+      "step": 3
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8e-05,
+      "loss": 3.0953,
+      "step": 4
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 2.1908,
+      "step": 5
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.996661101836394e-05,
+      "loss": 2.466,
+      "step": 6
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.996661101836394e-05,
+      "loss": 2.2803,
+      "step": 7
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.993322203672788e-05,
+      "loss": 2.0984,
+      "step": 8
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.993322203672788e-05,
+      "loss": 2.0903,
+      "step": 9
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.989983305509183e-05,
+      "loss": 2.0065,
+      "step": 10
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.986644407345576e-05,
+      "loss": 1.8661,
+      "step": 11
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.98330550918197e-05,
+      "loss": 1.8533,
+      "step": 12
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.979966611018364e-05,
+      "loss": 1.8543,
+      "step": 13
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.976627712854757e-05,
+      "loss": 1.9578,
+      "step": 14
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.973288814691152e-05,
+      "loss": 1.7081,
+      "step": 15
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.969949916527546e-05,
+      "loss": 1.7393,
+      "step": 16
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.966611018363941e-05,
+      "loss": 1.6086,
+      "step": 17
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.963272120200335e-05,
+      "loss": 1.6879,
+      "step": 18
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.959933222036728e-05,
+      "loss": 1.6181,
+      "step": 19
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.956594323873122e-05,
+      "loss": 1.603,
+      "step": 20
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.953255425709517e-05,
+      "loss": 1.6168,
+      "step": 21
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.949916527545911e-05,
+      "loss": 1.6051,
+      "step": 22
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.946577629382304e-05,
+      "loss": 1.9057,
+      "step": 23
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.943238731218698e-05,
+      "loss": 1.5057,
+      "step": 24
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.939899833055092e-05,
+      "loss": 1.6293,
+      "step": 25
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.936560934891487e-05,
+      "loss": 1.6007,
+      "step": 26
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.93322203672788e-05,
+      "loss": 1.6205,
+      "step": 27
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.929883138564274e-05,
+      "loss": 1.6105,
+      "step": 28
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.926544240400668e-05,
+      "loss": 1.7133,
+      "step": 29
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.923205342237061e-05,
+      "loss": 1.4201,
+      "step": 30
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.919866444073456e-05,
+      "loss": 1.566,
+      "step": 31
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.91652754590985e-05,
+      "loss": 1.7912,
+      "step": 32
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.913188647746244e-05,
+      "loss": 1.7245,
+      "step": 33
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.909849749582639e-05,
+      "loss": 1.3152,
+      "step": 34
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.906510851419032e-05,
+      "loss": 1.5055,
+      "step": 35
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.903171953255426e-05,
+      "loss": 1.5059,
+      "step": 36
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.899833055091821e-05,
+      "loss": 1.6442,
+      "step": 37
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.896494156928215e-05,
+      "loss": 1.432,
+      "step": 38
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.893155258764608e-05,
+      "loss": 1.6411,
+      "step": 39
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.889816360601002e-05,
+      "loss": 1.6542,
+      "step": 40
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.886477462437396e-05,
+      "loss": 1.4936,
+      "step": 41
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.883138564273791e-05,
+      "loss": 1.4636,
+      "step": 42
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.879799666110184e-05,
+      "loss": 1.6588,
+      "step": 43
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.876460767946578e-05,
+      "loss": 1.4784,
+      "step": 44
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.873121869782972e-05,
+      "loss": 1.7978,
+      "step": 45
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.869782971619365e-05,
+      "loss": 1.4792,
+      "step": 46
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.86644407345576e-05,
+      "loss": 1.8932,
+      "step": 47
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.863105175292154e-05,
+      "loss": 1.4424,
+      "step": 48
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.859766277128548e-05,
+      "loss": 1.8545,
+      "step": 49
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.856427378964941e-05,
+      "loss": 1.5141,
+      "step": 50
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.853088480801336e-05,
+      "loss": 1.5515,
+      "step": 51
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.84974958263773e-05,
+      "loss": 1.5579,
+      "step": 52
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.846410684474125e-05,
+      "loss": 1.5297,
+      "step": 53
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.843071786310519e-05,
+      "loss": 1.4063,
+      "step": 54
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.839732888146912e-05,
+      "loss": 1.4876,
+      "step": 55
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.836393989983306e-05,
+      "loss": 1.5968,
+      "step": 56
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.8330550918197e-05,
+      "loss": 1.702,
+      "step": 57
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.829716193656095e-05,
+      "loss": 1.4334,
+      "step": 58
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.826377295492488e-05,
+      "loss": 1.6162,
+      "step": 59
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.823038397328882e-05,
+      "loss": 1.6897,
+      "step": 60
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.819699499165276e-05,
+      "loss": 1.3947,
+      "step": 61
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.816360601001669e-05,
+      "loss": 1.5239,
+      "step": 62
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.813021702838064e-05,
+      "loss": 1.5205,
+      "step": 63
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.809682804674458e-05,
+      "loss": 1.4992,
+      "step": 64
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.806343906510852e-05,
+      "loss": 1.3619,
+      "step": 65
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.803005008347245e-05,
+      "loss": 1.5914,
+      "step": 66
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.799666110183639e-05,
+      "loss": 1.5196,
+      "step": 67
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.796327212020034e-05,
+      "loss": 1.4615,
+      "step": 68
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.792988313856428e-05,
+      "loss": 1.7073,
+      "step": 69
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.789649415692823e-05,
+      "loss": 1.4674,
+      "step": 70
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.786310517529216e-05,
+      "loss": 1.3527,
+      "step": 71
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.78297161936561e-05,
+      "loss": 1.4926,
+      "step": 72
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.779632721202004e-05,
+      "loss": 1.5819,
+      "step": 73
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.776293823038399e-05,
+      "loss": 1.5851,
+      "step": 74
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.772954924874792e-05,
+      "loss": 1.5397,
+      "step": 75
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.769616026711186e-05,
+      "loss": 1.6758,
+      "step": 76
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.76627712854758e-05,
+      "loss": 1.332,
+      "step": 77
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.762938230383973e-05,
+      "loss": 1.7195,
+      "step": 78
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.759599332220368e-05,
+      "loss": 1.5733,
+      "step": 79
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.756260434056762e-05,
+      "loss": 1.6188,
+      "step": 80
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.752921535893156e-05,
+      "loss": 1.6966,
+      "step": 81
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.749582637729549e-05,
+      "loss": 1.3633,
+      "step": 82
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.746243739565943e-05,
+      "loss": 1.6326,
+      "step": 83
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.742904841402337e-05,
+      "loss": 1.4259,
+      "step": 84
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.739565943238732e-05,
+      "loss": 1.4006,
+      "step": 85
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.736227045075125e-05,
+      "loss": 1.5051,
+      "step": 86
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.73288814691152e-05,
+      "loss": 1.5654,
+      "step": 87
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.729549248747914e-05,
+      "loss": 1.4776,
+      "step": 88
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.726210350584308e-05,
+      "loss": 1.3615,
+      "step": 89
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.722871452420703e-05,
+      "loss": 1.4514,
+      "step": 90
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.719532554257096e-05,
+      "loss": 1.5547,
+      "step": 91
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.71619365609349e-05,
+      "loss": 1.4926,
+      "step": 92
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.712854757929884e-05,
+      "loss": 1.5664,
+      "step": 93
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.709515859766277e-05,
+      "loss": 1.2536,
+      "step": 94
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.706176961602672e-05,
+      "loss": 1.5664,
+      "step": 95
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.702838063439066e-05,
+      "loss": 1.6009,
+      "step": 96
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.69949916527546e-05,
+      "loss": 1.5207,
+      "step": 97
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.696160267111853e-05,
+      "loss": 1.3776,
+      "step": 98
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.692821368948247e-05,
+      "loss": 1.5374,
+      "step": 99
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.68948247078464e-05,
+      "loss": 1.2995,
+      "step": 100
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.686143572621035e-05,
+      "loss": 1.5631,
+      "step": 101
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.682804674457429e-05,
+      "loss": 1.5,
+      "step": 102
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.679465776293823e-05,
+      "loss": 1.5648,
+      "step": 103
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.676126878130218e-05,
+      "loss": 1.4789,
+      "step": 104
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.672787979966611e-05,
+      "loss": 1.5384,
+      "step": 105
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.669449081803006e-05,
+      "loss": 1.3885,
+      "step": 106
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.6661101836394e-05,
+      "loss": 1.4513,
+      "step": 107
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.662771285475794e-05,
+      "loss": 1.4676,
+      "step": 108
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.659432387312187e-05,
+      "loss": 1.4148,
+      "step": 109
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.656093489148581e-05,
+      "loss": 1.4491,
+      "step": 110
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.652754590984976e-05,
+      "loss": 1.525,
+      "step": 111
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.64941569282137e-05,
+      "loss": 1.4044,
+      "step": 112
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.646076794657763e-05,
+      "loss": 1.4083,
+      "step": 113
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.642737896494157e-05,
+      "loss": 1.4291,
+      "step": 114
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.639398998330551e-05,
+      "loss": 1.2803,
+      "step": 115
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.636060100166944e-05,
+      "loss": 1.531,
+      "step": 116
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.63272120200334e-05,
+      "loss": 1.5171,
+      "step": 117
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.629382303839733e-05,
+      "loss": 1.4076,
+      "step": 118
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.626043405676127e-05,
+      "loss": 1.4701,
+      "step": 119
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.62270450751252e-05,
+      "loss": 1.3906,
+      "step": 120
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.619365609348915e-05,
+      "loss": 1.6288,
+      "step": 121
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.616026711185309e-05,
+      "loss": 1.3735,
+      "step": 122
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.612687813021704e-05,
+      "loss": 1.2921,
+      "step": 123
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.609348914858098e-05,
+      "loss": 1.5322,
+      "step": 124
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.606010016694491e-05,
+      "loss": 1.4944,
+      "step": 125
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.602671118530885e-05,
+      "loss": 1.503,
+      "step": 126
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.59933222036728e-05,
+      "loss": 1.616,
+      "step": 127
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.595993322203674e-05,
+      "loss": 1.5755,
+      "step": 128
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.592654424040067e-05,
+      "loss": 1.6295,
+      "step": 129
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.589315525876461e-05,
+      "loss": 1.4764,
+      "step": 130
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.585976627712855e-05,
+      "loss": 1.5089,
+      "step": 131
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 9.582637729549248e-05,
+      "loss": 1.3362,
+      "step": 132
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.579298831385643e-05,
+      "loss": 1.3674,
+      "step": 133
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.575959933222037e-05,
+      "loss": 1.6924,
+      "step": 134
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.57262103505843e-05,
+      "loss": 1.7297,
+      "step": 135
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.569282136894824e-05,
+      "loss": 1.5391,
+      "step": 136
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.56594323873122e-05,
+      "loss": 1.7875,
+      "step": 137
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 9.562604340567613e-05,
+      "loss": 1.2755,
+      "step": 138
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 9.559265442404008e-05,
+      "loss": 1.5863,
+      "step": 139
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 9.555926544240402e-05,
+      "loss": 1.5498,
+      "step": 140
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.552587646076795e-05,
+      "loss": 1.46,
+      "step": 141
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.549248747913189e-05,
+      "loss": 1.56,
+      "step": 142
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.545909849749584e-05,
+      "loss": 1.4328,
+      "step": 143
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.542570951585978e-05,
+      "loss": 1.5301,
+      "step": 144
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.539232053422371e-05,
+      "loss": 1.2963,
+      "step": 145
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.535893155258765e-05,
+      "loss": 1.3577,
+      "step": 146
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.532554257095159e-05,
+      "loss": 1.3924,
+      "step": 147
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.529215358931554e-05,
+      "loss": 1.361,
+      "step": 148
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.525876460767947e-05,
+      "loss": 1.5379,
+      "step": 149
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.522537562604341e-05,
+      "loss": 1.5397,
+      "step": 150
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.468184232711792,
+      "eval_runtime": 20.2235,
+      "eval_samples_per_second": 9.89,
+      "eval_steps_per_second": 1.236,
+      "step": 150
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.519198664440735e-05,
+      "loss": 1.2553,
+      "step": 151
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.515859766277128e-05,
+      "loss": 1.1346,
+      "step": 152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.512520868113522e-05,
+      "loss": 1.0466,
+      "step": 153
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.509181969949917e-05,
+      "loss": 1.4557,
+      "step": 154
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.50584307178631e-05,
+      "loss": 1.2697,
+      "step": 155
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.502504173622706e-05,
+      "loss": 1.2929,
+      "step": 156
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.499165275459099e-05,
+      "loss": 1.2913,
+      "step": 157
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.495826377295493e-05,
+      "loss": 1.2227,
+      "step": 158
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.492487479131888e-05,
+      "loss": 1.3694,
+      "step": 159
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.489148580968282e-05,
+      "loss": 1.3926,
+      "step": 160
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.485809682804675e-05,
+      "loss": 1.3215,
+      "step": 161
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.482470784641069e-05,
+      "loss": 1.1937,
+      "step": 162
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.479131886477463e-05,
+      "loss": 1.2244,
+      "step": 163
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.475792988313858e-05,
+      "loss": 1.6335,
+      "step": 164
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.472454090150251e-05,
+      "loss": 1.2531,
+      "step": 165
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.469115191986645e-05,
+      "loss": 1.4972,
+      "step": 166
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.465776293823039e-05,
+      "loss": 1.1256,
+      "step": 167
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.462437395659432e-05,
+      "loss": 1.1715,
+      "step": 168
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.459098497495826e-05,
+      "loss": 1.2325,
+      "step": 169
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.455759599332221e-05,
+      "loss": 1.2474,
+      "step": 170
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.452420701168615e-05,
+      "loss": 1.269,
+      "step": 171
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.449081803005008e-05,
+      "loss": 1.0786,
+      "step": 172
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.445742904841403e-05,
+      "loss": 1.3829,
+      "step": 173
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.442404006677797e-05,
+      "loss": 1.3959,
+      "step": 174
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 9.439065108514192e-05,
+      "loss": 1.0559,
+      "step": 175
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 9.435726210350586e-05,
+      "loss": 1.3253,
+      "step": 176
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 9.432387312186979e-05,
+      "loss": 1.145,
+      "step": 177
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 9.429048414023373e-05,
+      "loss": 1.2605,
+      "step": 178
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 9.425709515859767e-05,
+      "loss": 1.1928,
+      "step": 179
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.422370617696162e-05,
+      "loss": 1.3051,
+      "step": 180
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.419031719532555e-05,
+      "loss": 1.2054,
+      "step": 181
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.415692821368949e-05,
+      "loss": 1.2047,
+      "step": 182
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.412353923205343e-05,
+      "loss": 1.1873,
+      "step": 183
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.409015025041736e-05,
+      "loss": 1.18,
+      "step": 184
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.40567612687813e-05,
+      "loss": 1.0805,
+      "step": 185
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.402337228714525e-05,
+      "loss": 1.3862,
+      "step": 186
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.398998330550918e-05,
+      "loss": 0.9582,
+      "step": 187
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.395659432387312e-05,
+      "loss": 1.3279,
+      "step": 188
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.392320534223706e-05,
+      "loss": 1.4812,
+      "step": 189
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.388981636060101e-05,
+      "loss": 1.4395,
+      "step": 190
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.385642737896494e-05,
+      "loss": 1.0285,
+      "step": 191
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.38230383973289e-05,
+      "loss": 1.1847,
+      "step": 192
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.378964941569283e-05,
+      "loss": 1.1528,
+      "step": 193
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.375626043405677e-05,
+      "loss": 1.2694,
+      "step": 194
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.37228714524207e-05,
+      "loss": 1.457,
+      "step": 195
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 9.368948247078465e-05,
+      "loss": 1.1153,
+      "step": 196
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 9.365609348914859e-05,
+      "loss": 1.0548,
+      "step": 197
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 9.362270450751253e-05,
+      "loss": 1.2857,
+      "step": 198
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.358931552587646e-05,
+      "loss": 1.0563,
+      "step": 199
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.35559265442404e-05,
+      "loss": 1.0842,
+      "step": 200
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 9.352253756260434e-05,
+      "loss": 1.3008,
+      "step": 201
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 9.348914858096829e-05,
+      "loss": 1.2448,
+      "step": 202
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 9.345575959933222e-05,
+      "loss": 1.3442,
+      "step": 203
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 9.342237061769616e-05,
+      "loss": 1.3122,
+      "step": 204
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.33889816360601e-05,
+      "loss": 1.2649,
+      "step": 205
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.335559265442403e-05,
+      "loss": 1.5701,
+      "step": 206
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 9.332220367278798e-05,
+      "loss": 1.3077,
+      "step": 207
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 9.328881469115192e-05,
+      "loss": 1.1245,
+      "step": 208
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 9.325542570951587e-05,
+      "loss": 1.0394,
+      "step": 209
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 9.322203672787981e-05,
+      "loss": 1.1921,
+      "step": 210
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 9.318864774624374e-05,
+      "loss": 1.1125,
+      "step": 211
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 9.31552587646077e-05,
+      "loss": 1.0952,
+      "step": 212
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 9.312186978297163e-05,
+      "loss": 1.2524,
+      "step": 213
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.308848080133557e-05,
+      "loss": 1.1199,
+      "step": 214
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.30550918196995e-05,
+      "loss": 1.5353,
+      "step": 215
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.302170283806344e-05,
+      "loss": 0.8928,
+      "step": 216
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.298831385642738e-05,
+      "loss": 1.4439,
+      "step": 217
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.295492487479133e-05,
+      "loss": 1.2655,
+      "step": 218
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.292153589315526e-05,
+      "loss": 1.1903,
+      "step": 219
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.28881469115192e-05,
+      "loss": 1.2366,
+      "step": 220
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.285475792988314e-05,
+      "loss": 1.2066,
+      "step": 221
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.282136894824707e-05,
+      "loss": 1.1515,
+      "step": 222
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.278797996661102e-05,
+      "loss": 1.2921,
+      "step": 223
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.275459098497496e-05,
+      "loss": 1.2262,
+      "step": 224
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.27212020033389e-05,
+      "loss": 1.1577,
+      "step": 225
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.268781302170285e-05,
+      "loss": 1.3499,
+      "step": 226
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.265442404006678e-05,
+      "loss": 1.1552,
+      "step": 227
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.262103505843073e-05,
+      "loss": 1.0747,
+      "step": 228
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.258764607679467e-05,
+      "loss": 1.0752,
+      "step": 229
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.255425709515861e-05,
+      "loss": 1.6491,
+      "step": 230
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.252086811352254e-05,
+      "loss": 1.1489,
+      "step": 231
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.248747913188648e-05,
+      "loss": 1.2443,
+      "step": 232
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.245409015025042e-05,
+      "loss": 1.2949,
+      "step": 233
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.242070116861437e-05,
+      "loss": 1.2026,
+      "step": 234
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.23873121869783e-05,
+      "loss": 1.1998,
+      "step": 235
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.235392320534224e-05,
+      "loss": 1.2852,
+      "step": 236
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.232053422370618e-05,
+      "loss": 1.3819,
+      "step": 237
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.228714524207011e-05,
+      "loss": 1.1417,
+      "step": 238
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.225375626043406e-05,
+      "loss": 1.3294,
+      "step": 239
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.2220367278798e-05,
+      "loss": 1.3981,
+      "step": 240
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.218697829716194e-05,
+      "loss": 1.2358,
+      "step": 241
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.215358931552587e-05,
+      "loss": 1.1872,
+      "step": 242
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.212020033388982e-05,
+      "loss": 1.0932,
+      "step": 243
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.208681135225376e-05,
+      "loss": 1.1129,
+      "step": 244
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.205342237061771e-05,
+      "loss": 1.2598,
+      "step": 245
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.202003338898165e-05,
+      "loss": 1.0025,
+      "step": 246
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.198664440734558e-05,
+      "loss": 1.3576,
+      "step": 247
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.195325542570952e-05,
+      "loss": 1.1838,
+      "step": 248
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.191986644407346e-05,
+      "loss": 1.2765,
+      "step": 249
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.18864774624374e-05,
+      "loss": 1.4302,
+      "step": 250
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.185308848080134e-05,
+      "loss": 1.0667,
+      "step": 251
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.181969949916528e-05,
+      "loss": 1.3159,
+      "step": 252
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.178631051752922e-05,
+      "loss": 1.4412,
+      "step": 253
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.175292153589315e-05,
+      "loss": 1.1478,
+      "step": 254
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.17195325542571e-05,
+      "loss": 1.107,
+      "step": 255
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.168614357262104e-05,
+      "loss": 1.1899,
+      "step": 256
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.165275459098498e-05,
+      "loss": 1.0595,
+      "step": 257
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.161936560934891e-05,
+      "loss": 1.1935,
+      "step": 258
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.158597662771285e-05,
+      "loss": 1.0552,
+      "step": 259
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.15525876460768e-05,
+      "loss": 1.3896,
+      "step": 260
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 9.151919866444074e-05,
+      "loss": 1.2509,
+      "step": 261
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.148580968280469e-05,
+      "loss": 1.185,
+      "step": 262
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.145242070116862e-05,
+      "loss": 1.5139,
+      "step": 263
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 9.141903171953256e-05,
+      "loss": 1.3051,
+      "step": 264
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.13856427378965e-05,
+      "loss": 1.1214,
+      "step": 265
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.135225375626045e-05,
+      "loss": 1.1103,
+      "step": 266
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 9.131886477462438e-05,
+      "loss": 1.4458,
+      "step": 267
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.128547579298832e-05,
+      "loss": 1.1427,
+      "step": 268
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.125208681135226e-05,
+      "loss": 1.0515,
+      "step": 269
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 9.121869782971619e-05,
+      "loss": 1.0916,
+      "step": 270
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 9.118530884808014e-05,
+      "loss": 1.1892,
+      "step": 271
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 9.115191986644408e-05,
+      "loss": 1.0862,
+      "step": 272
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 9.111853088480802e-05,
+      "loss": 1.23,
+      "step": 273
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.108514190317195e-05,
+      "loss": 1.1937,
+      "step": 274
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.105175292153589e-05,
+      "loss": 1.2344,
+      "step": 275
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.101836393989984e-05,
+      "loss": 1.105,
+      "step": 276
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.098497495826378e-05,
+      "loss": 1.2532,
+      "step": 277
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.095158597662771e-05,
+      "loss": 1.0368,
+      "step": 278
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.091819699499166e-05,
+      "loss": 1.2594,
+      "step": 279
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.08848080133556e-05,
+      "loss": 1.3224,
+      "step": 280
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.085141903171955e-05,
+      "loss": 1.1844,
+      "step": 281
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.081803005008348e-05,
+      "loss": 1.1668,
+      "step": 282
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.078464106844742e-05,
+      "loss": 1.2816,
+      "step": 283
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.075125208681136e-05,
+      "loss": 1.0698,
+      "step": 284
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.07178631051753e-05,
+      "loss": 1.1851,
+      "step": 285
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.068447412353923e-05,
+      "loss": 1.5333,
+      "step": 286
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.065108514190318e-05,
+      "loss": 1.1068,
+      "step": 287
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.061769616026712e-05,
+      "loss": 1.1968,
+      "step": 288
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.058430717863105e-05,
+      "loss": 1.243,
+      "step": 289
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.055091819699499e-05,
+      "loss": 1.1143,
+      "step": 290
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 9.051752921535893e-05,
+      "loss": 1.1447,
+      "step": 291
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.048414023372288e-05,
+      "loss": 1.2419,
+      "step": 292
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.045075125208681e-05,
+      "loss": 1.1498,
+      "step": 293
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 9.041736227045075e-05,
+      "loss": 1.3696,
+      "step": 294
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.038397328881469e-05,
+      "loss": 1.315,
+      "step": 295
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.035058430717864e-05,
+      "loss": 1.1067,
+      "step": 296
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 9.031719532554257e-05,
+      "loss": 1.2535,
+      "step": 297
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.028380634390652e-05,
+      "loss": 1.2362,
+      "step": 298
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.025041736227046e-05,
+      "loss": 1.3246,
+      "step": 299
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.02170283806344e-05,
+      "loss": 1.2842,
+      "step": 300
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.50325345993042,
+      "eval_runtime": 20.2144,
+      "eval_samples_per_second": 9.894,
+      "eval_steps_per_second": 1.237,
+      "step": 300
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.018363939899833e-05,
+      "loss": 0.7094,
+      "step": 301
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.015025041736227e-05,
+      "loss": 0.8285,
+      "step": 302
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.011686143572622e-05,
+      "loss": 0.8762,
+      "step": 303
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.008347245409016e-05,
+      "loss": 0.8812,
+      "step": 304
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.00500834724541e-05,
+      "loss": 0.689,
+      "step": 305
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 9.001669449081803e-05,
+      "loss": 0.7822,
+      "step": 306
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 8.998330550918197e-05,
+      "loss": 0.9094,
+      "step": 307
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 8.994991652754592e-05,
+      "loss": 0.8986,
+      "step": 308
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 8.991652754590985e-05,
+      "loss": 0.9939,
+      "step": 309
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 8.988313856427379e-05,
+      "loss": 1.1576,
+      "step": 310
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 8.984974958263773e-05,
+      "loss": 0.9138,
+      "step": 311
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 8.981636060100166e-05,
+      "loss": 0.8072,
+      "step": 312
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 8.978297161936561e-05,
+      "loss": 0.8031,
+      "step": 313
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 8.974958263772955e-05,
+      "loss": 0.7345,
+      "step": 314
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 8.97161936560935e-05,
+      "loss": 0.7264,
+      "step": 315
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 8.968280467445744e-05,
+      "loss": 0.8883,
+      "step": 316
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 8.964941569282137e-05,
+      "loss": 0.7055,
+      "step": 317
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 8.961602671118531e-05,
+      "loss": 0.8926,
+      "step": 318
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 8.958263772954926e-05,
+      "loss": 0.8066,
+      "step": 319
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 8.95492487479132e-05,
+      "loss": 0.8135,
+      "step": 320
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 8.951585976627713e-05,
+      "loss": 0.944,
+      "step": 321
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 8.948247078464107e-05,
+      "loss": 1.0089,
+      "step": 322
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 8.9449081803005e-05,
+      "loss": 0.7468,
+      "step": 323
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 8.941569282136896e-05,
+      "loss": 0.8235,
+      "step": 324
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 8.93823038397329e-05,
+      "loss": 0.7939,
+      "step": 325
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 8.934891485809683e-05,
+      "loss": 0.8033,
+      "step": 326
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 8.931552587646077e-05,
+      "loss": 0.7192,
+      "step": 327
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.92821368948247e-05,
+      "loss": 0.8564,
+      "step": 328
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.924874791318865e-05,
+      "loss": 0.8625,
+      "step": 329
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 8.921535893155259e-05,
+      "loss": 0.7422,
+      "step": 330
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.918196994991653e-05,
+      "loss": 0.6323,
+      "step": 331
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.914858096828048e-05,
+      "loss": 0.6927,
+      "step": 332
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.911519198664441e-05,
+      "loss": 0.7876,
+      "step": 333
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.908180300500835e-05,
+      "loss": 0.7004,
+      "step": 334
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.90484140233723e-05,
+      "loss": 0.7949,
+      "step": 335
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.901502504173624e-05,
+      "loss": 0.6753,
+      "step": 336
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.898163606010017e-05,
+      "loss": 0.7862,
+      "step": 337
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.894824707846411e-05,
+      "loss": 0.7915,
+      "step": 338
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.891485809682805e-05,
+      "loss": 0.8013,
+      "step": 339
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.8881469115192e-05,
+      "loss": 0.7224,
+      "step": 340
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.884808013355593e-05,
+      "loss": 0.8865,
+      "step": 341
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.881469115191987e-05,
+      "loss": 0.7721,
+      "step": 342
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.87813021702838e-05,
+      "loss": 0.7692,
+      "step": 343
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.874791318864774e-05,
+      "loss": 0.908,
+      "step": 344
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.871452420701169e-05,
+      "loss": 0.9616,
+      "step": 345
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.868113522537563e-05,
+      "loss": 0.7713,
+      "step": 346
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.864774624373957e-05,
+      "loss": 0.7991,
+      "step": 347
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.86143572621035e-05,
+      "loss": 0.6489,
+      "step": 348
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.858096828046745e-05,
+      "loss": 0.943,
+      "step": 349
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.854757929883139e-05,
+      "loss": 0.7388,
+      "step": 350
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 8.851419031719534e-05,
+      "loss": 0.8667,
+      "step": 351
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.848080133555928e-05,
+      "loss": 0.6081,
+      "step": 352
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.844741235392321e-05,
+      "loss": 0.7616,
+      "step": 353
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.841402337228715e-05,
+      "loss": 0.6489,
+      "step": 354
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.838063439065109e-05,
+      "loss": 0.8773,
+      "step": 355
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.834724540901504e-05,
+      "loss": 0.6121,
+      "step": 356
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.831385642737897e-05,
+      "loss": 0.8003,
+      "step": 357
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.828046744574291e-05,
+      "loss": 0.7821,
+      "step": 358
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.824707846410685e-05,
+      "loss": 0.7301,
+      "step": 359
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.821368948247078e-05,
+      "loss": 0.9243,
+      "step": 360
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.818030050083473e-05,
+      "loss": 0.8572,
+      "step": 361
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.814691151919867e-05,
+      "loss": 0.6719,
+      "step": 362
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 8.81135225375626e-05,
+      "loss": 0.7846,
+      "step": 363
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.808013355592654e-05,
+      "loss": 0.6821,
+      "step": 364
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.804674457429048e-05,
+      "loss": 0.9221,
+      "step": 365
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.801335559265443e-05,
+      "loss": 0.6662,
+      "step": 366
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.797996661101837e-05,
+      "loss": 0.7393,
+      "step": 367
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.794657762938232e-05,
+      "loss": 0.7646,
+      "step": 368
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.791318864774625e-05,
+      "loss": 0.7247,
+      "step": 369
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.787979966611019e-05,
+      "loss": 0.7161,
+      "step": 370
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.784641068447412e-05,
+      "loss": 0.9244,
+      "step": 371
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.781302170283808e-05,
+      "loss": 0.7404,
+      "step": 372
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.777963272120201e-05,
+      "loss": 0.9106,
+      "step": 373
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.774624373956595e-05,
+      "loss": 0.818,
+      "step": 374
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.771285475792988e-05,
+      "loss": 0.806,
+      "step": 375
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.767946577629382e-05,
+      "loss": 0.7554,
+      "step": 376
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.764607679465777e-05,
+      "loss": 0.7073,
+      "step": 377
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.761268781302171e-05,
+      "loss": 0.8769,
+      "step": 378
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.757929883138564e-05,
+      "loss": 0.8011,
+      "step": 379
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.754590984974958e-05,
+      "loss": 0.6235,
+      "step": 380
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.751252086811352e-05,
+      "loss": 0.6809,
+      "step": 381
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.747913188647745e-05,
+      "loss": 0.7779,
+      "step": 382
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.74457429048414e-05,
+      "loss": 0.8489,
+      "step": 383
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 8.741235392320535e-05,
+      "loss": 0.8759,
+      "step": 384
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 8.737896494156929e-05,
+      "loss": 0.8279,
+      "step": 385
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 8.734557595993323e-05,
+      "loss": 0.8737,
+      "step": 386
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 8.731218697829716e-05,
+      "loss": 0.8872,
+      "step": 387
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 8.727879799666111e-05,
+      "loss": 0.7139,
+      "step": 388
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 8.724540901502505e-05,
+      "loss": 0.7173,
+      "step": 389
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.721202003338899e-05,
+      "loss": 0.9551,
+      "step": 390
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.717863105175292e-05,
+      "loss": 0.7762,
+      "step": 391
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.714524207011686e-05,
+      "loss": 0.8352,
+      "step": 392
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.711185308848081e-05,
+      "loss": 0.8245,
+      "step": 393
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.707846410684475e-05,
+      "loss": 0.797,
+      "step": 394
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.704507512520868e-05,
+      "loss": 0.6881,
+      "step": 395
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.701168614357262e-05,
+      "loss": 0.7838,
+      "step": 396
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.697829716193656e-05,
+      "loss": 0.7802,
+      "step": 397
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.694490818030051e-05,
+      "loss": 0.9414,
+      "step": 398
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.691151919866444e-05,
+      "loss": 0.8689,
+      "step": 399
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 8.687813021702838e-05,
+      "loss": 0.7113,
+      "step": 400
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 8.684474123539233e-05,
+      "loss": 0.8172,
+      "step": 401
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 8.681135225375627e-05,
+      "loss": 0.8622,
+      "step": 402
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 8.67779632721202e-05,
+      "loss": 0.7979,
+      "step": 403
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 8.674457429048415e-05,
+      "loss": 0.8046,
+      "step": 404
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 8.671118530884809e-05,
+      "loss": 0.9507,
+      "step": 405
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 8.667779632721203e-05,
+      "loss": 0.7133,
+      "step": 406
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 8.664440734557596e-05,
+      "loss": 0.7464,
+      "step": 407
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 8.66110183639399e-05,
+      "loss": 0.9278,
+      "step": 408
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 8.657762938230385e-05,
+      "loss": 0.7427,
+      "step": 409
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 8.654424040066779e-05,
+      "loss": 0.844,
+      "step": 410
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 8.651085141903172e-05,
+      "loss": 0.7771,
+      "step": 411
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.647746243739566e-05,
+      "loss": 0.8845,
+      "step": 412
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.64440734557596e-05,
+      "loss": 0.7343,
+      "step": 413
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.641068447412355e-05,
+      "loss": 0.6679,
+      "step": 414
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.637729549248748e-05,
+      "loss": 0.8522,
+      "step": 415
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.634390651085142e-05,
+      "loss": 0.7722,
+      "step": 416
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.631051752921536e-05,
+      "loss": 0.8624,
+      "step": 417
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.62771285475793e-05,
+      "loss": 0.7453,
+      "step": 418
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.624373956594324e-05,
+      "loss": 0.686,
+      "step": 419
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.62103505843072e-05,
+      "loss": 0.8317,
+      "step": 420
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.617696160267113e-05,
+      "loss": 0.7581,
+      "step": 421
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.614357262103507e-05,
+      "loss": 0.6663,
+      "step": 422
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.6110183639399e-05,
+      "loss": 0.7611,
+      "step": 423
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.607679465776294e-05,
+      "loss": 0.6768,
+      "step": 424
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.604340567612689e-05,
+      "loss": 0.6899,
+      "step": 425
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 8.601001669449083e-05,
+      "loss": 0.8832,
+      "step": 426
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.597662771285476e-05,
+      "loss": 0.7896,
+      "step": 427
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.59432387312187e-05,
+      "loss": 0.6868,
+      "step": 428
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.590984974958264e-05,
+      "loss": 0.764,
+      "step": 429
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.587646076794659e-05,
+      "loss": 0.806,
+      "step": 430
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.584307178631052e-05,
+      "loss": 0.7853,
+      "step": 431
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.580968280467446e-05,
+      "loss": 0.776,
+      "step": 432
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.57762938230384e-05,
+      "loss": 0.7085,
+      "step": 433
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.574290484140233e-05,
+      "loss": 0.8246,
+      "step": 434
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.570951585976628e-05,
+      "loss": 1.043,
+      "step": 435
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.567612687813022e-05,
+      "loss": 0.7452,
+      "step": 436
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.564273789649417e-05,
+      "loss": 0.7348,
+      "step": 437
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.56093489148581e-05,
+      "loss": 0.8915,
+      "step": 438
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.557595993322204e-05,
+      "loss": 0.809,
+      "step": 439
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.554257095158598e-05,
+      "loss": 0.7115,
+      "step": 440
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 8.550918196994993e-05,
+      "loss": 0.8034,
+      "step": 441
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.547579298831387e-05,
+      "loss": 0.758,
+      "step": 442
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.54424040066778e-05,
+      "loss": 1.0122,
+      "step": 443
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.540901502504174e-05,
+      "loss": 0.6621,
+      "step": 444
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.537562604340568e-05,
+      "loss": 0.8691,
+      "step": 445
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.534223706176963e-05,
+      "loss": 1.0762,
+      "step": 446
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 8.530884808013356e-05,
+      "loss": 0.7786,
+      "step": 447
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.52754590984975e-05,
+      "loss": 0.8041,
+      "step": 448
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.524207011686144e-05,
+      "loss": 0.9271,
+      "step": 449
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.520868113522537e-05,
+      "loss": 1.0418,
+      "step": 450
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.714505910873413,
+      "eval_runtime": 20.1969,
+      "eval_samples_per_second": 9.903,
+      "eval_steps_per_second": 1.238,
+      "step": 450
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 8.517529215358931e-05,
+      "loss": 0.4172,
+      "step": 451
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 8.514190317195326e-05,
+      "loss": 0.4209,
+      "step": 452
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 8.51085141903172e-05,
+      "loss": 0.4754,
+      "step": 453
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 8.507512520868115e-05,
+      "loss": 0.479,
+      "step": 454
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 8.504173622704508e-05,
+      "loss": 0.4212,
+      "step": 455
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 8.500834724540902e-05,
+      "loss": 0.4185,
+      "step": 456
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 8.497495826377297e-05,
+      "loss": 0.3844,
+      "step": 457
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 8.49415692821369e-05,
+      "loss": 0.4066,
+      "step": 458
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 8.490818030050084e-05,
+      "loss": 0.4049,
+      "step": 459
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 8.487479131886478e-05,
+      "loss": 0.4916,
+      "step": 460
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 8.484140233722872e-05,
+      "loss": 0.425,
+      "step": 461
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 8.480801335559267e-05,
+      "loss": 0.2846,
+      "step": 462
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 8.47746243739566e-05,
+      "loss": 0.4491,
+      "step": 463
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 8.474123539232054e-05,
+      "loss": 0.438,
+      "step": 464
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 8.470784641068447e-05,
+      "loss": 0.4006,
+      "step": 465
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 8.467445742904841e-05,
+      "loss": 0.425,
+      "step": 466
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 8.464106844741235e-05,
+      "loss": 0.4095,
+      "step": 467
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 8.46076794657763e-05,
+      "loss": 0.4774,
+      "step": 468
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.457429048414023e-05,
+      "loss": 0.391,
+      "step": 469
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.454090150250417e-05,
+      "loss": 0.426,
+      "step": 470
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.450751252086812e-05,
+      "loss": 0.3671,
+      "step": 471
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.447412353923206e-05,
+      "loss": 0.3543,
+      "step": 472
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 8.444073455759601e-05,
+      "loss": 0.4055,
+      "step": 473
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.440734557595994e-05,
+      "loss": 0.4263,
+      "step": 474
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 8.437395659432388e-05,
+      "loss": 0.5765,
+      "step": 475
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 8.434056761268782e-05,
+      "loss": 0.4089,
+      "step": 476
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 8.430717863105175e-05,
+      "loss": 0.3865,
+      "step": 477
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 8.42737896494157e-05,
+      "loss": 0.4338,
+      "step": 478
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 8.424040066777964e-05,
+      "loss": 0.5126,
+      "step": 479
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 8.420701168614358e-05,
+      "loss": 0.3717,
+      "step": 480
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 8.417362270450751e-05,
+      "loss": 0.3904,
+      "step": 481
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 8.414023372287145e-05,
+      "loss": 0.3888,
+      "step": 482
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 8.410684474123539e-05,
+      "loss": 0.4159,
+      "step": 483
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 8.407345575959934e-05,
+      "loss": 0.4122,
+      "step": 484
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 8.404006677796327e-05,
+      "loss": 0.4365,
+      "step": 485
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 8.400667779632721e-05,
+      "loss": 0.4171,
+      "step": 486
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 8.397328881469115e-05,
+      "loss": 0.4226,
+      "step": 487
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 8.39398998330551e-05,
+      "loss": 0.4007,
+      "step": 488
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 8.390651085141903e-05,
+      "loss": 0.4156,
+      "step": 489
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 8.387312186978298e-05,
+      "loss": 0.3744,
+      "step": 490
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 8.383973288814692e-05,
+      "loss": 0.473,
+      "step": 491
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 8.380634390651086e-05,
+      "loss": 0.4256,
+      "step": 492
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.37729549248748e-05,
+      "loss": 0.4776,
+      "step": 493
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.373956594323874e-05,
+      "loss": 0.4321,
+      "step": 494
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.370617696160268e-05,
+      "loss": 0.4447,
+      "step": 495
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.367278797996662e-05,
+      "loss": 0.4506,
+      "step": 496
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.363939899833055e-05,
+      "loss": 0.3308,
+      "step": 497
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 8.360601001669449e-05,
+      "loss": 0.4118,
+      "step": 498
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.357262103505843e-05,
+      "loss": 0.5638,
+      "step": 499
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.353923205342238e-05,
+      "loss": 0.4621,
+      "step": 500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.350584307178631e-05,
+      "loss": 0.3824,
+      "step": 501
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.347245409015025e-05,
+      "loss": 0.4758,
+      "step": 502
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.343906510851419e-05,
+      "loss": 0.519,
+      "step": 503
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 8.340567612687812e-05,
+      "loss": 0.4614,
+      "step": 504
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.337228714524207e-05,
+      "loss": 0.4266,
+      "step": 505
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.333889816360601e-05,
+      "loss": 0.4941,
+      "step": 506
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 8.330550918196996e-05,
+      "loss": 0.4634,
+      "step": 507
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 8.32721202003339e-05,
+      "loss": 0.4257,
+      "step": 508
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 8.323873121869783e-05,
+      "loss": 0.3832,
+      "step": 509
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 8.320534223706178e-05,
+      "loss": 0.4093,
+      "step": 510
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 8.317195325542572e-05,
+      "loss": 0.5708,
+      "step": 511
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 8.313856427378966e-05,
+      "loss": 0.478,
+      "step": 512
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 8.310517529215359e-05,
+      "loss": 0.4112,
+      "step": 513
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.307178631051753e-05,
+      "loss": 0.4735,
+      "step": 514
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.303839732888147e-05,
+      "loss": 0.4188,
+      "step": 515
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.300500834724542e-05,
+      "loss": 0.3841,
+      "step": 516
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.297161936560935e-05,
+      "loss": 0.3967,
+      "step": 517
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.293823038397329e-05,
+      "loss": 0.5181,
+      "step": 518
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.290484140233723e-05,
+      "loss": 0.4953,
+      "step": 519
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.287145242070116e-05,
+      "loss": 0.4028,
+      "step": 520
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.283806343906511e-05,
+      "loss": 0.4284,
+      "step": 521
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 8.280467445742905e-05,
+      "loss": 0.5057,
+      "step": 522
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.277128547579299e-05,
+      "loss": 0.4965,
+      "step": 523
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.273789649415694e-05,
+      "loss": 0.4355,
+      "step": 524
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.270450751252087e-05,
+      "loss": 0.3379,
+      "step": 525
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.267111853088482e-05,
+      "loss": 0.5018,
+      "step": 526
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.263772954924876e-05,
+      "loss": 0.6183,
+      "step": 527
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 8.26043405676127e-05,
+      "loss": 0.465,
+      "step": 528
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 8.257095158597663e-05,
+      "loss": 0.4067,
+      "step": 529
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 8.253756260434057e-05,
+      "loss": 0.5059,
+      "step": 530
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 8.250417362270452e-05,
+      "loss": 0.4045,
+      "step": 531
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.247078464106846e-05,
+      "loss": 0.3993,
+      "step": 532
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.243739565943239e-05,
+      "loss": 0.471,
+      "step": 533
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 8.240400667779633e-05,
+      "loss": 0.5435,
+      "step": 534
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.237061769616027e-05,
+      "loss": 0.4713,
+      "step": 535
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.23372287145242e-05,
+      "loss": 0.4613,
+      "step": 536
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 8.230383973288815e-05,
+      "loss": 0.3777,
+      "step": 537
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 8.227045075125209e-05,
+      "loss": 0.5862,
+      "step": 538
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 8.223706176961603e-05,
+      "loss": 0.3067,
+      "step": 539
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 8.220367278797996e-05,
+      "loss": 0.3664,
+      "step": 540
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.217028380634391e-05,
+      "loss": 0.418,
+      "step": 541
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.213689482470785e-05,
+      "loss": 0.4795,
+      "step": 542
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 8.21035058430718e-05,
+      "loss": 0.6091,
+      "step": 543
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.207011686143574e-05,
+      "loss": 0.3558,
+      "step": 544
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.203672787979967e-05,
+      "loss": 0.3612,
+      "step": 545
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 8.200333889816361e-05,
+      "loss": 0.3594,
+      "step": 546
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.196994991652756e-05,
+      "loss": 0.5483,
+      "step": 547
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.19365609348915e-05,
+      "loss": 0.5171,
+      "step": 548
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.190317195325543e-05,
+      "loss": 0.4557,
+      "step": 549
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.186978297161937e-05,
+      "loss": 0.4991,
+      "step": 550
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.18363939899833e-05,
+      "loss": 0.4524,
+      "step": 551
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.180300500834724e-05,
+      "loss": 0.4316,
+      "step": 552
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.176961602671119e-05,
+      "loss": 0.5156,
+      "step": 553
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.173622704507513e-05,
+      "loss": 0.5611,
+      "step": 554
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.170283806343906e-05,
+      "loss": 0.4714,
+      "step": 555
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.1669449081803e-05,
+      "loss": 0.4958,
+      "step": 556
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.163606010016694e-05,
+      "loss": 0.4076,
+      "step": 557
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 8.160267111853089e-05,
+      "loss": 0.5583,
+      "step": 558
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.156928213689482e-05,
+      "loss": 0.5442,
+      "step": 559
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.153589315525877e-05,
+      "loss": 0.4534,
+      "step": 560
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.150250417362271e-05,
+      "loss": 0.3997,
+      "step": 561
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.146911519198665e-05,
+      "loss": 0.5131,
+      "step": 562
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.14357262103506e-05,
+      "loss": 0.374,
+      "step": 563
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.140233722871453e-05,
+      "loss": 0.4829,
+      "step": 564
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.136894824707847e-05,
+      "loss": 0.4136,
+      "step": 565
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.133555926544241e-05,
+      "loss": 0.4642,
+      "step": 566
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.130217028380634e-05,
+      "loss": 0.4895,
+      "step": 567
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.126878130217028e-05,
+      "loss": 0.4626,
+      "step": 568
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.123539232053423e-05,
+      "loss": 0.3744,
+      "step": 569
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 8.120200333889817e-05,
+      "loss": 0.3952,
+      "step": 570
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 8.11686143572621e-05,
+      "loss": 0.4186,
+      "step": 571
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 8.113522537562604e-05,
+      "loss": 0.3456,
+      "step": 572
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 8.110183639398998e-05,
+      "loss": 0.4711,
+      "step": 573
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.106844741235393e-05,
+      "loss": 0.3889,
+      "step": 574
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.103505843071786e-05,
+      "loss": 0.4845,
+      "step": 575
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.10016694490818e-05,
+      "loss": 0.5356,
+      "step": 576
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 8.096828046744575e-05,
+      "loss": 0.3698,
+      "step": 577
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 8.093489148580969e-05,
+      "loss": 0.511,
+      "step": 578
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 8.090150250417364e-05,
+      "loss": 0.3399,
+      "step": 579
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 8.086811352253757e-05,
+      "loss": 0.5413,
+      "step": 580
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 8.083472454090151e-05,
+      "loss": 0.4001,
+      "step": 581
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 8.080133555926545e-05,
+      "loss": 0.576,
+      "step": 582
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 8.076794657762938e-05,
+      "loss": 0.4579,
+      "step": 583
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 8.073455759599332e-05,
+      "loss": 0.4344,
+      "step": 584
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 8.070116861435727e-05,
+      "loss": 0.5124,
+      "step": 585
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 8.066777963272121e-05,
+      "loss": 0.3768,
+      "step": 586
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 8.063439065108514e-05,
+      "loss": 0.3999,
+      "step": 587
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 8.060100166944908e-05,
+      "loss": 0.457,
+      "step": 588
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 8.056761268781302e-05,
+      "loss": 0.4201,
+      "step": 589
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 8.053422370617697e-05,
+      "loss": 0.3962,
+      "step": 590
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 8.05008347245409e-05,
+      "loss": 0.4161,
+      "step": 591
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 8.046744574290484e-05,
+      "loss": 0.5802,
+      "step": 592
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 8.043405676126878e-05,
+      "loss": 0.5097,
+      "step": 593
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 8.040066777963273e-05,
+      "loss": 0.4063,
+      "step": 594
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 8.036727879799666e-05,
+      "loss": 0.4117,
+      "step": 595
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 8.033388981636061e-05,
+      "loss": 0.4932,
+      "step": 596
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 8.030050083472455e-05,
+      "loss": 0.4116,
+      "step": 597
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.026711185308849e-05,
+      "loss": 0.5246,
+      "step": 598
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.023372287145242e-05,
+      "loss": 0.5365,
+      "step": 599
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 8.020033388981636e-05,
+      "loss": 0.3202,
+      "step": 600
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.9553784132003784,
+      "eval_runtime": 20.2128,
+      "eval_samples_per_second": 9.895,
+      "eval_steps_per_second": 1.237,
+      "step": 600
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 8.016694490818031e-05,
+      "loss": 0.2603,
+      "step": 601
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 8.013355592654425e-05,
+      "loss": 0.2906,
+      "step": 602
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 8.010016694490818e-05,
+      "loss": 0.3142,
+      "step": 603
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 8.006677796327212e-05,
+      "loss": 0.2783,
+      "step": 604
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 8.003338898163606e-05,
+      "loss": 0.283,
+      "step": 605
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8e-05,
+      "loss": 0.2307,
+      "step": 606
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 7.996661101836394e-05,
+      "loss": 0.2492,
+      "step": 607
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 7.993322203672788e-05,
+      "loss": 0.2869,
+      "step": 608
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 7.989983305509182e-05,
+      "loss": 0.2451,
+      "step": 609
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.986644407345575e-05,
+      "loss": 0.2783,
+      "step": 610
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.98330550918197e-05,
+      "loss": 0.2497,
+      "step": 611
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 7.979966611018364e-05,
+      "loss": 0.2205,
+      "step": 612
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 7.976627712854759e-05,
+      "loss": 0.1803,
+      "step": 613
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 7.973288814691153e-05,
+      "loss": 0.304,
+      "step": 614
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 7.969949916527546e-05,
+      "loss": 0.2192,
+      "step": 615
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 7.96661101836394e-05,
+      "loss": 0.3601,
+      "step": 616
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 7.963272120200335e-05,
+      "loss": 0.1909,
+      "step": 617
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 7.959933222036729e-05,
+      "loss": 0.2619,
+      "step": 618
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 7.956594323873122e-05,
+      "loss": 0.298,
+      "step": 619
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 7.953255425709516e-05,
+      "loss": 0.2777,
+      "step": 620
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 7.94991652754591e-05,
+      "loss": 0.2655,
+      "step": 621
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 7.946577629382305e-05,
+      "loss": 0.3785,
+      "step": 622
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 7.943238731218698e-05,
+      "loss": 0.2422,
+      "step": 623
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.939899833055092e-05,
+      "loss": 0.2284,
+      "step": 624
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.936560934891486e-05,
+      "loss": 0.2843,
+      "step": 625
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.933222036727879e-05,
+      "loss": 0.177,
+      "step": 626
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.929883138564274e-05,
+      "loss": 0.2894,
+      "step": 627
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.926544240400668e-05,
+      "loss": 0.2172,
+      "step": 628
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.923205342237062e-05,
+      "loss": 0.264,
+      "step": 629
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 7.919866444073457e-05,
+      "loss": 0.2595,
+      "step": 630
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 7.91652754590985e-05,
+      "loss": 0.2299,
+      "step": 631
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 7.913188647746244e-05,
+      "loss": 0.2649,
+      "step": 632
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 7.909849749582639e-05,
+      "loss": 0.3105,
+      "step": 633
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 7.906510851419033e-05,
+      "loss": 0.2411,
+      "step": 634
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 7.903171953255426e-05,
+      "loss": 0.2138,
+      "step": 635
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 7.89983305509182e-05,
+      "loss": 0.2531,
+      "step": 636
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 7.896494156928214e-05,
+      "loss": 0.3214,
+      "step": 637
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 7.893155258764609e-05,
+      "loss": 0.2932,
+      "step": 638
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 7.889816360601002e-05,
+      "loss": 0.3028,
+      "step": 639
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.886477462437396e-05,
+      "loss": 0.3011,
+      "step": 640
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.88313856427379e-05,
+      "loss": 0.2424,
+      "step": 641
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 7.879799666110183e-05,
+      "loss": 0.257,
+      "step": 642
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.876460767946578e-05,
+      "loss": 0.2756,
+      "step": 643
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.873121869782972e-05,
+      "loss": 0.1906,
+      "step": 644
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 7.869782971619366e-05,
+      "loss": 0.2591,
+      "step": 645
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 7.86644407345576e-05,
+      "loss": 0.3239,
+      "step": 646
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 7.863105175292154e-05,
+      "loss": 0.2301,
+      "step": 647
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 7.859766277128548e-05,
+      "loss": 0.3172,
+      "step": 648
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 7.856427378964943e-05,
+      "loss": 0.2436,
+      "step": 649
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 7.853088480801337e-05,
+      "loss": 0.2327,
+      "step": 650
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 7.84974958263773e-05,
+      "loss": 0.2508,
+      "step": 651
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 7.846410684474124e-05,
+      "loss": 0.2944,
+      "step": 652
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 7.843071786310517e-05,
+      "loss": 0.2065,
+      "step": 653
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 7.839732888146912e-05,
+      "loss": 0.2155,
+      "step": 654
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 7.836393989983306e-05,
+      "loss": 0.3273,
+      "step": 655
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 7.8330550918197e-05,
+      "loss": 0.2869,
+      "step": 656
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 7.829716193656093e-05,
+      "loss": 0.2275,
+      "step": 657
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.826377295492487e-05,
+      "loss": 0.3173,
+      "step": 658
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.823038397328882e-05,
+      "loss": 0.2881,
+      "step": 659
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 7.819699499165276e-05,
+      "loss": 0.2681,
+      "step": 660
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.81636060100167e-05,
+      "loss": 0.2823,
+      "step": 661
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.813021702838063e-05,
+      "loss": 0.3213,
+      "step": 662
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 7.809682804674458e-05,
+      "loss": 0.3241,
+      "step": 663
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.806343906510852e-05,
+      "loss": 0.3379,
+      "step": 664
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.803005008347247e-05,
+      "loss": 0.3505,
+      "step": 665
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 7.79966611018364e-05,
+      "loss": 0.3336,
+      "step": 666
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 7.796327212020034e-05,
+      "loss": 0.2596,
+      "step": 667
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 7.792988313856428e-05,
+      "loss": 0.2873,
+      "step": 668
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 7.789649415692821e-05,
+      "loss": 0.2387,
+      "step": 669
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 7.786310517529216e-05,
+      "loss": 0.2758,
+      "step": 670
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 7.78297161936561e-05,
+      "loss": 0.2483,
+      "step": 671
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 7.779632721202004e-05,
+      "loss": 0.3107,
+      "step": 672
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 7.776293823038397e-05,
+      "loss": 0.2713,
+      "step": 673
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 7.772954924874791e-05,
+      "loss": 0.2263,
+      "step": 674
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 7.769616026711186e-05,
+      "loss": 0.2408,
+      "step": 675
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 7.76627712854758e-05,
+      "loss": 0.2399,
+      "step": 676
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 7.762938230383973e-05,
+      "loss": 0.3011,
+      "step": 677
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 7.759599332220367e-05,
+      "loss": 0.3638,
+      "step": 678
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 7.756260434056761e-05,
+      "loss": 0.2126,
+      "step": 679
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 7.752921535893156e-05,
+      "loss": 0.2825,
+      "step": 680
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 7.74958263772955e-05,
+      "loss": 0.2718,
+      "step": 681
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 7.746243739565944e-05,
+      "loss": 0.1944,
+      "step": 682
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 7.742904841402338e-05,
+      "loss": 0.2262,
+      "step": 683
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 7.739565943238732e-05,
+      "loss": 0.3614,
+      "step": 684
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 7.736227045075125e-05,
+      "loss": 0.2296,
+      "step": 685
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 7.73288814691152e-05,
+      "loss": 0.2769,
+      "step": 686
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 7.729549248747914e-05,
+      "loss": 0.2912,
+      "step": 687
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 7.726210350584308e-05,
+      "loss": 0.2583,
+      "step": 688
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 7.722871452420701e-05,
+      "loss": 0.2557,
+      "step": 689
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 7.719532554257095e-05,
+      "loss": 0.2725,
+      "step": 690
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 7.71619365609349e-05,
+      "loss": 0.2423,
+      "step": 691
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 7.712854757929884e-05,
+      "loss": 0.2601,
+      "step": 692
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.709515859766277e-05,
+      "loss": 0.2469,
+      "step": 693
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.706176961602671e-05,
+      "loss": 0.2462,
+      "step": 694
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.702838063439065e-05,
+      "loss": 0.275,
+      "step": 695
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.69949916527546e-05,
+      "loss": 0.214,
+      "step": 696
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.696160267111853e-05,
+      "loss": 0.327,
+      "step": 697
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.692821368948247e-05,
+      "loss": 0.3124,
+      "step": 698
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.689482470784642e-05,
+      "loss": 0.2721,
+      "step": 699
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.686143572621036e-05,
+      "loss": 0.2742,
+      "step": 700
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.682804674457429e-05,
+      "loss": 0.3206,
+      "step": 701
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 7.679465776293824e-05,
+      "loss": 0.2922,
+      "step": 702
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.676126878130218e-05,
+      "loss": 0.2757,
+      "step": 703
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.672787979966612e-05,
+      "loss": 0.2496,
+      "step": 704
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.669449081803005e-05,
+      "loss": 0.3071,
+      "step": 705
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.666110183639399e-05,
+      "loss": 0.2277,
+      "step": 706
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.662771285475794e-05,
+      "loss": 0.215,
+      "step": 707
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.659432387312188e-05,
+      "loss": 0.2452,
+      "step": 708
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.656093489148581e-05,
+      "loss": 0.2582,
+      "step": 709
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.652754590984975e-05,
+      "loss": 0.2371,
+      "step": 710
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.649415692821369e-05,
+      "loss": 0.2859,
+      "step": 711
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.646076794657764e-05,
+      "loss": 0.2556,
+      "step": 712
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.642737896494157e-05,
+      "loss": 0.2146,
+      "step": 713
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.639398998330551e-05,
+      "loss": 0.2698,
+      "step": 714
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.636060100166945e-05,
+      "loss": 0.2572,
+      "step": 715
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.63272120200334e-05,
+      "loss": 0.271,
+      "step": 716
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 7.629382303839733e-05,
+      "loss": 0.2393,
+      "step": 717
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.626043405676128e-05,
+      "loss": 0.2575,
+      "step": 718
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.622704507512522e-05,
+      "loss": 0.2318,
+      "step": 719
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.619365609348916e-05,
+      "loss": 0.3243,
+      "step": 720
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.616026711185309e-05,
+      "loss": 0.2589,
+      "step": 721
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.612687813021703e-05,
+      "loss": 0.2785,
+      "step": 722
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.609348914858098e-05,
+      "loss": 0.3601,
+      "step": 723
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.606010016694492e-05,
+      "loss": 0.2406,
+      "step": 724
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.602671118530885e-05,
+      "loss": 0.2853,
+      "step": 725
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.599332220367279e-05,
+      "loss": 0.333,
+      "step": 726
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.595993322203673e-05,
+      "loss": 0.2478,
+      "step": 727
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.592654424040068e-05,
+      "loss": 0.2418,
+      "step": 728
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.589315525876461e-05,
+      "loss": 0.272,
+      "step": 729
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.585976627712855e-05,
+      "loss": 0.314,
+      "step": 730
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.582637729549249e-05,
+      "loss": 0.2993,
+      "step": 731
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.579298831385642e-05,
+      "loss": 0.3551,
+      "step": 732
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.575959933222037e-05,
+      "loss": 0.3423,
+      "step": 733
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.572621035058431e-05,
+      "loss": 0.2462,
+      "step": 734
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.569282136894826e-05,
+      "loss": 0.3203,
+      "step": 735
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.56594323873122e-05,
+      "loss": 0.3026,
+      "step": 736
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.562604340567613e-05,
+      "loss": 0.2601,
+      "step": 737
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.559265442404007e-05,
+      "loss": 0.2327,
+      "step": 738
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.555926544240402e-05,
+      "loss": 0.3003,
+      "step": 739
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.552587646076796e-05,
+      "loss": 0.2622,
+      "step": 740
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.549248747913189e-05,
+      "loss": 0.2454,
+      "step": 741
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.545909849749583e-05,
+      "loss": 0.3481,
+      "step": 742
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.542570951585976e-05,
+      "loss": 0.3586,
+      "step": 743
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.539232053422371e-05,
+      "loss": 0.2833,
+      "step": 744
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.535893155258765e-05,
+      "loss": 0.2644,
+      "step": 745
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.532554257095159e-05,
+      "loss": 0.2322,
+      "step": 746
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.529215358931552e-05,
+      "loss": 0.2239,
+      "step": 747
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.525876460767946e-05,
+      "loss": 0.3035,
+      "step": 748
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.52253756260434e-05,
+      "loss": 0.3679,
+      "step": 749
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.519198664440735e-05,
+      "loss": 0.3109,
+      "step": 750
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 2.223313093185425,
+      "eval_runtime": 20.2294,
+      "eval_samples_per_second": 9.887,
+      "eval_steps_per_second": 1.236,
+      "step": 750
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.515859766277128e-05,
+      "loss": 0.138,
+      "step": 751
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.512520868113523e-05,
+      "loss": 0.185,
+      "step": 752
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.509181969949917e-05,
+      "loss": 0.1429,
+      "step": 753
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.505843071786311e-05,
+      "loss": 0.1948,
+      "step": 754
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.502504173622706e-05,
+      "loss": 0.2563,
+      "step": 755
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.4991652754591e-05,
+      "loss": 0.1463,
+      "step": 756
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.495826377295493e-05,
+      "loss": 0.1368,
+      "step": 757
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.492487479131887e-05,
+      "loss": 0.1922,
+      "step": 758
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.48914858096828e-05,
+      "loss": 0.1592,
+      "step": 759
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.485809682804675e-05,
+      "loss": 0.1574,
+      "step": 760
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.482470784641069e-05,
+      "loss": 0.1497,
+      "step": 761
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.479131886477463e-05,
+      "loss": 0.1925,
+      "step": 762
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.475792988313856e-05,
+      "loss": 0.2088,
+      "step": 763
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.47245409015025e-05,
+      "loss": 0.2189,
+      "step": 764
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.469115191986644e-05,
+      "loss": 0.1963,
+      "step": 765
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.465776293823039e-05,
+      "loss": 0.1987,
+      "step": 766
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.462437395659432e-05,
+      "loss": 0.235,
+      "step": 767
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.459098497495826e-05,
+      "loss": 0.2005,
+      "step": 768
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.455759599332221e-05,
+      "loss": 0.1407,
+      "step": 769
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.452420701168615e-05,
+      "loss": 0.2443,
+      "step": 770
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.44908180300501e-05,
+      "loss": 0.2454,
+      "step": 771
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.445742904841403e-05,
+      "loss": 0.1579,
+      "step": 772
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.442404006677797e-05,
+      "loss": 0.1704,
+      "step": 773
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.439065108514191e-05,
+      "loss": 0.1553,
+      "step": 774
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.435726210350584e-05,
+      "loss": 0.1853,
+      "step": 775
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.43238731218698e-05,
+      "loss": 0.1906,
+      "step": 776
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.429048414023373e-05,
+      "loss": 0.1623,
+      "step": 777
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.425709515859767e-05,
+      "loss": 0.1572,
+      "step": 778
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.42237061769616e-05,
+      "loss": 0.1863,
+      "step": 779
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 7.419031719532554e-05,
+      "loss": 0.2126,
+      "step": 780
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 7.415692821368948e-05,
+      "loss": 0.164,
+      "step": 781
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 7.412353923205343e-05,
+      "loss": 0.2037,
+      "step": 782
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 7.409015025041736e-05,
+      "loss": 0.2271,
+      "step": 783
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 7.40567612687813e-05,
+      "loss": 0.1585,
+      "step": 784
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 7.402337228714524e-05,
+      "loss": 0.1759,
+      "step": 785
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 7.398998330550919e-05,
+      "loss": 0.2356,
+      "step": 786
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 7.395659432387312e-05,
+      "loss": 0.2885,
+      "step": 787
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 7.392320534223707e-05,
+      "loss": 0.1904,
+      "step": 788
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 7.388981636060101e-05,
+      "loss": 0.1712,
+      "step": 789
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 7.385642737896495e-05,
+      "loss": 0.1842,
+      "step": 790
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 7.382303839732888e-05,
+      "loss": 0.1819,
+      "step": 791
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 7.378964941569283e-05,
+      "loss": 0.2327,
+      "step": 792
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 7.375626043405677e-05,
+      "loss": 0.1538,
+      "step": 793
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 7.37228714524207e-05,
+      "loss": 0.1759,
+      "step": 794
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 7.368948247078464e-05,
+      "loss": 0.1902,
+      "step": 795
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 7.365609348914858e-05,
+      "loss": 0.2768,
+      "step": 796
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 7.362270450751253e-05,
+      "loss": 0.1851,
+      "step": 797
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 7.358931552587647e-05,
+      "loss": 0.1482,
+      "step": 798
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 7.35559265442404e-05,
+      "loss": 0.1974,
+      "step": 799
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 7.352253756260434e-05,
+      "loss": 0.188,
+      "step": 800
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 7.348914858096828e-05,
+      "loss": 0.204,
+      "step": 801
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 7.345575959933221e-05,
+      "loss": 0.1463,
+      "step": 802
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 7.342237061769616e-05,
+      "loss": 0.1127,
+      "step": 803
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 7.33889816360601e-05,
+      "loss": 0.2108,
+      "step": 804
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 7.335559265442405e-05,
+      "loss": 0.1522,
+      "step": 805
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 7.332220367278799e-05,
+      "loss": 0.2224,
+      "step": 806
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 7.328881469115192e-05,
+      "loss": 0.1892,
+      "step": 807
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 7.325542570951587e-05,
+      "loss": 0.2045,
+      "step": 808
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 7.322203672787981e-05,
+      "loss": 0.1638,
+      "step": 809
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 7.318864774624375e-05,
+      "loss": 0.2112,
+      "step": 810
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 7.315525876460768e-05,
+      "loss": 0.3326,
+      "step": 811
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 7.312186978297162e-05,
+      "loss": 0.2403,
+      "step": 812
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 7.308848080133557e-05,
+      "loss": 0.1884,
+      "step": 813
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 7.30550918196995e-05,
+      "loss": 0.1898,
+      "step": 814
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 7.302170283806344e-05,
+      "loss": 0.1884,
+      "step": 815
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 7.298831385642738e-05,
+      "loss": 0.1843,
+      "step": 816
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 7.295492487479132e-05,
+      "loss": 0.1464,
+      "step": 817
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 7.292153589315525e-05,
+      "loss": 0.182,
+      "step": 818
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 7.28881469115192e-05,
+      "loss": 0.166,
+      "step": 819
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 7.285475792988314e-05,
+      "loss": 0.1591,
+      "step": 820
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 7.282136894824708e-05,
+      "loss": 0.1942,
+      "step": 821
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 7.278797996661103e-05,
+      "loss": 0.1668,
+      "step": 822
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 7.275459098497496e-05,
+      "loss": 0.1335,
+      "step": 823
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 7.272120200333891e-05,
+      "loss": 0.168,
+      "step": 824
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 7.268781302170285e-05,
+      "loss": 0.196,
+      "step": 825
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 7.265442404006679e-05,
+      "loss": 0.1719,
+      "step": 826
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 7.262103505843072e-05,
+      "loss": 0.2467,
+      "step": 827
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 7.258764607679466e-05,
+      "loss": 0.1921,
+      "step": 828
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 7.255425709515861e-05,
+      "loss": 0.2083,
+      "step": 829
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 7.252086811352255e-05,
+      "loss": 0.11,
+      "step": 830
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 7.248747913188648e-05,
+      "loss": 0.1381,
+      "step": 831
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 7.245409015025042e-05,
+      "loss": 0.1485,
+      "step": 832
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 7.242070116861435e-05,
+      "loss": 0.199,
+      "step": 833
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 7.238731218697829e-05,
+      "loss": 0.2068,
+      "step": 834
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 7.235392320534224e-05,
+      "loss": 0.1755,
+      "step": 835
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 7.232053422370618e-05,
+      "loss": 0.1794,
+      "step": 836
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 7.228714524207011e-05,
+      "loss": 0.1942,
+      "step": 837
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 7.225375626043405e-05,
+      "loss": 0.1795,
+      "step": 838
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 7.2220367278798e-05,
+      "loss": 0.1267,
+      "step": 839
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 7.218697829716194e-05,
+      "loss": 0.1189,
+      "step": 840
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 7.215358931552589e-05,
+      "loss": 0.2513,
+      "step": 841
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 7.212020033388982e-05,
+      "loss": 0.212,
+      "step": 842
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 7.208681135225376e-05,
+      "loss": 0.1647,
+      "step": 843
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 7.20534223706177e-05,
+      "loss": 0.1696,
+      "step": 844
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 7.202003338898165e-05,
+      "loss": 0.1966,
+      "step": 845
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 7.198664440734558e-05,
+      "loss": 0.1984,
+      "step": 846
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 7.195325542570952e-05,
+      "loss": 0.1401,
+      "step": 847
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 7.191986644407346e-05,
+      "loss": 0.2334,
+      "step": 848
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 7.18864774624374e-05,
+      "loss": 0.2344,
+      "step": 849
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 7.185308848080133e-05,
+      "loss": 0.1682,
+      "step": 850
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 7.181969949916528e-05,
+      "loss": 0.2078,
+      "step": 851
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 7.178631051752922e-05,
+      "loss": 0.2075,
+      "step": 852
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 7.175292153589315e-05,
+      "loss": 0.1479,
+      "step": 853
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 7.171953255425709e-05,
+      "loss": 0.149,
+      "step": 854
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 7.168614357262103e-05,
+      "loss": 0.1662,
+      "step": 855
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 7.165275459098498e-05,
+      "loss": 0.2011,
+      "step": 856
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 7.161936560934891e-05,
+      "loss": 0.2364,
+      "step": 857
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 7.158597662771286e-05,
+      "loss": 0.14,
+      "step": 858
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 7.15525876460768e-05,
+      "loss": 0.1892,
+      "step": 859
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 7.151919866444074e-05,
+      "loss": 0.1721,
+      "step": 860
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 7.148580968280469e-05,
+      "loss": 0.1855,
+      "step": 861
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 7.145242070116862e-05,
+      "loss": 0.1724,
+      "step": 862
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 7.141903171953256e-05,
+      "loss": 0.1932,
+      "step": 863
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 7.13856427378965e-05,
+      "loss": 0.2193,
+      "step": 864
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 7.135225375626043e-05,
+      "loss": 0.3077,
+      "step": 865
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 7.131886477462437e-05,
+      "loss": 0.1436,
+      "step": 866
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 7.128547579298832e-05,
+      "loss": 0.1511,
+      "step": 867
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 7.125208681135226e-05,
+      "loss": 0.1352,
+      "step": 868
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 7.12186978297162e-05,
+      "loss": 0.1789,
+      "step": 869
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 7.118530884808013e-05,
+      "loss": 0.1802,
+      "step": 870
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 7.115191986644407e-05,
+      "loss": 0.1598,
+      "step": 871
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 7.111853088480802e-05,
+      "loss": 0.2463,
+      "step": 872
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 7.108514190317195e-05,
+      "loss": 0.1816,
+      "step": 873
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 7.105175292153589e-05,
+      "loss": 0.1863,
+      "step": 874
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 7.101836393989984e-05,
+      "loss": 0.2699,
+      "step": 875
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 7.098497495826378e-05,
+      "loss": 0.1886,
+      "step": 876
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 7.095158597662773e-05,
+      "loss": 0.2246,
+      "step": 877
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 7.091819699499166e-05,
+      "loss": 0.2031,
+      "step": 878
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 7.08848080133556e-05,
+      "loss": 0.2034,
+      "step": 879
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 7.085141903171954e-05,
+      "loss": 0.1832,
+      "step": 880
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 7.081803005008347e-05,
+      "loss": 0.1635,
+      "step": 881
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 7.078464106844741e-05,
+      "loss": 0.1441,
+      "step": 882
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 7.075125208681136e-05,
+      "loss": 0.1771,
+      "step": 883
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 7.07178631051753e-05,
+      "loss": 0.1827,
+      "step": 884
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 7.068447412353923e-05,
+      "loss": 0.1773,
+      "step": 885
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 7.065108514190317e-05,
+      "loss": 0.1818,
+      "step": 886
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 7.06176961602671e-05,
+      "loss": 0.226,
+      "step": 887
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 7.058430717863106e-05,
+      "loss": 0.1773,
+      "step": 888
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 7.055091819699499e-05,
+      "loss": 0.1517,
+      "step": 889
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 7.051752921535893e-05,
+      "loss": 0.1699,
+      "step": 890
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 7.048414023372288e-05,
+      "loss": 0.2006,
+      "step": 891
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 7.045075125208682e-05,
+      "loss": 0.18,
+      "step": 892
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 7.041736227045077e-05,
+      "loss": 0.2269,
+      "step": 893
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 7.03839732888147e-05,
+      "loss": 0.1715,
+      "step": 894
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 7.035058430717864e-05,
+      "loss": 0.2013,
+      "step": 895
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 7.031719532554258e-05,
+      "loss": 0.1713,
+      "step": 896
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 7.028380634390651e-05,
+      "loss": 0.1979,
+      "step": 897
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 7.025041736227045e-05,
+      "loss": 0.1819,
+      "step": 898
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 7.02170283806344e-05,
+      "loss": 0.1922,
+      "step": 899
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 7.02170283806344e-05,
+      "loss": 0.3597,
+      "step": 900
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 2.419039011001587,
+      "eval_runtime": 20.2231,
+      "eval_samples_per_second": 9.89,
+      "eval_steps_per_second": 1.236,
+      "step": 900
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 7.018363939899834e-05,
+      "loss": 0.1246,
+      "step": 901
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 7.015025041736227e-05,
+      "loss": 0.1756,
+      "step": 902
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 7.011686143572621e-05,
+      "loss": 0.104,
+      "step": 903
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 7.008347245409015e-05,
+      "loss": 0.0997,
+      "step": 904
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 7.00500834724541e-05,
+      "loss": 0.0808,
+      "step": 905
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 7.001669449081803e-05,
+      "loss": 0.1626,
+      "step": 906
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 6.998330550918197e-05,
+      "loss": 0.228,
+      "step": 907
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 6.99499165275459e-05,
+      "loss": 0.1065,
+      "step": 908
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 6.991652754590986e-05,
+      "loss": 0.1429,
+      "step": 909
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 6.988313856427379e-05,
+      "loss": 0.0912,
+      "step": 910
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 6.984974958263774e-05,
+      "loss": 0.141,
+      "step": 911
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 6.981636060100168e-05,
+      "loss": 0.1569,
+      "step": 912
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 6.978297161936562e-05,
+      "loss": 0.1208,
+      "step": 913
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 6.974958263772955e-05,
+      "loss": 0.1076,
+      "step": 914
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 6.971619365609349e-05,
+      "loss": 0.1243,
+      "step": 915
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 6.968280467445744e-05,
+      "loss": 0.1322,
+      "step": 916
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 6.964941569282138e-05,
+      "loss": 0.1545,
+      "step": 917
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 6.961602671118531e-05,
+      "loss": 0.1286,
+      "step": 918
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 6.958263772954925e-05,
+      "loss": 0.1412,
+      "step": 919
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 6.954924874791319e-05,
+      "loss": 0.1597,
+      "step": 920
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 6.951585976627714e-05,
+      "loss": 0.1296,
+      "step": 921
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 6.948247078464107e-05,
+      "loss": 0.1411,
+      "step": 922
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 6.944908180300501e-05,
+      "loss": 0.1468,
+      "step": 923
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 6.941569282136894e-05,
+      "loss": 0.1253,
+      "step": 924
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 6.938230383973288e-05,
+      "loss": 0.1094,
+      "step": 925
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 6.934891485809683e-05,
+      "loss": 0.2346,
+      "step": 926
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 6.931552587646077e-05,
+      "loss": 0.1321,
+      "step": 927
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 6.928213689482472e-05,
+      "loss": 0.108,
+      "step": 928
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 6.924874791318865e-05,
+      "loss": 0.1208,
+      "step": 929
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 6.921535893155259e-05,
+      "loss": 0.0887,
+      "step": 930
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 6.918196994991654e-05,
+      "loss": 0.152,
+      "step": 931
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 6.914858096828048e-05,
+      "loss": 0.1201,
+      "step": 932
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 6.911519198664441e-05,
+      "loss": 0.2021,
+      "step": 933
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 6.908180300500835e-05,
+      "loss": 0.1545,
+      "step": 934
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 6.904841402337229e-05,
+      "loss": 0.1039,
+      "step": 935
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 6.901502504173622e-05,
+      "loss": 0.1611,
+      "step": 936
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 6.898163606010017e-05,
+      "loss": 0.1404,
+      "step": 937
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 6.894824707846411e-05,
+      "loss": 0.1301,
+      "step": 938
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 6.891485809682805e-05,
+      "loss": 0.114,
+      "step": 939
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 6.888146911519198e-05,
+      "loss": 0.1888,
+      "step": 940
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 6.884808013355592e-05,
+      "loss": 0.1347,
+      "step": 941
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 6.881469115191987e-05,
+      "loss": 0.1129,
+      "step": 942
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 6.878130217028381e-05,
+      "loss": 0.136,
+      "step": 943
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 6.874791318864774e-05,
+      "loss": 0.1386,
+      "step": 944
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 6.87145242070117e-05,
+      "loss": 0.15,
+      "step": 945
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 6.868113522537563e-05,
+      "loss": 0.0888,
+      "step": 946
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 6.864774624373958e-05,
+      "loss": 0.1191,
+      "step": 947
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 6.861435726210352e-05,
+      "loss": 0.1299,
+      "step": 948
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 6.858096828046745e-05,
+      "loss": 0.1237,
+      "step": 949
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 6.854757929883139e-05,
+      "loss": 0.1058,
+      "step": 950
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 6.851419031719533e-05,
+      "loss": 0.1415,
+      "step": 951
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 6.848080133555926e-05,
+      "loss": 0.1115,
+      "step": 952
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 6.844741235392321e-05,
+      "loss": 0.0917,
+      "step": 953
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 6.841402337228715e-05,
+      "loss": 0.1047,
+      "step": 954
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 6.838063439065109e-05,
+      "loss": 0.1209,
+      "step": 955
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 6.834724540901502e-05,
+      "loss": 0.1407,
+      "step": 956
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 6.831385642737896e-05,
+      "loss": 0.1926,
+      "step": 957
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 6.828046744574291e-05,
+      "loss": 0.1151,
+      "step": 958
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 6.824707846410685e-05,
+      "loss": 0.1028,
+      "step": 959
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 6.821368948247078e-05,
+      "loss": 0.1233,
+      "step": 960
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 6.818030050083472e-05,
+      "loss": 0.1358,
+      "step": 961
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 6.814691151919867e-05,
+      "loss": 0.1522,
+      "step": 962
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 6.811352253756261e-05,
+      "loss": 0.1565,
+      "step": 963
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 6.808013355592656e-05,
+      "loss": 0.1372,
+      "step": 964
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 6.80467445742905e-05,
+      "loss": 0.1076,
+      "step": 965
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 6.801335559265443e-05,
+      "loss": 0.1798,
+      "step": 966
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 6.797996661101837e-05,
+      "loss": 0.1613,
+      "step": 967
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 6.79465776293823e-05,
+      "loss": 0.1107,
+      "step": 968
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 6.791318864774625e-05,
+      "loss": 0.1879,
+      "step": 969
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 6.787979966611019e-05,
+      "loss": 0.1229,
+      "step": 970
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 6.784641068447413e-05,
+      "loss": 0.1809,
+      "step": 971
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 6.781302170283806e-05,
+      "loss": 0.1397,
+      "step": 972
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 6.7779632721202e-05,
+      "loss": 0.0973,
+      "step": 973
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 6.774624373956595e-05,
+      "loss": 0.1263,
+      "step": 974
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 6.771285475792989e-05,
+      "loss": 0.1651,
+      "step": 975
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 6.767946577629382e-05,
+      "loss": 0.1511,
+      "step": 976
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 6.764607679465776e-05,
+      "loss": 0.1207,
+      "step": 977
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 6.76126878130217e-05,
+      "loss": 0.1206,
+      "step": 978
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 6.757929883138565e-05,
+      "loss": 0.1226,
+      "step": 979
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 6.754590984974958e-05,
+      "loss": 0.187,
+      "step": 980
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 6.751252086811353e-05,
+      "loss": 0.1194,
+      "step": 981
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 6.747913188647747e-05,
+      "loss": 0.1702,
+      "step": 982
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 6.74457429048414e-05,
+      "loss": 0.2073,
+      "step": 983
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 6.741235392320534e-05,
+      "loss": 0.2356,
+      "step": 984
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 6.737896494156929e-05,
+      "loss": 0.1152,
+      "step": 985
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 6.734557595993323e-05,
+      "loss": 0.1391,
+      "step": 986
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 6.731218697829717e-05,
+      "loss": 0.1788,
+      "step": 987
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 6.72787979966611e-05,
+      "loss": 0.1251,
+      "step": 988
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 6.724540901502504e-05,
+      "loss": 0.2079,
+      "step": 989
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 6.721202003338899e-05,
+      "loss": 0.1448,
+      "step": 990
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 6.717863105175293e-05,
+      "loss": 0.127,
+      "step": 991
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 6.714524207011686e-05,
+      "loss": 0.0955,
+      "step": 992
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 6.71118530884808e-05,
+      "loss": 0.1439,
+      "step": 993
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 6.707846410684474e-05,
+      "loss": 0.1316,
+      "step": 994
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 6.704507512520869e-05,
+      "loss": 0.1125,
+      "step": 995
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 6.701168614357262e-05,
+      "loss": 0.0949,
+      "step": 996
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 6.697829716193656e-05,
+      "loss": 0.1724,
+      "step": 997
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 6.694490818030051e-05,
+      "loss": 0.1675,
+      "step": 998
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 6.691151919866445e-05,
+      "loss": 0.1628,
+      "step": 999
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 6.687813021702838e-05,
+      "loss": 0.1516,
+      "step": 1000
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 6.684474123539233e-05,
+      "loss": 0.1099,
+      "step": 1001
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 6.681135225375627e-05,
+      "loss": 0.2315,
+      "step": 1002
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 6.67779632721202e-05,
+      "loss": 0.1193,
+      "step": 1003
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 6.674457429048414e-05,
+      "loss": 0.1284,
+      "step": 1004
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 6.671118530884808e-05,
+      "loss": 0.1563,
+      "step": 1005
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 6.667779632721203e-05,
+      "loss": 0.1186,
+      "step": 1006
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 6.664440734557597e-05,
+      "loss": 0.1514,
+      "step": 1007
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 6.66110183639399e-05,
+      "loss": 0.1587,
+      "step": 1008
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 6.657762938230384e-05,
+      "loss": 0.1407,
+      "step": 1009
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 6.654424040066778e-05,
+      "loss": 0.1346,
+      "step": 1010
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 6.651085141903173e-05,
+      "loss": 0.1628,
+      "step": 1011
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 6.647746243739566e-05,
+      "loss": 0.0908,
+      "step": 1012
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 6.64440734557596e-05,
+      "loss": 0.2343,
+      "step": 1013
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 6.641068447412354e-05,
+      "loss": 0.134,
+      "step": 1014
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 6.637729549248749e-05,
+      "loss": 0.2266,
+      "step": 1015
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 6.634390651085142e-05,
+      "loss": 0.132,
+      "step": 1016
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 6.631051752921537e-05,
+      "loss": 0.2634,
+      "step": 1017
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 6.627712854757931e-05,
+      "loss": 0.2212,
+      "step": 1018
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 6.624373956594325e-05,
+      "loss": 0.1178,
+      "step": 1019
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 6.621035058430718e-05,
+      "loss": 0.1277,
+      "step": 1020
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 6.617696160267112e-05,
+      "loss": 0.1233,
+      "step": 1021
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 6.614357262103507e-05,
+      "loss": 0.1725,
+      "step": 1022
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 6.6110183639399e-05,
+      "loss": 0.1391,
+      "step": 1023
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 6.607679465776294e-05,
+      "loss": 0.0933,
+      "step": 1024
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 6.604340567612688e-05,
+      "loss": 0.1442,
+      "step": 1025
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 6.601001669449081e-05,
+      "loss": 0.1516,
+      "step": 1026
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 6.597662771285476e-05,
+      "loss": 0.1296,
+      "step": 1027
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 6.59432387312187e-05,
+      "loss": 0.1342,
+      "step": 1028
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.590984974958264e-05,
+      "loss": 0.1108,
+      "step": 1029
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 6.587646076794657e-05,
+      "loss": 0.156,
+      "step": 1030
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 6.584307178631051e-05,
+      "loss": 0.1275,
+      "step": 1031
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 6.580968280467446e-05,
+      "loss": 0.1572,
+      "step": 1032
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.57762938230384e-05,
+      "loss": 0.199,
+      "step": 1033
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.574290484140235e-05,
+      "loss": 0.1132,
+      "step": 1034
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 6.570951585976628e-05,
+      "loss": 0.1663,
+      "step": 1035
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 6.567612687813022e-05,
+      "loss": 0.1285,
+      "step": 1036
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 6.564273789649416e-05,
+      "loss": 0.1356,
+      "step": 1037
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 6.560934891485811e-05,
+      "loss": 0.136,
+      "step": 1038
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 6.557595993322204e-05,
+      "loss": 0.1271,
+      "step": 1039
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 6.554257095158598e-05,
+      "loss": 0.132,
+      "step": 1040
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 6.550918196994992e-05,
+      "loss": 0.1507,
+      "step": 1041
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 6.547579298831385e-05,
+      "loss": 0.141,
+      "step": 1042
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 6.54424040066778e-05,
+      "loss": 0.1748,
+      "step": 1043
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 6.540901502504174e-05,
+      "loss": 0.1925,
+      "step": 1044
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 6.537562604340568e-05,
+      "loss": 0.1393,
+      "step": 1045
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 6.534223706176961e-05,
+      "loss": 0.1712,
+      "step": 1046
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 6.530884808013355e-05,
+      "loss": 0.0836,
+      "step": 1047
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 6.527545909849749e-05,
+      "loss": 0.1988,
+      "step": 1048
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 6.524207011686144e-05,
+      "loss": 0.123,
+      "step": 1049
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6.520868113522537e-05,
+      "loss": 0.1135,
+      "step": 1050
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 2.490980625152588,
+      "eval_runtime": 20.1976,
+      "eval_samples_per_second": 9.902,
+      "eval_steps_per_second": 1.238,
+      "step": 1050
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 6.517529215358932e-05,
+      "loss": 0.1075,
+      "step": 1051
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 6.514190317195326e-05,
+      "loss": 0.105,
+      "step": 1052
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 6.51085141903172e-05,
+      "loss": 0.1062,
+      "step": 1053
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 6.507512520868115e-05,
+      "loss": 0.0815,
+      "step": 1054
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 6.504173622704508e-05,
+      "loss": 0.1287,
+      "step": 1055
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 6.500834724540902e-05,
+      "loss": 0.09,
+      "step": 1056
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 6.497495826377296e-05,
+      "loss": 0.1253,
+      "step": 1057
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 6.49415692821369e-05,
+      "loss": 0.1612,
+      "step": 1058
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 6.490818030050084e-05,
+      "loss": 0.1157,
+      "step": 1059
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 6.487479131886478e-05,
+      "loss": 0.1443,
+      "step": 1060
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 6.484140233722872e-05,
+      "loss": 0.1381,
+      "step": 1061
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 6.480801335559265e-05,
+      "loss": 0.1318,
+      "step": 1062
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 6.477462437395659e-05,
+      "loss": 0.1138,
+      "step": 1063
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 6.474123539232054e-05,
+      "loss": 0.1387,
+      "step": 1064
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 6.470784641068448e-05,
+      "loss": 0.0843,
+      "step": 1065
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 6.467445742904841e-05,
+      "loss": 0.096,
+      "step": 1066
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 6.464106844741235e-05,
+      "loss": 0.0904,
+      "step": 1067
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 6.46076794657763e-05,
+      "loss": 0.1055,
+      "step": 1068
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 6.457429048414024e-05,
+      "loss": 0.1797,
+      "step": 1069
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 6.454090150250419e-05,
+      "loss": 0.0975,
+      "step": 1070
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 6.450751252086812e-05,
+      "loss": 0.2322,
+      "step": 1071
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 6.447412353923206e-05,
+      "loss": 0.1108,
+      "step": 1072
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 6.4440734557596e-05,
+      "loss": 0.2056,
+      "step": 1073
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 6.440734557595993e-05,
+      "loss": 0.1914,
+      "step": 1074
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 6.437395659432388e-05,
+      "loss": 0.1136,
+      "step": 1075
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 6.434056761268782e-05,
+      "loss": 0.1185,
+      "step": 1076
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 6.430717863105176e-05,
+      "loss": 0.1111,
+      "step": 1077
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 6.427378964941569e-05,
+      "loss": 0.0994,
+      "step": 1078
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 6.424040066777963e-05,
+      "loss": 0.1171,
+      "step": 1079
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 6.420701168614358e-05,
+      "loss": 0.1677,
+      "step": 1080
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 6.417362270450752e-05,
+      "loss": 0.1171,
+      "step": 1081
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 6.414023372287145e-05,
+      "loss": 0.0849,
+      "step": 1082
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 6.410684474123539e-05,
+      "loss": 0.1105,
+      "step": 1083
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 6.407345575959933e-05,
+      "loss": 0.1154,
+      "step": 1084
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 6.404006677796328e-05,
+      "loss": 0.1011,
+      "step": 1085
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 6.400667779632721e-05,
+      "loss": 0.1053,
+      "step": 1086
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 6.397328881469116e-05,
+      "loss": 0.1619,
+      "step": 1087
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 6.39398998330551e-05,
+      "loss": 0.126,
+      "step": 1088
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 6.390651085141904e-05,
+      "loss": 0.1238,
+      "step": 1089
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 6.387312186978297e-05,
+      "loss": 0.108,
+      "step": 1090
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 6.383973288814692e-05,
+      "loss": 0.0917,
+      "step": 1091
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 6.380634390651086e-05,
+      "loss": 0.1446,
+      "step": 1092
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 6.37729549248748e-05,
+      "loss": 0.086,
+      "step": 1093
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 6.373956594323873e-05,
+      "loss": 0.1193,
+      "step": 1094
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 6.370617696160267e-05,
+      "loss": 0.184,
+      "step": 1095
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 6.367278797996662e-05,
+      "loss": 0.089,
+      "step": 1096
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 6.363939899833056e-05,
+      "loss": 0.0939,
+      "step": 1097
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 6.360601001669449e-05,
+      "loss": 0.1368,
+      "step": 1098
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 6.357262103505843e-05,
+      "loss": 0.1339,
+      "step": 1099
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 6.353923205342237e-05,
+      "loss": 0.0802,
+      "step": 1100
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 6.35058430717863e-05,
+      "loss": 0.1255,
+      "step": 1101
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 6.347245409015025e-05,
+      "loss": 0.095,
+      "step": 1102
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 6.343906510851419e-05,
+      "loss": 0.1209,
+      "step": 1103
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 6.340567612687814e-05,
+      "loss": 0.0751,
+      "step": 1104
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 6.337228714524208e-05,
+      "loss": 0.1022,
+      "step": 1105
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 6.333889816360601e-05,
+      "loss": 0.1121,
+      "step": 1106
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 6.330550918196996e-05,
+      "loss": 0.1087,
+      "step": 1107
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 6.32721202003339e-05,
+      "loss": 0.106,
+      "step": 1108
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 6.323873121869784e-05,
+      "loss": 0.0749,
+      "step": 1109
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 6.320534223706177e-05,
+      "loss": 0.131,
+      "step": 1110
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 6.317195325542571e-05,
+      "loss": 0.1294,
+      "step": 1111
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 6.313856427378966e-05,
+      "loss": 0.1243,
+      "step": 1112
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 6.31051752921536e-05,
+      "loss": 0.1148,
+      "step": 1113
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 6.307178631051753e-05,
+      "loss": 0.1347,
+      "step": 1114
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 6.303839732888147e-05,
+      "loss": 0.1425,
+      "step": 1115
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 6.30050083472454e-05,
+      "loss": 0.1244,
+      "step": 1116
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 6.297161936560934e-05,
+      "loss": 0.1333,
+      "step": 1117
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 6.293823038397329e-05,
+      "loss": 0.0764,
+      "step": 1118
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 6.290484140233723e-05,
+      "loss": 0.0826,
+      "step": 1119
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 6.287145242070116e-05,
+      "loss": 0.1987,
+      "step": 1120
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 6.283806343906511e-05,
+      "loss": 0.097,
+      "step": 1121
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 6.280467445742905e-05,
+      "loss": 0.1926,
+      "step": 1122
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 6.2771285475793e-05,
+      "loss": 0.0747,
+      "step": 1123
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 6.273789649415694e-05,
+      "loss": 0.0837,
+      "step": 1124
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 6.270450751252087e-05,
+      "loss": 0.135,
+      "step": 1125
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 6.267111853088481e-05,
+      "loss": 0.1676,
+      "step": 1126
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 6.263772954924875e-05,
+      "loss": 0.1443,
+      "step": 1127
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 6.26043405676127e-05,
+      "loss": 0.1023,
+      "step": 1128
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 6.257095158597663e-05,
+      "loss": 0.1154,
+      "step": 1129
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 6.253756260434057e-05,
+      "loss": 0.0844,
+      "step": 1130
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 6.250417362270451e-05,
+      "loss": 0.1259,
+      "step": 1131
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 6.247078464106844e-05,
+      "loss": 0.1008,
+      "step": 1132
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 6.243739565943238e-05,
+      "loss": 0.1531,
+      "step": 1133
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 6.240400667779633e-05,
+      "loss": 0.1602,
+      "step": 1134
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 6.237061769616027e-05,
+      "loss": 0.0833,
+      "step": 1135
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 6.23372287145242e-05,
+      "loss": 0.1255,
+      "step": 1136
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 6.230383973288815e-05,
+      "loss": 0.1125,
+      "step": 1137
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 6.227045075125209e-05,
+      "loss": 0.1376,
+      "step": 1138
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 6.223706176961604e-05,
+      "loss": 0.1113,
+      "step": 1139
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 6.220367278797998e-05,
+      "loss": 0.1369,
+      "step": 1140
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 6.217028380634391e-05,
+      "loss": 0.1017,
+      "step": 1141
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 6.213689482470785e-05,
+      "loss": 0.1287,
+      "step": 1142
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 6.210350584307179e-05,
+      "loss": 0.1365,
+      "step": 1143
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 6.207011686143574e-05,
+      "loss": 0.1254,
+      "step": 1144
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 6.203672787979967e-05,
+      "loss": 0.0881,
+      "step": 1145
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 6.200333889816361e-05,
+      "loss": 0.0939,
+      "step": 1146
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 6.196994991652755e-05,
+      "loss": 0.1074,
+      "step": 1147
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 6.193656093489148e-05,
+      "loss": 0.1445,
+      "step": 1148
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 6.190317195325542e-05,
+      "loss": 0.0958,
+      "step": 1149
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 6.186978297161937e-05,
+      "loss": 0.1541,
+      "step": 1150
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 6.183639398998331e-05,
+      "loss": 0.1072,
+      "step": 1151
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 6.180300500834724e-05,
+      "loss": 0.0983,
+      "step": 1152
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 6.176961602671118e-05,
+      "loss": 0.078,
+      "step": 1153
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 6.173622704507513e-05,
+      "loss": 0.1348,
+      "step": 1154
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 6.170283806343907e-05,
+      "loss": 0.0992,
+      "step": 1155
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 6.166944908180302e-05,
+      "loss": 0.1029,
+      "step": 1156
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 6.163606010016695e-05,
+      "loss": 0.081,
+      "step": 1157
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 6.160267111853089e-05,
+      "loss": 0.1595,
+      "step": 1158
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 6.156928213689483e-05,
+      "loss": 0.1562,
+      "step": 1159
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 6.153589315525878e-05,
+      "loss": 0.0913,
+      "step": 1160
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 6.150250417362271e-05,
+      "loss": 0.1144,
+      "step": 1161
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 6.146911519198665e-05,
+      "loss": 0.087,
+      "step": 1162
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 6.143572621035059e-05,
+      "loss": 0.115,
+      "step": 1163
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 6.140233722871452e-05,
+      "loss": 0.123,
+      "step": 1164
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 6.136894824707846e-05,
+      "loss": 0.1183,
+      "step": 1165
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 6.133555926544241e-05,
+      "loss": 0.1005,
+      "step": 1166
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 6.130217028380635e-05,
+      "loss": 0.1826,
+      "step": 1167
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 6.126878130217028e-05,
+      "loss": 0.1141,
+      "step": 1168
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 6.123539232053422e-05,
+      "loss": 0.131,
+      "step": 1169
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 6.120200333889816e-05,
+      "loss": 0.1216,
+      "step": 1170
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 6.11686143572621e-05,
+      "loss": 0.125,
+      "step": 1171
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 6.113522537562604e-05,
+      "loss": 0.1269,
+      "step": 1172
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 6.110183639398999e-05,
+      "loss": 0.1665,
+      "step": 1173
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 6.106844741235393e-05,
+      "loss": 0.1475,
+      "step": 1174
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 6.103505843071786e-05,
+      "loss": 0.1294,
+      "step": 1175
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 6.100166944908181e-05,
+      "loss": 0.1002,
+      "step": 1176
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 6.0968280467445746e-05,
+      "loss": 0.0933,
+      "step": 1177
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 6.093489148580969e-05,
+      "loss": 0.0931,
+      "step": 1178
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 6.0901502504173626e-05,
+      "loss": 0.1622,
+      "step": 1179
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 6.086811352253756e-05,
+      "loss": 0.118,
+      "step": 1180
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 6.083472454090151e-05,
+      "loss": 0.1172,
+      "step": 1181
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 6.080133555926545e-05,
+      "loss": 0.1763,
+      "step": 1182
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 6.0767946577629386e-05,
+      "loss": 0.0954,
+      "step": 1183
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 6.073455759599332e-05,
+      "loss": 0.12,
+      "step": 1184
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 6.070116861435726e-05,
+      "loss": 0.1272,
+      "step": 1185
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 6.06677796327212e-05,
+      "loss": 0.0989,
+      "step": 1186
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 6.0634390651085146e-05,
+      "loss": 0.093,
+      "step": 1187
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 6.060100166944909e-05,
+      "loss": 0.0992,
+      "step": 1188
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 6.0567612687813026e-05,
+      "loss": 0.1805,
+      "step": 1189
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 6.053422370617696e-05,
+      "loss": 0.1137,
+      "step": 1190
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 6.05008347245409e-05,
+      "loss": 0.1642,
+      "step": 1191
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 6.046744574290485e-05,
+      "loss": 0.0915,
+      "step": 1192
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 6.0434056761268785e-05,
+      "loss": 0.0956,
+      "step": 1193
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 6.040066777963272e-05,
+      "loss": 0.0842,
+      "step": 1194
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 6.0367278797996665e-05,
+      "loss": 0.1124,
+      "step": 1195
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 6.03338898163606e-05,
+      "loss": 0.1232,
+      "step": 1196
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 6.030050083472455e-05,
+      "loss": 0.0869,
+      "step": 1197
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 6.026711185308849e-05,
+      "loss": 0.1186,
+      "step": 1198
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 6.0233722871452425e-05,
+      "loss": 0.1542,
+      "step": 1199
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 6.020033388981636e-05,
+      "loss": 0.0863,
+      "step": 1200
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 2.574554443359375,
+      "eval_runtime": 20.1935,
+      "eval_samples_per_second": 9.904,
+      "eval_steps_per_second": 1.238,
+      "step": 1200
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 6.01669449081803e-05,
+      "loss": 0.081,
+      "step": 1201
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 6.013355592654424e-05,
+      "loss": 0.0968,
+      "step": 1202
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 6.0100166944908185e-05,
+      "loss": 0.0828,
+      "step": 1203
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 6.006677796327213e-05,
+      "loss": 0.1276,
+      "step": 1204
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 6.0033388981636065e-05,
+      "loss": 0.1452,
+      "step": 1205
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 6e-05,
+      "loss": 0.0991,
+      "step": 1206
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 5.996661101836394e-05,
+      "loss": 0.1093,
+      "step": 1207
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 5.993322203672789e-05,
+      "loss": 0.0682,
+      "step": 1208
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 5.9899833055091825e-05,
+      "loss": 0.1016,
+      "step": 1209
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 5.986644407345576e-05,
+      "loss": 0.0704,
+      "step": 1210
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 5.98330550918197e-05,
+      "loss": 0.0843,
+      "step": 1211
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 5.979966611018364e-05,
+      "loss": 0.0558,
+      "step": 1212
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 5.9766277128547585e-05,
+      "loss": 0.0607,
+      "step": 1213
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 5.973288814691153e-05,
+      "loss": 0.0677,
+      "step": 1214
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 5.9699499165275465e-05,
+      "loss": 0.1046,
+      "step": 1215
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 5.96661101836394e-05,
+      "loss": 0.0636,
+      "step": 1216
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 5.963272120200334e-05,
+      "loss": 0.0725,
+      "step": 1217
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 5.9599332220367274e-05,
+      "loss": 0.0964,
+      "step": 1218
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 5.9565943238731224e-05,
+      "loss": 0.074,
+      "step": 1219
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 5.953255425709516e-05,
+      "loss": 0.1122,
+      "step": 1220
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 5.9499165275459104e-05,
+      "loss": 0.1759,
+      "step": 1221
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 5.946577629382304e-05,
+      "loss": 0.0686,
+      "step": 1222
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 5.943238731218698e-05,
+      "loss": 0.1163,
+      "step": 1223
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 5.939899833055093e-05,
+      "loss": 0.0714,
+      "step": 1224
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 5.9365609348914864e-05,
+      "loss": 0.0777,
+      "step": 1225
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 5.93322203672788e-05,
+      "loss": 0.0838,
+      "step": 1226
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 5.929883138564274e-05,
+      "loss": 0.1643,
+      "step": 1227
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 5.9265442404006674e-05,
+      "loss": 0.0892,
+      "step": 1228
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 5.9232053422370624e-05,
+      "loss": 0.0872,
+      "step": 1229
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 5.919866444073456e-05,
+      "loss": 0.1411,
+      "step": 1230
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 5.9165275459098504e-05,
+      "loss": 0.1415,
+      "step": 1231
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 5.913188647746244e-05,
+      "loss": 0.0581,
+      "step": 1232
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 5.909849749582638e-05,
+      "loss": 0.0904,
+      "step": 1233
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 5.9065108514190314e-05,
+      "loss": 0.0812,
+      "step": 1234
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 5.9031719532554264e-05,
+      "loss": 0.0991,
+      "step": 1235
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 5.89983305509182e-05,
+      "loss": 0.0748,
+      "step": 1236
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 5.896494156928214e-05,
+      "loss": 0.0915,
+      "step": 1237
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 5.893155258764608e-05,
+      "loss": 0.1335,
+      "step": 1238
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 5.889816360601002e-05,
+      "loss": 0.136,
+      "step": 1239
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 5.886477462437397e-05,
+      "loss": 0.0829,
+      "step": 1240
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 5.8831385642737904e-05,
+      "loss": 0.1187,
+      "step": 1241
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 5.879799666110184e-05,
+      "loss": 0.0805,
+      "step": 1242
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 5.876460767946578e-05,
+      "loss": 0.0868,
+      "step": 1243
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 5.8731218697829713e-05,
+      "loss": 0.0892,
+      "step": 1244
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 5.8697829716193664e-05,
+      "loss": 0.0849,
+      "step": 1245
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 5.86644407345576e-05,
+      "loss": 0.1042,
+      "step": 1246
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 5.863105175292154e-05,
+      "loss": 0.0777,
+      "step": 1247
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 5.859766277128548e-05,
+      "loss": 0.1488,
+      "step": 1248
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 5.856427378964942e-05,
+      "loss": 0.1146,
+      "step": 1249
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 5.853088480801335e-05,
+      "loss": 0.0738,
+      "step": 1250
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 5.84974958263773e-05,
+      "loss": 0.0916,
+      "step": 1251
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 5.846410684474124e-05,
+      "loss": 0.0789,
+      "step": 1252
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 5.8430717863105176e-05,
+      "loss": 0.0801,
+      "step": 1253
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 5.839732888146911e-05,
+      "loss": 0.0843,
+      "step": 1254
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 5.8363939899833056e-05,
+      "loss": 0.0951,
+      "step": 1255
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 5.8330550918197e-05,
+      "loss": 0.0775,
+      "step": 1256
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 5.829716193656094e-05,
+      "loss": 0.09,
+      "step": 1257
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 5.826377295492488e-05,
+      "loss": 0.1013,
+      "step": 1258
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 5.8230383973288816e-05,
+      "loss": 0.1448,
+      "step": 1259
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 5.819699499165275e-05,
+      "loss": 0.0724,
+      "step": 1260
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 5.81636060100167e-05,
+      "loss": 0.1045,
+      "step": 1261
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 5.813021702838064e-05,
+      "loss": 0.0916,
+      "step": 1262
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 5.8096828046744576e-05,
+      "loss": 0.0938,
+      "step": 1263
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 5.806343906510851e-05,
+      "loss": 0.1661,
+      "step": 1264
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 5.8030050083472456e-05,
+      "loss": 0.1286,
+      "step": 1265
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 5.799666110183639e-05,
+      "loss": 0.1163,
+      "step": 1266
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 5.796327212020034e-05,
+      "loss": 0.0624,
+      "step": 1267
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 5.792988313856428e-05,
+      "loss": 0.0925,
+      "step": 1268
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 5.7896494156928216e-05,
+      "loss": 0.088,
+      "step": 1269
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.786310517529215e-05,
+      "loss": 0.096,
+      "step": 1270
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.782971619365609e-05,
+      "loss": 0.0939,
+      "step": 1271
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 5.779632721202004e-05,
+      "loss": 0.1196,
+      "step": 1272
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.7762938230383976e-05,
+      "loss": 0.0731,
+      "step": 1273
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.772954924874792e-05,
+      "loss": 0.1054,
+      "step": 1274
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 5.7696160267111856e-05,
+      "loss": 0.1047,
+      "step": 1275
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.766277128547579e-05,
+      "loss": 0.1099,
+      "step": 1276
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.762938230383974e-05,
+      "loss": 0.1522,
+      "step": 1277
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5.759599332220368e-05,
+      "loss": 0.1257,
+      "step": 1278
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5.7562604340567616e-05,
+      "loss": 0.1052,
+      "step": 1279
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5.752921535893155e-05,
+      "loss": 0.0778,
+      "step": 1280
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5.749582637729549e-05,
+      "loss": 0.0849,
+      "step": 1281
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5.746243739565943e-05,
+      "loss": 0.1077,
+      "step": 1282
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5.7429048414023375e-05,
+      "loss": 0.2259,
+      "step": 1283
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5.739565943238732e-05,
+      "loss": 0.1956,
+      "step": 1284
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.7362270450751255e-05,
+      "loss": 0.1682,
+      "step": 1285
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.732888146911519e-05,
+      "loss": 0.0739,
+      "step": 1286
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 5.729549248747913e-05,
+      "loss": 0.0928,
+      "step": 1287
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 5.726210350584308e-05,
+      "loss": 0.0756,
+      "step": 1288
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 5.7228714524207015e-05,
+      "loss": 0.1143,
+      "step": 1289
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5.719532554257095e-05,
+      "loss": 0.0734,
+      "step": 1290
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.7161936560934895e-05,
+      "loss": 0.0849,
+      "step": 1291
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.712854757929883e-05,
+      "loss": 0.1068,
+      "step": 1292
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 5.709515859766278e-05,
+      "loss": 0.0739,
+      "step": 1293
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.706176961602672e-05,
+      "loss": 0.0922,
+      "step": 1294
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.7028380634390655e-05,
+      "loss": 0.0834,
+      "step": 1295
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 5.699499165275459e-05,
+      "loss": 0.1136,
+      "step": 1296
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 5.696160267111853e-05,
+      "loss": 0.1268,
+      "step": 1297
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 5.6928213689482465e-05,
+      "loss": 0.1105,
+      "step": 1298
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 5.6894824707846415e-05,
+      "loss": 0.1145,
+      "step": 1299
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5.686143572621035e-05,
+      "loss": 0.1192,
+      "step": 1300
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5.6828046744574295e-05,
+      "loss": 0.0799,
+      "step": 1301
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 5.679465776293823e-05,
+      "loss": 0.0803,
+      "step": 1302
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 5.676126878130217e-05,
+      "loss": 0.0812,
+      "step": 1303
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 5.672787979966612e-05,
+      "loss": 0.1145,
+      "step": 1304
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5.6694490818030055e-05,
+      "loss": 0.1102,
+      "step": 1305
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5.666110183639399e-05,
+      "loss": 0.0734,
+      "step": 1306
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5.662771285475793e-05,
+      "loss": 0.0851,
+      "step": 1307
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 5.659432387312187e-05,
+      "loss": 0.0723,
+      "step": 1308
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 5.6560934891485815e-05,
+      "loss": 0.0695,
+      "step": 1309
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 5.652754590984976e-05,
+      "loss": 0.1234,
+      "step": 1310
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 5.6494156928213694e-05,
+      "loss": 0.0886,
+      "step": 1311
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 5.646076794657763e-05,
+      "loss": 0.1418,
+      "step": 1312
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 5.642737896494157e-05,
+      "loss": 0.1193,
+      "step": 1313
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 5.639398998330552e-05,
+      "loss": 0.185,
+      "step": 1314
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 5.6360601001669454e-05,
+      "loss": 0.0782,
+      "step": 1315
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 5.632721202003339e-05,
+      "loss": 0.1569,
+      "step": 1316
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 5.629382303839733e-05,
+      "loss": 0.1235,
+      "step": 1317
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 5.626043405676127e-05,
+      "loss": 0.0838,
+      "step": 1318
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 5.622704507512521e-05,
+      "loss": 0.1498,
+      "step": 1319
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 5.619365609348916e-05,
+      "loss": 0.0935,
+      "step": 1320
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 5.6160267111853094e-05,
+      "loss": 0.0998,
+      "step": 1321
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 5.612687813021703e-05,
+      "loss": 0.11,
+      "step": 1322
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 5.609348914858097e-05,
+      "loss": 0.1364,
+      "step": 1323
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 5.6060100166944904e-05,
+      "loss": 0.1525,
+      "step": 1324
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 5.6026711185308854e-05,
+      "loss": 0.0778,
+      "step": 1325
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 5.599332220367279e-05,
+      "loss": 0.1182,
+      "step": 1326
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 5.5959933222036734e-05,
+      "loss": 0.1297,
+      "step": 1327
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 5.592654424040067e-05,
+      "loss": 0.0827,
+      "step": 1328
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 5.589315525876461e-05,
+      "loss": 0.0949,
+      "step": 1329
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 5.585976627712856e-05,
+      "loss": 0.1161,
+      "step": 1330
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 5.5826377295492494e-05,
+      "loss": 0.1045,
+      "step": 1331
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 5.579298831385643e-05,
+      "loss": 0.129,
+      "step": 1332
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 5.575959933222037e-05,
+      "loss": 0.2137,
+      "step": 1333
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 5.572621035058431e-05,
+      "loss": 0.1177,
+      "step": 1334
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 5.569282136894825e-05,
+      "loss": 0.123,
+      "step": 1335
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5.56594323873122e-05,
+      "loss": 0.0791,
+      "step": 1336
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5.5626043405676134e-05,
+      "loss": 0.1026,
+      "step": 1337
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 5.559265442404007e-05,
+      "loss": 0.0908,
+      "step": 1338
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 5.555926544240401e-05,
+      "loss": 0.0888,
+      "step": 1339
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 5.552587646076794e-05,
+      "loss": 0.1193,
+      "step": 1340
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.5492487479131893e-05,
+      "loss": 0.1622,
+      "step": 1341
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 5.545909849749583e-05,
+      "loss": 0.118,
+      "step": 1342
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 5.5425709515859767e-05,
+      "loss": 0.1121,
+      "step": 1343
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5.539232053422371e-05,
+      "loss": 0.1145,
+      "step": 1344
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.5358931552587646e-05,
+      "loss": 0.0867,
+      "step": 1345
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.53255425709516e-05,
+      "loss": 0.1445,
+      "step": 1346
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 5.529215358931553e-05,
+      "loss": 0.0967,
+      "step": 1347
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.525876460767947e-05,
+      "loss": 0.1388,
+      "step": 1348
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.5225375626043406e-05,
+      "loss": 0.093,
+      "step": 1349
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.519198664440734e-05,
+      "loss": 0.1082,
+      "step": 1350
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 2.6888976097106934,
+      "eval_runtime": 20.1924,
+      "eval_samples_per_second": 9.905,
+      "eval_steps_per_second": 1.238,
+      "step": 1350
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 5.5158597662771286e-05,
+      "loss": 0.0944,
+      "step": 1351
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 5.512520868113523e-05,
+      "loss": 0.0899,
+      "step": 1352
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 5.509181969949917e-05,
+      "loss": 0.1218,
+      "step": 1353
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5.505843071786311e-05,
+      "loss": 0.0789,
+      "step": 1354
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5.5025041736227046e-05,
+      "loss": 0.1704,
+      "step": 1355
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 5.499165275459098e-05,
+      "loss": 0.0783,
+      "step": 1356
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 5.495826377295493e-05,
+      "loss": 0.0703,
+      "step": 1357
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 5.492487479131887e-05,
+      "loss": 0.0925,
+      "step": 1358
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 5.4891485809682806e-05,
+      "loss": 0.1553,
+      "step": 1359
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5.485809682804674e-05,
+      "loss": 0.087,
+      "step": 1360
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5.4824707846410686e-05,
+      "loss": 0.0588,
+      "step": 1361
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 5.479131886477463e-05,
+      "loss": 0.0951,
+      "step": 1362
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 5.475792988313857e-05,
+      "loss": 0.0697,
+      "step": 1363
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 5.472454090150251e-05,
+      "loss": 0.12,
+      "step": 1364
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 5.4691151919866446e-05,
+      "loss": 0.0718,
+      "step": 1365
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 5.465776293823038e-05,
+      "loss": 0.1032,
+      "step": 1366
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 5.462437395659432e-05,
+      "loss": 0.1095,
+      "step": 1367
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 5.459098497495827e-05,
+      "loss": 0.0811,
+      "step": 1368
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 5.4557595993322206e-05,
+      "loss": 0.0737,
+      "step": 1369
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 5.452420701168615e-05,
+      "loss": 0.0649,
+      "step": 1370
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 5.4490818030050086e-05,
+      "loss": 0.1846,
+      "step": 1371
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 5.445742904841402e-05,
+      "loss": 0.0893,
+      "step": 1372
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 5.442404006677797e-05,
+      "loss": 0.0853,
+      "step": 1373
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 5.439065108514191e-05,
+      "loss": 0.1136,
+      "step": 1374
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 5.4357262103505845e-05,
+      "loss": 0.1714,
+      "step": 1375
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 5.432387312186978e-05,
+      "loss": 0.083,
+      "step": 1376
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 5.429048414023372e-05,
+      "loss": 0.0566,
+      "step": 1377
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 5.425709515859767e-05,
+      "loss": 0.1038,
+      "step": 1378
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 5.4223706176961605e-05,
+      "loss": 0.0833,
+      "step": 1379
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 5.419031719532555e-05,
+      "loss": 0.0988,
+      "step": 1380
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 5.4156928213689485e-05,
+      "loss": 0.1318,
+      "step": 1381
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 5.412353923205342e-05,
+      "loss": 0.0771,
+      "step": 1382
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 5.409015025041736e-05,
+      "loss": 0.0705,
+      "step": 1383
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 5.405676126878131e-05,
+      "loss": 0.0629,
+      "step": 1384
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 5.4023372287145245e-05,
+      "loss": 0.0848,
+      "step": 1385
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 5.398998330550918e-05,
+      "loss": 0.0865,
+      "step": 1386
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 5.3956594323873125e-05,
+      "loss": 0.0819,
+      "step": 1387
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 5.392320534223706e-05,
+      "loss": 0.1647,
+      "step": 1388
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 5.388981636060101e-05,
+      "loss": 0.0918,
+      "step": 1389
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 5.385642737896495e-05,
+      "loss": 0.0703,
+      "step": 1390
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 5.3823038397328885e-05,
+      "loss": 0.086,
+      "step": 1391
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 5.378964941569282e-05,
+      "loss": 0.0937,
+      "step": 1392
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 5.375626043405676e-05,
+      "loss": 0.1346,
+      "step": 1393
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 5.372287145242071e-05,
+      "loss": 0.0569,
+      "step": 1394
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 5.3689482470784645e-05,
+      "loss": 0.0737,
+      "step": 1395
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 5.365609348914858e-05,
+      "loss": 0.0885,
+      "step": 1396
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 5.3622704507512525e-05,
+      "loss": 0.0954,
+      "step": 1397
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 5.358931552587646e-05,
+      "loss": 0.0808,
+      "step": 1398
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 5.35559265442404e-05,
+      "loss": 0.0545,
+      "step": 1399
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 5.352253756260435e-05,
+      "loss": 0.0743,
+      "step": 1400
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 5.3489148580968285e-05,
+      "loss": 0.0959,
+      "step": 1401
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 5.345575959933222e-05,
+      "loss": 0.0801,
+      "step": 1402
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 5.342237061769616e-05,
+      "loss": 0.0786,
+      "step": 1403
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 5.33889816360601e-05,
+      "loss": 0.0681,
+      "step": 1404
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 5.3355592654424044e-05,
+      "loss": 0.0887,
+      "step": 1405
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 5.332220367278799e-05,
+      "loss": 0.0892,
+      "step": 1406
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 5.3288814691151924e-05,
+      "loss": 0.0723,
+      "step": 1407
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 5.325542570951586e-05,
+      "loss": 0.2277,
+      "step": 1408
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 5.32220367278798e-05,
+      "loss": 0.0978,
+      "step": 1409
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 5.318864774624375e-05,
+      "loss": 0.0531,
+      "step": 1410
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 5.3155258764607684e-05,
+      "loss": 0.0651,
+      "step": 1411
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 5.312186978297162e-05,
+      "loss": 0.0647,
+      "step": 1412
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 5.308848080133556e-05,
+      "loss": 0.1715,
+      "step": 1413
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 5.30550918196995e-05,
+      "loss": 0.0857,
+      "step": 1414
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 5.302170283806344e-05,
+      "loss": 0.0748,
+      "step": 1415
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 5.298831385642739e-05,
+      "loss": 0.1137,
+      "step": 1416
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 5.2954924874791324e-05,
+      "loss": 0.064,
+      "step": 1417
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 5.292153589315526e-05,
+      "loss": 0.0645,
+      "step": 1418
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 5.28881469115192e-05,
+      "loss": 0.0763,
+      "step": 1419
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 5.2854757929883134e-05,
+      "loss": 0.0896,
+      "step": 1420
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 5.2821368948247084e-05,
+      "loss": 0.0838,
+      "step": 1421
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 5.278797996661102e-05,
+      "loss": 0.1307,
+      "step": 1422
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 5.2754590984974964e-05,
+      "loss": 0.0873,
+      "step": 1423
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 5.27212020033389e-05,
+      "loss": 0.0842,
+      "step": 1424
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 5.268781302170284e-05,
+      "loss": 0.1416,
+      "step": 1425
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 5.265442404006679e-05,
+      "loss": 0.0836,
+      "step": 1426
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 5.2621035058430724e-05,
+      "loss": 0.1236,
+      "step": 1427
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 5.258764607679466e-05,
+      "loss": 0.1036,
+      "step": 1428
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5.25542570951586e-05,
+      "loss": 0.0704,
+      "step": 1429
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5.252086811352253e-05,
+      "loss": 0.099,
+      "step": 1430
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 5.248747913188648e-05,
+      "loss": 0.1462,
+      "step": 1431
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5.245409015025042e-05,
+      "loss": 0.1642,
+      "step": 1432
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5.2420701168614363e-05,
+      "loss": 0.0685,
+      "step": 1433
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 5.23873121869783e-05,
+      "loss": 0.0772,
+      "step": 1434
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 5.2353923205342237e-05,
+      "loss": 0.0794,
+      "step": 1435
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 5.232053422370617e-05,
+      "loss": 0.0932,
+      "step": 1436
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 5.228714524207012e-05,
+      "loss": 0.0732,
+      "step": 1437
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 5.225375626043406e-05,
+      "loss": 0.0897,
+      "step": 1438
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 5.2220367278797996e-05,
+      "loss": 0.0776,
+      "step": 1439
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 5.218697829716194e-05,
+      "loss": 0.1354,
+      "step": 1440
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 5.2153589315525876e-05,
+      "loss": 0.1273,
+      "step": 1441
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 5.2120200333889826e-05,
+      "loss": 0.0787,
+      "step": 1442
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 5.208681135225376e-05,
+      "loss": 0.1313,
+      "step": 1443
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 5.20534223706177e-05,
+      "loss": 0.0654,
+      "step": 1444
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 5.2020033388981636e-05,
+      "loss": 0.0695,
+      "step": 1445
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 5.198664440734557e-05,
+      "loss": 0.0897,
+      "step": 1446
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 5.195325542570952e-05,
+      "loss": 0.1457,
+      "step": 1447
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 5.191986644407346e-05,
+      "loss": 0.1034,
+      "step": 1448
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 5.18864774624374e-05,
+      "loss": 0.1138,
+      "step": 1449
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 5.185308848080134e-05,
+      "loss": 0.0792,
+      "step": 1450
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 5.1819699499165276e-05,
+      "loss": 0.0805,
+      "step": 1451
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 5.178631051752921e-05,
+      "loss": 0.1278,
+      "step": 1452
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 5.175292153589316e-05,
+      "loss": 0.0774,
+      "step": 1453
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 5.17195325542571e-05,
+      "loss": 0.1191,
+      "step": 1454
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 5.1686143572621036e-05,
+      "loss": 0.0742,
+      "step": 1455
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 5.165275459098497e-05,
+      "loss": 0.1201,
+      "step": 1456
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 5.1619365609348916e-05,
+      "loss": 0.107,
+      "step": 1457
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 5.158597662771286e-05,
+      "loss": 0.0671,
+      "step": 1458
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 5.15525876460768e-05,
+      "loss": 0.0874,
+      "step": 1459
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 5.151919866444074e-05,
+      "loss": 0.1096,
+      "step": 1460
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 5.1485809682804676e-05,
+      "loss": 0.2839,
+      "step": 1461
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 5.145242070116861e-05,
+      "loss": 0.0772,
+      "step": 1462
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 5.141903171953256e-05,
+      "loss": 0.1039,
+      "step": 1463
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 5.13856427378965e-05,
+      "loss": 0.1099,
+      "step": 1464
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 5.1352253756260436e-05,
+      "loss": 0.1056,
+      "step": 1465
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 5.131886477462438e-05,
+      "loss": 0.0753,
+      "step": 1466
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 5.1285475792988315e-05,
+      "loss": 0.0758,
+      "step": 1467
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 5.125208681135225e-05,
+      "loss": 0.0926,
+      "step": 1468
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 5.12186978297162e-05,
+      "loss": 0.067,
+      "step": 1469
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 5.118530884808014e-05,
+      "loss": 0.0838,
+      "step": 1470
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 5.1151919866444075e-05,
+      "loss": 0.0909,
+      "step": 1471
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 5.111853088480801e-05,
+      "loss": 0.0966,
+      "step": 1472
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 5.108514190317195e-05,
+      "loss": 0.0835,
+      "step": 1473
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 5.10517529215359e-05,
+      "loss": 0.076,
+      "step": 1474
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 5.1018363939899835e-05,
+      "loss": 0.0963,
+      "step": 1475
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 5.098497495826378e-05,
+      "loss": 0.1062,
+      "step": 1476
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 5.0951585976627715e-05,
+      "loss": 0.0733,
+      "step": 1477
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 5.091819699499165e-05,
+      "loss": 0.0964,
+      "step": 1478
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 5.08848080133556e-05,
+      "loss": 0.0843,
+      "step": 1479
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5.085141903171954e-05,
+      "loss": 0.0871,
+      "step": 1480
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5.0818030050083475e-05,
+      "loss": 0.1116,
+      "step": 1481
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 5.078464106844741e-05,
+      "loss": 0.083,
+      "step": 1482
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5.0751252086811355e-05,
+      "loss": 0.0942,
+      "step": 1483
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5.071786310517529e-05,
+      "loss": 0.0762,
+      "step": 1484
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5.068447412353924e-05,
+      "loss": 0.1233,
+      "step": 1485
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5.065108514190318e-05,
+      "loss": 0.085,
+      "step": 1486
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5.0617696160267115e-05,
+      "loss": 0.1096,
+      "step": 1487
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 5.058430717863105e-05,
+      "loss": 0.0629,
+      "step": 1488
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 5.055091819699499e-05,
+      "loss": 0.0772,
+      "step": 1489
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 5.051752921535894e-05,
+      "loss": 0.0868,
+      "step": 1490
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 5.0484140233722875e-05,
+      "loss": 0.2028,
+      "step": 1491
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 5.045075125208681e-05,
+      "loss": 0.1043,
+      "step": 1492
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 5.0417362270450755e-05,
+      "loss": 0.0976,
+      "step": 1493
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 5.038397328881469e-05,
+      "loss": 0.0647,
+      "step": 1494
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 5.035058430717864e-05,
+      "loss": 0.0658,
+      "step": 1495
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 5.031719532554258e-05,
+      "loss": 0.1305,
+      "step": 1496
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 5.0283806343906514e-05,
+      "loss": 0.0907,
+      "step": 1497
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 5.025041736227045e-05,
+      "loss": 0.1389,
+      "step": 1498
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 5.021702838063439e-05,
+      "loss": 0.0855,
+      "step": 1499
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 5.018363939899833e-05,
+      "loss": 0.0757,
+      "step": 1500
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 2.72354793548584,
+      "eval_runtime": 20.2033,
+      "eval_samples_per_second": 9.899,
+      "eval_steps_per_second": 1.237,
+      "step": 1500
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 5.0150250417362274e-05,
+      "loss": 0.0826,
+      "step": 1501
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 5.011686143572622e-05,
+      "loss": 0.0783,
+      "step": 1502
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 5.0083472454090154e-05,
+      "loss": 0.0434,
+      "step": 1503
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 5.005008347245409e-05,
+      "loss": 0.0782,
+      "step": 1504
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 5.001669449081803e-05,
+      "loss": 0.062,
+      "step": 1505
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 4.998330550918197e-05,
+      "loss": 0.0612,
+      "step": 1506
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 4.9949916527545914e-05,
+      "loss": 0.0632,
+      "step": 1507
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 4.991652754590985e-05,
+      "loss": 0.0571,
+      "step": 1508
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 4.988313856427379e-05,
+      "loss": 0.0618,
+      "step": 1509
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 4.984974958263773e-05,
+      "loss": 0.076,
+      "step": 1510
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 4.9816360601001674e-05,
+      "loss": 0.0697,
+      "step": 1511
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 4.978297161936561e-05,
+      "loss": 0.064,
+      "step": 1512
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 4.9749582637729554e-05,
+      "loss": 0.0607,
+      "step": 1513
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 4.971619365609349e-05,
+      "loss": 0.0744,
+      "step": 1514
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 4.9682804674457434e-05,
+      "loss": 0.0704,
+      "step": 1515
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 4.964941569282137e-05,
+      "loss": 0.055,
+      "step": 1516
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 4.961602671118531e-05,
+      "loss": 0.128,
+      "step": 1517
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 4.958263772954925e-05,
+      "loss": 0.0711,
+      "step": 1518
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 4.9549248747913194e-05,
+      "loss": 0.1125,
+      "step": 1519
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 4.951585976627713e-05,
+      "loss": 0.0868,
+      "step": 1520
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 4.9482470784641074e-05,
+      "loss": 0.1132,
+      "step": 1521
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 4.944908180300501e-05,
+      "loss": 0.0875,
+      "step": 1522
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 4.9415692821368953e-05,
+      "loss": 0.1102,
+      "step": 1523
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 4.938230383973289e-05,
+      "loss": 0.0813,
+      "step": 1524
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 4.934891485809683e-05,
+      "loss": 0.0786,
+      "step": 1525
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 4.931552587646077e-05,
+      "loss": 0.0676,
+      "step": 1526
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 4.9282136894824707e-05,
+      "loss": 0.0876,
+      "step": 1527
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 4.924874791318865e-05,
+      "loss": 0.1606,
+      "step": 1528
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 4.921535893155259e-05,
+      "loss": 0.0654,
+      "step": 1529
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 4.918196994991653e-05,
+      "loss": 0.07,
+      "step": 1530
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 4.914858096828047e-05,
+      "loss": 0.0652,
+      "step": 1531
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 4.911519198664441e-05,
+      "loss": 0.1077,
+      "step": 1532
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 4.9081803005008346e-05,
+      "loss": 0.0685,
+      "step": 1533
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 4.904841402337229e-05,
+      "loss": 0.0997,
+      "step": 1534
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 4.9015025041736226e-05,
+      "loss": 0.0982,
+      "step": 1535
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 4.898163606010017e-05,
+      "loss": 0.1061,
+      "step": 1536
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 4.894824707846411e-05,
+      "loss": 0.1088,
+      "step": 1537
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 4.891485809682805e-05,
+      "loss": 0.0745,
+      "step": 1538
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 4.888146911519199e-05,
+      "loss": 0.0821,
+      "step": 1539
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 4.884808013355593e-05,
+      "loss": 0.1292,
+      "step": 1540
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 4.8814691151919866e-05,
+      "loss": 0.0803,
+      "step": 1541
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 4.878130217028381e-05,
+      "loss": 0.0622,
+      "step": 1542
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 4.8747913188647746e-05,
+      "loss": 0.0998,
+      "step": 1543
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 4.871452420701168e-05,
+      "loss": 0.0705,
+      "step": 1544
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 4.8681135225375626e-05,
+      "loss": 0.1369,
+      "step": 1545
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 4.864774624373957e-05,
+      "loss": 0.0636,
+      "step": 1546
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 4.861435726210351e-05,
+      "loss": 0.0828,
+      "step": 1547
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 4.858096828046745e-05,
+      "loss": 0.0535,
+      "step": 1548
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 4.8547579298831386e-05,
+      "loss": 0.0755,
+      "step": 1549
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 4.851419031719533e-05,
+      "loss": 0.0497,
+      "step": 1550
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 4.8480801335559266e-05,
+      "loss": 0.0619,
+      "step": 1551
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 4.84474123539232e-05,
+      "loss": 0.0668,
+      "step": 1552
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 4.8414023372287146e-05,
+      "loss": 0.1061,
+      "step": 1553
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 4.838063439065109e-05,
+      "loss": 0.0648,
+      "step": 1554
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 4.834724540901503e-05,
+      "loss": 0.0558,
+      "step": 1555
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 4.831385642737897e-05,
+      "loss": 0.1611,
+      "step": 1556
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 4.8280467445742906e-05,
+      "loss": 0.1042,
+      "step": 1557
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 4.824707846410685e-05,
+      "loss": 0.0654,
+      "step": 1558
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 4.8213689482470785e-05,
+      "loss": 0.0827,
+      "step": 1559
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 4.818030050083472e-05,
+      "loss": 0.0884,
+      "step": 1560
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 4.8146911519198665e-05,
+      "loss": 0.063,
+      "step": 1561
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 4.81135225375626e-05,
+      "loss": 0.0944,
+      "step": 1562
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 4.8080133555926545e-05,
+      "loss": 0.0713,
+      "step": 1563
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 4.804674457429049e-05,
+      "loss": 0.1146,
+      "step": 1564
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 4.8013355592654425e-05,
+      "loss": 0.0532,
+      "step": 1565
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 4.797996661101837e-05,
+      "loss": 0.0696,
+      "step": 1566
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 4.7946577629382305e-05,
+      "loss": 0.0937,
+      "step": 1567
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 4.791318864774624e-05,
+      "loss": 0.1009,
+      "step": 1568
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 4.7879799666110185e-05,
+      "loss": 0.11,
+      "step": 1569
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 4.784641068447412e-05,
+      "loss": 0.096,
+      "step": 1570
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 4.7813021702838065e-05,
+      "loss": 0.0606,
+      "step": 1571
+    },
+    {
+      "epoch": 10.48,
+      "learning_rate": 4.777963272120201e-05,
+      "loss": 0.0571,
+      "step": 1572
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 4.7746243739565945e-05,
+      "loss": 0.0503,
+      "step": 1573
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 4.771285475792989e-05,
+      "loss": 0.1222,
+      "step": 1574
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 4.7679465776293825e-05,
+      "loss": 0.0595,
+      "step": 1575
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 4.764607679465777e-05,
+      "loss": 0.0779,
+      "step": 1576
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 4.7612687813021705e-05,
+      "loss": 0.0968,
+      "step": 1577
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 4.757929883138564e-05,
+      "loss": 0.121,
+      "step": 1578
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 4.7545909849749585e-05,
+      "loss": 0.082,
+      "step": 1579
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 4.751252086811353e-05,
+      "loss": 0.0628,
+      "step": 1580
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 4.7479131886477465e-05,
+      "loss": 0.1219,
+      "step": 1581
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 4.744574290484141e-05,
+      "loss": 0.089,
+      "step": 1582
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 4.7412353923205345e-05,
+      "loss": 0.0598,
+      "step": 1583
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 4.737896494156929e-05,
+      "loss": 0.0665,
+      "step": 1584
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 4.7345575959933225e-05,
+      "loss": 0.1042,
+      "step": 1585
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 4.731218697829716e-05,
+      "loss": 0.0762,
+      "step": 1586
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 4.7278797996661104e-05,
+      "loss": 0.1173,
+      "step": 1587
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 4.724540901502504e-05,
+      "loss": 0.0634,
+      "step": 1588
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 4.7212020033388984e-05,
+      "loss": 0.0832,
+      "step": 1589
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 4.717863105175293e-05,
+      "loss": 0.0667,
+      "step": 1590
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 4.7145242070116864e-05,
+      "loss": 0.0665,
+      "step": 1591
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 4.711185308848081e-05,
+      "loss": 0.0661,
+      "step": 1592
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 4.7078464106844744e-05,
+      "loss": 0.1077,
+      "step": 1593
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 4.704507512520868e-05,
+      "loss": 0.0969,
+      "step": 1594
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 4.7011686143572624e-05,
+      "loss": 0.1112,
+      "step": 1595
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 4.697829716193656e-05,
+      "loss": 0.0745,
+      "step": 1596
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 4.6944908180300504e-05,
+      "loss": 0.0685,
+      "step": 1597
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 4.691151919866445e-05,
+      "loss": 0.0648,
+      "step": 1598
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 4.6878130217028384e-05,
+      "loss": 0.0759,
+      "step": 1599
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 4.684474123539233e-05,
+      "loss": 0.0888,
+      "step": 1600
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 4.6811352253756264e-05,
+      "loss": 0.0854,
+      "step": 1601
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 4.67779632721202e-05,
+      "loss": 0.0684,
+      "step": 1602
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 4.6744574290484144e-05,
+      "loss": 0.0883,
+      "step": 1603
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 4.671118530884808e-05,
+      "loss": 0.0983,
+      "step": 1604
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 4.667779632721202e-05,
+      "loss": 0.0713,
+      "step": 1605
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 4.664440734557596e-05,
+      "loss": 0.086,
+      "step": 1606
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 4.6611018363939904e-05,
+      "loss": 0.055,
+      "step": 1607
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 4.657762938230385e-05,
+      "loss": 0.1109,
+      "step": 1608
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 4.6544240400667784e-05,
+      "loss": 0.0839,
+      "step": 1609
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 4.651085141903172e-05,
+      "loss": 0.083,
+      "step": 1610
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 4.6477462437395664e-05,
+      "loss": 0.1516,
+      "step": 1611
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 4.64440734557596e-05,
+      "loss": 0.085,
+      "step": 1612
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 4.641068447412354e-05,
+      "loss": 0.1528,
+      "step": 1613
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 4.637729549248748e-05,
+      "loss": 0.0852,
+      "step": 1614
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 4.6343906510851423e-05,
+      "loss": 0.0746,
+      "step": 1615
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 4.631051752921537e-05,
+      "loss": 0.1015,
+      "step": 1616
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 4.6277128547579303e-05,
+      "loss": 0.0751,
+      "step": 1617
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 4.624373956594324e-05,
+      "loss": 0.0652,
+      "step": 1618
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 4.621035058430718e-05,
+      "loss": 0.1116,
+      "step": 1619
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 4.617696160267112e-05,
+      "loss": 0.1024,
+      "step": 1620
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 4.6143572621035056e-05,
+      "loss": 0.084,
+      "step": 1621
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 4.6110183639399e-05,
+      "loss": 0.0965,
+      "step": 1622
+    },
+    {
+      "epoch": 10.82,
+      "learning_rate": 4.6076794657762936e-05,
+      "loss": 0.1286,
+      "step": 1623
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 4.604340567612688e-05,
+      "loss": 0.0696,
+      "step": 1624
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 4.601001669449082e-05,
+      "loss": 0.0748,
+      "step": 1625
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 4.597662771285476e-05,
+      "loss": 0.1328,
+      "step": 1626
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 4.59432387312187e-05,
+      "loss": 0.1491,
+      "step": 1627
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 4.590984974958264e-05,
+      "loss": 0.0698,
+      "step": 1628
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 4.5876460767946576e-05,
+      "loss": 0.1065,
+      "step": 1629
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 4.584307178631052e-05,
+      "loss": 0.0785,
+      "step": 1630
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 4.5809682804674456e-05,
+      "loss": 0.0809,
+      "step": 1631
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 4.57762938230384e-05,
+      "loss": 0.0736,
+      "step": 1632
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 4.574290484140234e-05,
+      "loss": 0.1473,
+      "step": 1633
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 4.570951585976628e-05,
+      "loss": 0.0561,
+      "step": 1634
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 4.567612687813022e-05,
+      "loss": 0.2131,
+      "step": 1635
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 4.564273789649416e-05,
+      "loss": 0.1013,
+      "step": 1636
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 4.5609348914858096e-05,
+      "loss": 0.0546,
+      "step": 1637
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 4.557595993322204e-05,
+      "loss": 0.0913,
+      "step": 1638
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 4.5542570951585976e-05,
+      "loss": 0.1481,
+      "step": 1639
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 4.550918196994992e-05,
+      "loss": 0.0767,
+      "step": 1640
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 4.5475792988313856e-05,
+      "loss": 0.0848,
+      "step": 1641
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 4.54424040066778e-05,
+      "loss": 0.0878,
+      "step": 1642
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 4.540901502504174e-05,
+      "loss": 0.0771,
+      "step": 1643
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 4.537562604340568e-05,
+      "loss": 0.1432,
+      "step": 1644
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 4.5342237061769616e-05,
+      "loss": 0.0726,
+      "step": 1645
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 4.530884808013356e-05,
+      "loss": 0.0733,
+      "step": 1646
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 4.5275459098497496e-05,
+      "loss": 0.1102,
+      "step": 1647
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 4.524207011686144e-05,
+      "loss": 0.0777,
+      "step": 1648
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 4.5208681135225376e-05,
+      "loss": 0.0573,
+      "step": 1649
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.517529215358932e-05,
+      "loss": 0.0818,
+      "step": 1650
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 2.8288543224334717,
+      "eval_runtime": 20.196,
+      "eval_samples_per_second": 9.903,
+      "eval_steps_per_second": 1.238,
+      "step": 1650
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.514190317195326e-05,
+      "loss": 0.1489,
+      "step": 1651
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.51085141903172e-05,
+      "loss": 0.0543,
+      "step": 1652
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 4.5075125208681135e-05,
+      "loss": 0.0532,
+      "step": 1653
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 4.504173622704508e-05,
+      "loss": 0.0506,
+      "step": 1654
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 4.5008347245409015e-05,
+      "loss": 0.0613,
+      "step": 1655
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 4.497495826377296e-05,
+      "loss": 0.05,
+      "step": 1656
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 4.4941569282136895e-05,
+      "loss": 0.0751,
+      "step": 1657
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 4.490818030050083e-05,
+      "loss": 0.0901,
+      "step": 1658
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 4.4874791318864775e-05,
+      "loss": 0.056,
+      "step": 1659
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 4.484140233722872e-05,
+      "loss": 0.0714,
+      "step": 1660
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 4.4808013355592655e-05,
+      "loss": 0.0636,
+      "step": 1661
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 4.47746243739566e-05,
+      "loss": 0.048,
+      "step": 1662
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 4.4741235392320535e-05,
+      "loss": 0.044,
+      "step": 1663
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 4.470784641068448e-05,
+      "loss": 0.15,
+      "step": 1664
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 4.4674457429048415e-05,
+      "loss": 0.0368,
+      "step": 1665
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 4.464106844741235e-05,
+      "loss": 0.0884,
+      "step": 1666
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 4.4607679465776295e-05,
+      "loss": 0.0468,
+      "step": 1667
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 4.457429048414024e-05,
+      "loss": 0.0755,
+      "step": 1668
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 4.4540901502504175e-05,
+      "loss": 0.1811,
+      "step": 1669
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 4.450751252086812e-05,
+      "loss": 0.0506,
+      "step": 1670
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 4.4474123539232055e-05,
+      "loss": 0.0766,
+      "step": 1671
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 4.4440734557596e-05,
+      "loss": 0.0778,
+      "step": 1672
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 4.4407345575959935e-05,
+      "loss": 0.0591,
+      "step": 1673
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 4.437395659432387e-05,
+      "loss": 0.0625,
+      "step": 1674
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 4.4340567612687815e-05,
+      "loss": 0.0814,
+      "step": 1675
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 4.430717863105175e-05,
+      "loss": 0.059,
+      "step": 1676
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 4.4273789649415695e-05,
+      "loss": 0.1031,
+      "step": 1677
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 4.424040066777964e-05,
+      "loss": 0.079,
+      "step": 1678
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 4.4207011686143574e-05,
+      "loss": 0.1085,
+      "step": 1679
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 4.417362270450752e-05,
+      "loss": 0.0777,
+      "step": 1680
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 4.4140233722871454e-05,
+      "loss": 0.0779,
+      "step": 1681
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 4.410684474123539e-05,
+      "loss": 0.0568,
+      "step": 1682
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 4.4073455759599334e-05,
+      "loss": 0.0548,
+      "step": 1683
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 4.404006677796327e-05,
+      "loss": 0.0773,
+      "step": 1684
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 4.4006677796327214e-05,
+      "loss": 0.1037,
+      "step": 1685
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 4.397328881469116e-05,
+      "loss": 0.0652,
+      "step": 1686
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 4.3939899833055094e-05,
+      "loss": 0.0595,
+      "step": 1687
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 4.390651085141904e-05,
+      "loss": 0.0754,
+      "step": 1688
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 4.3873121869782974e-05,
+      "loss": 0.0859,
+      "step": 1689
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 4.383973288814691e-05,
+      "loss": 0.1052,
+      "step": 1690
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 4.3806343906510854e-05,
+      "loss": 0.0442,
+      "step": 1691
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 4.377295492487479e-05,
+      "loss": 0.0472,
+      "step": 1692
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 4.373956594323873e-05,
+      "loss": 0.0719,
+      "step": 1693
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 4.370617696160268e-05,
+      "loss": 0.07,
+      "step": 1694
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 4.3672787979966614e-05,
+      "loss": 0.1261,
+      "step": 1695
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 4.363939899833056e-05,
+      "loss": 0.07,
+      "step": 1696
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 4.3606010016694494e-05,
+      "loss": 0.06,
+      "step": 1697
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 4.357262103505843e-05,
+      "loss": 0.127,
+      "step": 1698
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 4.3539232053422374e-05,
+      "loss": 0.0479,
+      "step": 1699
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 4.350584307178631e-05,
+      "loss": 0.0666,
+      "step": 1700
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 4.3472454090150254e-05,
+      "loss": 0.0623,
+      "step": 1701
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 4.343906510851419e-05,
+      "loss": 0.0591,
+      "step": 1702
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 4.3405676126878134e-05,
+      "loss": 0.075,
+      "step": 1703
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 4.337228714524208e-05,
+      "loss": 0.1015,
+      "step": 1704
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 4.3338898163606014e-05,
+      "loss": 0.0712,
+      "step": 1705
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 4.330550918196995e-05,
+      "loss": 0.0907,
+      "step": 1706
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 4.3272120200333893e-05,
+      "loss": 0.0571,
+      "step": 1707
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 4.323873121869783e-05,
+      "loss": 0.1533,
+      "step": 1708
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 4.3205342237061773e-05,
+      "loss": 0.0781,
+      "step": 1709
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 4.317195325542571e-05,
+      "loss": 0.0695,
+      "step": 1710
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 4.313856427378965e-05,
+      "loss": 0.1025,
+      "step": 1711
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 4.31051752921536e-05,
+      "loss": 0.0773,
+      "step": 1712
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 4.307178631051753e-05,
+      "loss": 0.1082,
+      "step": 1713
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 4.303839732888147e-05,
+      "loss": 0.0509,
+      "step": 1714
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 4.300500834724541e-05,
+      "loss": 0.1186,
+      "step": 1715
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 4.297161936560935e-05,
+      "loss": 0.0518,
+      "step": 1716
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 4.293823038397329e-05,
+      "loss": 0.1107,
+      "step": 1717
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 4.290484140233723e-05,
+      "loss": 0.0679,
+      "step": 1718
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 4.2871452420701166e-05,
+      "loss": 0.0671,
+      "step": 1719
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 4.283806343906511e-05,
+      "loss": 0.0986,
+      "step": 1720
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 4.280467445742905e-05,
+      "loss": 0.0521,
+      "step": 1721
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 4.277128547579299e-05,
+      "loss": 0.0844,
+      "step": 1722
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 4.273789649415693e-05,
+      "loss": 0.0928,
+      "step": 1723
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 4.270450751252087e-05,
+      "loss": 0.095,
+      "step": 1724
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 4.267111853088481e-05,
+      "loss": 0.0813,
+      "step": 1725
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 4.263772954924875e-05,
+      "loss": 0.0968,
+      "step": 1726
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 4.2604340567612686e-05,
+      "loss": 0.0668,
+      "step": 1727
+    },
+    {
+      "epoch": 11.52,
+      "learning_rate": 4.257095158597663e-05,
+      "loss": 0.0828,
+      "step": 1728
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 4.253756260434057e-05,
+      "loss": 0.0737,
+      "step": 1729
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 4.250417362270451e-05,
+      "loss": 0.0544,
+      "step": 1730
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 4.247078464106845e-05,
+      "loss": 0.0968,
+      "step": 1731
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 4.243739565943239e-05,
+      "loss": 0.0843,
+      "step": 1732
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 4.240400667779633e-05,
+      "loss": 0.065,
+      "step": 1733
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 4.237061769616027e-05,
+      "loss": 0.0739,
+      "step": 1734
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 4.2337228714524206e-05,
+      "loss": 0.0952,
+      "step": 1735
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 4.230383973288815e-05,
+      "loss": 0.107,
+      "step": 1736
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 4.2270450751252086e-05,
+      "loss": 0.0929,
+      "step": 1737
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 4.223706176961603e-05,
+      "loss": 0.1177,
+      "step": 1738
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 4.220367278797997e-05,
+      "loss": 0.1158,
+      "step": 1739
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 4.217028380634391e-05,
+      "loss": 0.0544,
+      "step": 1740
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 4.213689482470785e-05,
+      "loss": 0.0819,
+      "step": 1741
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 4.210350584307179e-05,
+      "loss": 0.0655,
+      "step": 1742
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 4.2070116861435725e-05,
+      "loss": 0.1546,
+      "step": 1743
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 4.203672787979967e-05,
+      "loss": 0.1567,
+      "step": 1744
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 4.2003338898163605e-05,
+      "loss": 0.0756,
+      "step": 1745
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 4.196994991652755e-05,
+      "loss": 0.0688,
+      "step": 1746
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 4.193656093489149e-05,
+      "loss": 0.061,
+      "step": 1747
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 4.190317195325543e-05,
+      "loss": 0.1023,
+      "step": 1748
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 4.186978297161937e-05,
+      "loss": 0.092,
+      "step": 1749
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 4.183639398998331e-05,
+      "loss": 0.0752,
+      "step": 1750
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 4.1803005008347245e-05,
+      "loss": 0.0623,
+      "step": 1751
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 4.176961602671119e-05,
+      "loss": 0.0884,
+      "step": 1752
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 4.1736227045075125e-05,
+      "loss": 0.0562,
+      "step": 1753
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 4.170283806343906e-05,
+      "loss": 0.0768,
+      "step": 1754
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 4.1669449081803005e-05,
+      "loss": 0.0553,
+      "step": 1755
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 4.163606010016695e-05,
+      "loss": 0.0804,
+      "step": 1756
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 4.160267111853089e-05,
+      "loss": 0.1252,
+      "step": 1757
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 4.156928213689483e-05,
+      "loss": 0.1102,
+      "step": 1758
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 4.1535893155258765e-05,
+      "loss": 0.0865,
+      "step": 1759
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 4.150250417362271e-05,
+      "loss": 0.078,
+      "step": 1760
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 4.1469115191986645e-05,
+      "loss": 0.0512,
+      "step": 1761
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 4.143572621035058e-05,
+      "loss": 0.1494,
+      "step": 1762
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 4.1402337228714525e-05,
+      "loss": 0.1002,
+      "step": 1763
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 4.136894824707847e-05,
+      "loss": 0.0856,
+      "step": 1764
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 4.133555926544241e-05,
+      "loss": 0.0514,
+      "step": 1765
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 4.130217028380635e-05,
+      "loss": 0.0672,
+      "step": 1766
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 4.1268781302170285e-05,
+      "loss": 0.1002,
+      "step": 1767
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 4.123539232053423e-05,
+      "loss": 0.0735,
+      "step": 1768
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 4.1202003338898165e-05,
+      "loss": 0.0713,
+      "step": 1769
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 4.11686143572621e-05,
+      "loss": 0.0921,
+      "step": 1770
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 4.1135225375626044e-05,
+      "loss": 0.0909,
+      "step": 1771
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 4.110183639398998e-05,
+      "loss": 0.0689,
+      "step": 1772
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 4.1068447412353924e-05,
+      "loss": 0.0723,
+      "step": 1773
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 4.103505843071787e-05,
+      "loss": 0.0817,
+      "step": 1774
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 4.1001669449081804e-05,
+      "loss": 0.0912,
+      "step": 1775
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 4.096828046744575e-05,
+      "loss": 0.0822,
+      "step": 1776
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 4.0934891485809684e-05,
+      "loss": 0.0753,
+      "step": 1777
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 4.090150250417362e-05,
+      "loss": 0.0589,
+      "step": 1778
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 4.0868113522537564e-05,
+      "loss": 0.0602,
+      "step": 1779
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 4.08347245409015e-05,
+      "loss": 0.0642,
+      "step": 1780
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 4.0801335559265444e-05,
+      "loss": 0.1306,
+      "step": 1781
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 4.076794657762939e-05,
+      "loss": 0.0782,
+      "step": 1782
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 4.0734557595993324e-05,
+      "loss": 0.0702,
+      "step": 1783
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 4.070116861435727e-05,
+      "loss": 0.0818,
+      "step": 1784
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 4.0667779632721204e-05,
+      "loss": 0.0927,
+      "step": 1785
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 4.063439065108514e-05,
+      "loss": 0.0811,
+      "step": 1786
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 4.0601001669449084e-05,
+      "loss": 0.0847,
+      "step": 1787
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 4.056761268781302e-05,
+      "loss": 0.0919,
+      "step": 1788
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 4.0534223706176964e-05,
+      "loss": 0.0704,
+      "step": 1789
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 4.05008347245409e-05,
+      "loss": 0.0624,
+      "step": 1790
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 4.0467445742904844e-05,
+      "loss": 0.1154,
+      "step": 1791
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 4.043405676126879e-05,
+      "loss": 0.0615,
+      "step": 1792
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 4.0400667779632724e-05,
+      "loss": 0.0944,
+      "step": 1793
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 4.036727879799666e-05,
+      "loss": 0.0613,
+      "step": 1794
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 4.0333889816360604e-05,
+      "loss": 0.0569,
+      "step": 1795
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 4.030050083472454e-05,
+      "loss": 0.0522,
+      "step": 1796
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 4.0267111853088484e-05,
+      "loss": 0.0721,
+      "step": 1797
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 4.023372287145242e-05,
+      "loss": 0.0978,
+      "step": 1798
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 4.0200333889816363e-05,
+      "loss": 0.0857,
+      "step": 1799
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.016694490818031e-05,
+      "loss": 0.0698,
+      "step": 1800
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 2.8434059619903564,
+      "eval_runtime": 20.197,
+      "eval_samples_per_second": 9.902,
+      "eval_steps_per_second": 1.238,
+      "step": 1800
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.0133555926544243e-05,
+      "loss": 0.0644,
+      "step": 1801
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.010016694490818e-05,
+      "loss": 0.0484,
+      "step": 1802
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.006677796327212e-05,
+      "loss": 0.0846,
+      "step": 1803
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 4.003338898163606e-05,
+      "loss": 0.0535,
+      "step": 1804
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 4e-05,
+      "loss": 0.0561,
+      "step": 1805
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 3.996661101836394e-05,
+      "loss": 0.0685,
+      "step": 1806
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 3.9933222036727876e-05,
+      "loss": 0.0795,
+      "step": 1807
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 3.989983305509182e-05,
+      "loss": 0.0486,
+      "step": 1808
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 3.986644407345576e-05,
+      "loss": 0.0444,
+      "step": 1809
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 3.98330550918197e-05,
+      "loss": 0.0692,
+      "step": 1810
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 3.979966611018364e-05,
+      "loss": 0.0784,
+      "step": 1811
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 3.976627712854758e-05,
+      "loss": 0.0417,
+      "step": 1812
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 3.973288814691152e-05,
+      "loss": 0.0878,
+      "step": 1813
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 3.969949916527546e-05,
+      "loss": 0.0586,
+      "step": 1814
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 3.9666110183639396e-05,
+      "loss": 0.0729,
+      "step": 1815
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 3.963272120200334e-05,
+      "loss": 0.0562,
+      "step": 1816
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 3.959933222036728e-05,
+      "loss": 0.1051,
+      "step": 1817
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 3.956594323873122e-05,
+      "loss": 0.0564,
+      "step": 1818
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 3.953255425709516e-05,
+      "loss": 0.0607,
+      "step": 1819
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 3.94991652754591e-05,
+      "loss": 0.092,
+      "step": 1820
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 3.946577629382304e-05,
+      "loss": 0.0671,
+      "step": 1821
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 3.943238731218698e-05,
+      "loss": 0.0411,
+      "step": 1822
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 3.9398998330550916e-05,
+      "loss": 0.0667,
+      "step": 1823
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 3.936560934891486e-05,
+      "loss": 0.1215,
+      "step": 1824
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 3.93322203672788e-05,
+      "loss": 0.0897,
+      "step": 1825
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 3.929883138564274e-05,
+      "loss": 0.0646,
+      "step": 1826
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 3.926544240400668e-05,
+      "loss": 0.0751,
+      "step": 1827
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 3.923205342237062e-05,
+      "loss": 0.052,
+      "step": 1828
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 3.919866444073456e-05,
+      "loss": 0.0584,
+      "step": 1829
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 3.91652754590985e-05,
+      "loss": 0.0854,
+      "step": 1830
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 3.9131886477462436e-05,
+      "loss": 0.0633,
+      "step": 1831
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 3.909849749582638e-05,
+      "loss": 0.116,
+      "step": 1832
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 3.9065108514190316e-05,
+      "loss": 0.0692,
+      "step": 1833
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 3.903171953255426e-05,
+      "loss": 0.0746,
+      "step": 1834
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 3.89983305509182e-05,
+      "loss": 0.0988,
+      "step": 1835
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 3.896494156928214e-05,
+      "loss": 0.1028,
+      "step": 1836
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 3.893155258764608e-05,
+      "loss": 0.0533,
+      "step": 1837
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 3.889816360601002e-05,
+      "loss": 0.0664,
+      "step": 1838
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 3.8864774624373955e-05,
+      "loss": 0.0684,
+      "step": 1839
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 3.88313856427379e-05,
+      "loss": 0.1442,
+      "step": 1840
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 3.8797996661101835e-05,
+      "loss": 0.0851,
+      "step": 1841
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 3.876460767946578e-05,
+      "loss": 0.0693,
+      "step": 1842
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 3.873121869782972e-05,
+      "loss": 0.0452,
+      "step": 1843
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 3.869782971619366e-05,
+      "loss": 0.0885,
+      "step": 1844
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 3.86644407345576e-05,
+      "loss": 0.0547,
+      "step": 1845
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 3.863105175292154e-05,
+      "loss": 0.0747,
+      "step": 1846
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 3.8597662771285475e-05,
+      "loss": 0.1514,
+      "step": 1847
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 3.856427378964942e-05,
+      "loss": 0.0681,
+      "step": 1848
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 3.8530884808013355e-05,
+      "loss": 0.0884,
+      "step": 1849
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 3.84974958263773e-05,
+      "loss": 0.0896,
+      "step": 1850
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 3.8464106844741235e-05,
+      "loss": 0.0706,
+      "step": 1851
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 3.843071786310518e-05,
+      "loss": 0.0836,
+      "step": 1852
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 3.839732888146912e-05,
+      "loss": 0.0707,
+      "step": 1853
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 3.836393989983306e-05,
+      "loss": 0.0718,
+      "step": 1854
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 3.8330550918196995e-05,
+      "loss": 0.0586,
+      "step": 1855
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 3.829716193656094e-05,
+      "loss": 0.0512,
+      "step": 1856
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 3.8263772954924875e-05,
+      "loss": 0.0511,
+      "step": 1857
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 3.823038397328882e-05,
+      "loss": 0.0791,
+      "step": 1858
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 3.8196994991652755e-05,
+      "loss": 0.0657,
+      "step": 1859
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 3.81636060100167e-05,
+      "loss": 0.0905,
+      "step": 1860
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 3.813021702838064e-05,
+      "loss": 0.0841,
+      "step": 1861
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 3.809682804674458e-05,
+      "loss": 0.1793,
+      "step": 1862
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 3.8063439065108514e-05,
+      "loss": 0.0979,
+      "step": 1863
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 3.803005008347246e-05,
+      "loss": 0.1061,
+      "step": 1864
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 3.7996661101836394e-05,
+      "loss": 0.0876,
+      "step": 1865
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 3.796327212020034e-05,
+      "loss": 0.0595,
+      "step": 1866
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 3.7929883138564274e-05,
+      "loss": 0.057,
+      "step": 1867
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 3.789649415692821e-05,
+      "loss": 0.05,
+      "step": 1868
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 3.7863105175292154e-05,
+      "loss": 0.0612,
+      "step": 1869
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 3.78297161936561e-05,
+      "loss": 0.1173,
+      "step": 1870
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 3.7796327212020034e-05,
+      "loss": 0.0538,
+      "step": 1871
+    },
+    {
+      "epoch": 12.48,
+      "learning_rate": 3.776293823038398e-05,
+      "loss": 0.0596,
+      "step": 1872
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 3.7729549248747914e-05,
+      "loss": 0.0975,
+      "step": 1873
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 3.769616026711186e-05,
+      "loss": 0.0767,
+      "step": 1874
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 3.7662771285475794e-05,
+      "loss": 0.0484,
+      "step": 1875
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 3.762938230383973e-05,
+      "loss": 0.0957,
+      "step": 1876
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 3.7595993322203674e-05,
+      "loss": 0.0515,
+      "step": 1877
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 3.756260434056762e-05,
+      "loss": 0.0676,
+      "step": 1878
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 3.7529215358931554e-05,
+      "loss": 0.099,
+      "step": 1879
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 3.74958263772955e-05,
+      "loss": 0.0649,
+      "step": 1880
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 3.7462437395659434e-05,
+      "loss": 0.0717,
+      "step": 1881
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 3.742904841402338e-05,
+      "loss": 0.0609,
+      "step": 1882
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 3.7395659432387314e-05,
+      "loss": 0.0743,
+      "step": 1883
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 3.736227045075125e-05,
+      "loss": 0.065,
+      "step": 1884
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 3.7328881469115194e-05,
+      "loss": 0.0568,
+      "step": 1885
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 3.729549248747913e-05,
+      "loss": 0.0709,
+      "step": 1886
+    },
+    {
+      "epoch": 12.58,
+      "learning_rate": 3.7262103505843074e-05,
+      "loss": 0.061,
+      "step": 1887
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 3.722871452420702e-05,
+      "loss": 0.088,
+      "step": 1888
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 3.7195325542570954e-05,
+      "loss": 0.054,
+      "step": 1889
+    },
+    {
+      "epoch": 12.6,
+      "learning_rate": 3.71619365609349e-05,
+      "loss": 0.0491,
+      "step": 1890
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 3.7128547579298833e-05,
+      "loss": 0.067,
+      "step": 1891
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 3.709515859766277e-05,
+      "loss": 0.0811,
+      "step": 1892
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 3.7061769616026713e-05,
+      "loss": 0.0565,
+      "step": 1893
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 3.702838063439065e-05,
+      "loss": 0.0987,
+      "step": 1894
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 3.699499165275459e-05,
+      "loss": 0.0564,
+      "step": 1895
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 3.696160267111854e-05,
+      "loss": 0.0506,
+      "step": 1896
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 3.692821368948247e-05,
+      "loss": 0.1077,
+      "step": 1897
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 3.689482470784642e-05,
+      "loss": 0.0783,
+      "step": 1898
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 3.686143572621035e-05,
+      "loss": 0.0601,
+      "step": 1899
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 3.682804674457429e-05,
+      "loss": 0.1124,
+      "step": 1900
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 3.679465776293823e-05,
+      "loss": 0.0626,
+      "step": 1901
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 3.676126878130217e-05,
+      "loss": 0.0674,
+      "step": 1902
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 3.6727879799666106e-05,
+      "loss": 0.0828,
+      "step": 1903
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 3.669449081803005e-05,
+      "loss": 0.0515,
+      "step": 1904
+    },
+    {
+      "epoch": 12.7,
+      "learning_rate": 3.666110183639399e-05,
+      "loss": 0.0921,
+      "step": 1905
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 3.6627712854757936e-05,
+      "loss": 0.0881,
+      "step": 1906
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 3.659432387312187e-05,
+      "loss": 0.0512,
+      "step": 1907
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 3.656093489148581e-05,
+      "loss": 0.1282,
+      "step": 1908
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 3.652754590984975e-05,
+      "loss": 0.0612,
+      "step": 1909
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 3.649415692821369e-05,
+      "loss": 0.0817,
+      "step": 1910
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 3.6460767946577626e-05,
+      "loss": 0.0798,
+      "step": 1911
+    },
+    {
+      "epoch": 12.75,
+      "learning_rate": 3.642737896494157e-05,
+      "loss": 0.0673,
+      "step": 1912
+    },
+    {
+      "epoch": 12.75,
+      "learning_rate": 3.639398998330551e-05,
+      "loss": 0.063,
+      "step": 1913
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 3.6360601001669456e-05,
+      "loss": 0.1064,
+      "step": 1914
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 3.632721202003339e-05,
+      "loss": 0.0509,
+      "step": 1915
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 3.629382303839733e-05,
+      "loss": 0.0713,
+      "step": 1916
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 3.626043405676127e-05,
+      "loss": 0.0658,
+      "step": 1917
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 3.622704507512521e-05,
+      "loss": 0.1155,
+      "step": 1918
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 3.6193656093489146e-05,
+      "loss": 0.0502,
+      "step": 1919
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 3.616026711185309e-05,
+      "loss": 0.0721,
+      "step": 1920
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 3.6126878130217026e-05,
+      "loss": 0.096,
+      "step": 1921
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 3.609348914858097e-05,
+      "loss": 0.0777,
+      "step": 1922
+    },
+    {
+      "epoch": 12.82,
+      "learning_rate": 3.606010016694491e-05,
+      "loss": 0.0689,
+      "step": 1923
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 3.602671118530885e-05,
+      "loss": 0.1065,
+      "step": 1924
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 3.599332220367279e-05,
+      "loss": 0.1012,
+      "step": 1925
+    },
+    {
+      "epoch": 12.84,
+      "learning_rate": 3.595993322203673e-05,
+      "loss": 0.0552,
+      "step": 1926
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 3.5926544240400665e-05,
+      "loss": 0.0579,
+      "step": 1927
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 3.589315525876461e-05,
+      "loss": 0.0816,
+      "step": 1928
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 3.5859766277128545e-05,
+      "loss": 0.0818,
+      "step": 1929
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 3.582637729549249e-05,
+      "loss": 0.052,
+      "step": 1930
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 3.579298831385643e-05,
+      "loss": 0.1012,
+      "step": 1931
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 3.575959933222037e-05,
+      "loss": 0.0878,
+      "step": 1932
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 3.572621035058431e-05,
+      "loss": 0.061,
+      "step": 1933
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 3.569282136894825e-05,
+      "loss": 0.1251,
+      "step": 1934
+    },
+    {
+      "epoch": 12.9,
+      "learning_rate": 3.5659432387312185e-05,
+      "loss": 0.1117,
+      "step": 1935
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 3.562604340567613e-05,
+      "loss": 0.0655,
+      "step": 1936
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 3.5592654424040065e-05,
+      "loss": 0.0603,
+      "step": 1937
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 3.555926544240401e-05,
+      "loss": 0.0758,
+      "step": 1938
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 3.5525876460767945e-05,
+      "loss": 0.0813,
+      "step": 1939
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 3.549248747913189e-05,
+      "loss": 0.0712,
+      "step": 1940
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 3.545909849749583e-05,
+      "loss": 0.058,
+      "step": 1941
+    },
+    {
+      "epoch": 12.95,
+      "learning_rate": 3.542570951585977e-05,
+      "loss": 0.0509,
+      "step": 1942
+    },
+    {
+      "epoch": 12.95,
+      "learning_rate": 3.5392320534223705e-05,
+      "loss": 0.083,
+      "step": 1943
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 3.535893155258765e-05,
+      "loss": 0.0952,
+      "step": 1944
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 3.5325542570951585e-05,
+      "loss": 0.0663,
+      "step": 1945
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 3.529215358931553e-05,
+      "loss": 0.0939,
+      "step": 1946
+    },
+    {
+      "epoch": 12.98,
+      "learning_rate": 3.5258764607679465e-05,
+      "loss": 0.0596,
+      "step": 1947
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 3.522537562604341e-05,
+      "loss": 0.0916,
+      "step": 1948
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 3.519198664440735e-05,
+      "loss": 0.0695,
+      "step": 1949
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.515859766277129e-05,
+      "loss": 0.0714,
+      "step": 1950
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 2.795626163482666,
+      "eval_runtime": 20.1831,
+      "eval_samples_per_second": 9.909,
+      "eval_steps_per_second": 1.239,
+      "step": 1950
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 3.5125208681135225e-05,
+      "loss": 0.1028,
+      "step": 1951
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 3.509181969949917e-05,
+      "loss": 0.1105,
+      "step": 1952
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 3.5058430717863105e-05,
+      "loss": 0.0815,
+      "step": 1953
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 3.502504173622705e-05,
+      "loss": 0.0696,
+      "step": 1954
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 3.4991652754590984e-05,
+      "loss": 0.0958,
+      "step": 1955
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 3.495826377295493e-05,
+      "loss": 0.0822,
+      "step": 1956
+    },
+    {
+      "epoch": 13.05,
+      "learning_rate": 3.492487479131887e-05,
+      "loss": 0.0613,
+      "step": 1957
+    },
+    {
+      "epoch": 13.05,
+      "learning_rate": 3.489148580968281e-05,
+      "loss": 0.0464,
+      "step": 1958
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 3.4858096828046744e-05,
+      "loss": 0.0516,
+      "step": 1959
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 3.482470784641069e-05,
+      "loss": 0.0891,
+      "step": 1960
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 3.4791318864774624e-05,
+      "loss": 0.0949,
+      "step": 1961
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 3.475792988313857e-05,
+      "loss": 0.0564,
+      "step": 1962
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 3.4724540901502504e-05,
+      "loss": 0.0726,
+      "step": 1963
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 3.469115191986644e-05,
+      "loss": 0.0722,
+      "step": 1964
+    },
+    {
+      "epoch": 13.1,
+      "learning_rate": 3.4657762938230384e-05,
+      "loss": 0.0571,
+      "step": 1965
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 3.462437395659433e-05,
+      "loss": 0.1212,
+      "step": 1966
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 3.459098497495827e-05,
+      "loss": 0.0882,
+      "step": 1967
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 3.455759599332221e-05,
+      "loss": 0.0676,
+      "step": 1968
+    },
+    {
+      "epoch": 13.13,
+      "learning_rate": 3.4524207011686144e-05,
+      "loss": 0.134,
+      "step": 1969
+    },
+    {
+      "epoch": 13.13,
+      "learning_rate": 3.449081803005009e-05,
+      "loss": 0.0534,
+      "step": 1970
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 3.4457429048414024e-05,
+      "loss": 0.0861,
+      "step": 1971
+    },
+    {
+      "epoch": 13.15,
+      "learning_rate": 3.442404006677796e-05,
+      "loss": 0.0685,
+      "step": 1972
+    },
+    {
+      "epoch": 13.15,
+      "learning_rate": 3.4390651085141904e-05,
+      "loss": 0.0659,
+      "step": 1973
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 3.435726210350585e-05,
+      "loss": 0.0801,
+      "step": 1974
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 3.432387312186979e-05,
+      "loss": 0.0624,
+      "step": 1975
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 3.429048414023373e-05,
+      "loss": 0.0636,
+      "step": 1976
+    },
+    {
+      "epoch": 13.18,
+      "learning_rate": 3.4257095158597664e-05,
+      "loss": 0.0581,
+      "step": 1977
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 3.422370617696161e-05,
+      "loss": 0.0423,
+      "step": 1978
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 3.4190317195325544e-05,
+      "loss": 0.0477,
+      "step": 1979
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 3.415692821368948e-05,
+      "loss": 0.0593,
+      "step": 1980
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 3.4123539232053424e-05,
+      "loss": 0.0636,
+      "step": 1981
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 3.409015025041736e-05,
+      "loss": 0.0831,
+      "step": 1982
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 3.4056761268781303e-05,
+      "loss": 0.1285,
+      "step": 1983
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 3.402337228714525e-05,
+      "loss": 0.0839,
+      "step": 1984
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 3.3989983305509183e-05,
+      "loss": 0.0959,
+      "step": 1985
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 3.395659432387313e-05,
+      "loss": 0.0593,
+      "step": 1986
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 3.392320534223706e-05,
+      "loss": 0.053,
+      "step": 1987
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 3.3889816360601e-05,
+      "loss": 0.0432,
+      "step": 1988
+    },
+    {
+      "epoch": 13.26,
+      "learning_rate": 3.385642737896494e-05,
+      "loss": 0.0734,
+      "step": 1989
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 3.382303839732888e-05,
+      "loss": 0.0614,
+      "step": 1990
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 3.378964941569282e-05,
+      "loss": 0.0691,
+      "step": 1991
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 3.3756260434056767e-05,
+      "loss": 0.0721,
+      "step": 1992
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 3.37228714524207e-05,
+      "loss": 0.0515,
+      "step": 1993
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 3.3689482470784646e-05,
+      "loss": 0.0575,
+      "step": 1994
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 3.365609348914858e-05,
+      "loss": 0.1009,
+      "step": 1995
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 3.362270450751252e-05,
+      "loss": 0.0526,
+      "step": 1996
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 3.358931552587646e-05,
+      "loss": 0.0636,
+      "step": 1997
+    },
+    {
+      "epoch": 13.32,
+      "learning_rate": 3.35559265442404e-05,
+      "loss": 0.064,
+      "step": 1998
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 3.352253756260434e-05,
+      "loss": 0.0675,
+      "step": 1999
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 3.348914858096828e-05,
+      "loss": 0.0539,
+      "step": 2000
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 3.345575959933222e-05,
+      "loss": 0.0683,
+      "step": 2001
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 3.3422370617696166e-05,
+      "loss": 0.1096,
+      "step": 2002
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 3.33889816360601e-05,
+      "loss": 0.0797,
+      "step": 2003
+    },
+    {
+      "epoch": 13.36,
+      "learning_rate": 3.335559265442404e-05,
+      "loss": 0.0803,
+      "step": 2004
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 3.332220367278798e-05,
+      "loss": 0.0511,
+      "step": 2005
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 3.328881469115192e-05,
+      "loss": 0.1385,
+      "step": 2006
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 3.325542570951586e-05,
+      "loss": 0.0725,
+      "step": 2007
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 3.32220367278798e-05,
+      "loss": 0.0718,
+      "step": 2008
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 3.318864774624374e-05,
+      "loss": 0.0675,
+      "step": 2009
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 3.3155258764607686e-05,
+      "loss": 0.0657,
+      "step": 2010
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 3.312186978297162e-05,
+      "loss": 0.1476,
+      "step": 2011
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 3.308848080133556e-05,
+      "loss": 0.0915,
+      "step": 2012
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 3.30550918196995e-05,
+      "loss": 0.0629,
+      "step": 2013
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 3.302170283806344e-05,
+      "loss": 0.0478,
+      "step": 2014
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 3.298831385642738e-05,
+      "loss": 0.0475,
+      "step": 2015
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 3.295492487479132e-05,
+      "loss": 0.0876,
+      "step": 2016
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 3.2921535893155256e-05,
+      "loss": 0.1503,
+      "step": 2017
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 3.28881469115192e-05,
+      "loss": 0.0762,
+      "step": 2018
+    },
+    {
+      "epoch": 13.46,
+      "learning_rate": 3.285475792988314e-05,
+      "loss": 0.0629,
+      "step": 2019
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 3.282136894824708e-05,
+      "loss": 0.0682,
+      "step": 2020
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 3.278797996661102e-05,
+      "loss": 0.0842,
+      "step": 2021
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 3.275459098497496e-05,
+      "loss": 0.0726,
+      "step": 2022
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 3.27212020033389e-05,
+      "loss": 0.062,
+      "step": 2023
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 3.268781302170284e-05,
+      "loss": 0.0456,
+      "step": 2024
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 3.2654424040066775e-05,
+      "loss": 0.0469,
+      "step": 2025
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 3.262103505843072e-05,
+      "loss": 0.046,
+      "step": 2026
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 3.258764607679466e-05,
+      "loss": 0.0517,
+      "step": 2027
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 3.25542570951586e-05,
+      "loss": 0.0536,
+      "step": 2028
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 3.252086811352254e-05,
+      "loss": 0.1152,
+      "step": 2029
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 3.248747913188648e-05,
+      "loss": 0.0607,
+      "step": 2030
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 3.245409015025042e-05,
+      "loss": 0.0969,
+      "step": 2031
+    },
+    {
+      "epoch": 13.55,
+      "learning_rate": 3.242070116861436e-05,
+      "loss": 0.0679,
+      "step": 2032
+    },
+    {
+      "epoch": 13.55,
+      "learning_rate": 3.2387312186978295e-05,
+      "loss": 0.0581,
+      "step": 2033
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 3.235392320534224e-05,
+      "loss": 0.0638,
+      "step": 2034
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 3.2320534223706175e-05,
+      "loss": 0.0518,
+      "step": 2035
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 3.228714524207012e-05,
+      "loss": 0.0751,
+      "step": 2036
+    },
+    {
+      "epoch": 13.58,
+      "learning_rate": 3.225375626043406e-05,
+      "loss": 0.0442,
+      "step": 2037
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 3.2220367278798e-05,
+      "loss": 0.0668,
+      "step": 2038
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 3.218697829716194e-05,
+      "loss": 0.0982,
+      "step": 2039
+    },
+    {
+      "epoch": 13.6,
+      "learning_rate": 3.215358931552588e-05,
+      "loss": 0.0572,
+      "step": 2040
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 3.2120200333889815e-05,
+      "loss": 0.0763,
+      "step": 2041
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 3.208681135225376e-05,
+      "loss": 0.0668,
+      "step": 2042
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 3.2053422370617695e-05,
+      "loss": 0.0607,
+      "step": 2043
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 3.202003338898164e-05,
+      "loss": 0.052,
+      "step": 2044
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 3.198664440734558e-05,
+      "loss": 0.0449,
+      "step": 2045
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 3.195325542570952e-05,
+      "loss": 0.0351,
+      "step": 2046
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 3.191986644407346e-05,
+      "loss": 0.0803,
+      "step": 2047
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 3.18864774624374e-05,
+      "loss": 0.1548,
+      "step": 2048
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 3.1853088480801334e-05,
+      "loss": 0.0721,
+      "step": 2049
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 3.181969949916528e-05,
+      "loss": 0.0933,
+      "step": 2050
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 3.1786310517529214e-05,
+      "loss": 0.1528,
+      "step": 2051
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 3.175292153589315e-05,
+      "loss": 0.1163,
+      "step": 2052
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 3.1719532554257094e-05,
+      "loss": 0.0595,
+      "step": 2053
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 3.168614357262104e-05,
+      "loss": 0.0463,
+      "step": 2054
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 3.165275459098498e-05,
+      "loss": 0.0549,
+      "step": 2055
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 3.161936560934892e-05,
+      "loss": 0.0549,
+      "step": 2056
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 3.1585976627712854e-05,
+      "loss": 0.0489,
+      "step": 2057
+    },
+    {
+      "epoch": 13.72,
+      "learning_rate": 3.15525876460768e-05,
+      "loss": 0.0765,
+      "step": 2058
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 3.1519198664440734e-05,
+      "loss": 0.0704,
+      "step": 2059
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 3.148580968280467e-05,
+      "loss": 0.0548,
+      "step": 2060
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 3.1452420701168614e-05,
+      "loss": 0.1045,
+      "step": 2061
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 3.141903171953256e-05,
+      "loss": 0.0797,
+      "step": 2062
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 3.13856427378965e-05,
+      "loss": 0.0707,
+      "step": 2063
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 3.135225375626044e-05,
+      "loss": 0.0613,
+      "step": 2064
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 3.1318864774624374e-05,
+      "loss": 0.0571,
+      "step": 2065
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 3.128547579298832e-05,
+      "loss": 0.0631,
+      "step": 2066
+    },
+    {
+      "epoch": 13.78,
+      "learning_rate": 3.1252086811352254e-05,
+      "loss": 0.0531,
+      "step": 2067
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 3.121869782971619e-05,
+      "loss": 0.1033,
+      "step": 2068
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 3.1185308848080134e-05,
+      "loss": 0.0605,
+      "step": 2069
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 3.115191986644408e-05,
+      "loss": 0.0529,
+      "step": 2070
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 3.111853088480802e-05,
+      "loss": 0.0535,
+      "step": 2071
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 3.108514190317196e-05,
+      "loss": 0.0857,
+      "step": 2072
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 3.1051752921535894e-05,
+      "loss": 0.0873,
+      "step": 2073
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 3.101836393989984e-05,
+      "loss": 0.076,
+      "step": 2074
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 3.0984974958263773e-05,
+      "loss": 0.0511,
+      "step": 2075
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 3.0984974958263773e-05,
+      "loss": 0.169,
+      "step": 2076
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 3.095158597662771e-05,
+      "loss": 0.0475,
+      "step": 2077
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 3.0918196994991653e-05,
+      "loss": 0.0665,
+      "step": 2078
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 3.088480801335559e-05,
+      "loss": 0.056,
+      "step": 2079
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 3.085141903171953e-05,
+      "loss": 0.1036,
+      "step": 2080
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 3.081803005008348e-05,
+      "loss": 0.0602,
+      "step": 2081
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 3.078464106844741e-05,
+      "loss": 0.0645,
+      "step": 2082
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 3.075125208681136e-05,
+      "loss": 0.0456,
+      "step": 2083
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 3.071786310517529e-05,
+      "loss": 0.0775,
+      "step": 2084
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 3.068447412353923e-05,
+      "loss": 0.1439,
+      "step": 2085
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 3.065108514190317e-05,
+      "loss": 0.053,
+      "step": 2086
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 3.061769616026711e-05,
+      "loss": 0.0633,
+      "step": 2087
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 3.058430717863105e-05,
+      "loss": 0.1637,
+      "step": 2088
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 3.0550918196994996e-05,
+      "loss": 0.1631,
+      "step": 2089
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 3.051752921535893e-05,
+      "loss": 0.067,
+      "step": 2090
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 3.0484140233722873e-05,
+      "loss": 0.0743,
+      "step": 2091
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 3.0450751252086813e-05,
+      "loss": 0.0813,
+      "step": 2092
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 3.0417362270450756e-05,
+      "loss": 0.0635,
+      "step": 2093
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 3.0383973288814693e-05,
+      "loss": 0.1171,
+      "step": 2094
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 3.035058430717863e-05,
+      "loss": 0.0789,
+      "step": 2095
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 3.0317195325542573e-05,
+      "loss": 0.0944,
+      "step": 2096
+    },
+    {
+      "epoch": 13.98,
+      "learning_rate": 3.0283806343906513e-05,
+      "loss": 0.0671,
+      "step": 2097
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 3.025041736227045e-05,
+      "loss": 0.0638,
+      "step": 2098
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 3.0217028380634393e-05,
+      "loss": 0.053,
+      "step": 2099
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.0183639398998333e-05,
+      "loss": 0.0779,
+      "step": 2100
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 2.9732532501220703,
+      "eval_runtime": 20.1977,
+      "eval_samples_per_second": 9.902,
+      "eval_steps_per_second": 1.238,
+      "step": 2100
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.0150250417362276e-05,
+      "loss": 0.0528,
+      "step": 2101
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.0116861435726213e-05,
+      "loss": 0.0682,
+      "step": 2102
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 3.008347245409015e-05,
+      "loss": 0.0872,
+      "step": 2103
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 3.0050083472454093e-05,
+      "loss": 0.0524,
+      "step": 2104
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 3.0016694490818032e-05,
+      "loss": 0.0735,
+      "step": 2105
+    },
+    {
+      "epoch": 14.04,
+      "learning_rate": 2.998330550918197e-05,
+      "loss": 0.0657,
+      "step": 2106
+    },
+    {
+      "epoch": 14.05,
+      "learning_rate": 2.9949916527545912e-05,
+      "loss": 0.0874,
+      "step": 2107
+    },
+    {
+      "epoch": 14.05,
+      "learning_rate": 2.991652754590985e-05,
+      "loss": 0.0788,
+      "step": 2108
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 2.9883138564273792e-05,
+      "loss": 0.0393,
+      "step": 2109
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 2.9849749582637732e-05,
+      "loss": 0.0825,
+      "step": 2110
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 2.981636060100167e-05,
+      "loss": 0.1275,
+      "step": 2111
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 2.9782971619365612e-05,
+      "loss": 0.0794,
+      "step": 2112
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 2.9749582637729552e-05,
+      "loss": 0.0994,
+      "step": 2113
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 2.971619365609349e-05,
+      "loss": 0.0475,
+      "step": 2114
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 2.9682804674457432e-05,
+      "loss": 0.0751,
+      "step": 2115
+    },
+    {
+      "epoch": 14.11,
+      "learning_rate": 2.964941569282137e-05,
+      "loss": 0.0464,
+      "step": 2116
+    },
+    {
+      "epoch": 14.11,
+      "learning_rate": 2.9616026711185312e-05,
+      "loss": 0.1074,
+      "step": 2117
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 2.9582637729549252e-05,
+      "loss": 0.0412,
+      "step": 2118
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 2.954924874791319e-05,
+      "loss": 0.0769,
+      "step": 2119
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 2.9515859766277132e-05,
+      "loss": 0.0661,
+      "step": 2120
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 2.948247078464107e-05,
+      "loss": 0.0716,
+      "step": 2121
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 2.944908180300501e-05,
+      "loss": 0.0998,
+      "step": 2122
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 2.9415692821368952e-05,
+      "loss": 0.0587,
+      "step": 2123
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 2.938230383973289e-05,
+      "loss": 0.07,
+      "step": 2124
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 2.9348914858096832e-05,
+      "loss": 0.065,
+      "step": 2125
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 2.931552587646077e-05,
+      "loss": 0.0737,
+      "step": 2126
+    },
+    {
+      "epoch": 14.18,
+      "learning_rate": 2.928213689482471e-05,
+      "loss": 0.0616,
+      "step": 2127
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 2.924874791318865e-05,
+      "loss": 0.0702,
+      "step": 2128
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 2.9215358931552588e-05,
+      "loss": 0.0448,
+      "step": 2129
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 2.9181969949916528e-05,
+      "loss": 0.0836,
+      "step": 2130
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 2.914858096828047e-05,
+      "loss": 0.0543,
+      "step": 2131
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 2.9115191986644408e-05,
+      "loss": 0.0954,
+      "step": 2132
+    },
+    {
+      "epoch": 14.22,
+      "learning_rate": 2.908180300500835e-05,
+      "loss": 0.0984,
+      "step": 2133
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 2.9048414023372288e-05,
+      "loss": 0.0686,
+      "step": 2134
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 2.9015025041736228e-05,
+      "loss": 0.0565,
+      "step": 2135
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 2.898163606010017e-05,
+      "loss": 0.0692,
+      "step": 2136
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 2.8948247078464108e-05,
+      "loss": 0.047,
+      "step": 2137
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 2.8914858096828045e-05,
+      "loss": 0.055,
+      "step": 2138
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 2.8881469115191988e-05,
+      "loss": 0.0985,
+      "step": 2139
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 2.8848080133555928e-05,
+      "loss": 0.0462,
+      "step": 2140
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 2.881469115191987e-05,
+      "loss": 0.0455,
+      "step": 2141
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 2.8781302170283808e-05,
+      "loss": 0.0681,
+      "step": 2142
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 2.8747913188647744e-05,
+      "loss": 0.0759,
+      "step": 2143
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 2.8714524207011688e-05,
+      "loss": 0.0589,
+      "step": 2144
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 2.8681135225375628e-05,
+      "loss": 0.0579,
+      "step": 2145
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 2.8647746243739564e-05,
+      "loss": 0.0779,
+      "step": 2146
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 2.8614357262103508e-05,
+      "loss": 0.0411,
+      "step": 2147
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 2.8580968280467448e-05,
+      "loss": 0.0666,
+      "step": 2148
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 2.854757929883139e-05,
+      "loss": 0.0478,
+      "step": 2149
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 2.8514190317195328e-05,
+      "loss": 0.0553,
+      "step": 2150
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 2.8480801335559264e-05,
+      "loss": 0.0558,
+      "step": 2151
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 2.8447412353923207e-05,
+      "loss": 0.0542,
+      "step": 2152
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 2.8414023372287147e-05,
+      "loss": 0.0456,
+      "step": 2153
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 2.8380634390651084e-05,
+      "loss": 0.0442,
+      "step": 2154
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 2.8347245409015027e-05,
+      "loss": 0.0717,
+      "step": 2155
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 2.8313856427378964e-05,
+      "loss": 0.0688,
+      "step": 2156
+    },
+    {
+      "epoch": 14.38,
+      "learning_rate": 2.8280467445742907e-05,
+      "loss": 0.0557,
+      "step": 2157
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 2.8247078464106847e-05,
+      "loss": 0.0593,
+      "step": 2158
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 2.8213689482470784e-05,
+      "loss": 0.0925,
+      "step": 2159
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 2.8180300500834727e-05,
+      "loss": 0.05,
+      "step": 2160
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 2.8146911519198664e-05,
+      "loss": 0.0952,
+      "step": 2161
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 2.8113522537562604e-05,
+      "loss": 0.0558,
+      "step": 2162
+    },
+    {
+      "epoch": 14.42,
+      "learning_rate": 2.8080133555926547e-05,
+      "loss": 0.0435,
+      "step": 2163
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 2.8046744574290484e-05,
+      "loss": 0.0649,
+      "step": 2164
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 2.8013355592654427e-05,
+      "loss": 0.0727,
+      "step": 2165
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 2.7979966611018367e-05,
+      "loss": 0.0794,
+      "step": 2166
+    },
+    {
+      "epoch": 14.45,
+      "learning_rate": 2.7946577629382304e-05,
+      "loss": 0.0589,
+      "step": 2167
+    },
+    {
+      "epoch": 14.45,
+      "learning_rate": 2.7913188647746247e-05,
+      "loss": 0.0736,
+      "step": 2168
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 2.7879799666110183e-05,
+      "loss": 0.0773,
+      "step": 2169
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 2.7846410684474123e-05,
+      "loss": 0.1565,
+      "step": 2170
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 2.7813021702838067e-05,
+      "loss": 0.0707,
+      "step": 2171
+    },
+    {
+      "epoch": 14.48,
+      "learning_rate": 2.7779632721202003e-05,
+      "loss": 0.0586,
+      "step": 2172
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 2.7746243739565947e-05,
+      "loss": 0.1109,
+      "step": 2173
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 2.7712854757929883e-05,
+      "loss": 0.0539,
+      "step": 2174
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 2.7679465776293823e-05,
+      "loss": 0.1163,
+      "step": 2175
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 2.7646076794657767e-05,
+      "loss": 0.0878,
+      "step": 2176
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 2.7612687813021703e-05,
+      "loss": 0.0674,
+      "step": 2177
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 2.7579298831385643e-05,
+      "loss": 0.0718,
+      "step": 2178
+    },
+    {
+      "epoch": 14.53,
+      "learning_rate": 2.7545909849749586e-05,
+      "loss": 0.0545,
+      "step": 2179
+    },
+    {
+      "epoch": 14.53,
+      "learning_rate": 2.7512520868113523e-05,
+      "loss": 0.0698,
+      "step": 2180
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 2.7479131886477466e-05,
+      "loss": 0.1067,
+      "step": 2181
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 2.7445742904841403e-05,
+      "loss": 0.0747,
+      "step": 2182
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 2.7412353923205343e-05,
+      "loss": 0.0951,
+      "step": 2183
+    },
+    {
+      "epoch": 14.56,
+      "learning_rate": 2.7378964941569286e-05,
+      "loss": 0.0743,
+      "step": 2184
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 2.7345575959933223e-05,
+      "loss": 0.0703,
+      "step": 2185
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 2.731218697829716e-05,
+      "loss": 0.0697,
+      "step": 2186
+    },
+    {
+      "epoch": 14.58,
+      "learning_rate": 2.7278797996661103e-05,
+      "loss": 0.0946,
+      "step": 2187
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 2.7245409015025043e-05,
+      "loss": 0.0461,
+      "step": 2188
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 2.7212020033388986e-05,
+      "loss": 0.0448,
+      "step": 2189
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 2.7178631051752923e-05,
+      "loss": 0.0794,
+      "step": 2190
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 2.714524207011686e-05,
+      "loss": 0.0963,
+      "step": 2191
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 2.7111853088480803e-05,
+      "loss": 0.0602,
+      "step": 2192
+    },
+    {
+      "epoch": 14.62,
+      "learning_rate": 2.7078464106844743e-05,
+      "loss": 0.048,
+      "step": 2193
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 2.704507512520868e-05,
+      "loss": 0.0668,
+      "step": 2194
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 2.7011686143572623e-05,
+      "loss": 0.0624,
+      "step": 2195
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 2.6978297161936563e-05,
+      "loss": 0.0447,
+      "step": 2196
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 2.6944908180300506e-05,
+      "loss": 0.0534,
+      "step": 2197
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 2.6911519198664442e-05,
+      "loss": 0.0505,
+      "step": 2198
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 2.687813021702838e-05,
+      "loss": 0.0613,
+      "step": 2199
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 2.6844741235392322e-05,
+      "loss": 0.0721,
+      "step": 2200
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 2.6811352253756262e-05,
+      "loss": 0.0523,
+      "step": 2201
+    },
+    {
+      "epoch": 14.68,
+      "learning_rate": 2.67779632721202e-05,
+      "loss": 0.0526,
+      "step": 2202
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 2.6744574290484142e-05,
+      "loss": 0.0643,
+      "step": 2203
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 2.671118530884808e-05,
+      "loss": 0.0579,
+      "step": 2204
+    },
+    {
+      "epoch": 14.7,
+      "learning_rate": 2.6677796327212022e-05,
+      "loss": 0.055,
+      "step": 2205
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 2.6644407345575962e-05,
+      "loss": 0.0597,
+      "step": 2206
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 2.66110183639399e-05,
+      "loss": 0.061,
+      "step": 2207
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 2.6577629382303842e-05,
+      "loss": 0.0757,
+      "step": 2208
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 2.654424040066778e-05,
+      "loss": 0.0489,
+      "step": 2209
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 2.651085141903172e-05,
+      "loss": 0.0623,
+      "step": 2210
+    },
+    {
+      "epoch": 14.74,
+      "learning_rate": 2.6477462437395662e-05,
+      "loss": 0.0869,
+      "step": 2211
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 2.64440734557596e-05,
+      "loss": 0.058,
+      "step": 2212
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 2.6410684474123542e-05,
+      "loss": 0.0661,
+      "step": 2213
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 2.6377295492487482e-05,
+      "loss": 0.0872,
+      "step": 2214
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 2.634390651085142e-05,
+      "loss": 0.0625,
+      "step": 2215
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 2.6310517529215362e-05,
+      "loss": 0.0957,
+      "step": 2216
+    },
+    {
+      "epoch": 14.78,
+      "learning_rate": 2.62771285475793e-05,
+      "loss": 0.0699,
+      "step": 2217
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 2.624373956594324e-05,
+      "loss": 0.0533,
+      "step": 2218
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 2.6210350584307182e-05,
+      "loss": 0.0581,
+      "step": 2219
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 2.6176961602671118e-05,
+      "loss": 0.0834,
+      "step": 2220
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 2.614357262103506e-05,
+      "loss": 0.0537,
+      "step": 2221
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 2.6110183639398998e-05,
+      "loss": 0.145,
+      "step": 2222
+    },
+    {
+      "epoch": 14.82,
+      "learning_rate": 2.6076794657762938e-05,
+      "loss": 0.0688,
+      "step": 2223
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 2.604340567612688e-05,
+      "loss": 0.0572,
+      "step": 2224
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 2.6010016694490818e-05,
+      "loss": 0.0576,
+      "step": 2225
+    },
+    {
+      "epoch": 14.84,
+      "learning_rate": 2.597662771285476e-05,
+      "loss": 0.0441,
+      "step": 2226
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 2.59432387312187e-05,
+      "loss": 0.0723,
+      "step": 2227
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 2.5909849749582638e-05,
+      "loss": 0.05,
+      "step": 2228
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 2.587646076794658e-05,
+      "loss": 0.063,
+      "step": 2229
+    },
+    {
+      "epoch": 14.87,
+      "learning_rate": 2.5843071786310518e-05,
+      "loss": 0.0576,
+      "step": 2230
+    },
+    {
+      "epoch": 14.87,
+      "learning_rate": 2.5809682804674458e-05,
+      "loss": 0.0506,
+      "step": 2231
+    },
+    {
+      "epoch": 14.88,
+      "learning_rate": 2.57762938230384e-05,
+      "loss": 0.0942,
+      "step": 2232
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 2.5742904841402338e-05,
+      "loss": 0.1494,
+      "step": 2233
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 2.570951585976628e-05,
+      "loss": 0.1083,
+      "step": 2234
+    },
+    {
+      "epoch": 14.9,
+      "learning_rate": 2.5676126878130218e-05,
+      "loss": 0.092,
+      "step": 2235
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 2.5642737896494158e-05,
+      "loss": 0.0861,
+      "step": 2236
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 2.56093489148581e-05,
+      "loss": 0.0737,
+      "step": 2237
+    },
+    {
+      "epoch": 14.92,
+      "learning_rate": 2.5575959933222038e-05,
+      "loss": 0.057,
+      "step": 2238
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 2.5542570951585974e-05,
+      "loss": 0.0904,
+      "step": 2239
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 2.5509181969949918e-05,
+      "loss": 0.0624,
+      "step": 2240
+    },
+    {
+      "epoch": 14.94,
+      "learning_rate": 2.5475792988313858e-05,
+      "loss": 0.0682,
+      "step": 2241
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 2.54424040066778e-05,
+      "loss": 0.0545,
+      "step": 2242
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 2.5409015025041737e-05,
+      "loss": 0.0583,
+      "step": 2243
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 2.5375626043405677e-05,
+      "loss": 0.067,
+      "step": 2244
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 2.534223706176962e-05,
+      "loss": 0.0522,
+      "step": 2245
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 2.5308848080133557e-05,
+      "loss": 0.1015,
+      "step": 2246
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 2.5275459098497494e-05,
+      "loss": 0.075,
+      "step": 2247
+    },
+    {
+      "epoch": 14.99,
+      "learning_rate": 2.5242070116861437e-05,
+      "loss": 0.0809,
+      "step": 2248
+    },
+    {
+      "epoch": 14.99,
+      "learning_rate": 2.5208681135225377e-05,
+      "loss": 0.0508,
+      "step": 2249
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 2.517529215358932e-05,
+      "loss": 0.067,
+      "step": 2250
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 3.047564744949341,
+      "eval_runtime": 20.1901,
+      "eval_samples_per_second": 9.906,
+      "eval_steps_per_second": 1.238,
+      "step": 2250
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 2.5141903171953257e-05,
+      "loss": 0.0908,
+      "step": 2251
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 2.5108514190317194e-05,
+      "loss": 0.1154,
+      "step": 2252
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 2.5075125208681137e-05,
+      "loss": 0.0454,
+      "step": 2253
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 2.5041736227045077e-05,
+      "loss": 0.045,
+      "step": 2254
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 2.5008347245409014e-05,
+      "loss": 0.0547,
+      "step": 2255
+    },
+    {
+      "epoch": 15.04,
+      "learning_rate": 2.4974958263772957e-05,
+      "loss": 0.0474,
+      "step": 2256
+    },
+    {
+      "epoch": 15.05,
+      "learning_rate": 2.4941569282136894e-05,
+      "loss": 0.0651,
+      "step": 2257
+    },
+    {
+      "epoch": 15.05,
+      "learning_rate": 2.4908180300500837e-05,
+      "loss": 0.0785,
+      "step": 2258
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 2.4874791318864777e-05,
+      "loss": 0.0431,
+      "step": 2259
+    },
+    {
+      "epoch": 15.07,
+      "learning_rate": 2.4841402337228717e-05,
+      "loss": 0.0635,
+      "step": 2260
+    },
+    {
+      "epoch": 15.07,
+      "learning_rate": 2.4808013355592653e-05,
+      "loss": 0.0374,
+      "step": 2261
+    },
+    {
+      "epoch": 15.08,
+      "learning_rate": 2.4774624373956597e-05,
+      "loss": 0.0437,
+      "step": 2262
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 2.4741235392320537e-05,
+      "loss": 0.0588,
+      "step": 2263
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 2.4707846410684477e-05,
+      "loss": 0.0431,
+      "step": 2264
+    },
+    {
+      "epoch": 15.1,
+      "learning_rate": 2.4674457429048413e-05,
+      "loss": 0.0746,
+      "step": 2265
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 2.4641068447412353e-05,
+      "loss": 0.0743,
+      "step": 2266
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 2.4607679465776297e-05,
+      "loss": 0.0575,
+      "step": 2267
+    },
+    {
+      "epoch": 15.12,
+      "learning_rate": 2.4574290484140237e-05,
+      "loss": 0.0584,
+      "step": 2268
+    },
+    {
+      "epoch": 15.13,
+      "learning_rate": 2.4540901502504173e-05,
+      "loss": 0.0513,
+      "step": 2269
+    },
+    {
+      "epoch": 15.13,
+      "learning_rate": 2.4507512520868113e-05,
+      "loss": 0.059,
+      "step": 2270
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 2.4474123539232056e-05,
+      "loss": 0.0533,
+      "step": 2271
+    },
+    {
+      "epoch": 15.15,
+      "learning_rate": 2.4440734557595996e-05,
+      "loss": 0.0597,
+      "step": 2272
+    },
+    {
+      "epoch": 15.15,
+      "learning_rate": 2.4407345575959933e-05,
+      "loss": 0.0518,
+      "step": 2273
+    },
+    {
+      "epoch": 15.16,
+      "learning_rate": 2.4373956594323873e-05,
+      "loss": 0.064,
+      "step": 2274
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 2.4340567612687813e-05,
+      "loss": 0.0514,
+      "step": 2275
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 2.4307178631051756e-05,
+      "loss": 0.1018,
+      "step": 2276
+    },
+    {
+      "epoch": 15.18,
+      "learning_rate": 2.4273789649415693e-05,
+      "loss": 0.1495,
+      "step": 2277
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 2.4240400667779633e-05,
+      "loss": 0.0641,
+      "step": 2278
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 2.4207011686143573e-05,
+      "loss": 0.0645,
+      "step": 2279
+    },
+    {
+      "epoch": 15.2,
+      "learning_rate": 2.4173622704507516e-05,
+      "loss": 0.0571,
+      "step": 2280
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 2.4140233722871453e-05,
+      "loss": 0.0854,
+      "step": 2281
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 2.4106844741235393e-05,
+      "loss": 0.1215,
+      "step": 2282
+    },
+    {
+      "epoch": 15.22,
+      "learning_rate": 2.4073455759599333e-05,
+      "loss": 0.0761,
+      "step": 2283
+    },
+    {
+      "epoch": 15.23,
+      "learning_rate": 2.4040066777963273e-05,
+      "loss": 0.041,
+      "step": 2284
+    },
+    {
+      "epoch": 15.23,
+      "learning_rate": 2.4006677796327213e-05,
+      "loss": 0.1029,
+      "step": 2285
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 2.3973288814691153e-05,
+      "loss": 0.0588,
+      "step": 2286
+    },
+    {
+      "epoch": 15.25,
+      "learning_rate": 2.3939899833055093e-05,
+      "loss": 0.0587,
+      "step": 2287
+    },
+    {
+      "epoch": 15.25,
+      "learning_rate": 2.3906510851419033e-05,
+      "loss": 0.0531,
+      "step": 2288
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 2.3873121869782972e-05,
+      "loss": 0.0835,
+      "step": 2289
+    },
+    {
+      "epoch": 15.27,
+      "learning_rate": 2.3839732888146912e-05,
+      "loss": 0.0477,
+      "step": 2290
+    },
+    {
+      "epoch": 15.27,
+      "learning_rate": 2.3806343906510852e-05,
+      "loss": 0.062,
+      "step": 2291
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 2.3772954924874792e-05,
+      "loss": 0.1009,
+      "step": 2292
+    },
+    {
+      "epoch": 15.29,
+      "learning_rate": 2.3739565943238732e-05,
+      "loss": 0.0599,
+      "step": 2293
+    },
+    {
+      "epoch": 15.29,
+      "learning_rate": 2.3706176961602672e-05,
+      "loss": 0.0517,
+      "step": 2294
+    },
+    {
+      "epoch": 15.3,
+      "learning_rate": 2.3672787979966612e-05,
+      "loss": 0.0471,
+      "step": 2295
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 2.3639398998330552e-05,
+      "loss": 0.0622,
+      "step": 2296
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 2.3606010016694492e-05,
+      "loss": 0.0825,
+      "step": 2297
+    },
+    {
+      "epoch": 15.32,
+      "learning_rate": 2.3572621035058432e-05,
+      "loss": 0.0657,
+      "step": 2298
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 2.3539232053422372e-05,
+      "loss": 0.0744,
+      "step": 2299
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 2.3505843071786312e-05,
+      "loss": 0.0652,
+      "step": 2300
+    },
+    {
+      "epoch": 15.34,
+      "learning_rate": 2.3472454090150252e-05,
+      "loss": 0.0993,
+      "step": 2301
+    },
+    {
+      "epoch": 15.35,
+      "learning_rate": 2.3439065108514192e-05,
+      "loss": 0.0473,
+      "step": 2302
+    },
+    {
+      "epoch": 15.35,
+      "learning_rate": 2.3405676126878132e-05,
+      "loss": 0.0971,
+      "step": 2303
+    },
+    {
+      "epoch": 15.36,
+      "learning_rate": 2.3372287145242072e-05,
+      "loss": 0.0539,
+      "step": 2304
+    },
+    {
+      "epoch": 15.37,
+      "learning_rate": 2.333889816360601e-05,
+      "loss": 0.0469,
+      "step": 2305
+    },
+    {
+      "epoch": 15.37,
+      "learning_rate": 2.3305509181969952e-05,
+      "loss": 0.0623,
+      "step": 2306
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 2.3272120200333892e-05,
+      "loss": 0.1135,
+      "step": 2307
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 2.3238731218697832e-05,
+      "loss": 0.0476,
+      "step": 2308
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 2.320534223706177e-05,
+      "loss": 0.0686,
+      "step": 2309
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 2.3171953255425712e-05,
+      "loss": 0.0788,
+      "step": 2310
+    },
+    {
+      "epoch": 15.41,
+      "learning_rate": 2.3138564273789652e-05,
+      "loss": 0.0619,
+      "step": 2311
+    },
+    {
+      "epoch": 15.41,
+      "learning_rate": 2.310517529215359e-05,
+      "loss": 0.063,
+      "step": 2312
+    },
+    {
+      "epoch": 15.42,
+      "learning_rate": 2.3071786310517528e-05,
+      "loss": 0.0646,
+      "step": 2313
+    },
+    {
+      "epoch": 15.43,
+      "learning_rate": 2.3038397328881468e-05,
+      "loss": 0.0603,
+      "step": 2314
+    },
+    {
+      "epoch": 15.43,
+      "learning_rate": 2.300500834724541e-05,
+      "loss": 0.062,
+      "step": 2315
+    },
+    {
+      "epoch": 15.44,
+      "learning_rate": 2.297161936560935e-05,
+      "loss": 0.0895,
+      "step": 2316
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 2.2938230383973288e-05,
+      "loss": 0.1395,
+      "step": 2317
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 2.2904841402337228e-05,
+      "loss": 0.0593,
+      "step": 2318
+    },
+    {
+      "epoch": 15.46,
+      "learning_rate": 2.287145242070117e-05,
+      "loss": 0.0489,
+      "step": 2319
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 2.283806343906511e-05,
+      "loss": 0.052,
+      "step": 2320
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 2.2804674457429048e-05,
+      "loss": 0.0516,
+      "step": 2321
+    },
+    {
+      "epoch": 15.48,
+      "learning_rate": 2.2771285475792988e-05,
+      "loss": 0.0572,
+      "step": 2322
+    },
+    {
+      "epoch": 15.49,
+      "learning_rate": 2.2737896494156928e-05,
+      "loss": 0.0506,
+      "step": 2323
+    },
+    {
+      "epoch": 15.49,
+      "learning_rate": 2.270450751252087e-05,
+      "loss": 0.1025,
+      "step": 2324
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 2.2671118530884808e-05,
+      "loss": 0.0865,
+      "step": 2325
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 2.2637729549248748e-05,
+      "loss": 0.0395,
+      "step": 2326
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 2.2604340567612688e-05,
+      "loss": 0.0644,
+      "step": 2327
+    },
+    {
+      "epoch": 15.52,
+      "learning_rate": 2.257095158597663e-05,
+      "loss": 0.0528,
+      "step": 2328
+    },
+    {
+      "epoch": 15.53,
+      "learning_rate": 2.2537562604340568e-05,
+      "loss": 0.0794,
+      "step": 2329
+    },
+    {
+      "epoch": 15.53,
+      "learning_rate": 2.2504173622704508e-05,
+      "loss": 0.0454,
+      "step": 2330
+    },
+    {
+      "epoch": 15.54,
+      "learning_rate": 2.2470784641068448e-05,
+      "loss": 0.0383,
+      "step": 2331
+    },
+    {
+      "epoch": 15.55,
+      "learning_rate": 2.2437395659432388e-05,
+      "loss": 0.0732,
+      "step": 2332
+    },
+    {
+      "epoch": 15.55,
+      "learning_rate": 2.2404006677796328e-05,
+      "loss": 0.0441,
+      "step": 2333
+    },
+    {
+      "epoch": 15.56,
+      "learning_rate": 2.2370617696160268e-05,
+      "loss": 0.0526,
+      "step": 2334
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 2.2337228714524207e-05,
+      "loss": 0.057,
+      "step": 2335
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 2.2303839732888147e-05,
+      "loss": 0.0448,
+      "step": 2336
+    },
+    {
+      "epoch": 15.58,
+      "learning_rate": 2.2270450751252087e-05,
+      "loss": 0.1354,
+      "step": 2337
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 2.2237061769616027e-05,
+      "loss": 0.0638,
+      "step": 2338
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 2.2203672787979967e-05,
+      "loss": 0.0579,
+      "step": 2339
+    },
+    {
+      "epoch": 15.6,
+      "learning_rate": 2.2170283806343907e-05,
+      "loss": 0.0561,
+      "step": 2340
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 2.2136894824707847e-05,
+      "loss": 0.0549,
+      "step": 2341
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 2.2103505843071787e-05,
+      "loss": 0.0633,
+      "step": 2342
+    },
+    {
+      "epoch": 15.62,
+      "learning_rate": 2.2070116861435727e-05,
+      "loss": 0.0858,
+      "step": 2343
+    },
+    {
+      "epoch": 15.63,
+      "learning_rate": 2.2036727879799667e-05,
+      "loss": 0.0673,
+      "step": 2344
+    },
+    {
+      "epoch": 15.63,
+      "learning_rate": 2.2003338898163607e-05,
+      "loss": 0.0728,
+      "step": 2345
+    },
+    {
+      "epoch": 15.64,
+      "learning_rate": 2.1969949916527547e-05,
+      "loss": 0.0565,
+      "step": 2346
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 2.1936560934891487e-05,
+      "loss": 0.0988,
+      "step": 2347
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 2.1903171953255427e-05,
+      "loss": 0.0678,
+      "step": 2348
+    },
+    {
+      "epoch": 15.66,
+      "learning_rate": 2.1869782971619364e-05,
+      "loss": 0.1142,
+      "step": 2349
+    },
+    {
+      "epoch": 15.67,
+      "learning_rate": 2.1836393989983307e-05,
+      "loss": 0.0712,
+      "step": 2350
+    },
+    {
+      "epoch": 15.67,
+      "learning_rate": 2.1803005008347247e-05,
+      "loss": 0.066,
+      "step": 2351
+    },
+    {
+      "epoch": 15.68,
+      "learning_rate": 2.1769616026711187e-05,
+      "loss": 0.0453,
+      "step": 2352
+    },
+    {
+      "epoch": 15.69,
+      "learning_rate": 2.1736227045075127e-05,
+      "loss": 0.0542,
+      "step": 2353
+    },
+    {
+      "epoch": 15.69,
+      "learning_rate": 2.1702838063439067e-05,
+      "loss": 0.0921,
+      "step": 2354
+    },
+    {
+      "epoch": 15.7,
+      "learning_rate": 2.1669449081803007e-05,
+      "loss": 0.0765,
+      "step": 2355
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 2.1636060100166947e-05,
+      "loss": 0.0873,
+      "step": 2356
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 2.1602671118530887e-05,
+      "loss": 0.0533,
+      "step": 2357
+    },
+    {
+      "epoch": 15.72,
+      "learning_rate": 2.1569282136894827e-05,
+      "loss": 0.072,
+      "step": 2358
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 2.1535893155258767e-05,
+      "loss": 0.0882,
+      "step": 2359
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 2.1502504173622707e-05,
+      "loss": 0.0638,
+      "step": 2360
+    },
+    {
+      "epoch": 15.74,
+      "learning_rate": 2.1469115191986647e-05,
+      "loss": 0.0577,
+      "step": 2361
+    },
+    {
+      "epoch": 15.75,
+      "learning_rate": 2.1435726210350583e-05,
+      "loss": 0.1565,
+      "step": 2362
+    },
+    {
+      "epoch": 15.75,
+      "learning_rate": 2.1402337228714526e-05,
+      "loss": 0.0908,
+      "step": 2363
+    },
+    {
+      "epoch": 15.76,
+      "learning_rate": 2.1368948247078466e-05,
+      "loss": 0.0612,
+      "step": 2364
+    },
+    {
+      "epoch": 15.77,
+      "learning_rate": 2.1335559265442406e-05,
+      "loss": 0.095,
+      "step": 2365
+    },
+    {
+      "epoch": 15.77,
+      "learning_rate": 2.1302170283806343e-05,
+      "loss": 0.0668,
+      "step": 2366
+    },
+    {
+      "epoch": 15.78,
+      "learning_rate": 2.1268781302170286e-05,
+      "loss": 0.0426,
+      "step": 2367
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 2.1235392320534226e-05,
+      "loss": 0.098,
+      "step": 2368
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 2.1202003338898166e-05,
+      "loss": 0.072,
+      "step": 2369
+    },
+    {
+      "epoch": 15.8,
+      "learning_rate": 2.1168614357262103e-05,
+      "loss": 0.0807,
+      "step": 2370
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 2.1135225375626043e-05,
+      "loss": 0.0682,
+      "step": 2371
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 2.1101836393989986e-05,
+      "loss": 0.0514,
+      "step": 2372
+    },
+    {
+      "epoch": 15.82,
+      "learning_rate": 2.1068447412353926e-05,
+      "loss": 0.0515,
+      "step": 2373
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 2.1035058430717863e-05,
+      "loss": 0.0749,
+      "step": 2374
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 2.1001669449081803e-05,
+      "loss": 0.064,
+      "step": 2375
+    },
+    {
+      "epoch": 15.84,
+      "learning_rate": 2.0968280467445746e-05,
+      "loss": 0.0576,
+      "step": 2376
+    },
+    {
+      "epoch": 15.85,
+      "learning_rate": 2.0934891485809686e-05,
+      "loss": 0.0657,
+      "step": 2377
+    },
+    {
+      "epoch": 15.85,
+      "learning_rate": 2.0901502504173623e-05,
+      "loss": 0.0944,
+      "step": 2378
+    },
+    {
+      "epoch": 15.86,
+      "learning_rate": 2.0868113522537563e-05,
+      "loss": 0.0847,
+      "step": 2379
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 2.0834724540901503e-05,
+      "loss": 0.0621,
+      "step": 2380
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 2.0801335559265446e-05,
+      "loss": 0.0561,
+      "step": 2381
+    },
+    {
+      "epoch": 15.88,
+      "learning_rate": 2.0767946577629382e-05,
+      "loss": 0.0576,
+      "step": 2382
+    },
+    {
+      "epoch": 15.89,
+      "learning_rate": 2.0734557595993322e-05,
+      "loss": 0.0629,
+      "step": 2383
+    },
+    {
+      "epoch": 15.89,
+      "learning_rate": 2.0701168614357262e-05,
+      "loss": 0.1094,
+      "step": 2384
+    },
+    {
+      "epoch": 15.9,
+      "learning_rate": 2.0667779632721206e-05,
+      "loss": 0.0724,
+      "step": 2385
+    },
+    {
+      "epoch": 15.91,
+      "learning_rate": 2.0634390651085142e-05,
+      "loss": 0.0747,
+      "step": 2386
+    },
+    {
+      "epoch": 15.91,
+      "learning_rate": 2.0601001669449082e-05,
+      "loss": 0.0728,
+      "step": 2387
+    },
+    {
+      "epoch": 15.92,
+      "learning_rate": 2.0567612687813022e-05,
+      "loss": 0.089,
+      "step": 2388
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 2.0534223706176962e-05,
+      "loss": 0.0572,
+      "step": 2389
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 2.0500834724540902e-05,
+      "loss": 0.0866,
+      "step": 2390
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 2.0467445742904842e-05,
+      "loss": 0.0771,
+      "step": 2391
+    },
+    {
+      "epoch": 15.95,
+      "learning_rate": 2.0434056761268782e-05,
+      "loss": 0.0472,
+      "step": 2392
+    },
+    {
+      "epoch": 15.95,
+      "learning_rate": 2.0400667779632722e-05,
+      "loss": 0.0596,
+      "step": 2393
+    },
+    {
+      "epoch": 15.96,
+      "learning_rate": 2.0367278797996662e-05,
+      "loss": 0.0618,
+      "step": 2394
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 2.0333889816360602e-05,
+      "loss": 0.0669,
+      "step": 2395
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 2.0300500834724542e-05,
+      "loss": 0.0509,
+      "step": 2396
+    },
+    {
+      "epoch": 15.98,
+      "learning_rate": 2.0267111853088482e-05,
+      "loss": 0.0681,
+      "step": 2397
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 2.0233722871452422e-05,
+      "loss": 0.0614,
+      "step": 2398
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 2.0200333889816362e-05,
+      "loss": 0.0993,
+      "step": 2399
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 2.0166944908180302e-05,
+      "loss": 0.065,
+      "step": 2400
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 3.1341705322265625,
+      "eval_runtime": 20.2079,
+      "eval_samples_per_second": 9.897,
+      "eval_steps_per_second": 1.237,
+      "step": 2400
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 2.0133555926544242e-05,
+      "loss": 0.0458,
+      "step": 2401
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 2.0100166944908182e-05,
+      "loss": 0.0449,
+      "step": 2402
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 2.0066777963272122e-05,
+      "loss": 0.0416,
+      "step": 2403
+    },
+    {
+      "epoch": 16.03,
+      "learning_rate": 2.003338898163606e-05,
+      "loss": 0.0586,
+      "step": 2404
+    },
+    {
+      "epoch": 16.03,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 2405
+    },
+    {
+      "epoch": 16.04,
+      "learning_rate": 1.9966611018363938e-05,
+      "loss": 0.0412,
+      "step": 2406
+    },
+    {
+      "epoch": 16.05,
+      "learning_rate": 1.993322203672788e-05,
+      "loss": 0.0408,
+      "step": 2407
+    },
+    {
+      "epoch": 16.05,
+      "learning_rate": 1.989983305509182e-05,
+      "loss": 0.0744,
+      "step": 2408
+    },
+    {
+      "epoch": 16.06,
+      "learning_rate": 1.986644407345576e-05,
+      "loss": 0.0568,
+      "step": 2409
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 1.9833055091819698e-05,
+      "loss": 0.0518,
+      "step": 2410
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 1.979966611018364e-05,
+      "loss": 0.0752,
+      "step": 2411
+    },
+    {
+      "epoch": 16.08,
+      "learning_rate": 1.976627712854758e-05,
+      "loss": 0.0919,
+      "step": 2412
+    },
+    {
+      "epoch": 16.09,
+      "learning_rate": 1.973288814691152e-05,
+      "loss": 0.0499,
+      "step": 2413
+    },
+    {
+      "epoch": 16.09,
+      "learning_rate": 1.9699499165275458e-05,
+      "loss": 0.117,
+      "step": 2414
+    },
+    {
+      "epoch": 16.1,
+      "learning_rate": 1.96661101836394e-05,
+      "loss": 0.0546,
+      "step": 2415
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 1.963272120200334e-05,
+      "loss": 0.0924,
+      "step": 2416
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 1.959933222036728e-05,
+      "loss": 0.0553,
+      "step": 2417
+    },
+    {
+      "epoch": 16.12,
+      "learning_rate": 1.9565943238731218e-05,
+      "loss": 0.0897,
+      "step": 2418
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 1.9532554257095158e-05,
+      "loss": 0.0606,
+      "step": 2419
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 1.94991652754591e-05,
+      "loss": 0.0581,
+      "step": 2420
+    },
+    {
+      "epoch": 16.14,
+      "learning_rate": 1.946577629382304e-05,
+      "loss": 0.0636,
+      "step": 2421
+    },
+    {
+      "epoch": 16.15,
+      "learning_rate": 1.9432387312186978e-05,
+      "loss": 0.0771,
+      "step": 2422
+    },
+    {
+      "epoch": 16.15,
+      "learning_rate": 1.9398998330550918e-05,
+      "loss": 0.0943,
+      "step": 2423
+    },
+    {
+      "epoch": 16.16,
+      "learning_rate": 1.936560934891486e-05,
+      "loss": 0.1236,
+      "step": 2424
+    },
+    {
+      "epoch": 16.17,
+      "learning_rate": 1.93322203672788e-05,
+      "loss": 0.0364,
+      "step": 2425
+    },
+    {
+      "epoch": 16.17,
+      "learning_rate": 1.9298831385642738e-05,
+      "loss": 0.0556,
+      "step": 2426
+    },
+    {
+      "epoch": 16.18,
+      "learning_rate": 1.9265442404006677e-05,
+      "loss": 0.0442,
+      "step": 2427
+    },
+    {
+      "epoch": 16.19,
+      "learning_rate": 1.9232053422370617e-05,
+      "loss": 0.0488,
+      "step": 2428
+    },
+    {
+      "epoch": 16.19,
+      "learning_rate": 1.919866444073456e-05,
+      "loss": 0.0475,
+      "step": 2429
+    },
+    {
+      "epoch": 16.2,
+      "learning_rate": 1.9165275459098497e-05,
+      "loss": 0.065,
+      "step": 2430
+    },
+    {
+      "epoch": 16.21,
+      "learning_rate": 1.9131886477462437e-05,
+      "loss": 0.0492,
+      "step": 2431
+    },
+    {
+      "epoch": 16.21,
+      "learning_rate": 1.9098497495826377e-05,
+      "loss": 0.0675,
+      "step": 2432
+    },
+    {
+      "epoch": 16.22,
+      "learning_rate": 1.906510851419032e-05,
+      "loss": 0.0494,
+      "step": 2433
+    },
+    {
+      "epoch": 16.23,
+      "learning_rate": 1.9031719532554257e-05,
+      "loss": 0.0687,
+      "step": 2434
+    },
+    {
+      "epoch": 16.23,
+      "learning_rate": 1.8998330550918197e-05,
+      "loss": 0.0523,
+      "step": 2435
+    },
+    {
+      "epoch": 16.24,
+      "learning_rate": 1.8964941569282137e-05,
+      "loss": 0.053,
+      "step": 2436
+    },
+    {
+      "epoch": 16.25,
+      "learning_rate": 1.8931552587646077e-05,
+      "loss": 0.051,
+      "step": 2437
+    },
+    {
+      "epoch": 16.25,
+      "learning_rate": 1.8898163606010017e-05,
+      "loss": 0.106,
+      "step": 2438
+    },
+    {
+      "epoch": 16.26,
+      "learning_rate": 1.8864774624373957e-05,
+      "loss": 0.0561,
+      "step": 2439
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 1.8831385642737897e-05,
+      "loss": 0.0737,
+      "step": 2440
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 1.8797996661101837e-05,
+      "loss": 0.0612,
+      "step": 2441
+    },
+    {
+      "epoch": 16.28,
+      "learning_rate": 1.8764607679465777e-05,
+      "loss": 0.0772,
+      "step": 2442
+    },
+    {
+      "epoch": 16.29,
+      "learning_rate": 1.8731218697829717e-05,
+      "loss": 0.0564,
+      "step": 2443
+    },
+    {
+      "epoch": 16.29,
+      "learning_rate": 1.8697829716193657e-05,
+      "loss": 0.0606,
+      "step": 2444
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 1.8664440734557597e-05,
+      "loss": 0.095,
+      "step": 2445
+    },
+    {
+      "epoch": 16.31,
+      "learning_rate": 1.8631051752921537e-05,
+      "loss": 0.044,
+      "step": 2446
+    },
+    {
+      "epoch": 16.31,
+      "learning_rate": 1.8597662771285477e-05,
+      "loss": 0.0377,
+      "step": 2447
+    },
+    {
+      "epoch": 16.32,
+      "learning_rate": 1.8564273789649417e-05,
+      "loss": 0.0719,
+      "step": 2448
+    },
+    {
+      "epoch": 16.33,
+      "learning_rate": 1.8530884808013357e-05,
+      "loss": 0.0511,
+      "step": 2449
+    },
+    {
+      "epoch": 16.33,
+      "learning_rate": 1.8497495826377297e-05,
+      "loss": 0.0804,
+      "step": 2450
+    },
+    {
+      "epoch": 16.34,
+      "learning_rate": 1.8464106844741237e-05,
+      "loss": 0.0427,
+      "step": 2451
+    },
+    {
+      "epoch": 16.35,
+      "learning_rate": 1.8430717863105177e-05,
+      "loss": 0.0986,
+      "step": 2452
+    },
+    {
+      "epoch": 16.35,
+      "learning_rate": 1.8397328881469117e-05,
+      "loss": 0.0687,
+      "step": 2453
+    },
+    {
+      "epoch": 16.36,
+      "learning_rate": 1.8363939899833053e-05,
+      "loss": 0.0585,
+      "step": 2454
+    },
+    {
+      "epoch": 16.37,
+      "learning_rate": 1.8330550918196996e-05,
+      "loss": 0.0603,
+      "step": 2455
+    },
+    {
+      "epoch": 16.37,
+      "learning_rate": 1.8297161936560936e-05,
+      "loss": 0.0865,
+      "step": 2456
+    },
+    {
+      "epoch": 16.38,
+      "learning_rate": 1.8263772954924876e-05,
+      "loss": 0.0725,
+      "step": 2457
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 1.8230383973288813e-05,
+      "loss": 0.0557,
+      "step": 2458
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 1.8196994991652756e-05,
+      "loss": 0.089,
+      "step": 2459
+    },
+    {
+      "epoch": 16.4,
+      "learning_rate": 1.8163606010016696e-05,
+      "loss": 0.0567,
+      "step": 2460
+    },
+    {
+      "epoch": 16.41,
+      "learning_rate": 1.8130217028380636e-05,
+      "loss": 0.0455,
+      "step": 2461
+    },
+    {
+      "epoch": 16.41,
+      "learning_rate": 1.8096828046744573e-05,
+      "loss": 0.0681,
+      "step": 2462
+    },
+    {
+      "epoch": 16.42,
+      "learning_rate": 1.8063439065108513e-05,
+      "loss": 0.0743,
+      "step": 2463
+    },
+    {
+      "epoch": 16.43,
+      "learning_rate": 1.8030050083472456e-05,
+      "loss": 0.0417,
+      "step": 2464
+    },
+    {
+      "epoch": 16.43,
+      "learning_rate": 1.7996661101836396e-05,
+      "loss": 0.1458,
+      "step": 2465
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 1.7963272120200333e-05,
+      "loss": 0.0423,
+      "step": 2466
+    },
+    {
+      "epoch": 16.45,
+      "learning_rate": 1.7929883138564273e-05,
+      "loss": 0.0388,
+      "step": 2467
+    },
+    {
+      "epoch": 16.45,
+      "learning_rate": 1.7896494156928216e-05,
+      "loss": 0.0429,
+      "step": 2468
+    },
+    {
+      "epoch": 16.46,
+      "learning_rate": 1.7863105175292156e-05,
+      "loss": 0.0621,
+      "step": 2469
+    },
+    {
+      "epoch": 16.47,
+      "learning_rate": 1.7829716193656093e-05,
+      "loss": 0.0652,
+      "step": 2470
+    },
+    {
+      "epoch": 16.47,
+      "learning_rate": 1.7796327212020033e-05,
+      "loss": 0.0555,
+      "step": 2471
+    },
+    {
+      "epoch": 16.48,
+      "learning_rate": 1.7762938230383973e-05,
+      "loss": 0.0907,
+      "step": 2472
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 1.7729549248747916e-05,
+      "loss": 0.0642,
+      "step": 2473
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 1.7696160267111852e-05,
+      "loss": 0.1355,
+      "step": 2474
+    },
+    {
+      "epoch": 16.5,
+      "learning_rate": 1.7662771285475792e-05,
+      "loss": 0.0618,
+      "step": 2475
+    },
+    {
+      "epoch": 16.51,
+      "learning_rate": 1.7629382303839732e-05,
+      "loss": 0.0915,
+      "step": 2476
+    },
+    {
+      "epoch": 16.51,
+      "learning_rate": 1.7595993322203676e-05,
+      "loss": 0.0508,
+      "step": 2477
+    },
+    {
+      "epoch": 16.52,
+      "learning_rate": 1.7562604340567612e-05,
+      "loss": 0.0754,
+      "step": 2478
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 1.7529215358931552e-05,
+      "loss": 0.0463,
+      "step": 2479
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 1.7495826377295492e-05,
+      "loss": 0.0418,
+      "step": 2480
+    },
+    {
+      "epoch": 16.54,
+      "learning_rate": 1.7462437395659436e-05,
+      "loss": 0.0866,
+      "step": 2481
+    },
+    {
+      "epoch": 16.55,
+      "learning_rate": 1.7429048414023372e-05,
+      "loss": 0.0699,
+      "step": 2482
+    },
+    {
+      "epoch": 16.55,
+      "learning_rate": 1.7395659432387312e-05,
+      "loss": 0.0927,
+      "step": 2483
+    },
+    {
+      "epoch": 16.56,
+      "learning_rate": 1.7362270450751252e-05,
+      "loss": 0.0544,
+      "step": 2484
+    },
+    {
+      "epoch": 16.57,
+      "learning_rate": 1.7328881469115192e-05,
+      "loss": 0.0837,
+      "step": 2485
+    },
+    {
+      "epoch": 16.57,
+      "learning_rate": 1.7295492487479135e-05,
+      "loss": 0.075,
+      "step": 2486
+    },
+    {
+      "epoch": 16.58,
+      "learning_rate": 1.7262103505843072e-05,
+      "loss": 0.0458,
+      "step": 2487
+    },
+    {
+      "epoch": 16.59,
+      "learning_rate": 1.7228714524207012e-05,
+      "loss": 0.0571,
+      "step": 2488
+    },
+    {
+      "epoch": 16.59,
+      "learning_rate": 1.7195325542570952e-05,
+      "loss": 0.095,
+      "step": 2489
+    },
+    {
+      "epoch": 16.6,
+      "learning_rate": 1.7161936560934895e-05,
+      "loss": 0.082,
+      "step": 2490
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 1.7128547579298832e-05,
+      "loss": 0.037,
+      "step": 2491
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 1.7095158597662772e-05,
+      "loss": 0.0694,
+      "step": 2492
+    },
+    {
+      "epoch": 16.62,
+      "learning_rate": 1.7061769616026712e-05,
+      "loss": 0.0638,
+      "step": 2493
+    },
+    {
+      "epoch": 16.63,
+      "learning_rate": 1.7028380634390652e-05,
+      "loss": 0.0746,
+      "step": 2494
+    },
+    {
+      "epoch": 16.63,
+      "learning_rate": 1.6994991652754592e-05,
+      "loss": 0.0463,
+      "step": 2495
+    },
+    {
+      "epoch": 16.64,
+      "learning_rate": 1.696160267111853e-05,
+      "loss": 0.1283,
+      "step": 2496
+    },
+    {
+      "epoch": 16.65,
+      "learning_rate": 1.692821368948247e-05,
+      "loss": 0.0633,
+      "step": 2497
+    },
+    {
+      "epoch": 16.65,
+      "learning_rate": 1.689482470784641e-05,
+      "loss": 0.0639,
+      "step": 2498
+    },
+    {
+      "epoch": 16.66,
+      "learning_rate": 1.686143572621035e-05,
+      "loss": 0.0755,
+      "step": 2499
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 1.682804674457429e-05,
+      "loss": 0.0727,
+      "step": 2500
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 1.679465776293823e-05,
+      "loss": 0.0504,
+      "step": 2501
+    },
+    {
+      "epoch": 16.68,
+      "learning_rate": 1.676126878130217e-05,
+      "loss": 0.0757,
+      "step": 2502
+    },
+    {
+      "epoch": 16.69,
+      "learning_rate": 1.672787979966611e-05,
+      "loss": 0.0793,
+      "step": 2503
+    },
+    {
+      "epoch": 16.69,
+      "learning_rate": 1.669449081803005e-05,
+      "loss": 0.0595,
+      "step": 2504
+    },
+    {
+      "epoch": 16.7,
+      "learning_rate": 1.666110183639399e-05,
+      "loss": 0.0821,
+      "step": 2505
+    },
+    {
+      "epoch": 16.71,
+      "learning_rate": 1.662771285475793e-05,
+      "loss": 0.1124,
+      "step": 2506
+    },
+    {
+      "epoch": 16.71,
+      "learning_rate": 1.659432387312187e-05,
+      "loss": 0.0757,
+      "step": 2507
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 1.656093489148581e-05,
+      "loss": 0.061,
+      "step": 2508
+    },
+    {
+      "epoch": 16.73,
+      "learning_rate": 1.652754590984975e-05,
+      "loss": 0.0869,
+      "step": 2509
+    },
+    {
+      "epoch": 16.73,
+      "learning_rate": 1.649415692821369e-05,
+      "loss": 0.0487,
+      "step": 2510
+    },
+    {
+      "epoch": 16.74,
+      "learning_rate": 1.6460767946577628e-05,
+      "loss": 0.0585,
+      "step": 2511
+    },
+    {
+      "epoch": 16.75,
+      "learning_rate": 1.642737896494157e-05,
+      "loss": 0.0603,
+      "step": 2512
+    },
+    {
+      "epoch": 16.75,
+      "learning_rate": 1.639398998330551e-05,
+      "loss": 0.0683,
+      "step": 2513
+    },
+    {
+      "epoch": 16.76,
+      "learning_rate": 1.636060100166945e-05,
+      "loss": 0.0623,
+      "step": 2514
+    },
+    {
+      "epoch": 16.77,
+      "learning_rate": 1.6327212020033388e-05,
+      "loss": 0.0648,
+      "step": 2515
+    },
+    {
+      "epoch": 16.77,
+      "learning_rate": 1.629382303839733e-05,
+      "loss": 0.0579,
+      "step": 2516
+    },
+    {
+      "epoch": 16.78,
+      "learning_rate": 1.626043405676127e-05,
+      "loss": 0.0554,
+      "step": 2517
+    },
+    {
+      "epoch": 16.79,
+      "learning_rate": 1.622704507512521e-05,
+      "loss": 0.0548,
+      "step": 2518
+    },
+    {
+      "epoch": 16.79,
+      "learning_rate": 1.6193656093489147e-05,
+      "loss": 0.0787,
+      "step": 2519
+    },
+    {
+      "epoch": 16.8,
+      "learning_rate": 1.6160267111853087e-05,
+      "loss": 0.0946,
+      "step": 2520
+    },
+    {
+      "epoch": 16.81,
+      "learning_rate": 1.612687813021703e-05,
+      "loss": 0.0634,
+      "step": 2521
+    },
+    {
+      "epoch": 16.81,
+      "learning_rate": 1.609348914858097e-05,
+      "loss": 0.0997,
+      "step": 2522
+    },
+    {
+      "epoch": 16.82,
+      "learning_rate": 1.6060100166944907e-05,
+      "loss": 0.0567,
+      "step": 2523
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 1.6026711185308847e-05,
+      "loss": 0.074,
+      "step": 2524
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 1.599332220367279e-05,
+      "loss": 0.0884,
+      "step": 2525
+    },
+    {
+      "epoch": 16.84,
+      "learning_rate": 1.595993322203673e-05,
+      "loss": 0.0687,
+      "step": 2526
+    },
+    {
+      "epoch": 16.85,
+      "learning_rate": 1.5926544240400667e-05,
+      "loss": 0.0898,
+      "step": 2527
+    },
+    {
+      "epoch": 16.85,
+      "learning_rate": 1.5893155258764607e-05,
+      "loss": 0.0682,
+      "step": 2528
+    },
+    {
+      "epoch": 16.86,
+      "learning_rate": 1.5859766277128547e-05,
+      "loss": 0.0541,
+      "step": 2529
+    },
+    {
+      "epoch": 16.87,
+      "learning_rate": 1.582637729549249e-05,
+      "loss": 0.0894,
+      "step": 2530
+    },
+    {
+      "epoch": 16.87,
+      "learning_rate": 1.5792988313856427e-05,
+      "loss": 0.1342,
+      "step": 2531
+    },
+    {
+      "epoch": 16.88,
+      "learning_rate": 1.5759599332220367e-05,
+      "loss": 0.0637,
+      "step": 2532
+    },
+    {
+      "epoch": 16.89,
+      "learning_rate": 1.5726210350584307e-05,
+      "loss": 0.0673,
+      "step": 2533
+    },
+    {
+      "epoch": 16.89,
+      "learning_rate": 1.569282136894825e-05,
+      "loss": 0.0484,
+      "step": 2534
+    },
+    {
+      "epoch": 16.9,
+      "learning_rate": 1.5659432387312187e-05,
+      "loss": 0.0519,
+      "step": 2535
+    },
+    {
+      "epoch": 16.91,
+      "learning_rate": 1.5626043405676127e-05,
+      "loss": 0.0374,
+      "step": 2536
+    },
+    {
+      "epoch": 16.91,
+      "learning_rate": 1.5592654424040067e-05,
+      "loss": 0.0383,
+      "step": 2537
+    },
+    {
+      "epoch": 16.92,
+      "learning_rate": 1.555926544240401e-05,
+      "loss": 0.0647,
+      "step": 2538
+    },
+    {
+      "epoch": 16.93,
+      "learning_rate": 1.5525876460767947e-05,
+      "loss": 0.0557,
+      "step": 2539
+    },
+    {
+      "epoch": 16.93,
+      "learning_rate": 1.5492487479131887e-05,
+      "loss": 0.0491,
+      "step": 2540
+    },
+    {
+      "epoch": 16.94,
+      "learning_rate": 1.5459098497495827e-05,
+      "loss": 0.0587,
+      "step": 2541
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 1.5425709515859767e-05,
+      "loss": 0.0799,
+      "step": 2542
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 1.5392320534223707e-05,
+      "loss": 0.0604,
+      "step": 2543
+    },
+    {
+      "epoch": 16.96,
+      "learning_rate": 1.5358931552587647e-05,
+      "loss": 0.0619,
+      "step": 2544
+    },
+    {
+      "epoch": 16.97,
+      "learning_rate": 1.5325542570951587e-05,
+      "loss": 0.0904,
+      "step": 2545
+    },
+    {
+      "epoch": 16.97,
+      "learning_rate": 1.5292153589315527e-05,
+      "loss": 0.0677,
+      "step": 2546
+    },
+    {
+      "epoch": 16.98,
+      "learning_rate": 1.5258764607679465e-05,
+      "loss": 0.0662,
+      "step": 2547
+    },
+    {
+      "epoch": 16.99,
+      "learning_rate": 1.5225375626043406e-05,
+      "loss": 0.0516,
+      "step": 2548
+    },
+    {
+      "epoch": 16.99,
+      "learning_rate": 1.5191986644407346e-05,
+      "loss": 0.0519,
+      "step": 2549
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 1.5158597662771286e-05,
+      "loss": 0.067,
+      "step": 2550
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 3.190051317214966,
+      "eval_runtime": 20.1962,
+      "eval_samples_per_second": 9.903,
+      "eval_steps_per_second": 1.238,
+      "step": 2550
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 1.5125208681135225e-05,
+      "loss": 0.0725,
+      "step": 2551
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 1.5091819699499166e-05,
+      "loss": 0.0703,
+      "step": 2552
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 1.5058430717863106e-05,
+      "loss": 0.0454,
+      "step": 2553
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 1.5025041736227046e-05,
+      "loss": 0.0481,
+      "step": 2554
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 1.4991652754590985e-05,
+      "loss": 0.0583,
+      "step": 2555
+    },
+    {
+      "epoch": 17.04,
+      "learning_rate": 1.4958263772954924e-05,
+      "loss": 0.0646,
+      "step": 2556
+    },
+    {
+      "epoch": 17.05,
+      "learning_rate": 1.4924874791318866e-05,
+      "loss": 0.0521,
+      "step": 2557
+    },
+    {
+      "epoch": 17.05,
+      "learning_rate": 1.4891485809682806e-05,
+      "loss": 0.0785,
+      "step": 2558
+    },
+    {
+      "epoch": 17.06,
+      "learning_rate": 1.4858096828046744e-05,
+      "loss": 0.0592,
+      "step": 2559
+    },
+    {
+      "epoch": 17.07,
+      "learning_rate": 1.4824707846410684e-05,
+      "loss": 0.062,
+      "step": 2560
+    },
+    {
+      "epoch": 17.07,
+      "learning_rate": 1.4791318864774626e-05,
+      "loss": 0.0813,
+      "step": 2561
+    },
+    {
+      "epoch": 17.08,
+      "learning_rate": 1.4757929883138566e-05,
+      "loss": 0.1176,
+      "step": 2562
+    },
+    {
+      "epoch": 17.09,
+      "learning_rate": 1.4724540901502504e-05,
+      "loss": 0.0651,
+      "step": 2563
+    },
+    {
+      "epoch": 17.09,
+      "learning_rate": 1.4691151919866444e-05,
+      "loss": 0.0616,
+      "step": 2564
+    },
+    {
+      "epoch": 17.1,
+      "learning_rate": 1.4657762938230384e-05,
+      "loss": 0.0828,
+      "step": 2565
+    },
+    {
+      "epoch": 17.11,
+      "learning_rate": 1.4624373956594326e-05,
+      "loss": 0.0402,
+      "step": 2566
+    },
+    {
+      "epoch": 17.11,
+      "learning_rate": 1.4590984974958264e-05,
+      "loss": 0.0355,
+      "step": 2567
+    },
+    {
+      "epoch": 17.12,
+      "learning_rate": 1.4557595993322204e-05,
+      "loss": 0.0523,
+      "step": 2568
+    },
+    {
+      "epoch": 17.13,
+      "learning_rate": 1.4524207011686144e-05,
+      "loss": 0.0592,
+      "step": 2569
+    },
+    {
+      "epoch": 17.13,
+      "learning_rate": 1.4490818030050086e-05,
+      "loss": 0.048,
+      "step": 2570
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 1.4457429048414022e-05,
+      "loss": 0.0543,
+      "step": 2571
+    },
+    {
+      "epoch": 17.15,
+      "learning_rate": 1.4424040066777964e-05,
+      "loss": 0.0453,
+      "step": 2572
+    },
+    {
+      "epoch": 17.15,
+      "learning_rate": 1.4390651085141904e-05,
+      "loss": 0.0521,
+      "step": 2573
+    },
+    {
+      "epoch": 17.16,
+      "learning_rate": 1.4357262103505844e-05,
+      "loss": 0.061,
+      "step": 2574
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 1.4323873121869782e-05,
+      "loss": 0.0619,
+      "step": 2575
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 1.4290484140233724e-05,
+      "loss": 0.0411,
+      "step": 2576
+    },
+    {
+      "epoch": 17.18,
+      "learning_rate": 1.4257095158597664e-05,
+      "loss": 0.0694,
+      "step": 2577
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 1.4223706176961604e-05,
+      "loss": 0.0618,
+      "step": 2578
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 1.4190317195325542e-05,
+      "loss": 0.0841,
+      "step": 2579
+    },
+    {
+      "epoch": 17.2,
+      "learning_rate": 1.4156928213689482e-05,
+      "loss": 0.107,
+      "step": 2580
+    },
+    {
+      "epoch": 17.21,
+      "learning_rate": 1.4123539232053424e-05,
+      "loss": 0.0791,
+      "step": 2581
+    },
+    {
+      "epoch": 17.21,
+      "learning_rate": 1.4090150250417364e-05,
+      "loss": 0.0667,
+      "step": 2582
+    },
+    {
+      "epoch": 17.22,
+      "learning_rate": 1.4056761268781302e-05,
+      "loss": 0.0579,
+      "step": 2583
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 1.4023372287145242e-05,
+      "loss": 0.0539,
+      "step": 2584
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 1.3989983305509183e-05,
+      "loss": 0.0702,
+      "step": 2585
+    },
+    {
+      "epoch": 17.24,
+      "learning_rate": 1.3956594323873123e-05,
+      "loss": 0.0959,
+      "step": 2586
+    },
+    {
+      "epoch": 17.25,
+      "learning_rate": 1.3923205342237062e-05,
+      "loss": 0.0734,
+      "step": 2587
+    },
+    {
+      "epoch": 17.25,
+      "learning_rate": 1.3889816360601002e-05,
+      "loss": 0.0432,
+      "step": 2588
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 1.3856427378964942e-05,
+      "loss": 0.0991,
+      "step": 2589
+    },
+    {
+      "epoch": 17.27,
+      "learning_rate": 1.3823038397328883e-05,
+      "loss": 0.0859,
+      "step": 2590
+    },
+    {
+      "epoch": 17.27,
+      "learning_rate": 1.3789649415692822e-05,
+      "loss": 0.0539,
+      "step": 2591
+    },
+    {
+      "epoch": 17.28,
+      "learning_rate": 1.3756260434056762e-05,
+      "loss": 0.0877,
+      "step": 2592
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 1.3722871452420701e-05,
+      "loss": 0.0448,
+      "step": 2593
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 1.3689482470784643e-05,
+      "loss": 0.1354,
+      "step": 2594
+    },
+    {
+      "epoch": 17.3,
+      "learning_rate": 1.365609348914858e-05,
+      "loss": 0.0577,
+      "step": 2595
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 1.3622704507512521e-05,
+      "loss": 0.0645,
+      "step": 2596
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 1.3589315525876461e-05,
+      "loss": 0.0742,
+      "step": 2597
+    },
+    {
+      "epoch": 17.32,
+      "learning_rate": 1.3555926544240401e-05,
+      "loss": 0.0438,
+      "step": 2598
+    },
+    {
+      "epoch": 17.33,
+      "learning_rate": 1.352253756260434e-05,
+      "loss": 0.0446,
+      "step": 2599
+    },
+    {
+      "epoch": 17.33,
+      "learning_rate": 1.3489148580968281e-05,
+      "loss": 0.1046,
+      "step": 2600
+    },
+    {
+      "epoch": 17.34,
+      "learning_rate": 1.3455759599332221e-05,
+      "loss": 0.0383,
+      "step": 2601
+    },
+    {
+      "epoch": 17.35,
+      "learning_rate": 1.3422370617696161e-05,
+      "loss": 0.0706,
+      "step": 2602
+    },
+    {
+      "epoch": 17.35,
+      "learning_rate": 1.33889816360601e-05,
+      "loss": 0.0431,
+      "step": 2603
+    },
+    {
+      "epoch": 17.36,
+      "learning_rate": 1.335559265442404e-05,
+      "loss": 0.0668,
+      "step": 2604
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 1.3322203672787981e-05,
+      "loss": 0.0646,
+      "step": 2605
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 1.3288814691151921e-05,
+      "loss": 0.0493,
+      "step": 2606
+    },
+    {
+      "epoch": 17.38,
+      "learning_rate": 1.325542570951586e-05,
+      "loss": 0.0509,
+      "step": 2607
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 1.32220367278798e-05,
+      "loss": 0.0573,
+      "step": 2608
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 1.3188647746243741e-05,
+      "loss": 0.0482,
+      "step": 2609
+    },
+    {
+      "epoch": 17.4,
+      "learning_rate": 1.3155258764607681e-05,
+      "loss": 0.1377,
+      "step": 2610
+    },
+    {
+      "epoch": 17.41,
+      "learning_rate": 1.312186978297162e-05,
+      "loss": 0.0649,
+      "step": 2611
+    },
+    {
+      "epoch": 17.41,
+      "learning_rate": 1.3088480801335559e-05,
+      "loss": 0.0799,
+      "step": 2612
+    },
+    {
+      "epoch": 17.42,
+      "learning_rate": 1.3055091819699499e-05,
+      "loss": 0.0534,
+      "step": 2613
+    },
+    {
+      "epoch": 17.43,
+      "learning_rate": 1.302170283806344e-05,
+      "loss": 0.0648,
+      "step": 2614
+    },
+    {
+      "epoch": 17.43,
+      "learning_rate": 1.298831385642738e-05,
+      "loss": 0.0644,
+      "step": 2615
+    },
+    {
+      "epoch": 17.44,
+      "learning_rate": 1.2954924874791319e-05,
+      "loss": 0.0989,
+      "step": 2616
+    },
+    {
+      "epoch": 17.45,
+      "learning_rate": 1.2921535893155259e-05,
+      "loss": 0.0794,
+      "step": 2617
+    },
+    {
+      "epoch": 17.45,
+      "learning_rate": 1.28881469115192e-05,
+      "loss": 0.101,
+      "step": 2618
+    },
+    {
+      "epoch": 17.46,
+      "learning_rate": 1.285475792988314e-05,
+      "loss": 0.0751,
+      "step": 2619
+    },
+    {
+      "epoch": 17.47,
+      "learning_rate": 1.2821368948247079e-05,
+      "loss": 0.0558,
+      "step": 2620
+    },
+    {
+      "epoch": 17.47,
+      "learning_rate": 1.2787979966611019e-05,
+      "loss": 0.0439,
+      "step": 2621
+    },
+    {
+      "epoch": 17.48,
+      "learning_rate": 1.2754590984974959e-05,
+      "loss": 0.0668,
+      "step": 2622
+    },
+    {
+      "epoch": 17.49,
+      "learning_rate": 1.27212020033389e-05,
+      "loss": 0.0607,
+      "step": 2623
+    },
+    {
+      "epoch": 17.49,
+      "learning_rate": 1.2687813021702839e-05,
+      "loss": 0.0504,
+      "step": 2624
+    },
+    {
+      "epoch": 17.5,
+      "learning_rate": 1.2654424040066779e-05,
+      "loss": 0.0618,
+      "step": 2625
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 1.2621035058430719e-05,
+      "loss": 0.0611,
+      "step": 2626
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 1.258764607679466e-05,
+      "loss": 0.0578,
+      "step": 2627
+    },
+    {
+      "epoch": 17.52,
+      "learning_rate": 1.2554257095158597e-05,
+      "loss": 0.0534,
+      "step": 2628
+    },
+    {
+      "epoch": 17.53,
+      "learning_rate": 1.2520868113522539e-05,
+      "loss": 0.1142,
+      "step": 2629
+    },
+    {
+      "epoch": 17.53,
+      "learning_rate": 1.2487479131886479e-05,
+      "loss": 0.0458,
+      "step": 2630
+    },
+    {
+      "epoch": 17.54,
+      "learning_rate": 1.2454090150250418e-05,
+      "loss": 0.0722,
+      "step": 2631
+    },
+    {
+      "epoch": 17.55,
+      "learning_rate": 1.2420701168614358e-05,
+      "loss": 0.0732,
+      "step": 2632
+    },
+    {
+      "epoch": 17.55,
+      "learning_rate": 1.2387312186978298e-05,
+      "loss": 0.0404,
+      "step": 2633
+    },
+    {
+      "epoch": 17.56,
+      "learning_rate": 1.2353923205342238e-05,
+      "loss": 0.062,
+      "step": 2634
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 1.2320534223706177e-05,
+      "loss": 0.0557,
+      "step": 2635
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 1.2287145242070118e-05,
+      "loss": 0.0797,
+      "step": 2636
+    },
+    {
+      "epoch": 17.58,
+      "learning_rate": 1.2253756260434057e-05,
+      "loss": 0.0432,
+      "step": 2637
+    },
+    {
+      "epoch": 17.59,
+      "learning_rate": 1.2220367278797998e-05,
+      "loss": 0.0608,
+      "step": 2638
+    },
+    {
+      "epoch": 17.59,
+      "learning_rate": 1.2186978297161937e-05,
+      "loss": 0.0717,
+      "step": 2639
+    },
+    {
+      "epoch": 17.6,
+      "learning_rate": 1.2153589315525878e-05,
+      "loss": 0.0573,
+      "step": 2640
+    },
+    {
+      "epoch": 17.61,
+      "learning_rate": 1.2120200333889816e-05,
+      "loss": 0.0452,
+      "step": 2641
+    },
+    {
+      "epoch": 17.61,
+      "learning_rate": 1.2086811352253758e-05,
+      "loss": 0.0667,
+      "step": 2642
+    },
+    {
+      "epoch": 17.62,
+      "learning_rate": 1.2053422370617696e-05,
+      "loss": 0.0431,
+      "step": 2643
+    },
+    {
+      "epoch": 17.63,
+      "learning_rate": 1.2020033388981636e-05,
+      "loss": 0.0763,
+      "step": 2644
+    },
+    {
+      "epoch": 17.63,
+      "learning_rate": 1.1986644407345576e-05,
+      "loss": 0.0412,
+      "step": 2645
+    },
+    {
+      "epoch": 17.64,
+      "learning_rate": 1.1953255425709516e-05,
+      "loss": 0.053,
+      "step": 2646
+    },
+    {
+      "epoch": 17.65,
+      "learning_rate": 1.1919866444073456e-05,
+      "loss": 0.0627,
+      "step": 2647
+    },
+    {
+      "epoch": 17.65,
+      "learning_rate": 1.1886477462437396e-05,
+      "loss": 0.0632,
+      "step": 2648
+    },
+    {
+      "epoch": 17.66,
+      "learning_rate": 1.1853088480801336e-05,
+      "loss": 0.0652,
+      "step": 2649
+    },
+    {
+      "epoch": 17.67,
+      "learning_rate": 1.1819699499165276e-05,
+      "loss": 0.0494,
+      "step": 2650
+    },
+    {
+      "epoch": 17.67,
+      "learning_rate": 1.1786310517529216e-05,
+      "loss": 0.0463,
+      "step": 2651
+    },
+    {
+      "epoch": 17.68,
+      "learning_rate": 1.1752921535893156e-05,
+      "loss": 0.0535,
+      "step": 2652
+    },
+    {
+      "epoch": 17.69,
+      "learning_rate": 1.1719532554257096e-05,
+      "loss": 0.0606,
+      "step": 2653
+    },
+    {
+      "epoch": 17.69,
+      "learning_rate": 1.1686143572621036e-05,
+      "loss": 0.0548,
+      "step": 2654
+    },
+    {
+      "epoch": 17.7,
+      "learning_rate": 1.1652754590984976e-05,
+      "loss": 0.0706,
+      "step": 2655
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 1.1619365609348916e-05,
+      "loss": 0.0562,
+      "step": 2656
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 1.1585976627712856e-05,
+      "loss": 0.0488,
+      "step": 2657
+    },
+    {
+      "epoch": 17.72,
+      "learning_rate": 1.1552587646076796e-05,
+      "loss": 0.0878,
+      "step": 2658
+    },
+    {
+      "epoch": 17.73,
+      "learning_rate": 1.1519198664440734e-05,
+      "loss": 0.066,
+      "step": 2659
+    },
+    {
+      "epoch": 17.73,
+      "learning_rate": 1.1485809682804676e-05,
+      "loss": 0.0667,
+      "step": 2660
+    },
+    {
+      "epoch": 17.74,
+      "learning_rate": 1.1452420701168614e-05,
+      "loss": 0.141,
+      "step": 2661
+    },
+    {
+      "epoch": 17.75,
+      "learning_rate": 1.1419031719532556e-05,
+      "loss": 0.0387,
+      "step": 2662
+    },
+    {
+      "epoch": 17.75,
+      "learning_rate": 1.1385642737896494e-05,
+      "loss": 0.0602,
+      "step": 2663
+    },
+    {
+      "epoch": 17.76,
+      "learning_rate": 1.1352253756260436e-05,
+      "loss": 0.0432,
+      "step": 2664
+    },
+    {
+      "epoch": 17.77,
+      "learning_rate": 1.1318864774624374e-05,
+      "loss": 0.1293,
+      "step": 2665
+    },
+    {
+      "epoch": 17.77,
+      "learning_rate": 1.1285475792988316e-05,
+      "loss": 0.0549,
+      "step": 2666
+    },
+    {
+      "epoch": 17.78,
+      "learning_rate": 1.1252086811352254e-05,
+      "loss": 0.0571,
+      "step": 2667
+    },
+    {
+      "epoch": 17.79,
+      "learning_rate": 1.1218697829716194e-05,
+      "loss": 0.0727,
+      "step": 2668
+    },
+    {
+      "epoch": 17.79,
+      "learning_rate": 1.1185308848080134e-05,
+      "loss": 0.0789,
+      "step": 2669
+    },
+    {
+      "epoch": 17.8,
+      "learning_rate": 1.1151919866444074e-05,
+      "loss": 0.0522,
+      "step": 2670
+    },
+    {
+      "epoch": 17.81,
+      "learning_rate": 1.1118530884808014e-05,
+      "loss": 0.0678,
+      "step": 2671
+    },
+    {
+      "epoch": 17.81,
+      "learning_rate": 1.1085141903171954e-05,
+      "loss": 0.0816,
+      "step": 2672
+    },
+    {
+      "epoch": 17.82,
+      "learning_rate": 1.1051752921535894e-05,
+      "loss": 0.0789,
+      "step": 2673
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 1.1018363939899834e-05,
+      "loss": 0.0452,
+      "step": 2674
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 1.0984974958263774e-05,
+      "loss": 0.0364,
+      "step": 2675
+    },
+    {
+      "epoch": 17.84,
+      "learning_rate": 1.0951585976627714e-05,
+      "loss": 0.0581,
+      "step": 2676
+    },
+    {
+      "epoch": 17.85,
+      "learning_rate": 1.0918196994991653e-05,
+      "loss": 0.0481,
+      "step": 2677
+    },
+    {
+      "epoch": 17.85,
+      "learning_rate": 1.0884808013355593e-05,
+      "loss": 0.0818,
+      "step": 2678
+    },
+    {
+      "epoch": 17.86,
+      "learning_rate": 1.0851419031719533e-05,
+      "loss": 0.05,
+      "step": 2679
+    },
+    {
+      "epoch": 17.87,
+      "learning_rate": 1.0818030050083473e-05,
+      "loss": 0.0422,
+      "step": 2680
+    },
+    {
+      "epoch": 17.87,
+      "learning_rate": 1.0784641068447413e-05,
+      "loss": 0.0432,
+      "step": 2681
+    },
+    {
+      "epoch": 17.88,
+      "learning_rate": 1.0751252086811353e-05,
+      "loss": 0.0562,
+      "step": 2682
+    },
+    {
+      "epoch": 17.89,
+      "learning_rate": 1.0717863105175292e-05,
+      "loss": 0.0615,
+      "step": 2683
+    },
+    {
+      "epoch": 17.89,
+      "learning_rate": 1.0684474123539233e-05,
+      "loss": 0.0534,
+      "step": 2684
+    },
+    {
+      "epoch": 17.9,
+      "learning_rate": 1.0651085141903172e-05,
+      "loss": 0.0491,
+      "step": 2685
+    },
+    {
+      "epoch": 17.91,
+      "learning_rate": 1.0617696160267113e-05,
+      "loss": 0.093,
+      "step": 2686
+    },
+    {
+      "epoch": 17.91,
+      "learning_rate": 1.0584307178631051e-05,
+      "loss": 0.0846,
+      "step": 2687
+    },
+    {
+      "epoch": 17.92,
+      "learning_rate": 1.0550918196994993e-05,
+      "loss": 0.0676,
+      "step": 2688
+    },
+    {
+      "epoch": 17.93,
+      "learning_rate": 1.0517529215358931e-05,
+      "loss": 0.1368,
+      "step": 2689
+    },
+    {
+      "epoch": 17.93,
+      "learning_rate": 1.0484140233722873e-05,
+      "loss": 0.1026,
+      "step": 2690
+    },
+    {
+      "epoch": 17.94,
+      "learning_rate": 1.0450751252086811e-05,
+      "loss": 0.0583,
+      "step": 2691
+    },
+    {
+      "epoch": 17.95,
+      "learning_rate": 1.0417362270450751e-05,
+      "loss": 0.0698,
+      "step": 2692
+    },
+    {
+      "epoch": 17.95,
+      "learning_rate": 1.0383973288814691e-05,
+      "loss": 0.0539,
+      "step": 2693
+    },
+    {
+      "epoch": 17.96,
+      "learning_rate": 1.0350584307178631e-05,
+      "loss": 0.0548,
+      "step": 2694
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 1.0317195325542571e-05,
+      "loss": 0.0594,
+      "step": 2695
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 1.0283806343906511e-05,
+      "loss": 0.0634,
+      "step": 2696
+    },
+    {
+      "epoch": 17.98,
+      "learning_rate": 1.0250417362270451e-05,
+      "loss": 0.0599,
+      "step": 2697
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 1.0217028380634391e-05,
+      "loss": 0.0637,
+      "step": 2698
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 1.0183639398998331e-05,
+      "loss": 0.0875,
+      "step": 2699
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 1.0150250417362271e-05,
+      "loss": 0.0777,
+      "step": 2700
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 3.249044179916382,
+      "eval_runtime": 20.2197,
+      "eval_samples_per_second": 9.891,
+      "eval_steps_per_second": 1.236,
+      "step": 2700
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 1.0116861435726211e-05,
+      "loss": 0.059,
+      "step": 2701
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 1.0083472454090151e-05,
+      "loss": 0.0596,
+      "step": 2702
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 1.0050083472454091e-05,
+      "loss": 0.0544,
+      "step": 2703
+    },
+    {
+      "epoch": 18.03,
+      "learning_rate": 1.001669449081803e-05,
+      "loss": 0.0959,
+      "step": 2704
+    },
+    {
+      "epoch": 18.03,
+      "learning_rate": 9.983305509181969e-06,
+      "loss": 0.0943,
+      "step": 2705
+    },
+    {
+      "epoch": 18.04,
+      "learning_rate": 9.94991652754591e-06,
+      "loss": 0.0561,
+      "step": 2706
+    },
+    {
+      "epoch": 18.05,
+      "learning_rate": 9.916527545909849e-06,
+      "loss": 0.0798,
+      "step": 2707
+    },
+    {
+      "epoch": 18.05,
+      "learning_rate": 9.88313856427379e-06,
+      "loss": 0.0627,
+      "step": 2708
+    },
+    {
+      "epoch": 18.06,
+      "learning_rate": 9.849749582637729e-06,
+      "loss": 0.0462,
+      "step": 2709
+    },
+    {
+      "epoch": 18.07,
+      "learning_rate": 9.81636060100167e-06,
+      "loss": 0.0498,
+      "step": 2710
+    },
+    {
+      "epoch": 18.07,
+      "learning_rate": 9.782971619365609e-06,
+      "loss": 0.0613,
+      "step": 2711
+    },
+    {
+      "epoch": 18.08,
+      "learning_rate": 9.74958263772955e-06,
+      "loss": 0.0735,
+      "step": 2712
+    },
+    {
+      "epoch": 18.09,
+      "learning_rate": 9.716193656093489e-06,
+      "loss": 0.0584,
+      "step": 2713
+    },
+    {
+      "epoch": 18.09,
+      "learning_rate": 9.68280467445743e-06,
+      "loss": 0.1254,
+      "step": 2714
+    },
+    {
+      "epoch": 18.1,
+      "learning_rate": 9.649415692821369e-06,
+      "loss": 0.0446,
+      "step": 2715
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 9.616026711185309e-06,
+      "loss": 0.0671,
+      "step": 2716
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 9.582637729549249e-06,
+      "loss": 0.0512,
+      "step": 2717
+    },
+    {
+      "epoch": 18.12,
+      "learning_rate": 9.549248747913189e-06,
+      "loss": 0.0432,
+      "step": 2718
+    },
+    {
+      "epoch": 18.13,
+      "learning_rate": 9.515859766277129e-06,
+      "loss": 0.0567,
+      "step": 2719
+    },
+    {
+      "epoch": 18.13,
+      "learning_rate": 9.482470784641069e-06,
+      "loss": 0.0826,
+      "step": 2720
+    },
+    {
+      "epoch": 18.14,
+      "learning_rate": 9.449081803005009e-06,
+      "loss": 0.0823,
+      "step": 2721
+    },
+    {
+      "epoch": 18.15,
+      "learning_rate": 9.415692821368949e-06,
+      "loss": 0.0578,
+      "step": 2722
+    },
+    {
+      "epoch": 18.15,
+      "learning_rate": 9.382303839732888e-06,
+      "loss": 0.0788,
+      "step": 2723
+    },
+    {
+      "epoch": 18.16,
+      "learning_rate": 9.348914858096828e-06,
+      "loss": 0.0781,
+      "step": 2724
+    },
+    {
+      "epoch": 18.17,
+      "learning_rate": 9.315525876460768e-06,
+      "loss": 0.0677,
+      "step": 2725
+    },
+    {
+      "epoch": 18.17,
+      "learning_rate": 9.282136894824708e-06,
+      "loss": 0.0464,
+      "step": 2726
+    },
+    {
+      "epoch": 18.18,
+      "learning_rate": 9.248747913188648e-06,
+      "loss": 0.0705,
+      "step": 2727
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 9.215358931552588e-06,
+      "loss": 0.0723,
+      "step": 2728
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 9.181969949916527e-06,
+      "loss": 0.0985,
+      "step": 2729
+    },
+    {
+      "epoch": 18.2,
+      "learning_rate": 9.148580968280468e-06,
+      "loss": 0.0592,
+      "step": 2730
+    },
+    {
+      "epoch": 18.21,
+      "learning_rate": 9.115191986644407e-06,
+      "loss": 0.0379,
+      "step": 2731
+    },
+    {
+      "epoch": 18.21,
+      "learning_rate": 9.081803005008348e-06,
+      "loss": 0.0458,
+      "step": 2732
+    },
+    {
+      "epoch": 18.22,
+      "learning_rate": 9.048414023372286e-06,
+      "loss": 0.0449,
+      "step": 2733
+    },
+    {
+      "epoch": 18.23,
+      "learning_rate": 9.015025041736228e-06,
+      "loss": 0.0474,
+      "step": 2734
+    },
+    {
+      "epoch": 18.23,
+      "learning_rate": 8.981636060100166e-06,
+      "loss": 0.0533,
+      "step": 2735
+    },
+    {
+      "epoch": 18.24,
+      "learning_rate": 8.948247078464108e-06,
+      "loss": 0.0402,
+      "step": 2736
+    },
+    {
+      "epoch": 18.25,
+      "learning_rate": 8.914858096828046e-06,
+      "loss": 0.0559,
+      "step": 2737
+    },
+    {
+      "epoch": 18.25,
+      "learning_rate": 8.881469115191986e-06,
+      "loss": 0.0616,
+      "step": 2738
+    },
+    {
+      "epoch": 18.26,
+      "learning_rate": 8.848080133555926e-06,
+      "loss": 0.0543,
+      "step": 2739
+    },
+    {
+      "epoch": 18.27,
+      "learning_rate": 8.814691151919866e-06,
+      "loss": 0.0655,
+      "step": 2740
+    },
+    {
+      "epoch": 18.27,
+      "learning_rate": 8.781302170283806e-06,
+      "loss": 0.047,
+      "step": 2741
+    },
+    {
+      "epoch": 18.28,
+      "learning_rate": 8.747913188647746e-06,
+      "loss": 0.0774,
+      "step": 2742
+    },
+    {
+      "epoch": 18.29,
+      "learning_rate": 8.714524207011686e-06,
+      "loss": 0.0788,
+      "step": 2743
+    },
+    {
+      "epoch": 18.29,
+      "learning_rate": 8.681135225375626e-06,
+      "loss": 0.0833,
+      "step": 2744
+    },
+    {
+      "epoch": 18.3,
+      "learning_rate": 8.647746243739568e-06,
+      "loss": 0.0516,
+      "step": 2745
+    },
+    {
+      "epoch": 18.31,
+      "learning_rate": 8.614357262103506e-06,
+      "loss": 0.0586,
+      "step": 2746
+    },
+    {
+      "epoch": 18.31,
+      "learning_rate": 8.580968280467448e-06,
+      "loss": 0.0429,
+      "step": 2747
+    },
+    {
+      "epoch": 18.32,
+      "learning_rate": 8.547579298831386e-06,
+      "loss": 0.0591,
+      "step": 2748
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 8.514190317195326e-06,
+      "loss": 0.0511,
+      "step": 2749
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 8.480801335559266e-06,
+      "loss": 0.0459,
+      "step": 2750
+    },
+    {
+      "epoch": 18.34,
+      "learning_rate": 8.447412353923206e-06,
+      "loss": 0.0796,
+      "step": 2751
+    },
+    {
+      "epoch": 18.35,
+      "learning_rate": 8.414023372287146e-06,
+      "loss": 0.0494,
+      "step": 2752
+    },
+    {
+      "epoch": 18.35,
+      "learning_rate": 8.380634390651086e-06,
+      "loss": 0.0337,
+      "step": 2753
+    },
+    {
+      "epoch": 18.36,
+      "learning_rate": 8.347245409015026e-06,
+      "loss": 0.0574,
+      "step": 2754
+    },
+    {
+      "epoch": 18.37,
+      "learning_rate": 8.313856427378966e-06,
+      "loss": 0.0877,
+      "step": 2755
+    },
+    {
+      "epoch": 18.37,
+      "learning_rate": 8.280467445742906e-06,
+      "loss": 0.0426,
+      "step": 2756
+    },
+    {
+      "epoch": 18.38,
+      "learning_rate": 8.247078464106846e-06,
+      "loss": 0.0514,
+      "step": 2757
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 8.213689482470786e-06,
+      "loss": 0.0521,
+      "step": 2758
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 8.180300500834726e-06,
+      "loss": 0.0615,
+      "step": 2759
+    },
+    {
+      "epoch": 18.4,
+      "learning_rate": 8.146911519198665e-06,
+      "loss": 0.0849,
+      "step": 2760
+    },
+    {
+      "epoch": 18.41,
+      "learning_rate": 8.113522537562605e-06,
+      "loss": 0.0441,
+      "step": 2761
+    },
+    {
+      "epoch": 18.41,
+      "learning_rate": 8.080133555926544e-06,
+      "loss": 0.0438,
+      "step": 2762
+    },
+    {
+      "epoch": 18.42,
+      "learning_rate": 8.046744574290485e-06,
+      "loss": 0.0614,
+      "step": 2763
+    },
+    {
+      "epoch": 18.43,
+      "learning_rate": 8.013355592654424e-06,
+      "loss": 0.1414,
+      "step": 2764
+    },
+    {
+      "epoch": 18.43,
+      "learning_rate": 7.979966611018365e-06,
+      "loss": 0.0498,
+      "step": 2765
+    },
+    {
+      "epoch": 18.44,
+      "learning_rate": 7.946577629382304e-06,
+      "loss": 0.0519,
+      "step": 2766
+    },
+    {
+      "epoch": 18.45,
+      "learning_rate": 7.913188647746245e-06,
+      "loss": 0.0552,
+      "step": 2767
+    },
+    {
+      "epoch": 18.45,
+      "learning_rate": 7.879799666110184e-06,
+      "loss": 0.0866,
+      "step": 2768
+    },
+    {
+      "epoch": 18.46,
+      "learning_rate": 7.846410684474125e-06,
+      "loss": 0.0624,
+      "step": 2769
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 7.813021702838063e-06,
+      "loss": 0.0408,
+      "step": 2770
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 7.779632721202005e-06,
+      "loss": 0.0813,
+      "step": 2771
+    },
+    {
+      "epoch": 18.48,
+      "learning_rate": 7.746243739565943e-06,
+      "loss": 0.0604,
+      "step": 2772
+    },
+    {
+      "epoch": 18.49,
+      "learning_rate": 7.712854757929883e-06,
+      "loss": 0.0763,
+      "step": 2773
+    },
+    {
+      "epoch": 18.49,
+      "learning_rate": 7.679465776293823e-06,
+      "loss": 0.0597,
+      "step": 2774
+    },
+    {
+      "epoch": 18.5,
+      "learning_rate": 7.646076794657763e-06,
+      "loss": 0.044,
+      "step": 2775
+    },
+    {
+      "epoch": 18.51,
+      "learning_rate": 7.612687813021703e-06,
+      "loss": 0.0349,
+      "step": 2776
+    },
+    {
+      "epoch": 18.51,
+      "learning_rate": 7.579298831385643e-06,
+      "loss": 0.0566,
+      "step": 2777
+    },
+    {
+      "epoch": 18.52,
+      "learning_rate": 7.545909849749583e-06,
+      "loss": 0.0608,
+      "step": 2778
+    },
+    {
+      "epoch": 18.53,
+      "learning_rate": 7.512520868113523e-06,
+      "loss": 0.0981,
+      "step": 2779
+    },
+    {
+      "epoch": 18.53,
+      "learning_rate": 7.479131886477462e-06,
+      "loss": 0.0408,
+      "step": 2780
+    },
+    {
+      "epoch": 18.54,
+      "learning_rate": 7.445742904841403e-06,
+      "loss": 0.0451,
+      "step": 2781
+    },
+    {
+      "epoch": 18.55,
+      "learning_rate": 7.412353923205342e-06,
+      "loss": 0.0643,
+      "step": 2782
+    },
+    {
+      "epoch": 18.55,
+      "learning_rate": 7.378964941569283e-06,
+      "loss": 0.0453,
+      "step": 2783
+    },
+    {
+      "epoch": 18.56,
+      "learning_rate": 7.345575959933222e-06,
+      "loss": 0.0581,
+      "step": 2784
+    },
+    {
+      "epoch": 18.57,
+      "learning_rate": 7.312186978297163e-06,
+      "loss": 0.0731,
+      "step": 2785
+    },
+    {
+      "epoch": 18.57,
+      "learning_rate": 7.278797996661102e-06,
+      "loss": 0.0619,
+      "step": 2786
+    },
+    {
+      "epoch": 18.58,
+      "learning_rate": 7.245409015025043e-06,
+      "loss": 0.0489,
+      "step": 2787
+    },
+    {
+      "epoch": 18.59,
+      "learning_rate": 7.212020033388982e-06,
+      "loss": 0.061,
+      "step": 2788
+    },
+    {
+      "epoch": 18.59,
+      "learning_rate": 7.178631051752922e-06,
+      "loss": 0.0713,
+      "step": 2789
+    },
+    {
+      "epoch": 18.6,
+      "learning_rate": 7.145242070116862e-06,
+      "loss": 0.0561,
+      "step": 2790
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 7.111853088480802e-06,
+      "loss": 0.0584,
+      "step": 2791
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 7.078464106844741e-06,
+      "loss": 0.0989,
+      "step": 2792
+    },
+    {
+      "epoch": 18.62,
+      "learning_rate": 7.045075125208682e-06,
+      "loss": 0.1002,
+      "step": 2793
+    },
+    {
+      "epoch": 18.63,
+      "learning_rate": 7.011686143572621e-06,
+      "loss": 0.0675,
+      "step": 2794
+    },
+    {
+      "epoch": 18.63,
+      "learning_rate": 6.978297161936562e-06,
+      "loss": 0.099,
+      "step": 2795
+    },
+    {
+      "epoch": 18.64,
+      "learning_rate": 6.944908180300501e-06,
+      "loss": 0.0477,
+      "step": 2796
+    },
+    {
+      "epoch": 18.65,
+      "learning_rate": 6.911519198664442e-06,
+      "loss": 0.0555,
+      "step": 2797
+    },
+    {
+      "epoch": 18.65,
+      "learning_rate": 6.878130217028381e-06,
+      "loss": 0.0771,
+      "step": 2798
+    },
+    {
+      "epoch": 18.66,
+      "learning_rate": 6.844741235392322e-06,
+      "loss": 0.0572,
+      "step": 2799
+    },
+    {
+      "epoch": 18.67,
+      "learning_rate": 6.811352253756261e-06,
+      "loss": 0.0898,
+      "step": 2800
+    },
+    {
+      "epoch": 18.67,
+      "learning_rate": 6.777963272120201e-06,
+      "loss": 0.0779,
+      "step": 2801
+    },
+    {
+      "epoch": 18.68,
+      "learning_rate": 6.744574290484141e-06,
+      "loss": 0.072,
+      "step": 2802
+    },
+    {
+      "epoch": 18.69,
+      "learning_rate": 6.711185308848081e-06,
+      "loss": 0.0933,
+      "step": 2803
+    },
+    {
+      "epoch": 18.69,
+      "learning_rate": 6.67779632721202e-06,
+      "loss": 0.0499,
+      "step": 2804
+    },
+    {
+      "epoch": 18.7,
+      "learning_rate": 6.6444073455759605e-06,
+      "loss": 0.054,
+      "step": 2805
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 6.6110183639399e-06,
+      "loss": 0.0549,
+      "step": 2806
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 6.5776293823038405e-06,
+      "loss": 0.0375,
+      "step": 2807
+    },
+    {
+      "epoch": 18.72,
+      "learning_rate": 6.5442404006677796e-06,
+      "loss": 0.0494,
+      "step": 2808
+    },
+    {
+      "epoch": 18.73,
+      "learning_rate": 6.51085141903172e-06,
+      "loss": 0.0499,
+      "step": 2809
+    },
+    {
+      "epoch": 18.73,
+      "learning_rate": 6.4774624373956595e-06,
+      "loss": 0.082,
+      "step": 2810
+    },
+    {
+      "epoch": 18.74,
+      "learning_rate": 6.4440734557596e-06,
+      "loss": 0.0673,
+      "step": 2811
+    },
+    {
+      "epoch": 18.75,
+      "learning_rate": 6.4106844741235394e-06,
+      "loss": 0.0405,
+      "step": 2812
+    },
+    {
+      "epoch": 18.75,
+      "learning_rate": 6.377295492487479e-06,
+      "loss": 0.0929,
+      "step": 2813
+    },
+    {
+      "epoch": 18.76,
+      "learning_rate": 6.343906510851419e-06,
+      "loss": 0.0694,
+      "step": 2814
+    },
+    {
+      "epoch": 18.77,
+      "learning_rate": 6.310517529215359e-06,
+      "loss": 0.0517,
+      "step": 2815
+    },
+    {
+      "epoch": 18.77,
+      "learning_rate": 6.2771285475792984e-06,
+      "loss": 0.0642,
+      "step": 2816
+    },
+    {
+      "epoch": 18.78,
+      "learning_rate": 6.243739565943239e-06,
+      "loss": 0.074,
+      "step": 2817
+    },
+    {
+      "epoch": 18.79,
+      "learning_rate": 6.210350584307179e-06,
+      "loss": 0.0463,
+      "step": 2818
+    },
+    {
+      "epoch": 18.79,
+      "learning_rate": 6.176961602671119e-06,
+      "loss": 0.058,
+      "step": 2819
+    },
+    {
+      "epoch": 18.8,
+      "learning_rate": 6.143572621035059e-06,
+      "loss": 0.0437,
+      "step": 2820
+    },
+    {
+      "epoch": 18.81,
+      "learning_rate": 6.110183639398999e-06,
+      "loss": 0.0775,
+      "step": 2821
+    },
+    {
+      "epoch": 18.81,
+      "learning_rate": 6.076794657762939e-06,
+      "loss": 0.085,
+      "step": 2822
+    },
+    {
+      "epoch": 18.82,
+      "learning_rate": 6.043405676126879e-06,
+      "loss": 0.0842,
+      "step": 2823
+    },
+    {
+      "epoch": 18.83,
+      "learning_rate": 6.010016694490818e-06,
+      "loss": 0.0422,
+      "step": 2824
+    },
+    {
+      "epoch": 18.83,
+      "learning_rate": 5.976627712854758e-06,
+      "loss": 0.0578,
+      "step": 2825
+    },
+    {
+      "epoch": 18.84,
+      "learning_rate": 5.943238731218698e-06,
+      "loss": 0.0789,
+      "step": 2826
+    },
+    {
+      "epoch": 18.85,
+      "learning_rate": 5.909849749582638e-06,
+      "loss": 0.0693,
+      "step": 2827
+    },
+    {
+      "epoch": 18.85,
+      "learning_rate": 5.876460767946578e-06,
+      "loss": 0.0716,
+      "step": 2828
+    },
+    {
+      "epoch": 18.86,
+      "learning_rate": 5.843071786310518e-06,
+      "loss": 0.0603,
+      "step": 2829
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 5.809682804674458e-06,
+      "loss": 0.0485,
+      "step": 2830
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 5.776293823038398e-06,
+      "loss": 0.0673,
+      "step": 2831
+    },
+    {
+      "epoch": 18.88,
+      "learning_rate": 5.742904841402338e-06,
+      "loss": 0.0419,
+      "step": 2832
+    },
+    {
+      "epoch": 18.89,
+      "learning_rate": 5.709515859766278e-06,
+      "loss": 0.0569,
+      "step": 2833
+    },
+    {
+      "epoch": 18.89,
+      "learning_rate": 5.676126878130218e-06,
+      "loss": 0.0546,
+      "step": 2834
+    },
+    {
+      "epoch": 18.9,
+      "learning_rate": 5.642737896494158e-06,
+      "loss": 0.0783,
+      "step": 2835
+    },
+    {
+      "epoch": 18.91,
+      "learning_rate": 5.609348914858097e-06,
+      "loss": 0.0603,
+      "step": 2836
+    },
+    {
+      "epoch": 18.91,
+      "learning_rate": 5.575959933222037e-06,
+      "loss": 0.0488,
+      "step": 2837
+    },
+    {
+      "epoch": 18.92,
+      "learning_rate": 5.542570951585977e-06,
+      "loss": 0.0407,
+      "step": 2838
+    },
+    {
+      "epoch": 18.93,
+      "learning_rate": 5.509181969949917e-06,
+      "loss": 0.0592,
+      "step": 2839
+    },
+    {
+      "epoch": 18.93,
+      "learning_rate": 5.475792988313857e-06,
+      "loss": 0.0663,
+      "step": 2840
+    },
+    {
+      "epoch": 18.94,
+      "learning_rate": 5.442404006677797e-06,
+      "loss": 0.0831,
+      "step": 2841
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 5.409015025041737e-06,
+      "loss": 0.0519,
+      "step": 2842
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 5.375626043405677e-06,
+      "loss": 0.0489,
+      "step": 2843
+    },
+    {
+      "epoch": 18.96,
+      "learning_rate": 5.342237061769617e-06,
+      "loss": 0.051,
+      "step": 2844
+    },
+    {
+      "epoch": 18.97,
+      "learning_rate": 5.308848080133557e-06,
+      "loss": 0.068,
+      "step": 2845
+    },
+    {
+      "epoch": 18.97,
+      "learning_rate": 5.2754590984974965e-06,
+      "loss": 0.0427,
+      "step": 2846
+    },
+    {
+      "epoch": 18.98,
+      "learning_rate": 5.2420701168614365e-06,
+      "loss": 0.0762,
+      "step": 2847
+    },
+    {
+      "epoch": 18.99,
+      "learning_rate": 5.208681135225376e-06,
+      "loss": 0.0657,
+      "step": 2848
+    },
+    {
+      "epoch": 18.99,
+      "learning_rate": 5.175292153589316e-06,
+      "loss": 0.0854,
+      "step": 2849
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 5.1419031719532556e-06,
+      "loss": 0.0647,
+      "step": 2850
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 3.3112998008728027,
+      "eval_runtime": 20.2079,
+      "eval_samples_per_second": 9.897,
+      "eval_steps_per_second": 1.237,
+      "step": 2850
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 5.1085141903171955e-06,
+      "loss": 0.0544,
+      "step": 2851
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 5.0751252086811355e-06,
+      "loss": 0.0543,
+      "step": 2852
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 5.0417362270450755e-06,
+      "loss": 0.0677,
+      "step": 2853
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 5.008347245409015e-06,
+      "loss": 0.0726,
+      "step": 2854
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 4.974958263772955e-06,
+      "loss": 0.0488,
+      "step": 2855
+    },
+    {
+      "epoch": 19.04,
+      "learning_rate": 4.941569282136895e-06,
+      "loss": 0.0423,
+      "step": 2856
+    },
+    {
+      "epoch": 19.05,
+      "learning_rate": 4.908180300500835e-06,
+      "loss": 0.0885,
+      "step": 2857
+    },
+    {
+      "epoch": 19.05,
+      "learning_rate": 4.874791318864775e-06,
+      "loss": 0.0562,
+      "step": 2858
+    },
+    {
+      "epoch": 19.06,
+      "learning_rate": 4.841402337228715e-06,
+      "loss": 0.0749,
+      "step": 2859
+    },
+    {
+      "epoch": 19.07,
+      "learning_rate": 4.808013355592654e-06,
+      "loss": 0.082,
+      "step": 2860
+    },
+    {
+      "epoch": 19.07,
+      "learning_rate": 4.774624373956594e-06,
+      "loss": 0.0982,
+      "step": 2861
+    },
+    {
+      "epoch": 19.08,
+      "learning_rate": 4.741235392320534e-06,
+      "loss": 0.0631,
+      "step": 2862
+    },
+    {
+      "epoch": 19.09,
+      "learning_rate": 4.707846410684474e-06,
+      "loss": 0.0468,
+      "step": 2863
+    },
+    {
+      "epoch": 19.09,
+      "learning_rate": 4.674457429048414e-06,
+      "loss": 0.0602,
+      "step": 2864
+    },
+    {
+      "epoch": 19.1,
+      "learning_rate": 4.641068447412354e-06,
+      "loss": 0.0567,
+      "step": 2865
+    },
+    {
+      "epoch": 19.11,
+      "learning_rate": 4.607679465776294e-06,
+      "loss": 0.0499,
+      "step": 2866
+    },
+    {
+      "epoch": 19.11,
+      "learning_rate": 4.574290484140234e-06,
+      "loss": 0.0936,
+      "step": 2867
+    },
+    {
+      "epoch": 19.12,
+      "learning_rate": 4.540901502504174e-06,
+      "loss": 0.0628,
+      "step": 2868
+    },
+    {
+      "epoch": 19.13,
+      "learning_rate": 4.507512520868114e-06,
+      "loss": 0.0603,
+      "step": 2869
+    },
+    {
+      "epoch": 19.13,
+      "learning_rate": 4.474123539232054e-06,
+      "loss": 0.0668,
+      "step": 2870
+    },
+    {
+      "epoch": 19.14,
+      "learning_rate": 4.440734557595993e-06,
+      "loss": 0.0349,
+      "step": 2871
+    },
+    {
+      "epoch": 19.15,
+      "learning_rate": 4.407345575959933e-06,
+      "loss": 0.0695,
+      "step": 2872
+    },
+    {
+      "epoch": 19.15,
+      "learning_rate": 4.373956594323873e-06,
+      "loss": 0.0538,
+      "step": 2873
+    },
+    {
+      "epoch": 19.16,
+      "learning_rate": 4.340567612687813e-06,
+      "loss": 0.0678,
+      "step": 2874
+    },
+    {
+      "epoch": 19.17,
+      "learning_rate": 4.307178631051753e-06,
+      "loss": 0.0543,
+      "step": 2875
+    },
+    {
+      "epoch": 19.17,
+      "learning_rate": 4.273789649415693e-06,
+      "loss": 0.0864,
+      "step": 2876
+    },
+    {
+      "epoch": 19.18,
+      "learning_rate": 4.240400667779633e-06,
+      "loss": 0.0475,
+      "step": 2877
+    },
+    {
+      "epoch": 19.19,
+      "learning_rate": 4.207011686143573e-06,
+      "loss": 0.0638,
+      "step": 2878
+    },
+    {
+      "epoch": 19.19,
+      "learning_rate": 4.173622704507513e-06,
+      "loss": 0.0712,
+      "step": 2879
+    },
+    {
+      "epoch": 19.2,
+      "learning_rate": 4.140233722871453e-06,
+      "loss": 0.0482,
+      "step": 2880
+    },
+    {
+      "epoch": 19.21,
+      "learning_rate": 4.106844741235393e-06,
+      "loss": 0.1119,
+      "step": 2881
+    },
+    {
+      "epoch": 19.21,
+      "learning_rate": 4.073455759599333e-06,
+      "loss": 0.0454,
+      "step": 2882
+    },
+    {
+      "epoch": 19.22,
+      "learning_rate": 4.040066777963272e-06,
+      "loss": 0.0505,
+      "step": 2883
+    },
+    {
+      "epoch": 19.23,
+      "learning_rate": 4.006677796327212e-06,
+      "loss": 0.053,
+      "step": 2884
+    },
+    {
+      "epoch": 19.23,
+      "learning_rate": 3.973288814691152e-06,
+      "loss": 0.0476,
+      "step": 2885
+    },
+    {
+      "epoch": 19.24,
+      "learning_rate": 3.939899833055092e-06,
+      "loss": 0.0477,
+      "step": 2886
+    },
+    {
+      "epoch": 19.25,
+      "learning_rate": 3.906510851419032e-06,
+      "loss": 0.0546,
+      "step": 2887
+    },
+    {
+      "epoch": 19.25,
+      "learning_rate": 3.873121869782972e-06,
+      "loss": 0.0612,
+      "step": 2888
+    },
+    {
+      "epoch": 19.26,
+      "learning_rate": 3.839732888146912e-06,
+      "loss": 0.0473,
+      "step": 2889
+    },
+    {
+      "epoch": 19.27,
+      "learning_rate": 3.8063439065108516e-06,
+      "loss": 0.0548,
+      "step": 2890
+    },
+    {
+      "epoch": 19.27,
+      "learning_rate": 3.7729549248747916e-06,
+      "loss": 0.0784,
+      "step": 2891
+    },
+    {
+      "epoch": 19.28,
+      "learning_rate": 3.739565943238731e-06,
+      "loss": 0.1181,
+      "step": 2892
+    },
+    {
+      "epoch": 19.29,
+      "learning_rate": 3.706176961602671e-06,
+      "loss": 0.0676,
+      "step": 2893
+    },
+    {
+      "epoch": 19.29,
+      "learning_rate": 3.672787979966611e-06,
+      "loss": 0.094,
+      "step": 2894
+    },
+    {
+      "epoch": 19.3,
+      "learning_rate": 3.639398998330551e-06,
+      "loss": 0.0914,
+      "step": 2895
+    },
+    {
+      "epoch": 19.31,
+      "learning_rate": 3.606010016694491e-06,
+      "loss": 0.0414,
+      "step": 2896
+    },
+    {
+      "epoch": 19.31,
+      "learning_rate": 3.572621035058431e-06,
+      "loss": 0.0586,
+      "step": 2897
+    },
+    {
+      "epoch": 19.32,
+      "learning_rate": 3.5392320534223705e-06,
+      "loss": 0.0471,
+      "step": 2898
+    },
+    {
+      "epoch": 19.33,
+      "learning_rate": 3.5058430717863105e-06,
+      "loss": 0.0583,
+      "step": 2899
+    },
+    {
+      "epoch": 19.33,
+      "learning_rate": 3.4724540901502504e-06,
+      "loss": 0.0665,
+      "step": 2900
+    },
+    {
+      "epoch": 19.34,
+      "learning_rate": 3.4390651085141904e-06,
+      "loss": 0.0695,
+      "step": 2901
+    },
+    {
+      "epoch": 19.35,
+      "learning_rate": 3.4056761268781303e-06,
+      "loss": 0.0382,
+      "step": 2902
+    },
+    {
+      "epoch": 19.35,
+      "learning_rate": 3.3722871452420703e-06,
+      "loss": 0.0638,
+      "step": 2903
+    },
+    {
+      "epoch": 19.36,
+      "learning_rate": 3.33889816360601e-06,
+      "loss": 0.0577,
+      "step": 2904
+    },
+    {
+      "epoch": 19.37,
+      "learning_rate": 3.30550918196995e-06,
+      "loss": 0.0833,
+      "step": 2905
+    },
+    {
+      "epoch": 19.37,
+      "learning_rate": 3.2721202003338898e-06,
+      "loss": 0.0653,
+      "step": 2906
+    },
+    {
+      "epoch": 19.38,
+      "learning_rate": 3.2387312186978297e-06,
+      "loss": 0.0861,
+      "step": 2907
+    },
+    {
+      "epoch": 19.39,
+      "learning_rate": 3.2053422370617697e-06,
+      "loss": 0.0586,
+      "step": 2908
+    },
+    {
+      "epoch": 19.39,
+      "learning_rate": 3.1719532554257097e-06,
+      "loss": 0.1046,
+      "step": 2909
+    },
+    {
+      "epoch": 19.4,
+      "learning_rate": 3.1385642737896492e-06,
+      "loss": 0.0585,
+      "step": 2910
+    },
+    {
+      "epoch": 19.41,
+      "learning_rate": 3.1051752921535896e-06,
+      "loss": 0.0475,
+      "step": 2911
+    },
+    {
+      "epoch": 19.41,
+      "learning_rate": 3.0717863105175296e-06,
+      "loss": 0.0483,
+      "step": 2912
+    },
+    {
+      "epoch": 19.42,
+      "learning_rate": 3.0383973288814695e-06,
+      "loss": 0.0534,
+      "step": 2913
+    },
+    {
+      "epoch": 19.43,
+      "learning_rate": 3.005008347245409e-06,
+      "loss": 0.0487,
+      "step": 2914
+    },
+    {
+      "epoch": 19.43,
+      "learning_rate": 2.971619365609349e-06,
+      "loss": 0.0623,
+      "step": 2915
+    },
+    {
+      "epoch": 19.44,
+      "learning_rate": 2.938230383973289e-06,
+      "loss": 0.0425,
+      "step": 2916
+    },
+    {
+      "epoch": 19.45,
+      "learning_rate": 2.904841402337229e-06,
+      "loss": 0.0674,
+      "step": 2917
+    },
+    {
+      "epoch": 19.45,
+      "learning_rate": 2.871452420701169e-06,
+      "loss": 0.0833,
+      "step": 2918
+    },
+    {
+      "epoch": 19.46,
+      "learning_rate": 2.838063439065109e-06,
+      "loss": 0.0414,
+      "step": 2919
+    },
+    {
+      "epoch": 19.47,
+      "learning_rate": 2.8046744574290484e-06,
+      "loss": 0.045,
+      "step": 2920
+    },
+    {
+      "epoch": 19.47,
+      "learning_rate": 2.7712854757929884e-06,
+      "loss": 0.0373,
+      "step": 2921
+    },
+    {
+      "epoch": 19.48,
+      "learning_rate": 2.7378964941569284e-06,
+      "loss": 0.0606,
+      "step": 2922
+    },
+    {
+      "epoch": 19.49,
+      "learning_rate": 2.7045075125208683e-06,
+      "loss": 0.0804,
+      "step": 2923
+    },
+    {
+      "epoch": 19.49,
+      "learning_rate": 2.6711185308848083e-06,
+      "loss": 0.0549,
+      "step": 2924
+    },
+    {
+      "epoch": 19.5,
+      "learning_rate": 2.6377295492487483e-06,
+      "loss": 0.0384,
+      "step": 2925
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 2.604340567612688e-06,
+      "loss": 0.0999,
+      "step": 2926
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 2.5709515859766278e-06,
+      "loss": 0.1181,
+      "step": 2927
+    },
+    {
+      "epoch": 19.52,
+      "learning_rate": 2.5375626043405677e-06,
+      "loss": 0.0719,
+      "step": 2928
+    },
+    {
+      "epoch": 19.53,
+      "learning_rate": 2.5041736227045077e-06,
+      "loss": 0.047,
+      "step": 2929
+    },
+    {
+      "epoch": 19.53,
+      "learning_rate": 2.4707846410684477e-06,
+      "loss": 0.0567,
+      "step": 2930
+    },
+    {
+      "epoch": 19.54,
+      "learning_rate": 2.4373956594323876e-06,
+      "loss": 0.0389,
+      "step": 2931
+    },
+    {
+      "epoch": 19.55,
+      "learning_rate": 2.404006677796327e-06,
+      "loss": 0.0611,
+      "step": 2932
+    },
+    {
+      "epoch": 19.55,
+      "learning_rate": 2.370617696160267e-06,
+      "loss": 0.042,
+      "step": 2933
+    },
+    {
+      "epoch": 19.56,
+      "learning_rate": 2.337228714524207e-06,
+      "loss": 0.0717,
+      "step": 2934
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 2.303839732888147e-06,
+      "loss": 0.0468,
+      "step": 2935
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 2.270450751252087e-06,
+      "loss": 0.0868,
+      "step": 2936
+    },
+    {
+      "epoch": 19.58,
+      "learning_rate": 2.237061769616027e-06,
+      "loss": 0.0542,
+      "step": 2937
+    },
+    {
+      "epoch": 19.59,
+      "learning_rate": 2.2036727879799665e-06,
+      "loss": 0.038,
+      "step": 2938
+    },
+    {
+      "epoch": 19.59,
+      "learning_rate": 2.1702838063439065e-06,
+      "loss": 0.053,
+      "step": 2939
+    },
+    {
+      "epoch": 19.6,
+      "learning_rate": 2.1368948247078465e-06,
+      "loss": 0.0693,
+      "step": 2940
+    },
+    {
+      "epoch": 19.61,
+      "learning_rate": 2.1035058430717864e-06,
+      "loss": 0.0672,
+      "step": 2941
+    },
+    {
+      "epoch": 19.61,
+      "learning_rate": 2.0701168614357264e-06,
+      "loss": 0.071,
+      "step": 2942
+    },
+    {
+      "epoch": 19.62,
+      "learning_rate": 2.0367278797996664e-06,
+      "loss": 0.07,
+      "step": 2943
+    },
+    {
+      "epoch": 19.63,
+      "learning_rate": 2.003338898163606e-06,
+      "loss": 0.0511,
+      "step": 2944
+    },
+    {
+      "epoch": 19.63,
+      "learning_rate": 1.969949916527546e-06,
+      "loss": 0.0698,
+      "step": 2945
+    },
+    {
+      "epoch": 19.64,
+      "learning_rate": 1.936560934891486e-06,
+      "loss": 0.0524,
+      "step": 2946
+    },
+    {
+      "epoch": 19.65,
+      "learning_rate": 1.9031719532554258e-06,
+      "loss": 0.0499,
+      "step": 2947
+    },
+    {
+      "epoch": 19.65,
+      "learning_rate": 1.8697829716193656e-06,
+      "loss": 0.0564,
+      "step": 2948
+    },
+    {
+      "epoch": 19.66,
+      "learning_rate": 1.8363939899833055e-06,
+      "loss": 0.0605,
+      "step": 2949
+    },
+    {
+      "epoch": 19.67,
+      "learning_rate": 1.8030050083472455e-06,
+      "loss": 0.0547,
+      "step": 2950
+    },
+    {
+      "epoch": 19.67,
+      "learning_rate": 1.7696160267111852e-06,
+      "loss": 0.0525,
+      "step": 2951
+    },
+    {
+      "epoch": 19.68,
+      "learning_rate": 1.7362270450751252e-06,
+      "loss": 0.0757,
+      "step": 2952
+    },
+    {
+      "epoch": 19.69,
+      "learning_rate": 1.7028380634390652e-06,
+      "loss": 0.0462,
+      "step": 2953
+    },
+    {
+      "epoch": 19.69,
+      "learning_rate": 1.669449081803005e-06,
+      "loss": 0.0623,
+      "step": 2954
+    },
+    {
+      "epoch": 19.7,
+      "learning_rate": 1.6360601001669449e-06,
+      "loss": 0.0568,
+      "step": 2955
+    },
+    {
+      "epoch": 19.71,
+      "learning_rate": 1.6026711185308849e-06,
+      "loss": 0.0684,
+      "step": 2956
+    },
+    {
+      "epoch": 19.71,
+      "learning_rate": 1.5692821368948246e-06,
+      "loss": 0.0369,
+      "step": 2957
+    },
+    {
+      "epoch": 19.72,
+      "learning_rate": 1.5358931552587648e-06,
+      "loss": 0.0593,
+      "step": 2958
+    },
+    {
+      "epoch": 19.73,
+      "learning_rate": 1.5025041736227045e-06,
+      "loss": 0.0398,
+      "step": 2959
+    },
+    {
+      "epoch": 19.73,
+      "learning_rate": 1.4691151919866445e-06,
+      "loss": 0.0588,
+      "step": 2960
+    },
+    {
+      "epoch": 19.74,
+      "learning_rate": 1.4357262103505845e-06,
+      "loss": 0.0412,
+      "step": 2961
+    },
+    {
+      "epoch": 19.75,
+      "learning_rate": 1.4023372287145242e-06,
+      "loss": 0.0655,
+      "step": 2962
+    },
+    {
+      "epoch": 19.75,
+      "learning_rate": 1.3689482470784642e-06,
+      "loss": 0.034,
+      "step": 2963
+    },
+    {
+      "epoch": 19.76,
+      "learning_rate": 1.3355592654424042e-06,
+      "loss": 0.0626,
+      "step": 2964
+    },
+    {
+      "epoch": 19.77,
+      "learning_rate": 1.302170283806344e-06,
+      "loss": 0.0494,
+      "step": 2965
+    },
+    {
+      "epoch": 19.77,
+      "learning_rate": 1.2687813021702839e-06,
+      "loss": 0.0802,
+      "step": 2966
+    },
+    {
+      "epoch": 19.78,
+      "learning_rate": 1.2353923205342238e-06,
+      "loss": 0.086,
+      "step": 2967
+    },
+    {
+      "epoch": 19.79,
+      "learning_rate": 1.2020033388981636e-06,
+      "loss": 0.0718,
+      "step": 2968
+    },
+    {
+      "epoch": 19.79,
+      "learning_rate": 1.1686143572621036e-06,
+      "loss": 0.0554,
+      "step": 2969
+    },
+    {
+      "epoch": 19.8,
+      "learning_rate": 1.1352253756260435e-06,
+      "loss": 0.0549,
+      "step": 2970
+    },
+    {
+      "epoch": 19.81,
+      "learning_rate": 1.1018363939899833e-06,
+      "loss": 0.0644,
+      "step": 2971
+    },
+    {
+      "epoch": 19.81,
+      "learning_rate": 1.0684474123539232e-06,
+      "loss": 0.0536,
+      "step": 2972
+    },
+    {
+      "epoch": 19.82,
+      "learning_rate": 1.0350584307178632e-06,
+      "loss": 0.0575,
+      "step": 2973
+    },
+    {
+      "epoch": 19.83,
+      "learning_rate": 1.001669449081803e-06,
+      "loss": 0.0877,
+      "step": 2974
+    },
+    {
+      "epoch": 19.83,
+      "learning_rate": 9.68280467445743e-07,
+      "loss": 0.0482,
+      "step": 2975
+    },
+    {
+      "epoch": 19.84,
+      "learning_rate": 9.348914858096828e-07,
+      "loss": 0.0481,
+      "step": 2976
+    },
+    {
+      "epoch": 19.85,
+      "learning_rate": 9.015025041736227e-07,
+      "loss": 0.0577,
+      "step": 2977
+    },
+    {
+      "epoch": 19.85,
+      "learning_rate": 8.681135225375626e-07,
+      "loss": 0.0515,
+      "step": 2978
+    },
+    {
+      "epoch": 19.86,
+      "learning_rate": 8.347245409015025e-07,
+      "loss": 0.0623,
+      "step": 2979
+    },
+    {
+      "epoch": 19.87,
+      "learning_rate": 8.013355592654424e-07,
+      "loss": 0.0643,
+      "step": 2980
+    },
+    {
+      "epoch": 19.87,
+      "learning_rate": 7.679465776293824e-07,
+      "loss": 0.0573,
+      "step": 2981
+    },
+    {
+      "epoch": 19.88,
+      "learning_rate": 7.345575959933223e-07,
+      "loss": 0.0495,
+      "step": 2982
+    },
+    {
+      "epoch": 19.89,
+      "learning_rate": 7.011686143572621e-07,
+      "loss": 0.1311,
+      "step": 2983
+    },
+    {
+      "epoch": 19.89,
+      "learning_rate": 6.677796327212021e-07,
+      "loss": 0.0609,
+      "step": 2984
+    },
+    {
+      "epoch": 19.9,
+      "learning_rate": 6.343906510851419e-07,
+      "loss": 0.0473,
+      "step": 2985
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 6.010016694490818e-07,
+      "loss": 0.0622,
+      "step": 2986
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 5.676126878130218e-07,
+      "loss": 0.0403,
+      "step": 2987
+    },
+    {
+      "epoch": 19.92,
+      "learning_rate": 5.342237061769616e-07,
+      "loss": 0.0867,
+      "step": 2988
+    },
+    {
+      "epoch": 19.93,
+      "learning_rate": 5.008347245409015e-07,
+      "loss": 0.0596,
+      "step": 2989
+    },
+    {
+      "epoch": 19.93,
+      "learning_rate": 4.674457429048414e-07,
+      "loss": 0.0497,
+      "step": 2990
+    },
+    {
+      "epoch": 19.94,
+      "learning_rate": 4.340567612687813e-07,
+      "loss": 0.0601,
+      "step": 2991
+    },
+    {
+      "epoch": 19.95,
+      "learning_rate": 4.006677796327212e-07,
+      "loss": 0.0595,
+      "step": 2992
+    },
+    {
+      "epoch": 19.95,
+      "learning_rate": 3.672787979966611e-07,
+      "loss": 0.1274,
+      "step": 2993
+    },
+    {
+      "epoch": 19.96,
+      "learning_rate": 3.3388981636060104e-07,
+      "loss": 0.0652,
+      "step": 2994
+    },
+    {
+      "epoch": 19.97,
+      "learning_rate": 3.005008347245409e-07,
+      "loss": 0.053,
+      "step": 2995
+    },
+    {
+      "epoch": 19.97,
+      "learning_rate": 2.671118530884808e-07,
+      "loss": 0.0547,
+      "step": 2996
+    },
+    {
+      "epoch": 19.98,
+      "learning_rate": 2.337228714524207e-07,
+      "loss": 0.0863,
+      "step": 2997
+    },
+    {
+      "epoch": 19.99,
+      "learning_rate": 2.003338898163606e-07,
+      "loss": 0.0505,
+      "step": 2998
+    },
+    {
+      "epoch": 19.99,
+      "learning_rate": 1.6694490818030052e-07,
+      "loss": 0.0738,
+      "step": 2999
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 1.335559265442404e-07,
+      "loss": 0.0661,
+      "step": 3000
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 3.312180757522583,
+      "eval_runtime": 20.1859,
+      "eval_samples_per_second": 9.908,
+      "eval_steps_per_second": 1.238,
+      "step": 3000
+    },
+    {
+      "epoch": 20.0,
+      "step": 3000,
+      "total_flos": 1.39753267789824e+17,
+      "train_loss": 0.2855387391100327,
+      "train_runtime": 5294.5765,
+      "train_samples_per_second": 2.266,
+      "train_steps_per_second": 0.567
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3000,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "total_flos": 1.39753267789824e+17,
+  "trial_name": null,
+  "trial_params": null
+}