diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12953 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 14.998144712430427,
+  "eval_steps": 20,
+  "global_step": 2021,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.4068,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.3929,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9e-07,
+      "loss": 1.4264,
+      "step": 3
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.3762,
+      "step": 4
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5e-06,
+      "loss": 1.4034,
+      "step": 5
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8e-06,
+      "loss": 1.3497,
+      "step": 6
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.1e-06,
+      "loss": 1.3295,
+      "step": 7
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.2797,
+      "step": 8
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.7e-06,
+      "loss": 1.2234,
+      "step": 9
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3e-06,
+      "loss": 1.2516,
+      "step": 10
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9999995396668603e-06,
+      "loss": 1.2266,
+      "step": 11
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9999981586677232e-06,
+      "loss": 1.2597,
+      "step": 12
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9999958570034366e-06,
+      "loss": 1.1746,
+      "step": 13
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9999926346754133e-06,
+      "loss": 1.2191,
+      "step": 14
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9999884916856308e-06,
+      "loss": 1.2596,
+      "step": 15
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.999983428036632e-06,
+      "loss": 1.1985,
+      "step": 16
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9999774437315253e-06,
+      "loss": 1.1989,
+      "step": 17
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9999705387739833e-06,
+      "loss": 1.1987,
+      "step": 18
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9999627131682443e-06,
+      "loss": 1.172,
+      "step": 19
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9999539669191114e-06,
+      "loss": 1.1761,
+      "step": 20
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 1.335253357887268,
+      "eval_runtime": 1.8159,
+      "eval_samples_per_second": 36.346,
+      "eval_steps_per_second": 3.304,
+      "step": 20
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9999443000319527e-06,
+      "loss": 1.1674,
+      "step": 21
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9999337125127016e-06,
+      "loss": 1.134,
+      "step": 22
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9999222043678566e-06,
+      "loss": 1.1289,
+      "step": 23
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9999097756044815e-06,
+      "loss": 1.1247,
+      "step": 24
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.999896426230204e-06,
+      "loss": 1.1139,
+      "step": 25
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.999882156253218e-06,
+      "loss": 1.1142,
+      "step": 26
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9998669656822823e-06,
+      "loss": 1.1345,
+      "step": 27
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9998508545267203e-06,
+      "loss": 1.136,
+      "step": 28
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9998338227964208e-06,
+      "loss": 1.1404,
+      "step": 29
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9998158705018376e-06,
+      "loss": 1.1195,
+      "step": 30
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.999796997653989e-06,
+      "loss": 1.0955,
+      "step": 31
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.999777204264459e-06,
+      "loss": 1.1165,
+      "step": 32
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.999756490345396e-06,
+      "loss": 1.1253,
+      "step": 33
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.999734855909514e-06,
+      "loss": 1.145,
+      "step": 34
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9997123009700924e-06,
+      "loss": 1.0867,
+      "step": 35
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.999688825540974e-06,
+      "loss": 1.1319,
+      "step": 36
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9996644296365676e-06,
+      "loss": 1.1444,
+      "step": 37
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.999639113271847e-06,
+      "loss": 1.113,
+      "step": 38
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.999612876462351e-06,
+      "loss": 1.0607,
+      "step": 39
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9995857192241834e-06,
+      "loss": 1.1323,
+      "step": 40
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.2668023109436035,
+      "eval_runtime": 1.8227,
+      "eval_samples_per_second": 36.211,
+      "eval_steps_per_second": 3.292,
+      "step": 40
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.999557641574012e-06,
+      "loss": 1.1621,
+      "step": 41
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.99952864352907e-06,
+      "loss": 1.1072,
+      "step": 42
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.999498725107157e-06,
+      "loss": 1.1117,
+      "step": 43
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9994678863266354e-06,
+      "loss": 1.1317,
+      "step": 44
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9994361272064333e-06,
+      "loss": 1.0689,
+      "step": 45
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9994034477660443e-06,
+      "loss": 1.1003,
+      "step": 46
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9993698480255257e-06,
+      "loss": 1.1086,
+      "step": 47
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9993353280055003e-06,
+      "loss": 1.1257,
+      "step": 48
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.999299887727156e-06,
+      "loss": 1.0776,
+      "step": 49
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.9992635272122455e-06,
+      "loss": 1.0661,
+      "step": 50
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.999226246483085e-06,
+      "loss": 1.0616,
+      "step": 51
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.9991880455625584e-06,
+      "loss": 1.073,
+      "step": 52
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.9991489244741104e-06,
+      "loss": 1.0927,
+      "step": 53
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.999108883241754e-06,
+      "loss": 1.0852,
+      "step": 54
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.999067921890066e-06,
+      "loss": 1.0932,
+      "step": 55
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9990260404441862e-06,
+      "loss": 1.0667,
+      "step": 56
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9989832389298215e-06,
+      "loss": 1.0796,
+      "step": 57
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.998939517373242e-06,
+      "loss": 1.1102,
+      "step": 58
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.998894875801283e-06,
+      "loss": 1.1272,
+      "step": 59
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.9988493142413446e-06,
+      "loss": 1.0271,
+      "step": 60
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.2398362159729004,
+      "eval_runtime": 1.8175,
+      "eval_samples_per_second": 36.314,
+      "eval_steps_per_second": 3.301,
+      "step": 60
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.9988028327213922e-06,
+      "loss": 1.0986,
+      "step": 61
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.998755431269954e-06,
+      "loss": 1.0927,
+      "step": 62
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.998707109916124e-06,
+      "loss": 1.0607,
+      "step": 63
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9986578686895617e-06,
+      "loss": 1.0706,
+      "step": 64
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9986077076204903e-06,
+      "loss": 1.083,
+      "step": 65
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.998556626739696e-06,
+      "loss": 1.1019,
+      "step": 66
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.9985046260785325e-06,
+      "loss": 1.09,
+      "step": 67
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.998451705668916e-06,
+      "loss": 1.0593,
+      "step": 68
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.9983978655433277e-06,
+      "loss": 1.0834,
+      "step": 69
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.9983431057348144e-06,
+      "loss": 1.0941,
+      "step": 70
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.9982874262769854e-06,
+      "loss": 1.0811,
+      "step": 71
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.9982308272040164e-06,
+      "loss": 1.0473,
+      "step": 72
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.9981733085506456e-06,
+      "loss": 1.0653,
+      "step": 73
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.9981148703521778e-06,
+      "loss": 1.0929,
+      "step": 74
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9980555126444804e-06,
+      "loss": 1.1257,
+      "step": 75
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9979952354639853e-06,
+      "loss": 1.0512,
+      "step": 76
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.9979340388476904e-06,
+      "loss": 1.0529,
+      "step": 77
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.997871922833156e-06,
+      "loss": 1.0494,
+      "step": 78
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9978088874585083e-06,
+      "loss": 1.0679,
+      "step": 79
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9977449327624364e-06,
+      "loss": 1.0835,
+      "step": 80
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 1.2232372760772705,
+      "eval_runtime": 1.8225,
+      "eval_samples_per_second": 36.215,
+      "eval_steps_per_second": 3.292,
+      "step": 80
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.997680058784194e-06,
+      "loss": 1.0601,
+      "step": 81
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.9976142655636e-06,
+      "loss": 1.0954,
+      "step": 82
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.997547553141036e-06,
+      "loss": 1.0848,
+      "step": 83
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.9974799215574497e-06,
+      "loss": 1.0514,
+      "step": 84
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.9974113708543503e-06,
+      "loss": 1.0502,
+      "step": 85
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.997341901073814e-06,
+      "loss": 1.0602,
+      "step": 86
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.99727151225848e-06,
+      "loss": 1.0845,
+      "step": 87
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.99720020445155e-06,
+      "loss": 1.0371,
+      "step": 88
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.9971279776967924e-06,
+      "loss": 1.0749,
+      "step": 89
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.997054832038537e-06,
+      "loss": 1.0695,
+      "step": 90
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.9969807675216805e-06,
+      "loss": 1.0792,
+      "step": 91
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.996905784191681e-06,
+      "loss": 1.0782,
+      "step": 92
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.996829882094562e-06,
+      "loss": 1.0735,
+      "step": 93
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.99675306127691e-06,
+      "loss": 1.1059,
+      "step": 94
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.996675321785877e-06,
+      "loss": 1.0736,
+      "step": 95
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.996596663669176e-06,
+      "loss": 1.0464,
+      "step": 96
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.9965170869750878e-06,
+      "loss": 1.1287,
+      "step": 97
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.996436591752453e-06,
+      "loss": 1.0405,
+      "step": 98
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.9963551780506785e-06,
+      "loss": 1.0556,
+      "step": 99
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.996272845919734e-06,
+      "loss": 1.0669,
+      "step": 100
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.2120617628097534,
+      "eval_runtime": 1.8172,
+      "eval_samples_per_second": 36.32,
+      "eval_steps_per_second": 3.302,
+      "step": 100
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.9961895954101527e-06,
+      "loss": 1.1008,
+      "step": 101
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.996105426573033e-06,
+      "loss": 1.0584,
+      "step": 102
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.996020339460035e-06,
+      "loss": 1.0757,
+      "step": 103
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.9959343341233837e-06,
+      "loss": 1.0573,
+      "step": 104
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.9958474106158662e-06,
+      "loss": 1.0408,
+      "step": 105
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.995759568990836e-06,
+      "loss": 1.0736,
+      "step": 106
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.995670809302207e-06,
+      "loss": 1.1007,
+      "step": 107
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.995581131604458e-06,
+      "loss": 1.0354,
+      "step": 108
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.9954905359526313e-06,
+      "loss": 1.0524,
+      "step": 109
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9953990224023332e-06,
+      "loss": 1.0354,
+      "step": 110
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9953065910097314e-06,
+      "loss": 1.0763,
+      "step": 111
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9952132418315587e-06,
+      "loss": 1.0249,
+      "step": 112
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.995118974925111e-06,
+      "loss": 1.0402,
+      "step": 113
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9950237903482472e-06,
+      "loss": 1.0615,
+      "step": 114
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9949276881593895e-06,
+      "loss": 1.0768,
+      "step": 115
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.994830668417523e-06,
+      "loss": 1.0536,
+      "step": 116
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.994732731182196e-06,
+      "loss": 1.0653,
+      "step": 117
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9946338765135202e-06,
+      "loss": 1.0439,
+      "step": 118
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9945341044721717e-06,
+      "loss": 1.0862,
+      "step": 119
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.994433415119386e-06,
+      "loss": 1.0388,
+      "step": 120
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.2034080028533936,
+      "eval_runtime": 1.8176,
+      "eval_samples_per_second": 36.312,
+      "eval_steps_per_second": 3.301,
+      "step": 120
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9943318085169663e-06,
+      "loss": 1.0952,
+      "step": 121
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.994229284727275e-06,
+      "loss": 1.1112,
+      "step": 122
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.9941258438132394e-06,
+      "loss": 1.0819,
+      "step": 123
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9940214858383487e-06,
+      "loss": 1.0623,
+      "step": 124
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9939162108666566e-06,
+      "loss": 1.0569,
+      "step": 125
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.993810018962777e-06,
+      "loss": 1.1112,
+      "step": 126
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9937029101918894e-06,
+      "loss": 1.0466,
+      "step": 127
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.993594884619734e-06,
+      "loss": 1.0011,
+      "step": 128
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.993485942312615e-06,
+      "loss": 1.0446,
+      "step": 129
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.9933760833373984e-06,
+      "loss": 1.0511,
+      "step": 130
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.993265307761513e-06,
+      "loss": 1.1086,
+      "step": 131
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.993153615652951e-06,
+      "loss": 1.046,
+      "step": 132
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.9930410070802654e-06,
+      "loss": 1.0486,
+      "step": 133
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.9929274821125738e-06,
+      "loss": 1.0632,
+      "step": 134
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.992813040819555e-06,
+      "loss": 1.0408,
+      "step": 135
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9926976832714507e-06,
+      "loss": 1.0186,
+      "step": 136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.992581409539064e-06,
+      "loss": 0.9954,
+      "step": 137
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9924642196937624e-06,
+      "loss": 0.9309,
+      "step": 138
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9923461138074736e-06,
+      "loss": 0.9568,
+      "step": 139
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9922270919526876e-06,
+      "loss": 0.9606,
+      "step": 140
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.197842001914978,
+      "eval_runtime": 1.8134,
+      "eval_samples_per_second": 36.396,
+      "eval_steps_per_second": 3.309,
+      "step": 140
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9921071542024587e-06,
+      "loss": 1.0029,
+      "step": 141
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.991986300630401e-06,
+      "loss": 1.0018,
+      "step": 142
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.991864531310693e-06,
+      "loss": 0.9866,
+      "step": 143
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.9917418463180715e-06,
+      "loss": 0.9859,
+      "step": 144
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.991618245727841e-06,
+      "loss": 0.9916,
+      "step": 145
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.991493729615862e-06,
+      "loss": 0.9824,
+      "step": 146
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.9913682980585616e-06,
+      "loss": 1.0042,
+      "step": 147
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.9912419511329254e-06,
+      "loss": 0.9981,
+      "step": 148
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.9911146889165036e-06,
+      "loss": 1.0002,
+      "step": 149
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.9909865114874058e-06,
+      "loss": 0.9574,
+      "step": 150
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.990857418924305e-06,
+      "loss": 0.9668,
+      "step": 151
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.990727411306436e-06,
+      "loss": 1.0108,
+      "step": 152
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.990596488713593e-06,
+      "loss": 0.9681,
+      "step": 153
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.9904646512261343e-06,
+      "loss": 0.9691,
+      "step": 154
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.9903318989249785e-06,
+      "loss": 1.003,
+      "step": 155
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.990198231891606e-06,
+      "loss": 0.9395,
+      "step": 156
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.990063650208059e-06,
+      "loss": 0.9838,
+      "step": 157
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.9899281539569405e-06,
+      "loss": 0.9882,
+      "step": 158
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.9897917432214144e-06,
+      "loss": 1.0172,
+      "step": 159
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.9896544180852077e-06,
+      "loss": 0.9634,
+      "step": 160
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.1941014528274536,
+      "eval_runtime": 1.8182,
+      "eval_samples_per_second": 36.3,
+      "eval_steps_per_second": 3.3,
+      "step": 160
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.9895161786326065e-06,
+      "loss": 0.9727,
+      "step": 161
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.98937702494846e-06,
+      "loss": 1.0213,
+      "step": 162
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.989236957118176e-06,
+      "loss": 0.9209,
+      "step": 163
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.989095975227727e-06,
+      "loss": 1.0163,
+      "step": 164
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.988954079363643e-06,
+      "loss": 0.9846,
+      "step": 165
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.988811269613018e-06,
+      "loss": 0.975,
+      "step": 166
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.988667546063504e-06,
+      "loss": 0.989,
+      "step": 167
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.988522908803315e-06,
+      "loss": 1.014,
+      "step": 168
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.9883773579212275e-06,
+      "loss": 0.9486,
+      "step": 169
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.988230893506577e-06,
+      "loss": 1.0202,
+      "step": 170
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9880835156492595e-06,
+      "loss": 0.9764,
+      "step": 171
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9879352244397326e-06,
+      "loss": 0.9589,
+      "step": 172
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9877860199690135e-06,
+      "loss": 0.9351,
+      "step": 173
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9876359023286814e-06,
+      "loss": 0.9954,
+      "step": 174
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9874848716108747e-06,
+      "loss": 0.9919,
+      "step": 175
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9873329279082928e-06,
+      "loss": 0.9629,
+      "step": 176
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9871800713141956e-06,
+      "loss": 0.9978,
+      "step": 177
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9870263019224023e-06,
+      "loss": 0.9322,
+      "step": 178
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.9868716198272935e-06,
+      "loss": 0.9769,
+      "step": 179
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.98671602512381e-06,
+      "loss": 0.9726,
+      "step": 180
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.1889244318008423,
+      "eval_runtime": 1.8198,
+      "eval_samples_per_second": 36.268,
+      "eval_steps_per_second": 3.297,
+      "step": 180
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.986559517907452e-06,
+      "loss": 0.9394,
+      "step": 181
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.9864020982742795e-06,
+      "loss": 0.97,
+      "step": 182
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.986243766320914e-06,
+      "loss": 1.0206,
+      "step": 183
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9860845221445356e-06,
+      "loss": 1.0069,
+      "step": 184
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9859243658428853e-06,
+      "loss": 0.9679,
+      "step": 185
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.985763297514263e-06,
+      "loss": 0.9403,
+      "step": 186
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9856013172575288e-06,
+      "loss": 0.9701,
+      "step": 187
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.9854384251721025e-06,
+      "loss": 1.0336,
+      "step": 188
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.985274621357964e-06,
+      "loss": 0.9859,
+      "step": 189
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.985109905915652e-06,
+      "loss": 0.998,
+      "step": 190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.984944278946266e-06,
+      "loss": 0.9998,
+      "step": 191
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.984777740551463e-06,
+      "loss": 0.9688,
+      "step": 192
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.9846102908334606e-06,
+      "loss": 0.9979,
+      "step": 193
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.9844419298950357e-06,
+      "loss": 0.94,
+      "step": 194
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.984272657839526e-06,
+      "loss": 0.97,
+      "step": 195
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.9841024747708248e-06,
+      "loss": 1.0064,
+      "step": 196
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.9839313807933876e-06,
+      "loss": 0.9905,
+      "step": 197
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.983759376012228e-06,
+      "loss": 0.9872,
+      "step": 198
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.9835864605329185e-06,
+      "loss": 0.9709,
+      "step": 199
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.983412634461591e-06,
+      "loss": 1.0163,
+      "step": 200
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.1839343309402466,
+      "eval_runtime": 1.8146,
+      "eval_samples_per_second": 36.371,
+      "eval_steps_per_second": 3.306,
+      "step": 200
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.9832378979049355e-06,
+      "loss": 0.9571,
+      "step": 201
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.983062250970202e-06,
+      "loss": 0.9708,
+      "step": 202
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.9828856937651985e-06,
+      "loss": 0.9476,
+      "step": 203
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.9827082263982912e-06,
+      "loss": 0.9975,
+      "step": 204
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.982529848978407e-06,
+      "loss": 1.0595,
+      "step": 205
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.9823505616150282e-06,
+      "loss": 0.9686,
+      "step": 206
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.982170364418199e-06,
+      "loss": 0.9426,
+      "step": 207
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.9819892574985186e-06,
+      "loss": 0.9972,
+      "step": 208
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.981807240967148e-06,
+      "loss": 0.9653,
+      "step": 209
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.9816243149358044e-06,
+      "loss": 0.9686,
+      "step": 210
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.981440479516763e-06,
+      "loss": 0.9471,
+      "step": 211
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.9812557348228585e-06,
+      "loss": 1.0193,
+      "step": 212
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.9810700809674828e-06,
+      "loss": 1.0111,
+      "step": 213
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.9808835180645865e-06,
+      "loss": 0.9878,
+      "step": 214
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.980696046228677e-06,
+      "loss": 0.9663,
+      "step": 215
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.9805076655748205e-06,
+      "loss": 0.9718,
+      "step": 216
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.9803183762186415e-06,
+      "loss": 1.0336,
+      "step": 217
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.9801281782763204e-06,
+      "loss": 0.9916,
+      "step": 218
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.979937071864598e-06,
+      "loss": 0.9927,
+      "step": 219
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.979745057100769e-06,
+      "loss": 1.0287,
+      "step": 220
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.1805609464645386,
+      "eval_runtime": 1.8158,
+      "eval_samples_per_second": 36.347,
+      "eval_steps_per_second": 3.304,
+      "step": 220
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.9795521341026896e-06,
+      "loss": 0.9554,
+      "step": 221
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.9793583029887707e-06,
+      "loss": 0.9743,
+      "step": 222
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.9791635638779817e-06,
+      "loss": 0.9771,
+      "step": 223
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.978967916889849e-06,
+      "loss": 0.9908,
+      "step": 224
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.9787713621444567e-06,
+      "loss": 0.9466,
+      "step": 225
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.9785738997624453e-06,
+      "loss": 1.004,
+      "step": 226
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.9783755298650127e-06,
+      "loss": 0.9519,
+      "step": 227
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.978176252573914e-06,
+      "loss": 1.0179,
+      "step": 228
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.977976068011461e-06,
+      "loss": 0.9777,
+      "step": 229
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.9777749763005226e-06,
+      "loss": 0.8937,
+      "step": 230
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.977572977564525e-06,
+      "loss": 1.0117,
+      "step": 231
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.97737007192745e-06,
+      "loss": 0.9482,
+      "step": 232
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.9771662595138358e-06,
+      "loss": 0.9396,
+      "step": 233
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.9769615404487782e-06,
+      "loss": 0.9414,
+      "step": 234
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.97675591485793e-06,
+      "loss": 1.0108,
+      "step": 235
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.976549382867499e-06,
+      "loss": 0.9767,
+      "step": 236
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.97634194460425e-06,
+      "loss": 0.9863,
+      "step": 237
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.9761336001955033e-06,
+      "loss": 0.9893,
+      "step": 238
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.975924349769137e-06,
+      "loss": 0.9742,
+      "step": 239
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.975714193453584e-06,
+      "loss": 0.9901,
+      "step": 240
+    },
+    {
+      "epoch": 1.78,
+      "eval_loss": 1.178310513496399,
+      "eval_runtime": 1.8219,
+      "eval_samples_per_second": 36.225,
+      "eval_steps_per_second": 3.293,
+      "step": 240
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.975503131377833e-06,
+      "loss": 0.9266,
+      "step": 241
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.97529116367143e-06,
+      "loss": 1.0126,
+      "step": 242
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.975078290464475e-06,
+      "loss": 0.9488,
+      "step": 243
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.974864511887626e-06,
+      "loss": 0.9814,
+      "step": 244
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.9746498280720946e-06,
+      "loss": 0.9781,
+      "step": 245
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.974434239149649e-06,
+      "loss": 0.9821,
+      "step": 246
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.974217745252613e-06,
+      "loss": 0.9941,
+      "step": 247
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.9740003465138657e-06,
+      "loss": 1.0028,
+      "step": 248
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.973782043066841e-06,
+      "loss": 0.9887,
+      "step": 249
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.97356283504553e-06,
+      "loss": 0.9887,
+      "step": 250
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.9733427225844766e-06,
+      "loss": 0.9529,
+      "step": 251
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.9731217058187812e-06,
+      "loss": 1.0188,
+      "step": 252
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.9728997848840987e-06,
+      "loss": 0.9591,
+      "step": 253
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.972676959916639e-06,
+      "loss": 0.9895,
+      "step": 254
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.9724532310531676e-06,
+      "loss": 0.9596,
+      "step": 255
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.9722285984310037e-06,
+      "loss": 0.9986,
+      "step": 256
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.9720030621880227e-06,
+      "loss": 0.9429,
+      "step": 257
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.9717766224626525e-06,
+      "loss": 0.9758,
+      "step": 258
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.971549279393878e-06,
+      "loss": 0.9795,
+      "step": 259
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.971321033121236e-06,
+      "loss": 0.9673,
+      "step": 260
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.1745336055755615,
+      "eval_runtime": 1.8221,
+      "eval_samples_per_second": 36.223,
+      "eval_steps_per_second": 3.293,
+      "step": 260
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.9710918837848193e-06,
+      "loss": 1.0123,
+      "step": 261
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.9708618315252747e-06,
+      "loss": 0.9913,
+      "step": 262
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.9706308764838034e-06,
+      "loss": 0.9619,
+      "step": 263
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.9703990188021605e-06,
+      "loss": 0.9914,
+      "step": 264
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.970166258622654e-06,
+      "loss": 0.9768,
+      "step": 265
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.969932596088148e-06,
+      "loss": 0.973,
+      "step": 266
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.969698031342059e-06,
+      "loss": 1.0215,
+      "step": 267
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.969462564528357e-06,
+      "loss": 0.9886,
+      "step": 268
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.969226195791567e-06,
+      "loss": 0.9781,
+      "step": 269
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.9689889252767666e-06,
+      "loss": 0.9462,
+      "step": 270
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9687507531295865e-06,
+      "loss": 0.8742,
+      "step": 271
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9685116794962124e-06,
+      "loss": 0.9217,
+      "step": 272
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.968271704523382e-06,
+      "loss": 0.9148,
+      "step": 273
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9680308283583862e-06,
+      "loss": 0.9101,
+      "step": 274
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9677890511490693e-06,
+      "loss": 0.8342,
+      "step": 275
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.9675463730438292e-06,
+      "loss": 0.9157,
+      "step": 276
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.967302794191616e-06,
+      "loss": 0.8935,
+      "step": 277
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.967058314741933e-06,
+      "loss": 0.8717,
+      "step": 278
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9668129348448364e-06,
+      "loss": 0.8871,
+      "step": 279
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.9665666546509343e-06,
+      "loss": 0.9255,
+      "step": 280
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 1.1810663938522339,
+      "eval_runtime": 1.8181,
+      "eval_samples_per_second": 36.301,
+      "eval_steps_per_second": 3.3,
+      "step": 280
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.966319474311389e-06,
+      "loss": 0.9082,
+      "step": 281
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.9660713939779124e-06,
+      "loss": 0.8678,
+      "step": 282
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.9658224138027724e-06,
+      "loss": 0.9226,
+      "step": 283
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.9655725339387863e-06,
+      "loss": 0.8845,
+      "step": 284
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.965321754539326e-06,
+      "loss": 0.9012,
+      "step": 285
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.9650700757583128e-06,
+      "loss": 0.8879,
+      "step": 286
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.964817497750222e-06,
+      "loss": 0.9228,
+      "step": 287
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.9645640206700813e-06,
+      "loss": 0.8736,
+      "step": 288
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.964309644673468e-06,
+      "loss": 0.9326,
+      "step": 289
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.964054369916512e-06,
+      "loss": 0.9154,
+      "step": 290
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.963798196555897e-06,
+      "loss": 0.8924,
+      "step": 291
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.9635411247488552e-06,
+      "loss": 0.9403,
+      "step": 292
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.9632831546531714e-06,
+      "loss": 0.9281,
+      "step": 293
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.963024286427182e-06,
+      "loss": 0.9068,
+      "step": 294
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.9627645202297753e-06,
+      "loss": 0.9293,
+      "step": 295
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.962503856220389e-06,
+      "loss": 0.8758,
+      "step": 296
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9622422945590135e-06,
+      "loss": 0.921,
+      "step": 297
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.9619798354061883e-06,
+      "loss": 0.8321,
+      "step": 298
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.9617164789230065e-06,
+      "loss": 0.9375,
+      "step": 299
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.9614522252711087e-06,
+      "loss": 0.8804,
+      "step": 300
+    },
+    {
+      "epoch": 2.23,
+      "eval_loss": 1.1792281866073608,
+      "eval_runtime": 1.8186,
+      "eval_samples_per_second": 36.291,
+      "eval_steps_per_second": 3.299,
+      "step": 300
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.9611870746126895e-06,
+      "loss": 0.8897,
+      "step": 301
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.9609210271104917e-06,
+      "loss": 0.9348,
+      "step": 302
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.9606540829278085e-06,
+      "loss": 0.9007,
+      "step": 303
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.960386242228486e-06,
+      "loss": 0.9271,
+      "step": 304
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.9601175051769174e-06,
+      "loss": 0.9047,
+      "step": 305
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9598478719380476e-06,
+      "loss": 0.9584,
+      "step": 306
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.9595773426773716e-06,
+      "loss": 0.959,
+      "step": 307
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.959305917560934e-06,
+      "loss": 0.8712,
+      "step": 308
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.9590335967553305e-06,
+      "loss": 0.973,
+      "step": 309
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.958760380427704e-06,
+      "loss": 0.9269,
+      "step": 310
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.9584862687457497e-06,
+      "loss": 0.8938,
+      "step": 311
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.9582112618777103e-06,
+      "loss": 0.9104,
+      "step": 312
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.9579353599923797e-06,
+      "loss": 0.8876,
+      "step": 313
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.9576585632590987e-06,
+      "loss": 0.9412,
+      "step": 314
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.9573808718477614e-06,
+      "loss": 0.9158,
+      "step": 315
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.9571022859288065e-06,
+      "loss": 0.8815,
+      "step": 316
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.9568228056732246e-06,
+      "loss": 0.8723,
+      "step": 317
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.9565424312525537e-06,
+      "loss": 0.9395,
+      "step": 318
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.9562611628388823e-06,
+      "loss": 0.8989,
+      "step": 319
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.955979000604846e-06,
+      "loss": 0.8772,
+      "step": 320
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 1.1800633668899536,
+      "eval_runtime": 1.8222,
+      "eval_samples_per_second": 36.221,
+      "eval_steps_per_second": 3.293,
+      "step": 320
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.9556959447236304e-06,
+      "loss": 0.9199,
+      "step": 321
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.9554119953689677e-06,
+      "loss": 0.9176,
+      "step": 322
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.95512715271514e-06,
+      "loss": 0.8644,
+      "step": 323
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.954841416936978e-06,
+      "loss": 0.8844,
+      "step": 324
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.9545547882098596e-06,
+      "loss": 0.9277,
+      "step": 325
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.95426726670971e-06,
+      "loss": 0.9435,
+      "step": 326
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.953978852613005e-06,
+      "loss": 0.9162,
+      "step": 327
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.9536895460967656e-06,
+      "loss": 0.9453,
+      "step": 328
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.9533993473385623e-06,
+      "loss": 0.9089,
+      "step": 329
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.9531082565165123e-06,
+      "loss": 0.8852,
+      "step": 330
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.9528162738092802e-06,
+      "loss": 0.9373,
+      "step": 331
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.952523399396079e-06,
+      "loss": 0.9493,
+      "step": 332
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.9522296334566686e-06,
+      "loss": 0.8737,
+      "step": 333
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.951934976171355e-06,
+      "loss": 0.8637,
+      "step": 334
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.9516394277209934e-06,
+      "loss": 0.8858,
+      "step": 335
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.9513429882869838e-06,
+      "loss": 0.9192,
+      "step": 336
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.9510456580512747e-06,
+      "loss": 0.9092,
+      "step": 337
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.9507474371963604e-06,
+      "loss": 0.916,
+      "step": 338
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.950448325905282e-06,
+      "loss": 0.8717,
+      "step": 339
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.950148324361628e-06,
+      "loss": 0.9222,
+      "step": 340
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 1.1759049892425537,
+      "eval_runtime": 1.8186,
+      "eval_samples_per_second": 36.292,
+      "eval_steps_per_second": 3.299,
+      "step": 340
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.9498474327495316e-06,
+      "loss": 0.9099,
+      "step": 341
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.949545651253674e-06,
+      "loss": 0.915,
+      "step": 342
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.949242980059282e-06,
+      "loss": 0.9461,
+      "step": 343
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.9489394193521275e-06,
+      "loss": 0.9263,
+      "step": 344
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9486349693185307e-06,
+      "loss": 0.8847,
+      "step": 345
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.9483296301453546e-06,
+      "loss": 0.8376,
+      "step": 346
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.9480234020200103e-06,
+      "loss": 0.9563,
+      "step": 347
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.947716285130454e-06,
+      "loss": 0.9451,
+      "step": 348
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.947408279665187e-06,
+      "loss": 0.8749,
+      "step": 349
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.9470993858132553e-06,
+      "loss": 0.9064,
+      "step": 350
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.946789603764252e-06,
+      "loss": 0.9351,
+      "step": 351
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.946478933708314e-06,
+      "loss": 0.9077,
+      "step": 352
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.9461673758361238e-06,
+      "loss": 0.8665,
+      "step": 353
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.9458549303389083e-06,
+      "loss": 0.9842,
+      "step": 354
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.9455415974084393e-06,
+      "loss": 0.9132,
+      "step": 355
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.945227377237034e-06,
+      "loss": 0.9119,
+      "step": 356
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.944912270017554e-06,
+      "loss": 0.9152,
+      "step": 357
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.9445962759434044e-06,
+      "loss": 0.9201,
+      "step": 358
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.944279395208535e-06,
+      "loss": 0.9165,
+      "step": 359
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.943961628007441e-06,
+      "loss": 0.9308,
+      "step": 360
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 1.1741434335708618,
+      "eval_runtime": 1.8188,
+      "eval_samples_per_second": 36.287,
+      "eval_steps_per_second": 3.299,
+      "step": 360
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.94364297453516e-06,
+      "loss": 0.9098,
+      "step": 361
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.943323434987275e-06,
+      "loss": 0.9378,
+      "step": 362
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.9430030095599115e-06,
+      "loss": 0.9383,
+      "step": 363
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.9426816984497395e-06,
+      "loss": 0.9176,
+      "step": 364
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.942359501853973e-06,
+      "loss": 0.8775,
+      "step": 365
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.9420364199703686e-06,
+      "loss": 0.8818,
+      "step": 366
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.9417124529972267e-06,
+      "loss": 0.9062,
+      "step": 367
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.9413876011333917e-06,
+      "loss": 0.8946,
+      "step": 368
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.9410618645782493e-06,
+      "loss": 0.8951,
+      "step": 369
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.94073524353173e-06,
+      "loss": 0.9068,
+      "step": 370
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.940407738194306e-06,
+      "loss": 0.8476,
+      "step": 371
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.9400793487669935e-06,
+      "loss": 0.9489,
+      "step": 372
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.9397500754513496e-06,
+      "loss": 0.9629,
+      "step": 373
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.9394199184494753e-06,
+      "loss": 0.9186,
+      "step": 374
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.9390888779640134e-06,
+      "loss": 0.8805,
+      "step": 375
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.938756954198149e-06,
+      "loss": 0.9264,
+      "step": 376
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.93842414735561e-06,
+      "loss": 0.9511,
+      "step": 377
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.9380904576406655e-06,
+      "loss": 0.8855,
+      "step": 378
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.937755885258126e-06,
+      "loss": 0.9003,
+      "step": 379
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.937420430413346e-06,
+      "loss": 0.9201,
+      "step": 380
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 1.1738792657852173,
+      "eval_runtime": 1.8202,
+      "eval_samples_per_second": 36.261,
+      "eval_steps_per_second": 3.296,
+      "step": 380
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.937084093312219e-06,
+      "loss": 0.8612,
+      "step": 381
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.9367468741611812e-06,
+      "loss": 0.9238,
+      "step": 382
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.9364087731672106e-06,
+      "loss": 0.9611,
+      "step": 383
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.9360697905378257e-06,
+      "loss": 0.9479,
+      "step": 384
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.9357299264810864e-06,
+      "loss": 0.9174,
+      "step": 385
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.935389181205594e-06,
+      "loss": 0.931,
+      "step": 386
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.9350475549204895e-06,
+      "loss": 0.8504,
+      "step": 387
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.9347050478354564e-06,
+      "loss": 0.883,
+      "step": 388
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.9343616601607174e-06,
+      "loss": 0.9005,
+      "step": 389
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.934017392107036e-06,
+      "loss": 0.9269,
+      "step": 390
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.9336722438857163e-06,
+      "loss": 0.9167,
+      "step": 391
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.933326215708602e-06,
+      "loss": 0.9332,
+      "step": 392
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.9329793077880787e-06,
+      "loss": 0.9023,
+      "step": 393
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.9326315203370696e-06,
+      "loss": 0.9156,
+      "step": 394
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.932282853569039e-06,
+      "loss": 0.9146,
+      "step": 395
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.931933307697991e-06,
+      "loss": 0.913,
+      "step": 396
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.9315828829384693e-06,
+      "loss": 0.9187,
+      "step": 397
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.9312315795055557e-06,
+      "loss": 0.9125,
+      "step": 398
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.9308793976148734e-06,
+      "loss": 0.8878,
+      "step": 399
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.9305263374825833e-06,
+      "loss": 0.8929,
+      "step": 400
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 1.1716059446334839,
+      "eval_runtime": 1.8168,
+      "eval_samples_per_second": 36.327,
+      "eval_steps_per_second": 3.302,
+      "step": 400
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.9301723993253854e-06,
+      "loss": 0.9183,
+      "step": 401
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.9298175833605195e-06,
+      "loss": 0.8823,
+      "step": 402
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.9294618898057636e-06,
+      "loss": 0.933,
+      "step": 403
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.929105318879434e-06,
+      "loss": 0.942,
+      "step": 404
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.9287478708003862e-06,
+      "loss": 0.8084,
+      "step": 405
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.9283895457880145e-06,
+      "loss": 0.8425,
+      "step": 406
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.9280303440622494e-06,
+      "loss": 0.8238,
+      "step": 407
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.9276702658435618e-06,
+      "loss": 0.8587,
+      "step": 408
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.9273093113529594e-06,
+      "loss": 0.8039,
+      "step": 409
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.9269474808119875e-06,
+      "loss": 0.8314,
+      "step": 410
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.9265847744427307e-06,
+      "loss": 0.7722,
+      "step": 411
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.9262211924678085e-06,
+      "loss": 0.8741,
+      "step": 412
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.9258567351103803e-06,
+      "loss": 0.8243,
+      "step": 413
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.9254914025941417e-06,
+      "loss": 0.8435,
+      "step": 414
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 2.9251251951433254e-06,
+      "loss": 0.8184,
+      "step": 415
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.924758112982702e-06,
+      "loss": 0.8221,
+      "step": 416
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.9243901563375767e-06,
+      "loss": 0.8289,
+      "step": 417
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 2.9240213254337944e-06,
+      "loss": 0.8236,
+      "step": 418
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.9236516204977347e-06,
+      "loss": 0.7968,
+      "step": 419
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.9232810417563146e-06,
+      "loss": 0.8567,
+      "step": 420
+    },
+    {
+      "epoch": 3.12,
+      "eval_loss": 1.183070182800293,
+      "eval_runtime": 1.8188,
+      "eval_samples_per_second": 36.289,
+      "eval_steps_per_second": 3.299,
+      "step": 420
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.9229095894369864e-06,
+      "loss": 0.8254,
+      "step": 421
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 2.92253726376774e-06,
+      "loss": 0.8831,
+      "step": 422
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 2.9221640649770993e-06,
+      "loss": 0.8665,
+      "step": 423
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.921789993294127e-06,
+      "loss": 0.8179,
+      "step": 424
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.9214150489484178e-06,
+      "loss": 0.8047,
+      "step": 425
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 2.921039232170106e-06,
+      "loss": 0.8421,
+      "step": 426
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 2.9206625431898586e-06,
+      "loss": 0.8374,
+      "step": 427
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.9202849822388787e-06,
+      "loss": 0.8681,
+      "step": 428
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.9199065495489046e-06,
+      "loss": 0.7846,
+      "step": 429
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 2.9195272453522113e-06,
+      "loss": 0.8603,
+      "step": 430
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 2.9191470698816055e-06,
+      "loss": 0.8215,
+      "step": 431
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.9187660233704304e-06,
+      "loss": 0.8055,
+      "step": 432
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.9183841060525648e-06,
+      "loss": 0.8492,
+      "step": 433
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 2.9180013181624204e-06,
+      "loss": 0.8341,
+      "step": 434
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 2.917617659934944e-06,
+      "loss": 0.8921,
+      "step": 435
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 2.917233131605616e-06,
+      "loss": 0.8003,
+      "step": 436
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 2.9168477334104514e-06,
+      "loss": 0.8381,
+      "step": 437
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 2.9164614655859992e-06,
+      "loss": 0.8521,
+      "step": 438
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 2.916074328369342e-06,
+      "loss": 0.8077,
+      "step": 439
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 2.915686321998095e-06,
+      "loss": 0.7917,
+      "step": 440
+    },
+    {
+      "epoch": 3.27,
+      "eval_loss": 1.187814712524414,
+      "eval_runtime": 1.8205,
+      "eval_samples_per_second": 36.254,
+      "eval_steps_per_second": 3.296,
+      "step": 440
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 2.915297446710409e-06,
+      "loss": 0.8763,
+      "step": 441
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 2.9149077027449668e-06,
+      "loss": 0.8379,
+      "step": 442
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 2.9145170903409834e-06,
+      "loss": 0.7889,
+      "step": 443
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 2.9141256097382085e-06,
+      "loss": 0.8686,
+      "step": 444
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 2.9137332611769244e-06,
+      "loss": 0.7992,
+      "step": 445
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 2.9133400448979453e-06,
+      "loss": 0.8446,
+      "step": 446
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 2.9129459611426186e-06,
+      "loss": 0.841,
+      "step": 447
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 2.912551010152825e-06,
+      "loss": 0.836,
+      "step": 448
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 2.9121551921709752e-06,
+      "loss": 0.8471,
+      "step": 449
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 2.9117585074400147e-06,
+      "loss": 0.8309,
+      "step": 450
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 2.911360956203418e-06,
+      "loss": 0.8481,
+      "step": 451
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 2.9109625387051947e-06,
+      "loss": 0.8503,
+      "step": 452
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 2.910563255189883e-06,
+      "loss": 0.8729,
+      "step": 453
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 2.9101631059025564e-06,
+      "loss": 0.7768,
+      "step": 454
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 2.909762091088815e-06,
+      "loss": 0.7945,
+      "step": 455
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 2.909360210994794e-06,
+      "loss": 0.8585,
+      "step": 456
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 2.9089574658671586e-06,
+      "loss": 0.861,
+      "step": 457
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 2.908553855953104e-06,
+      "loss": 0.8237,
+      "step": 458
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.9081493815003572e-06,
+      "loss": 0.8691,
+      "step": 459
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.9077440427571755e-06,
+      "loss": 0.8339,
+      "step": 460
+    },
+    {
+      "epoch": 3.41,
+      "eval_loss": 1.1889363527297974,
+      "eval_runtime": 1.8181,
+      "eval_samples_per_second": 36.302,
+      "eval_steps_per_second": 3.3,
+      "step": 460
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.9073378399723466e-06,
+      "loss": 0.8687,
+      "step": 461
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 2.906930773395189e-06,
+      "loss": 0.8403,
+      "step": 462
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.906522843275551e-06,
+      "loss": 0.8369,
+      "step": 463
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.90611404986381e-06,
+      "loss": 0.8479,
+      "step": 464
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.905704393410875e-06,
+      "loss": 0.8457,
+      "step": 465
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.905293874168184e-06,
+      "loss": 0.8487,
+      "step": 466
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.9048824923877047e-06,
+      "loss": 0.8231,
+      "step": 467
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.9044702483219335e-06,
+      "loss": 0.8109,
+      "step": 468
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.904057142223896e-06,
+      "loss": 0.8741,
+      "step": 469
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.9036431743471486e-06,
+      "loss": 0.8368,
+      "step": 470
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.9032283449457744e-06,
+      "loss": 0.8265,
+      "step": 471
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.9028126542743878e-06,
+      "loss": 0.8066,
+      "step": 472
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.902396102588129e-06,
+      "loss": 0.8475,
+      "step": 473
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.9019786901426686e-06,
+      "loss": 0.8404,
+      "step": 474
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.9015604171942054e-06,
+      "loss": 0.8632,
+      "step": 475
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.9011412839994646e-06,
+      "loss": 0.8267,
+      "step": 476
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.900721290815702e-06,
+      "loss": 0.8433,
+      "step": 477
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.9003004379007e-06,
+      "loss": 0.8693,
+      "step": 478
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.8998787255127677e-06,
+      "loss": 0.8366,
+      "step": 479
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.899456153910743e-06,
+      "loss": 0.8501,
+      "step": 480
+    },
+    {
+      "epoch": 3.56,
+      "eval_loss": 1.1901904344558716,
+      "eval_runtime": 1.8181,
+      "eval_samples_per_second": 36.302,
+      "eval_steps_per_second": 3.3,
+      "step": 480
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.899032723353992e-06,
+      "loss": 0.8661,
+      "step": 481
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.898608434102405e-06,
+      "loss": 0.8229,
+      "step": 482
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.8981832864164024e-06,
+      "loss": 0.8757,
+      "step": 483
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.89775728055693e-06,
+      "loss": 0.8651,
+      "step": 484
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.8973304167854613e-06,
+      "loss": 0.8368,
+      "step": 485
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.896902695363994e-06,
+      "loss": 0.8438,
+      "step": 486
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.896474116555056e-06,
+      "loss": 0.8822,
+      "step": 487
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.8960446806216974e-06,
+      "loss": 0.8484,
+      "step": 488
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 2.8956143878274975e-06,
+      "loss": 0.8135,
+      "step": 489
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.89518323843656e-06,
+      "loss": 0.863,
+      "step": 490
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.8947512327135146e-06,
+      "loss": 0.788,
+      "step": 491
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 2.894318370923517e-06,
+      "loss": 0.8344,
+      "step": 492
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 2.893884653332248e-06,
+      "loss": 0.8803,
+      "step": 493
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 2.8934500802059134e-06,
+      "loss": 0.8398,
+      "step": 494
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 2.893014651811245e-06,
+      "loss": 0.8352,
+      "step": 495
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 2.8925783684154975e-06,
+      "loss": 0.8328,
+      "step": 496
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 2.892141230286454e-06,
+      "loss": 0.8397,
+      "step": 497
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.891703237692418e-06,
+      "loss": 0.8407,
+      "step": 498
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.8912643909022218e-06,
+      "loss": 0.8391,
+      "step": 499
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 2.8908246901852172e-06,
+      "loss": 0.9057,
+      "step": 500
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 1.187972068786621,
+      "eval_runtime": 1.8205,
+      "eval_samples_per_second": 36.253,
+      "eval_steps_per_second": 3.296,
+      "step": 500
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 2.890384135811284e-06,
+      "loss": 0.8627,
+      "step": 501
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 2.8899427280508246e-06,
+      "loss": 0.8467,
+      "step": 502
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 2.8895004671747646e-06,
+      "loss": 0.8586,
+      "step": 503
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 2.889057353454554e-06,
+      "loss": 0.8447,
+      "step": 504
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 2.8886133871621663e-06,
+      "loss": 0.8052,
+      "step": 505
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 2.888168568570098e-06,
+      "loss": 0.8176,
+      "step": 506
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 2.8877228979513686e-06,
+      "loss": 0.8516,
+      "step": 507
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 2.8872763755795207e-06,
+      "loss": 0.863,
+      "step": 508
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.88682900172862e-06,
+      "loss": 0.8713,
+      "step": 509
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.8863807766732537e-06,
+      "loss": 0.8343,
+      "step": 510
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 2.8859317006885327e-06,
+      "loss": 0.8339,
+      "step": 511
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 2.88548177405009e-06,
+      "loss": 0.8455,
+      "step": 512
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.8850309970340804e-06,
+      "loss": 0.8231,
+      "step": 513
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.8845793699171807e-06,
+      "loss": 0.8068,
+      "step": 514
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 2.884126892976589e-06,
+      "loss": 0.8556,
+      "step": 515
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 2.883673566490026e-06,
+      "loss": 0.7879,
+      "step": 516
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.8832193907357333e-06,
+      "loss": 0.8943,
+      "step": 517
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.8827643659924736e-06,
+      "loss": 0.8528,
+      "step": 518
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 2.8823084925395306e-06,
+      "loss": 0.8567,
+      "step": 519
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.8818517706567094e-06,
+      "loss": 0.8054,
+      "step": 520
+    },
+    {
+      "epoch": 3.86,
+      "eval_loss": 1.1861604452133179,
+      "eval_runtime": 1.8192,
+      "eval_samples_per_second": 36.28,
+      "eval_steps_per_second": 3.298,
+      "step": 520
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.8813942006243363e-06,
+      "loss": 0.8382,
+      "step": 521
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.8809357827232562e-06,
+      "loss": 0.8485,
+      "step": 522
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.880476517234836e-06,
+      "loss": 0.8438,
+      "step": 523
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.880016404440964e-06,
+      "loss": 0.85,
+      "step": 524
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.8795554446240453e-06,
+      "loss": 0.8226,
+      "step": 525
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.8790936380670076e-06,
+      "loss": 0.8737,
+      "step": 526
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.878630985053297e-06,
+      "loss": 0.8645,
+      "step": 527
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 2.8781674858668797e-06,
+      "loss": 0.8486,
+      "step": 528
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.877703140792241e-06,
+      "loss": 0.8135,
+      "step": 529
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.8772379501143854e-06,
+      "loss": 0.8088,
+      "step": 530
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.8767719141188367e-06,
+      "loss": 0.82,
+      "step": 531
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.8763050330916374e-06,
+      "loss": 0.8318,
+      "step": 532
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.8758373073193483e-06,
+      "loss": 0.8846,
+      "step": 533
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.8753687370890486e-06,
+      "loss": 0.8377,
+      "step": 534
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 2.874899322688337e-06,
+      "loss": 0.842,
+      "step": 535
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.8744290644053293e-06,
+      "loss": 0.8193,
+      "step": 536
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.873957962528659e-06,
+      "loss": 0.8064,
+      "step": 537
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.8734860173474784e-06,
+      "loss": 0.8503,
+      "step": 538
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.8730132291514566e-06,
+      "loss": 0.8151,
+      "step": 539
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 2.8725395982307804e-06,
+      "loss": 0.7569,
+      "step": 540
+    },
+    {
+      "epoch": 4.01,
+      "eval_loss": 1.1860952377319336,
+      "eval_runtime": 1.8071,
+      "eval_samples_per_second": 36.523,
+      "eval_steps_per_second": 3.32,
+      "step": 540
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 2.872065124876153e-06,
+      "loss": 0.7851,
+      "step": 541
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 2.8715898093787973e-06,
+      "loss": 0.7192,
+      "step": 542
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 2.87111365203045e-06,
+      "loss": 0.7889,
+      "step": 543
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 2.8706366531233652e-06,
+      "loss": 0.785,
+      "step": 544
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 2.8701588129503156e-06,
+      "loss": 0.8169,
+      "step": 545
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 2.8696801318045875e-06,
+      "loss": 0.8021,
+      "step": 546
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 2.8692006099799858e-06,
+      "loss": 0.7271,
+      "step": 547
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 2.8687202477708292e-06,
+      "loss": 0.7666,
+      "step": 548
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 2.8682390454719536e-06,
+      "loss": 0.7613,
+      "step": 549
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 2.8677570033787104e-06,
+      "loss": 0.7508,
+      "step": 550
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 2.8672741217869657e-06,
+      "loss": 0.735,
+      "step": 551
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.8667904009931022e-06,
+      "loss": 0.7721,
+      "step": 552
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.866305841294015e-06,
+      "loss": 0.7428,
+      "step": 553
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.8658204429871186e-06,
+      "loss": 0.7504,
+      "step": 554
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.8653342063703377e-06,
+      "loss": 0.7682,
+      "step": 555
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.8648471317421136e-06,
+      "loss": 0.7479,
+      "step": 556
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.864359219401402e-06,
+      "loss": 0.7153,
+      "step": 557
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.8638704696476737e-06,
+      "loss": 0.7623,
+      "step": 558
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.86338088278091e-06,
+      "loss": 0.7506,
+      "step": 559
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.86289045910161e-06,
+      "loss": 0.7859,
+      "step": 560
+    },
+    {
+      "epoch": 4.16,
+      "eval_loss": 1.2153370380401611,
+      "eval_runtime": 1.8176,
+      "eval_samples_per_second": 36.311,
+      "eval_steps_per_second": 3.301,
+      "step": 560
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.862399198910784e-06,
+      "loss": 0.7643,
+      "step": 561
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.8619071025099572e-06,
+      "loss": 0.778,
+      "step": 562
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.8614141702011663e-06,
+      "loss": 0.736,
+      "step": 563
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.8609204022869626e-06,
+      "loss": 0.7119,
+      "step": 564
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.8604257990704103e-06,
+      "loss": 0.768,
+      "step": 565
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.8599303608550844e-06,
+      "loss": 0.7578,
+      "step": 566
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.859434087945075e-06,
+      "loss": 0.7777,
+      "step": 567
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.8589369806449827e-06,
+      "loss": 0.7424,
+      "step": 568
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.858439039259921e-06,
+      "loss": 0.7431,
+      "step": 569
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.8579402640955145e-06,
+      "loss": 0.7619,
+      "step": 570
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.8574406554579007e-06,
+      "loss": 0.7321,
+      "step": 571
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.8569402136537284e-06,
+      "loss": 0.7867,
+      "step": 572
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.8564389389901573e-06,
+      "loss": 0.7177,
+      "step": 573
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.855936831774858e-06,
+      "loss": 0.7104,
+      "step": 574
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.8554338923160135e-06,
+      "loss": 0.7831,
+      "step": 575
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.854930120922316e-06,
+      "loss": 0.7475,
+      "step": 576
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.8544255179029696e-06,
+      "loss": 0.7741,
+      "step": 577
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.853920083567688e-06,
+      "loss": 0.6904,
+      "step": 578
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.8534138182266953e-06,
+      "loss": 0.7327,
+      "step": 579
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.8529067221907263e-06,
+      "loss": 0.6936,
+      "step": 580
+    },
+    {
+      "epoch": 4.3,
+      "eval_loss": 1.2218281030654907,
+      "eval_runtime": 1.8216,
+      "eval_samples_per_second": 36.231,
+      "eval_steps_per_second": 3.294,
+      "step": 580
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.8523987957710244e-06,
+      "loss": 0.7582,
+      "step": 581
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 2.8518900392793444e-06,
+      "loss": 0.7545,
+      "step": 582
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.8513804530279488e-06,
+      "loss": 0.7001,
+      "step": 583
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.8508700373296103e-06,
+      "loss": 0.7773,
+      "step": 584
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.850358792497611e-06,
+      "loss": 0.7796,
+      "step": 585
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.8498467188457405e-06,
+      "loss": 0.7229,
+      "step": 586
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.8493338166882996e-06,
+      "loss": 0.7839,
+      "step": 587
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.8488200863400947e-06,
+      "loss": 0.7641,
+      "step": 588
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.848305528116443e-06,
+      "loss": 0.779,
+      "step": 589
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.8477901423331682e-06,
+      "loss": 0.7724,
+      "step": 590
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.8472739293066022e-06,
+      "loss": 0.7466,
+      "step": 591
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.8467568893535856e-06,
+      "loss": 0.7203,
+      "step": 592
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.8462390227914657e-06,
+      "loss": 0.7844,
+      "step": 593
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 2.8457203299380975e-06,
+      "loss": 0.7788,
+      "step": 594
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.8452008111118426e-06,
+      "loss": 0.7802,
+      "step": 595
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.8446804666315703e-06,
+      "loss": 0.8017,
+      "step": 596
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.8441592968166566e-06,
+      "loss": 0.7466,
+      "step": 597
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.843637301986983e-06,
+      "loss": 0.7519,
+      "step": 598
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.843114482462939e-06,
+      "loss": 0.7728,
+      "step": 599
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.8425908385654183e-06,
+      "loss": 0.7902,
+      "step": 600
+    },
+    {
+      "epoch": 4.45,
+      "eval_loss": 1.2204654216766357,
+      "eval_runtime": 1.8205,
+      "eval_samples_per_second": 36.255,
+      "eval_steps_per_second": 3.296,
+      "step": 600
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.8420663706158235e-06,
+      "loss": 0.7691,
+      "step": 601
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8415410789360603e-06,
+      "loss": 0.7997,
+      "step": 602
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8410149638485404e-06,
+      "loss": 0.7391,
+      "step": 603
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.8404880256761824e-06,
+      "loss": 0.791,
+      "step": 604
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.8399602647424085e-06,
+      "loss": 0.7822,
+      "step": 605
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.8394316813711465e-06,
+      "loss": 0.7549,
+      "step": 606
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.8389022758868295e-06,
+      "loss": 0.7556,
+      "step": 607
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.838372048614394e-06,
+      "loss": 0.7729,
+      "step": 608
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.8378409998792824e-06,
+      "loss": 0.7575,
+      "step": 609
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.8373091300074393e-06,
+      "loss": 0.7573,
+      "step": 610
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.836776439325316e-06,
+      "loss": 0.7338,
+      "step": 611
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.8362429281598647e-06,
+      "loss": 0.7335,
+      "step": 612
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.835708596838543e-06,
+      "loss": 0.8197,
+      "step": 613
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.8351734456893117e-06,
+      "loss": 0.7576,
+      "step": 614
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.8346374750406343e-06,
+      "loss": 0.7791,
+      "step": 615
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.834100685221478e-06,
+      "loss": 0.7716,
+      "step": 616
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.833563076561312e-06,
+      "loss": 0.7862,
+      "step": 617
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.833024649390108e-06,
+      "loss": 0.8044,
+      "step": 618
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.8324854040383405e-06,
+      "loss": 0.7762,
+      "step": 619
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.8319453408369872e-06,
+      "loss": 0.7401,
+      "step": 620
+    },
+    {
+      "epoch": 4.6,
+      "eval_loss": 1.2208995819091797,
+      "eval_runtime": 1.8182,
+      "eval_samples_per_second": 36.3,
+      "eval_steps_per_second": 3.3,
+      "step": 620
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.8314044601175256e-06,
+      "loss": 0.7116,
+      "step": 621
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.830862762211937e-06,
+      "loss": 0.7519,
+      "step": 622
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.8303202474527024e-06,
+      "loss": 0.7387,
+      "step": 623
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.8297769161728063e-06,
+      "loss": 0.7334,
+      "step": 624
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.829232768705733e-06,
+      "loss": 0.7841,
+      "step": 625
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.828687805385467e-06,
+      "loss": 0.7611,
+      "step": 626
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.8281420265464958e-06,
+      "loss": 0.7973,
+      "step": 627
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.827595432523805e-06,
+      "loss": 0.7236,
+      "step": 628
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.827048023652883e-06,
+      "loss": 0.7949,
+      "step": 629
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.8264998002697157e-06,
+      "loss": 0.8072,
+      "step": 630
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.825950762710791e-06,
+      "loss": 0.7858,
+      "step": 631
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.825400911313096e-06,
+      "loss": 0.7373,
+      "step": 632
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.824850246414117e-06,
+      "loss": 0.7478,
+      "step": 633
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.8242987683518393e-06,
+      "loss": 0.7974,
+      "step": 634
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.8237464774647484e-06,
+      "loss": 0.7588,
+      "step": 635
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.823193374091828e-06,
+      "loss": 0.7821,
+      "step": 636
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.8226394585725596e-06,
+      "loss": 0.7155,
+      "step": 637
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.822084731246925e-06,
+      "loss": 0.7535,
+      "step": 638
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.8215291924554026e-06,
+      "loss": 0.773,
+      "step": 639
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.820972842538971e-06,
+      "loss": 0.7765,
+      "step": 640
+    },
+    {
+      "epoch": 4.75,
+      "eval_loss": 1.2198340892791748,
+      "eval_runtime": 1.8178,
+      "eval_samples_per_second": 36.307,
+      "eval_steps_per_second": 3.301,
+      "step": 640
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.8204156818391036e-06,
+      "loss": 0.7209,
+      "step": 641
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.8198577106977747e-06,
+      "loss": 0.8059,
+      "step": 642
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.8192989294574525e-06,
+      "loss": 0.798,
+      "step": 643
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.8187393384611065e-06,
+      "loss": 0.8152,
+      "step": 644
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.8181789380521997e-06,
+      "loss": 0.7751,
+      "step": 645
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.8176177285746935e-06,
+      "loss": 0.7562,
+      "step": 646
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.8170557103730458e-06,
+      "loss": 0.7846,
+      "step": 647
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.816492883792211e-06,
+      "loss": 0.7539,
+      "step": 648
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.815929249177639e-06,
+      "loss": 0.7744,
+      "step": 649
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.8153648068752758e-06,
+      "loss": 0.7163,
+      "step": 650
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.8147995572315645e-06,
+      "loss": 0.7658,
+      "step": 651
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.8142335005934415e-06,
+      "loss": 0.7707,
+      "step": 652
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.81366663730834e-06,
+      "loss": 0.7529,
+      "step": 653
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.8130989677241886e-06,
+      "loss": 0.7494,
+      "step": 654
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.8125304921894087e-06,
+      "loss": 0.7971,
+      "step": 655
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.811961211052919e-06,
+      "loss": 0.7513,
+      "step": 656
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.8113911246641313e-06,
+      "loss": 0.7658,
+      "step": 657
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.8108202333729514e-06,
+      "loss": 0.8192,
+      "step": 658
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.8102485375297797e-06,
+      "loss": 0.7819,
+      "step": 659
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.8096760374855106e-06,
+      "loss": 0.7259,
+      "step": 660
+    },
+    {
+      "epoch": 4.9,
+      "eval_loss": 1.221872091293335,
+      "eval_runtime": 1.8224,
+      "eval_samples_per_second": 36.217,
+      "eval_steps_per_second": 3.292,
+      "step": 660
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.809102733591531e-06,
+      "loss": 0.7747,
+      "step": 661
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.8085286261997225e-06,
+      "loss": 0.8154,
+      "step": 662
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.8079537156624594e-06,
+      "loss": 0.7603,
+      "step": 663
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.8073780023326083e-06,
+      "loss": 0.7799,
+      "step": 664
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.80680148656353e-06,
+      "loss": 0.7635,
+      "step": 665
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.806224168709076e-06,
+      "loss": 0.7436,
+      "step": 666
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.805646049123592e-06,
+      "loss": 0.7616,
+      "step": 667
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.8050671281619135e-06,
+      "loss": 0.7917,
+      "step": 668
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.80448740617937e-06,
+      "loss": 0.7962,
+      "step": 669
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.803906883531782e-06,
+      "loss": 0.7094,
+      "step": 670
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.8033255605754614e-06,
+      "loss": 0.7495,
+      "step": 671
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.8027434376672097e-06,
+      "loss": 0.7805,
+      "step": 672
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.8021605151643228e-06,
+      "loss": 0.7491,
+      "step": 673
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.801576793424584e-06,
+      "loss": 0.6993,
+      "step": 674
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.800992272806269e-06,
+      "loss": 0.6452,
+      "step": 675
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.800406953668144e-06,
+      "loss": 0.6885,
+      "step": 676
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.7998208363694635e-06,
+      "loss": 0.6973,
+      "step": 677
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.7992339212699743e-06,
+      "loss": 0.6237,
+      "step": 678
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.7986462087299113e-06,
+      "loss": 0.6818,
+      "step": 679
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.798057699109999e-06,
+      "loss": 0.6637,
+      "step": 680
+    },
+    {
+      "epoch": 5.05,
+      "eval_loss": 1.2875888347625732,
+      "eval_runtime": 1.8174,
+      "eval_samples_per_second": 36.316,
+      "eval_steps_per_second": 3.301,
+      "step": 680
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.7974683927714518e-06,
+      "loss": 0.6542,
+      "step": 681
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.796878290075972e-06,
+      "loss": 0.6758,
+      "step": 682
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.7962873913857527e-06,
+      "loss": 0.6425,
+      "step": 683
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.795695697063473e-06,
+      "loss": 0.6434,
+      "step": 684
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.7951032074723014e-06,
+      "loss": 0.6296,
+      "step": 685
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 2.7945099229758958e-06,
+      "loss": 0.6997,
+      "step": 686
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.7939158439384e-06,
+      "loss": 0.7066,
+      "step": 687
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.7933209707244466e-06,
+      "loss": 0.6288,
+      "step": 688
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.7927253036991558e-06,
+      "loss": 0.627,
+      "step": 689
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 2.7921288432281344e-06,
+      "loss": 0.6843,
+      "step": 690
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.7915315896774756e-06,
+      "loss": 0.6574,
+      "step": 691
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.7909335434137617e-06,
+      "loss": 0.684,
+      "step": 692
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.790334704804059e-06,
+      "loss": 0.6099,
+      "step": 693
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.789735074215921e-06,
+      "loss": 0.6704,
+      "step": 694
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.7891346520173888e-06,
+      "loss": 0.7008,
+      "step": 695
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.7885334385769866e-06,
+      "loss": 0.6867,
+      "step": 696
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.7879314342637266e-06,
+      "loss": 0.6706,
+      "step": 697
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.787328639447105e-06,
+      "loss": 0.6566,
+      "step": 698
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.7867250544971037e-06,
+      "loss": 0.6831,
+      "step": 699
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.78612067978419e-06,
+      "loss": 0.674,
+      "step": 700
+    },
+    {
+      "epoch": 5.19,
+      "eval_loss": 1.2744817733764648,
+      "eval_runtime": 1.8223,
+      "eval_samples_per_second": 36.217,
+      "eval_steps_per_second": 3.292,
+      "step": 700
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.7855155156793152e-06,
+      "loss": 0.6779,
+      "step": 701
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.784909562553916e-06,
+      "loss": 0.6918,
+      "step": 702
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.7843028207799113e-06,
+      "loss": 0.6872,
+      "step": 703
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.7836952907297076e-06,
+      "loss": 0.6717,
+      "step": 704
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.7830869727761916e-06,
+      "loss": 0.6695,
+      "step": 705
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 2.782477867292736e-06,
+      "loss": 0.6827,
+      "step": 706
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.7818679746531955e-06,
+      "loss": 0.6796,
+      "step": 707
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.7812572952319093e-06,
+      "loss": 0.6141,
+      "step": 708
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.780645829403698e-06,
+      "loss": 0.6737,
+      "step": 709
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 2.7800335775438657e-06,
+      "loss": 0.6008,
+      "step": 710
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.7794205400281986e-06,
+      "loss": 0.6289,
+      "step": 711
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.7788067172329663e-06,
+      "loss": 0.6684,
+      "step": 712
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 2.7781921095349187e-06,
+      "loss": 0.6731,
+      "step": 713
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.7775767173112876e-06,
+      "loss": 0.6688,
+      "step": 714
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.776960540939788e-06,
+      "loss": 0.6223,
+      "step": 715
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.7763435807986145e-06,
+      "loss": 0.6793,
+      "step": 716
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 2.775725837266444e-06,
+      "loss": 0.6708,
+      "step": 717
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.775107310722433e-06,
+      "loss": 0.6351,
+      "step": 718
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.7744880015462192e-06,
+      "loss": 0.7171,
+      "step": 719
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.7738679101179206e-06,
+      "loss": 0.6523,
+      "step": 720
+    },
+    {
+      "epoch": 5.34,
+      "eval_loss": 1.2763301134109497,
+      "eval_runtime": 1.8243,
+      "eval_samples_per_second": 36.177,
+      "eval_steps_per_second": 3.289,
+      "step": 720
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.7732470368181355e-06,
+      "loss": 0.6248,
+      "step": 721
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 2.7726253820279427e-06,
+      "loss": 0.6511,
+      "step": 722
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.7720029461288987e-06,
+      "loss": 0.6585,
+      "step": 723
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.7713797295030418e-06,
+      "loss": 0.6567,
+      "step": 724
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.7707557325328872e-06,
+      "loss": 0.6394,
+      "step": 725
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.7701309556014308e-06,
+      "loss": 0.6797,
+      "step": 726
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.7695053990921467e-06,
+      "loss": 0.6857,
+      "step": 727
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.768879063388988e-06,
+      "loss": 0.6931,
+      "step": 728
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.768251948876384e-06,
+      "loss": 0.656,
+      "step": 729
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.7676240559392444e-06,
+      "loss": 0.6659,
+      "step": 730
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.7669953849629563e-06,
+      "loss": 0.6589,
+      "step": 731
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.766365936333383e-06,
+      "loss": 0.7067,
+      "step": 732
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.7657357104368655e-06,
+      "loss": 0.6925,
+      "step": 733
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.7651047076602235e-06,
+      "loss": 0.6485,
+      "step": 734
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.7644729283907516e-06,
+      "loss": 0.6772,
+      "step": 735
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 2.763840373016222e-06,
+      "loss": 0.6834,
+      "step": 736
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.7632070419248826e-06,
+      "loss": 0.6799,
+      "step": 737
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.7625729355054582e-06,
+      "loss": 0.663,
+      "step": 738
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.7619380541471486e-06,
+      "loss": 0.6453,
+      "step": 739
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.7613023982396304e-06,
+      "loss": 0.7129,
+      "step": 740
+    },
+    {
+      "epoch": 5.49,
+      "eval_loss": 1.2798060178756714,
+      "eval_runtime": 1.8197,
+      "eval_samples_per_second": 36.27,
+      "eval_steps_per_second": 3.297,
+      "step": 740
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.7606659681730542e-06,
+      "loss": 0.6825,
+      "step": 741
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.7600287643380466e-06,
+      "loss": 0.6911,
+      "step": 742
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.7593907871257094e-06,
+      "loss": 0.6526,
+      "step": 743
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 2.758752036927618e-06,
+      "loss": 0.675,
+      "step": 744
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.758112514135824e-06,
+      "loss": 0.6495,
+      "step": 745
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.7574722191428512e-06,
+      "loss": 0.69,
+      "step": 746
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.756831152341698e-06,
+      "loss": 0.682,
+      "step": 747
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.756189314125838e-06,
+      "loss": 0.6962,
+      "step": 748
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.7555467048892158e-06,
+      "loss": 0.6268,
+      "step": 749
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.7549033250262514e-06,
+      "loss": 0.6925,
+      "step": 750
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.7542591749318366e-06,
+      "loss": 0.6918,
+      "step": 751
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.7536142550013357e-06,
+      "loss": 0.672,
+      "step": 752
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.752968565630587e-06,
+      "loss": 0.6737,
+      "step": 753
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.7523221072159e-06,
+      "loss": 0.6884,
+      "step": 754
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.7516748801540556e-06,
+      "loss": 0.659,
+      "step": 755
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.7510268848423075e-06,
+      "loss": 0.6645,
+      "step": 756
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.750378121678381e-06,
+      "loss": 0.6648,
+      "step": 757
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.749728591060472e-06,
+      "loss": 0.6639,
+      "step": 758
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.7490782933872486e-06,
+      "loss": 0.6808,
+      "step": 759
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.7484272290578475e-06,
+      "loss": 0.6416,
+      "step": 760
+    },
+    {
+      "epoch": 5.64,
+      "eval_loss": 1.286723017692566,
+      "eval_runtime": 1.8239,
+      "eval_samples_per_second": 36.187,
+      "eval_steps_per_second": 3.29,
+      "step": 760
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.7477753984718783e-06,
+      "loss": 0.6545,
+      "step": 761
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.74712280202942e-06,
+      "loss": 0.7052,
+      "step": 762
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 2.746469440131021e-06,
+      "loss": 0.6579,
+      "step": 763
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.7458153131777007e-06,
+      "loss": 0.6702,
+      "step": 764
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.7451604215709475e-06,
+      "loss": 0.6889,
+      "step": 765
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.744504765712719e-06,
+      "loss": 0.6695,
+      "step": 766
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 2.7438483460054415e-06,
+      "loss": 0.6752,
+      "step": 767
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.7431911628520116e-06,
+      "loss": 0.645,
+      "step": 768
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.742533216655793e-06,
+      "loss": 0.6963,
+      "step": 769
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.7418745078206186e-06,
+      "loss": 0.6819,
+      "step": 770
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.7412150367507884e-06,
+      "loss": 0.677,
+      "step": 771
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 2.7405548038510717e-06,
+      "loss": 0.6759,
+      "step": 772
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.7398938095267043e-06,
+      "loss": 0.6697,
+      "step": 773
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.73923205418339e-06,
+      "loss": 0.6677,
+      "step": 774
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.7385695382272985e-06,
+      "loss": 0.7083,
+      "step": 775
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.737906262065068e-06,
+      "loss": 0.6992,
+      "step": 776
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.737242226103802e-06,
+      "loss": 0.6839,
+      "step": 777
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.736577430751071e-06,
+      "loss": 0.694,
+      "step": 778
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.7359118764149117e-06,
+      "loss": 0.686,
+      "step": 779
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.735245563503826e-06,
+      "loss": 0.7072,
+      "step": 780
+    },
+    {
+      "epoch": 5.79,
+      "eval_loss": 1.2781226634979248,
+      "eval_runtime": 1.8242,
+      "eval_samples_per_second": 36.18,
+      "eval_steps_per_second": 3.289,
+      "step": 780
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.734578492426782e-06,
+      "loss": 0.6937,
+      "step": 781
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.733910663593213e-06,
+      "loss": 0.6864,
+      "step": 782
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.7332420774130175e-06,
+      "loss": 0.6836,
+      "step": 783
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 2.7325727342965584e-06,
+      "loss": 0.7127,
+      "step": 784
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.731902634654663e-06,
+      "loss": 0.6954,
+      "step": 785
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.731231778898624e-06,
+      "loss": 0.6698,
+      "step": 786
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.7305601674401977e-06,
+      "loss": 0.6326,
+      "step": 787
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.7298878006916038e-06,
+      "loss": 0.6747,
+      "step": 788
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.7292146790655257e-06,
+      "loss": 0.6645,
+      "step": 789
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.7285408029751105e-06,
+      "loss": 0.6694,
+      "step": 790
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.7278661728339686e-06,
+      "loss": 0.6771,
+      "step": 791
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.7271907890561725e-06,
+      "loss": 0.676,
+      "step": 792
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.7265146520562572e-06,
+      "loss": 0.6413,
+      "step": 793
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.725837762249221e-06,
+      "loss": 0.658,
+      "step": 794
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.7251601200505238e-06,
+      "loss": 0.6885,
+      "step": 795
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 2.7244817258760858e-06,
+      "loss": 0.6742,
+      "step": 796
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 2.7238025801422915e-06,
+      "loss": 0.668,
+      "step": 797
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.723122683265985e-06,
+      "loss": 0.665,
+      "step": 798
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.722442035664471e-06,
+      "loss": 0.6594,
+      "step": 799
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.7217606377555167e-06,
+      "loss": 0.676,
+      "step": 800
+    },
+    {
+      "epoch": 5.94,
+      "eval_loss": 1.2843071222305298,
+      "eval_runtime": 1.8183,
+      "eval_samples_per_second": 36.297,
+      "eval_steps_per_second": 3.3,
+      "step": 800
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.7210784899573487e-06,
+      "loss": 0.6634,
+      "step": 801
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.720395592688653e-06,
+      "loss": 0.6837,
+      "step": 802
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.719711946368578e-06,
+      "loss": 0.6965,
+      "step": 803
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.719027551416729e-06,
+      "loss": 0.6632,
+      "step": 804
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.718342408253173e-06,
+      "loss": 0.672,
+      "step": 805
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.7176565172984355e-06,
+      "loss": 0.6788,
+      "step": 806
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.716969878973501e-06,
+      "loss": 0.6605,
+      "step": 807
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.716282493699812e-06,
+      "loss": 0.6989,
+      "step": 808
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.7155943618992712e-06,
+      "loss": 0.5779,
+      "step": 809
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.7149054839942377e-06,
+      "loss": 0.5798,
+      "step": 810
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.714215860407529e-06,
+      "loss": 0.6028,
+      "step": 811
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.713525491562421e-06,
+      "loss": 0.5597,
+      "step": 812
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.712834377882647e-06,
+      "loss": 0.5772,
+      "step": 813
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.712142519792397e-06,
+      "loss": 0.5998,
+      "step": 814
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 2.7114499177163164e-06,
+      "loss": 0.5763,
+      "step": 815
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 2.710756572079511e-06,
+      "loss": 0.5854,
+      "step": 816
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 2.710062483307539e-06,
+      "loss": 0.5574,
+      "step": 817
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 2.7093676518264174e-06,
+      "loss": 0.5377,
+      "step": 818
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 2.708672078062618e-06,
+      "loss": 0.5905,
+      "step": 819
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 2.7079757624430682e-06,
+      "loss": 0.5667,
+      "step": 820
+    },
+    {
+      "epoch": 6.09,
+      "eval_loss": 1.3376994132995605,
+      "eval_runtime": 1.8203,
+      "eval_samples_per_second": 36.259,
+      "eval_steps_per_second": 3.296,
+      "step": 820
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 2.707278705395151e-06,
+      "loss": 0.5404,
+      "step": 821
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 2.706580907346704e-06,
+      "loss": 0.5712,
+      "step": 822
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 2.7058823687260205e-06,
+      "loss": 0.5629,
+      "step": 823
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 2.7051830899618473e-06,
+      "loss": 0.4945,
+      "step": 824
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 2.7044830714833867e-06,
+      "loss": 0.5529,
+      "step": 825
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 2.703782313720294e-06,
+      "loss": 0.5552,
+      "step": 826
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 2.7030808171026775e-06,
+      "loss": 0.5416,
+      "step": 827
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 2.7023785820611018e-06,
+      "loss": 0.5451,
+      "step": 828
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 2.7016756090265814e-06,
+      "loss": 0.5744,
+      "step": 829
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 2.7009718984305866e-06,
+      "loss": 0.5541,
+      "step": 830
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 2.7002674507050386e-06,
+      "loss": 0.5757,
+      "step": 831
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 2.6995622662823115e-06,
+      "loss": 0.585,
+      "step": 832
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 2.6988563455952317e-06,
+      "loss": 0.5221,
+      "step": 833
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 2.6981496890770772e-06,
+      "loss": 0.5189,
+      "step": 834
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 2.6974422971615785e-06,
+      "loss": 0.5553,
+      "step": 835
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 2.6967341702829166e-06,
+      "loss": 0.5602,
+      "step": 836
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 2.6960253088757237e-06,
+      "loss": 0.6091,
+      "step": 837
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 2.695315713375083e-06,
+      "loss": 0.5613,
+      "step": 838
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 2.6946053842165284e-06,
+      "loss": 0.561,
+      "step": 839
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 2.693894321836044e-06,
+      "loss": 0.5139,
+      "step": 840
+    },
+    {
+      "epoch": 6.23,
+      "eval_loss": 1.3582661151885986,
+      "eval_runtime": 1.8184,
+      "eval_samples_per_second": 36.296,
+      "eval_steps_per_second": 3.3,
+      "step": 840
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 2.693182526670064e-06,
+      "loss": 0.5701,
+      "step": 841
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 2.692469999155472e-06,
+      "loss": 0.5691,
+      "step": 842
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 2.6917567397296012e-06,
+      "loss": 0.5357,
+      "step": 843
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 2.6910427488302344e-06,
+      "loss": 0.6018,
+      "step": 844
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 2.690328026895604e-06,
+      "loss": 0.5629,
+      "step": 845
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 2.6896125743643886e-06,
+      "loss": 0.5498,
+      "step": 846
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 2.6888963916757183e-06,
+      "loss": 0.6137,
+      "step": 847
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 2.688179479269169e-06,
+      "loss": 0.5461,
+      "step": 848
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 2.687461837584766e-06,
+      "loss": 0.5557,
+      "step": 849
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 2.6867434670629814e-06,
+      "loss": 0.5568,
+      "step": 850
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 2.6860243681447356e-06,
+      "loss": 0.6014,
+      "step": 851
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 2.6853045412713942e-06,
+      "loss": 0.582,
+      "step": 852
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 2.684583986884771e-06,
+      "loss": 0.58,
+      "step": 853
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 2.6838627054271265e-06,
+      "loss": 0.5808,
+      "step": 854
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 2.683140697341167e-06,
+      "loss": 0.5486,
+      "step": 855
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 2.6824179630700445e-06,
+      "loss": 0.5573,
+      "step": 856
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 2.6816945030573574e-06,
+      "loss": 0.5449,
+      "step": 857
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 2.6809703177471487e-06,
+      "loss": 0.5901,
+      "step": 858
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 2.6802454075839084e-06,
+      "loss": 0.5711,
+      "step": 859
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 2.679519773012568e-06,
+      "loss": 0.6025,
+      "step": 860
+    },
+    {
+      "epoch": 6.38,
+      "eval_loss": 1.3735154867172241,
+      "eval_runtime": 1.8222,
+      "eval_samples_per_second": 36.221,
+      "eval_steps_per_second": 3.293,
+      "step": 860
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 2.6787934144785072e-06,
+      "loss": 0.6337,
+      "step": 861
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 2.678066332427548e-06,
+      "loss": 0.6011,
+      "step": 862
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 2.677338527305957e-06,
+      "loss": 0.569,
+      "step": 863
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 2.6766099995604447e-06,
+      "loss": 0.5556,
+      "step": 864
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 2.6758807496381657e-06,
+      "loss": 0.5513,
+      "step": 865
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 2.6751507779867158e-06,
+      "loss": 0.6017,
+      "step": 866
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 2.6744200850541365e-06,
+      "loss": 0.5419,
+      "step": 867
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 2.6736886712889102e-06,
+      "loss": 0.5353,
+      "step": 868
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 2.672956537139962e-06,
+      "loss": 0.5642,
+      "step": 869
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 2.6722236830566594e-06,
+      "loss": 0.5572,
+      "step": 870
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 2.6714901094888123e-06,
+      "loss": 0.6033,
+      "step": 871
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 2.670755816886671e-06,
+      "loss": 0.5547,
+      "step": 872
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 2.6700208057009285e-06,
+      "loss": 0.5517,
+      "step": 873
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 2.669285076382717e-06,
+      "loss": 0.5852,
+      "step": 874
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 2.6685486293836124e-06,
+      "loss": 0.576,
+      "step": 875
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 2.6678114651556276e-06,
+      "loss": 0.5667,
+      "step": 876
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 2.6670735841512184e-06,
+      "loss": 0.6067,
+      "step": 877
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 2.6663349868232795e-06,
+      "loss": 0.5667,
+      "step": 878
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 2.6655956736251444e-06,
+      "loss": 0.5305,
+      "step": 879
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 2.6648556450105884e-06,
+      "loss": 0.5787,
+      "step": 880
+    },
+    {
+      "epoch": 6.53,
+      "eval_loss": 1.3643921613693237,
+      "eval_runtime": 1.8226,
+      "eval_samples_per_second": 36.212,
+      "eval_steps_per_second": 3.292,
+      "step": 880
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 2.6641149014338234e-06,
+      "loss": 0.5446,
+      "step": 881
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 2.663373443349501e-06,
+      "loss": 0.5425,
+      "step": 882
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 2.6626312712127126e-06,
+      "loss": 0.549,
+      "step": 883
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 2.6618883854789857e-06,
+      "loss": 0.5452,
+      "step": 884
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 2.661144786604287e-06,
+      "loss": 0.5709,
+      "step": 885
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 2.6604004750450216e-06,
+      "loss": 0.5767,
+      "step": 886
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 2.6596554512580304e-06,
+      "loss": 0.5658,
+      "step": 887
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 2.6589097157005922e-06,
+      "loss": 0.5761,
+      "step": 888
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 2.658163268830424e-06,
+      "loss": 0.5537,
+      "step": 889
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 2.657416111105676e-06,
+      "loss": 0.5692,
+      "step": 890
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 2.6566682429849385e-06,
+      "loss": 0.5946,
+      "step": 891
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 2.6559196649272356e-06,
+      "loss": 0.5728,
+      "step": 892
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.6551703773920276e-06,
+      "loss": 0.5734,
+      "step": 893
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.6544203808392104e-06,
+      "loss": 0.5612,
+      "step": 894
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 2.653669675729115e-06,
+      "loss": 0.5841,
+      "step": 895
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 2.652918262522507e-06,
+      "loss": 0.5568,
+      "step": 896
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.6521661416805877e-06,
+      "loss": 0.5367,
+      "step": 897
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.6514133136649916e-06,
+      "loss": 0.5875,
+      "step": 898
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 2.6506597789377873e-06,
+      "loss": 0.5775,
+      "step": 899
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 2.649905537961478e-06,
+      "loss": 0.5193,
+      "step": 900
+    },
+    {
+      "epoch": 6.68,
+      "eval_loss": 1.378929853439331,
+      "eval_runtime": 1.8224,
+      "eval_samples_per_second": 36.217,
+      "eval_steps_per_second": 3.292,
+      "step": 900
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 2.6491505911989994e-06,
+      "loss": 0.5896,
+      "step": 901
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 2.6483949391137212e-06,
+      "loss": 0.601,
+      "step": 902
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 2.6476385821694454e-06,
+      "loss": 0.5965,
+      "step": 903
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 2.6468815208304073e-06,
+      "loss": 0.5814,
+      "step": 904
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 2.646123755561273e-06,
+      "loss": 0.5663,
+      "step": 905
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 2.6453652868271435e-06,
+      "loss": 0.5717,
+      "step": 906
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 2.6446061150935488e-06,
+      "loss": 0.5447,
+      "step": 907
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 2.6438462408264513e-06,
+      "loss": 0.5529,
+      "step": 908
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 2.6430856644922454e-06,
+      "loss": 0.5519,
+      "step": 909
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 2.642324386557755e-06,
+      "loss": 0.593,
+      "step": 910
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 2.6415624074902364e-06,
+      "loss": 0.6138,
+      "step": 911
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 2.640799727757374e-06,
+      "loss": 0.5844,
+      "step": 912
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 2.6400363478272843e-06,
+      "loss": 0.5831,
+      "step": 913
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 2.6392722681685126e-06,
+      "loss": 0.5754,
+      "step": 914
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 2.638507489250034e-06,
+      "loss": 0.5756,
+      "step": 915
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 2.637742011541252e-06,
+      "loss": 0.5794,
+      "step": 916
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 2.636975835512e-06,
+      "loss": 0.5519,
+      "step": 917
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 2.6362089616325397e-06,
+      "loss": 0.5837,
+      "step": 918
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 2.6354413903735612e-06,
+      "loss": 0.56,
+      "step": 919
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.634673122206182e-06,
+      "loss": 0.6151,
+      "step": 920
+    },
+    {
+      "epoch": 6.83,
+      "eval_loss": 1.3711761236190796,
+      "eval_runtime": 1.8205,
+      "eval_samples_per_second": 36.254,
+      "eval_steps_per_second": 3.296,
+      "step": 920
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.633904157601948e-06,
+      "loss": 0.5351,
+      "step": 921
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 2.6331344970328326e-06,
+      "loss": 0.5536,
+      "step": 922
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 2.6323641409712354e-06,
+      "loss": 0.5539,
+      "step": 923
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 2.6315930898899853e-06,
+      "loss": 0.6096,
+      "step": 924
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 2.6308213442623344e-06,
+      "loss": 0.5581,
+      "step": 925
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 2.6300489045619638e-06,
+      "loss": 0.5632,
+      "step": 926
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 2.629275771262979e-06,
+      "loss": 0.5902,
+      "step": 927
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 2.6285019448399117e-06,
+      "loss": 0.5789,
+      "step": 928
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 2.6277274257677203e-06,
+      "loss": 0.5865,
+      "step": 929
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 2.626952214521786e-06,
+      "loss": 0.5542,
+      "step": 930
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 2.626176311577917e-06,
+      "loss": 0.5905,
+      "step": 931
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 2.6253997174123446e-06,
+      "loss": 0.5926,
+      "step": 932
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 2.624622432501724e-06,
+      "loss": 0.5436,
+      "step": 933
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 2.623844457323137e-06,
+      "loss": 0.5904,
+      "step": 934
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 2.623065792354086e-06,
+      "loss": 0.5983,
+      "step": 935
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 2.622286438072499e-06,
+      "loss": 0.5559,
+      "step": 936
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 2.6215063949567245e-06,
+      "loss": 0.5933,
+      "step": 937
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 2.620725663485537e-06,
+      "loss": 0.5633,
+      "step": 938
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 2.6199442441381313e-06,
+      "loss": 0.5625,
+      "step": 939
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.6191621373941253e-06,
+      "loss": 0.5591,
+      "step": 940
+    },
+    {
+      "epoch": 6.98,
+      "eval_loss": 1.3724634647369385,
+      "eval_runtime": 1.8212,
+      "eval_samples_per_second": 36.239,
+      "eval_steps_per_second": 3.294,
+      "step": 940
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.618379343733558e-06,
+      "loss": 0.5889,
+      "step": 941
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 2.617595863636891e-06,
+      "loss": 0.5483,
+      "step": 942
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.6168116975850068e-06,
+      "loss": 0.6035,
+      "step": 943
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 2.6160268460592087e-06,
+      "loss": 0.5332,
+      "step": 944
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 2.615241309541221e-06,
+      "loss": 0.49,
+      "step": 945
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 2.6144550885131883e-06,
+      "loss": 0.452,
+      "step": 946
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 2.613668183457675e-06,
+      "loss": 0.4837,
+      "step": 947
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 2.612880594857667e-06,
+      "loss": 0.4496,
+      "step": 948
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 2.6120923231965673e-06,
+      "loss": 0.4687,
+      "step": 949
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 2.6113033689581995e-06,
+      "loss": 0.4798,
+      "step": 950
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 2.610513732626806e-06,
+      "loss": 0.4322,
+      "step": 951
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 2.6097234146870485e-06,
+      "loss": 0.4622,
+      "step": 952
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 2.6089324156240054e-06,
+      "loss": 0.4724,
+      "step": 953
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 2.6081407359231754e-06,
+      "loss": 0.4452,
+      "step": 954
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 2.607348376070472e-06,
+      "loss": 0.4486,
+      "step": 955
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 2.6065553365522294e-06,
+      "loss": 0.458,
+      "step": 956
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 2.605761617855196e-06,
+      "loss": 0.4541,
+      "step": 957
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 2.60496722046654e-06,
+      "loss": 0.4377,
+      "step": 958
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 2.6041721448738434e-06,
+      "loss": 0.4401,
+      "step": 959
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 2.603376391565107e-06,
+      "loss": 0.4609,
+      "step": 960
+    },
+    {
+      "epoch": 7.12,
+      "eval_loss": 1.4474340677261353,
+      "eval_runtime": 1.8192,
+      "eval_samples_per_second": 36.28,
+      "eval_steps_per_second": 3.298,
+      "step": 960
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 2.602579961028745e-06,
+      "loss": 0.4532,
+      "step": 961
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 2.6017828537535893e-06,
+      "loss": 0.4371,
+      "step": 962
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 2.6009850702288865e-06,
+      "loss": 0.4779,
+      "step": 963
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 2.6001866109442973e-06,
+      "loss": 0.4644,
+      "step": 964
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 2.5993874763898994e-06,
+      "loss": 0.4784,
+      "step": 965
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 2.5985876670561827e-06,
+      "loss": 0.4523,
+      "step": 966
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 2.5977871834340524e-06,
+      "loss": 0.47,
+      "step": 967
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 2.596986026014827e-06,
+      "loss": 0.453,
+      "step": 968
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 2.59618419529024e-06,
+      "loss": 0.4554,
+      "step": 969
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 2.595381691752436e-06,
+      "loss": 0.473,
+      "step": 970
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 2.594578515893974e-06,
+      "loss": 0.4144,
+      "step": 971
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 2.593774668207825e-06,
+      "loss": 0.4667,
+      "step": 972
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 2.592970149187373e-06,
+      "loss": 0.4425,
+      "step": 973
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 2.5921649593264143e-06,
+      "loss": 0.4555,
+      "step": 974
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 2.591359099119155e-06,
+      "loss": 0.4694,
+      "step": 975
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 2.5905525690602147e-06,
+      "loss": 0.4436,
+      "step": 976
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 2.589745369644623e-06,
+      "loss": 0.4434,
+      "step": 977
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 2.588937501367822e-06,
+      "loss": 0.4506,
+      "step": 978
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 2.5881289647256615e-06,
+      "loss": 0.4839,
+      "step": 979
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 2.587319760214404e-06,
+      "loss": 0.4813,
+      "step": 980
+    },
+    {
+      "epoch": 7.27,
+      "eval_loss": 1.4841115474700928,
+      "eval_runtime": 1.8223,
+      "eval_samples_per_second": 36.217,
+      "eval_steps_per_second": 3.292,
+      "step": 980
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 2.5865098883307206e-06,
+      "loss": 0.4544,
+      "step": 981
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 2.585699349571693e-06,
+      "loss": 0.4752,
+      "step": 982
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 2.5848881444348114e-06,
+      "loss": 0.4576,
+      "step": 983
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 2.584076273417975e-06,
+      "loss": 0.4824,
+      "step": 984
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 2.5832637370194923e-06,
+      "loss": 0.426,
+      "step": 985
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 2.5824505357380796e-06,
+      "loss": 0.4507,
+      "step": 986
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 2.5816366700728624e-06,
+      "loss": 0.4647,
+      "step": 987
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 2.580822140523372e-06,
+      "loss": 0.4231,
+      "step": 988
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 2.5800069475895487e-06,
+      "loss": 0.4484,
+      "step": 989
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 2.57919109177174e-06,
+      "loss": 0.4522,
+      "step": 990
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 2.5783745735707e-06,
+      "loss": 0.4917,
+      "step": 991
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 2.577557393487588e-06,
+      "loss": 0.4764,
+      "step": 992
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 2.576739552023972e-06,
+      "loss": 0.489,
+      "step": 993
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 2.5759210496818245e-06,
+      "loss": 0.4859,
+      "step": 994
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 2.575101886963523e-06,
+      "loss": 0.4346,
+      "step": 995
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 2.5742820643718524e-06,
+      "loss": 0.4622,
+      "step": 996
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 2.57346158241e-06,
+      "loss": 0.4596,
+      "step": 997
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 2.5726404415815607e-06,
+      "loss": 0.4706,
+      "step": 998
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 2.571818642390531e-06,
+      "loss": 0.4699,
+      "step": 999
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 2.5709961853413137e-06,
+      "loss": 0.4405,
+      "step": 1000
+    },
+    {
+      "epoch": 7.42,
+      "eval_loss": 1.47830331325531,
+      "eval_runtime": 1.8167,
+      "eval_samples_per_second": 36.329,
+      "eval_steps_per_second": 3.303,
+      "step": 1000
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 2.570173070938714e-06,
+      "loss": 0.442,
+      "step": 1001
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 2.5693492996879406e-06,
+      "loss": 0.4839,
+      "step": 1002
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 2.568524872094606e-06,
+      "loss": 0.4767,
+      "step": 1003
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 2.5676997886647265e-06,
+      "loss": 0.4349,
+      "step": 1004
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 2.5668740499047186e-06,
+      "loss": 0.4232,
+      "step": 1005
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 2.5660476563214017e-06,
+      "loss": 0.4719,
+      "step": 1006
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 2.565220608421999e-06,
+      "loss": 0.4452,
+      "step": 1007
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 2.564392906714133e-06,
+      "loss": 0.4669,
+      "step": 1008
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 2.563564551705829e-06,
+      "loss": 0.4654,
+      "step": 1009
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 2.5627355439055116e-06,
+      "loss": 0.4398,
+      "step": 1010
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 2.561905883822008e-06,
+      "loss": 0.5014,
+      "step": 1011
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 2.561075571964545e-06,
+      "loss": 0.4504,
+      "step": 1012
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 2.5602446088427488e-06,
+      "loss": 0.4698,
+      "step": 1013
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 2.5594129949666466e-06,
+      "loss": 0.4508,
+      "step": 1014
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 2.5585807308466638e-06,
+      "loss": 0.5029,
+      "step": 1015
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 2.5577478169936255e-06,
+      "loss": 0.4828,
+      "step": 1016
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 2.5569142539187554e-06,
+      "loss": 0.5017,
+      "step": 1017
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 2.5560800421336756e-06,
+      "loss": 0.5018,
+      "step": 1018
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 2.5552451821504075e-06,
+      "loss": 0.4633,
+      "step": 1019
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 2.554409674481368e-06,
+      "loss": 0.4595,
+      "step": 1020
+    },
+    {
+      "epoch": 7.57,
+      "eval_loss": 1.484065055847168,
+      "eval_runtime": 1.8214,
+      "eval_samples_per_second": 36.235,
+      "eval_steps_per_second": 3.294,
+      "step": 1020
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 2.5535735196393745e-06,
+      "loss": 0.4908,
+      "step": 1021
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 2.5527367181376383e-06,
+      "loss": 0.4617,
+      "step": 1022
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 2.5518992704897707e-06,
+      "loss": 0.4132,
+      "step": 1023
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 2.5510611772097776e-06,
+      "loss": 0.4701,
+      "step": 1024
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 2.550222438812062e-06,
+      "loss": 0.4317,
+      "step": 1025
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 2.549383055811423e-06,
+      "loss": 0.4523,
+      "step": 1026
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 2.5485430287230536e-06,
+      "loss": 0.4198,
+      "step": 1027
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 2.547702358062546e-06,
+      "loss": 0.4903,
+      "step": 1028
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 2.5468610443458828e-06,
+      "loss": 0.4588,
+      "step": 1029
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 2.5460190880894446e-06,
+      "loss": 0.4857,
+      "step": 1030
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 2.5451764898100044e-06,
+      "loss": 0.4492,
+      "step": 1031
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 2.544333250024731e-06,
+      "loss": 0.4886,
+      "step": 1032
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 2.5434893692511852e-06,
+      "loss": 0.4711,
+      "step": 1033
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 2.5426448480073222e-06,
+      "loss": 0.4753,
+      "step": 1034
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 2.541799686811491e-06,
+      "loss": 0.4542,
+      "step": 1035
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 2.540953886182431e-06,
+      "loss": 0.465,
+      "step": 1036
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 2.540107446639277e-06,
+      "loss": 0.4238,
+      "step": 1037
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 2.5392603687015544e-06,
+      "loss": 0.4804,
+      "step": 1038
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 2.53841265288918e-06,
+      "loss": 0.4618,
+      "step": 1039
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 2.537564299722463e-06,
+      "loss": 0.4528,
+      "step": 1040
+    },
+    {
+      "epoch": 7.72,
+      "eval_loss": 1.4873430728912354,
+      "eval_runtime": 1.8209,
+      "eval_samples_per_second": 36.246,
+      "eval_steps_per_second": 3.295,
+      "step": 1040
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 2.5367153097221034e-06,
+      "loss": 0.4318,
+      "step": 1041
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 2.535865683409192e-06,
+      "loss": 0.4791,
+      "step": 1042
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 2.5350154213052108e-06,
+      "loss": 0.4654,
+      "step": 1043
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 2.5341645239320314e-06,
+      "loss": 0.5033,
+      "step": 1044
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 2.533312991811915e-06,
+      "loss": 0.4971,
+      "step": 1045
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 2.5324608254675143e-06,
+      "loss": 0.445,
+      "step": 1046
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 2.531608025421868e-06,
+      "loss": 0.4641,
+      "step": 1047
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 2.5307545921984065e-06,
+      "loss": 0.4929,
+      "step": 1048
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 2.529900526320948e-06,
+      "loss": 0.4621,
+      "step": 1049
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 2.529045828313699e-06,
+      "loss": 0.4756,
+      "step": 1050
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 2.5281904987012537e-06,
+      "loss": 0.4709,
+      "step": 1051
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 2.527334538008594e-06,
+      "loss": 0.4437,
+      "step": 1052
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 2.5264779467610897e-06,
+      "loss": 0.5041,
+      "step": 1053
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 2.525620725484497e-06,
+      "loss": 0.4772,
+      "step": 1054
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 2.5247628747049594e-06,
+      "loss": 0.4679,
+      "step": 1055
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 2.523904394949006e-06,
+      "loss": 0.4421,
+      "step": 1056
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 2.5230452867435526e-06,
+      "loss": 0.481,
+      "step": 1057
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 2.5221855506159e-06,
+      "loss": 0.4383,
+      "step": 1058
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 2.5213251870937364e-06,
+      "loss": 0.4294,
+      "step": 1059
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 2.520464196705131e-06,
+      "loss": 0.4796,
+      "step": 1060
+    },
+    {
+      "epoch": 7.87,
+      "eval_loss": 1.4874656200408936,
+      "eval_runtime": 1.8214,
+      "eval_samples_per_second": 36.236,
+      "eval_steps_per_second": 3.294,
+      "step": 1060
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 2.519602579978543e-06,
+      "loss": 0.5022,
+      "step": 1061
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 2.518740337442812e-06,
+      "loss": 0.4733,
+      "step": 1062
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 2.517877469627164e-06,
+      "loss": 0.4805,
+      "step": 1063
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 2.517013977061206e-06,
+      "loss": 0.4511,
+      "step": 1064
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 2.516149860274932e-06,
+      "loss": 0.5196,
+      "step": 1065
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 2.5152851197987175e-06,
+      "loss": 0.4754,
+      "step": 1066
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 2.5144197561633203e-06,
+      "loss": 0.4542,
+      "step": 1067
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 2.5135537698998805e-06,
+      "loss": 0.4928,
+      "step": 1068
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 2.5126871615399217e-06,
+      "loss": 0.4921,
+      "step": 1069
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 2.511819931615349e-06,
+      "loss": 0.497,
+      "step": 1070
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 2.510952080658448e-06,
+      "loss": 0.4647,
+      "step": 1071
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 2.510083609201886e-06,
+      "loss": 0.4666,
+      "step": 1072
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 2.5092145177787114e-06,
+      "loss": 0.4851,
+      "step": 1073
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 2.5083448069223535e-06,
+      "loss": 0.4692,
+      "step": 1074
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 2.507474477166621e-06,
+      "loss": 0.4513,
+      "step": 1075
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 2.5066035290457024e-06,
+      "loss": 0.4697,
+      "step": 1076
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 2.5057319630941668e-06,
+      "loss": 0.4379,
+      "step": 1077
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 2.5048597798469608e-06,
+      "loss": 0.4864,
+      "step": 1078
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 2.503986979839411e-06,
+      "loss": 0.3764,
+      "step": 1079
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 2.5031135636072235e-06,
+      "loss": 0.3684,
+      "step": 1080
+    },
+    {
+      "epoch": 8.01,
+      "eval_loss": 1.5208909511566162,
+      "eval_runtime": 1.8077,
+      "eval_samples_per_second": 36.51,
+      "eval_steps_per_second": 3.319,
+      "step": 1080
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 2.5022395316864807e-06,
+      "loss": 0.386,
+      "step": 1081
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 2.5013648846136436e-06,
+      "loss": 0.3499,
+      "step": 1082
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 2.5004896229255515e-06,
+      "loss": 0.3993,
+      "step": 1083
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 2.4996137471594202e-06,
+      "loss": 0.4017,
+      "step": 1084
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 2.4987372578528417e-06,
+      "loss": 0.3781,
+      "step": 1085
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 2.497860155543787e-06,
+      "loss": 0.3808,
+      "step": 1086
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 2.4969824407706003e-06,
+      "loss": 0.3584,
+      "step": 1087
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 2.496104114072004e-06,
+      "loss": 0.3541,
+      "step": 1088
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 2.4952251759870947e-06,
+      "loss": 0.3612,
+      "step": 1089
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 2.4943456270553456e-06,
+      "loss": 0.3724,
+      "step": 1090
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 2.493465467816604e-06,
+      "loss": 0.3719,
+      "step": 1091
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 2.4925846988110915e-06,
+      "loss": 0.3824,
+      "step": 1092
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 2.491703320579404e-06,
+      "loss": 0.3804,
+      "step": 1093
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 2.490821333662512e-06,
+      "loss": 0.3824,
+      "step": 1094
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 2.4899387386017595e-06,
+      "loss": 0.3856,
+      "step": 1095
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 2.489055535938863e-06,
+      "loss": 0.3734,
+      "step": 1096
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 2.488171726215913e-06,
+      "loss": 0.3671,
+      "step": 1097
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 2.4872873099753715e-06,
+      "loss": 0.3545,
+      "step": 1098
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 2.4864022877600735e-06,
+      "loss": 0.3603,
+      "step": 1099
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 2.485516660113226e-06,
+      "loss": 0.3312,
+      "step": 1100
+    },
+    {
+      "epoch": 8.16,
+      "eval_loss": 1.6160022020339966,
+      "eval_runtime": 1.8196,
+      "eval_samples_per_second": 36.272,
+      "eval_steps_per_second": 3.297,
+      "step": 1100
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 2.484630427578407e-06,
+      "loss": 0.3647,
+      "step": 1101
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 2.483743590699566e-06,
+      "loss": 0.3551,
+      "step": 1102
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 2.4828561500210246e-06,
+      "loss": 0.3718,
+      "step": 1103
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 2.481968106087472e-06,
+      "loss": 0.3429,
+      "step": 1104
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 2.481079459443972e-06,
+      "loss": 0.3425,
+      "step": 1105
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 2.4801902106359534e-06,
+      "loss": 0.3383,
+      "step": 1106
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 2.4793003602092194e-06,
+      "loss": 0.3377,
+      "step": 1107
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 2.478409908709938e-06,
+      "loss": 0.3895,
+      "step": 1108
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 2.4775188566846506e-06,
+      "loss": 0.4008,
+      "step": 1109
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 2.476627204680263e-06,
+      "loss": 0.3958,
+      "step": 1110
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 2.475734953244052e-06,
+      "loss": 0.352,
+      "step": 1111
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 2.474842102923662e-06,
+      "loss": 0.3624,
+      "step": 1112
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 2.473948654267103e-06,
+      "loss": 0.366,
+      "step": 1113
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 2.4730546078227556e-06,
+      "loss": 0.3793,
+      "step": 1114
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 2.472159964139364e-06,
+      "loss": 0.364,
+      "step": 1115
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 2.4712647237660405e-06,
+      "loss": 0.3476,
+      "step": 1116
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 2.4703688872522638e-06,
+      "loss": 0.4006,
+      "step": 1117
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 2.469472455147878e-06,
+      "loss": 0.3412,
+      "step": 1118
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 2.4685754280030936e-06,
+      "loss": 0.3248,
+      "step": 1119
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 2.467677806368485e-06,
+      "loss": 0.3526,
+      "step": 1120
+    },
+    {
+      "epoch": 8.31,
+      "eval_loss": 1.5985159873962402,
+      "eval_runtime": 1.8219,
+      "eval_samples_per_second": 36.226,
+      "eval_steps_per_second": 3.293,
+      "step": 1120
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 2.4667795907949926e-06,
+      "loss": 0.3776,
+      "step": 1121
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 2.4658807818339205e-06,
+      "loss": 0.3469,
+      "step": 1122
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 2.4649813800369377e-06,
+      "loss": 0.3491,
+      "step": 1123
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 2.4640813859560774e-06,
+      "loss": 0.3595,
+      "step": 1124
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 2.4631808001437345e-06,
+      "loss": 0.3871,
+      "step": 1125
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 2.4622796231526692e-06,
+      "loss": 0.3751,
+      "step": 1126
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 2.461377855536003e-06,
+      "loss": 0.3814,
+      "step": 1127
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 2.4604754978472213e-06,
+      "loss": 0.3824,
+      "step": 1128
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 2.4595725506401713e-06,
+      "loss": 0.3691,
+      "step": 1129
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 2.4586690144690605e-06,
+      "loss": 0.3816,
+      "step": 1130
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 2.4577648898884596e-06,
+      "loss": 0.3296,
+      "step": 1131
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 2.4568601774533003e-06,
+      "loss": 0.3613,
+      "step": 1132
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 2.4559548777188743e-06,
+      "loss": 0.3462,
+      "step": 1133
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 2.4550489912408345e-06,
+      "loss": 0.3816,
+      "step": 1134
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 2.4541425185751936e-06,
+      "loss": 0.379,
+      "step": 1135
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 2.453235460278324e-06,
+      "loss": 0.3544,
+      "step": 1136
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 2.4523278169069575e-06,
+      "loss": 0.3733,
+      "step": 1137
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 2.451419589018186e-06,
+      "loss": 0.3786,
+      "step": 1138
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 2.450510777169459e-06,
+      "loss": 0.3515,
+      "step": 1139
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 2.4496013819185844e-06,
+      "loss": 0.3606,
+      "step": 1140
+    },
+    {
+      "epoch": 8.46,
+      "eval_loss": 1.6050316095352173,
+      "eval_runtime": 1.8209,
+      "eval_samples_per_second": 36.246,
+      "eval_steps_per_second": 3.295,
+      "step": 1140
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 2.4486914038237287e-06,
+      "loss": 0.3357,
+      "step": 1141
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 2.447780843443416e-06,
+      "loss": 0.3654,
+      "step": 1142
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 2.446869701336528e-06,
+      "loss": 0.3565,
+      "step": 1143
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 2.445957978062303e-06,
+      "loss": 0.354,
+      "step": 1144
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 2.4450456741803365e-06,
+      "loss": 0.3545,
+      "step": 1145
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 2.44413279025058e-06,
+      "loss": 0.4163,
+      "step": 1146
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 2.4432193268333413e-06,
+      "loss": 0.3348,
+      "step": 1147
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 2.4423052844892835e-06,
+      "loss": 0.4024,
+      "step": 1148
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 2.441390663779425e-06,
+      "loss": 0.3542,
+      "step": 1149
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 2.44047546526514e-06,
+      "loss": 0.3706,
+      "step": 1150
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 2.4395596895081563e-06,
+      "loss": 0.3559,
+      "step": 1151
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 2.438643337070557e-06,
+      "loss": 0.3351,
+      "step": 1152
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 2.4377264085147775e-06,
+      "loss": 0.3719,
+      "step": 1153
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 2.4368089044036093e-06,
+      "loss": 0.3577,
+      "step": 1154
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 2.4358908253001956e-06,
+      "loss": 0.3754,
+      "step": 1155
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 2.4349721717680314e-06,
+      "loss": 0.3604,
+      "step": 1156
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 2.4340529443709675e-06,
+      "loss": 0.3764,
+      "step": 1157
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 2.4331331436732033e-06,
+      "loss": 0.3719,
+      "step": 1158
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 2.432212770239293e-06,
+      "loss": 0.3597,
+      "step": 1159
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 2.4312918246341404e-06,
+      "loss": 0.3563,
+      "step": 1160
+    },
+    {
+      "epoch": 8.61,
+      "eval_loss": 1.61266028881073,
+      "eval_runtime": 1.8216,
+      "eval_samples_per_second": 36.233,
+      "eval_steps_per_second": 3.294,
+      "step": 1160
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 2.4303703074230015e-06,
+      "loss": 0.3738,
+      "step": 1161
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 2.4294482191714828e-06,
+      "loss": 0.3747,
+      "step": 1162
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 2.428525560445541e-06,
+      "loss": 0.3386,
+      "step": 1163
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 2.427602331811484e-06,
+      "loss": 0.3661,
+      "step": 1164
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 2.4266785338359685e-06,
+      "loss": 0.3741,
+      "step": 1165
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 2.4257541670860005e-06,
+      "loss": 0.3709,
+      "step": 1166
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 2.4248292321289364e-06,
+      "loss": 0.3444,
+      "step": 1167
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 2.4239037295324797e-06,
+      "loss": 0.3596,
+      "step": 1168
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 2.4229776598646833e-06,
+      "loss": 0.394,
+      "step": 1169
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 2.422051023693948e-06,
+      "loss": 0.4066,
+      "step": 1170
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 2.4211238215890224e-06,
+      "loss": 0.367,
+      "step": 1171
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 2.420196054119002e-06,
+      "loss": 0.3711,
+      "step": 1172
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 2.4192677218533298e-06,
+      "loss": 0.3836,
+      "step": 1173
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 2.418338825361795e-06,
+      "loss": 0.376,
+      "step": 1174
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 2.417409365214534e-06,
+      "loss": 0.3567,
+      "step": 1175
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 2.416479341982028e-06,
+      "loss": 0.3872,
+      "step": 1176
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 2.4155487562351044e-06,
+      "loss": 0.376,
+      "step": 1177
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 2.414617608544936e-06,
+      "loss": 0.3564,
+      "step": 1178
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 2.41368589948304e-06,
+      "loss": 0.3487,
+      "step": 1179
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 2.4127536296212794e-06,
+      "loss": 0.391,
+      "step": 1180
+    },
+    {
+      "epoch": 8.76,
+      "eval_loss": 1.6107513904571533,
+      "eval_runtime": 1.8159,
+      "eval_samples_per_second": 36.346,
+      "eval_steps_per_second": 3.304,
+      "step": 1180
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 2.4118207995318593e-06,
+      "loss": 0.3619,
+      "step": 1181
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 2.4108874097873307e-06,
+      "loss": 0.3689,
+      "step": 1182
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 2.409953460960586e-06,
+      "loss": 0.3726,
+      "step": 1183
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 2.409018953624864e-06,
+      "loss": 0.3814,
+      "step": 1184
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 2.408083888353743e-06,
+      "loss": 0.3452,
+      "step": 1185
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 2.4071482657211446e-06,
+      "loss": 0.3561,
+      "step": 1186
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 2.406212086301334e-06,
+      "loss": 0.3903,
+      "step": 1187
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 2.4052753506689166e-06,
+      "loss": 0.3552,
+      "step": 1188
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 2.40433805939884e-06,
+      "loss": 0.3796,
+      "step": 1189
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 2.4034002130663917e-06,
+      "loss": 0.3726,
+      "step": 1190
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 2.4024618122472013e-06,
+      "loss": 0.3641,
+      "step": 1191
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 2.401522857517238e-06,
+      "loss": 0.3746,
+      "step": 1192
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 2.400583349452812e-06,
+      "loss": 0.368,
+      "step": 1193
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 2.399643288630571e-06,
+      "loss": 0.3937,
+      "step": 1194
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 2.3987026756275033e-06,
+      "loss": 0.3945,
+      "step": 1195
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 2.397761511020936e-06,
+      "loss": 0.376,
+      "step": 1196
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 2.3968197953885357e-06,
+      "loss": 0.3622,
+      "step": 1197
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 2.395877529308305e-06,
+      "loss": 0.3771,
+      "step": 1198
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 2.3949347133585873e-06,
+      "loss": 0.3764,
+      "step": 1199
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 2.39399134811806e-06,
+      "loss": 0.3547,
+      "step": 1200
+    },
+    {
+      "epoch": 8.91,
+      "eval_loss": 1.6076542139053345,
+      "eval_runtime": 1.8173,
+      "eval_samples_per_second": 36.318,
+      "eval_steps_per_second": 3.302,
+      "step": 1200
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 2.3930474341657408e-06,
+      "loss": 0.36,
+      "step": 1201
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 2.3921029720809824e-06,
+      "loss": 0.3652,
+      "step": 1202
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 2.391157962443474e-06,
+      "loss": 0.3595,
+      "step": 1203
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 2.390212405833242e-06,
+      "loss": 0.3787,
+      "step": 1204
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 2.3892663028306476e-06,
+      "loss": 0.3654,
+      "step": 1205
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 2.3883196540163867e-06,
+      "loss": 0.3384,
+      "step": 1206
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 2.3873724599714925e-06,
+      "loss": 0.3558,
+      "step": 1207
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 2.3864247212773304e-06,
+      "loss": 0.3718,
+      "step": 1208
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 2.385476438515601e-06,
+      "loss": 0.3561,
+      "step": 1209
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 2.3845276122683396e-06,
+      "loss": 0.3616,
+      "step": 1210
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 2.383578243117914e-06,
+      "loss": 0.3786,
+      "step": 1211
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 2.3826283316470263e-06,
+      "loss": 0.3704,
+      "step": 1212
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2.3816778784387097e-06,
+      "loss": 0.3698,
+      "step": 1213
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 2.3807268840763317e-06,
+      "loss": 0.2599,
+      "step": 1214
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 2.3797753491435915e-06,
+      "loss": 0.2801,
+      "step": 1215
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 2.378823274224519e-06,
+      "loss": 0.2859,
+      "step": 1216
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 2.3778706599034776e-06,
+      "loss": 0.2748,
+      "step": 1217
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 2.3769175067651594e-06,
+      "loss": 0.2651,
+      "step": 1218
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 2.375963815394589e-06,
+      "loss": 0.3137,
+      "step": 1219
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 2.37500958637712e-06,
+      "loss": 0.2897,
+      "step": 1220
+    },
+    {
+      "epoch": 9.05,
+      "eval_loss": 1.751111388206482,
+      "eval_runtime": 1.8169,
+      "eval_samples_per_second": 36.325,
+      "eval_steps_per_second": 3.302,
+      "step": 1220
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 2.374054820298438e-06,
+      "loss": 0.2666,
+      "step": 1221
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 2.3730995177445555e-06,
+      "loss": 0.2604,
+      "step": 1222
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 2.3721436793018176e-06,
+      "loss": 0.2989,
+      "step": 1223
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 2.371187305556895e-06,
+      "loss": 0.2918,
+      "step": 1224
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 2.3702303970967877e-06,
+      "loss": 0.2717,
+      "step": 1225
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 2.3692729545088265e-06,
+      "loss": 0.2666,
+      "step": 1226
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 2.3683149783806665e-06,
+      "loss": 0.2808,
+      "step": 1227
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 2.3673564693002934e-06,
+      "loss": 0.2623,
+      "step": 1228
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 2.366397427856017e-06,
+      "loss": 0.2959,
+      "step": 1229
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 2.365437854636477e-06,
+      "loss": 0.3101,
+      "step": 1230
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 2.3644777502306364e-06,
+      "loss": 0.2894,
+      "step": 1231
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 2.3635171152277865e-06,
+      "loss": 0.3014,
+      "step": 1232
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 2.362555950217543e-06,
+      "loss": 0.2642,
+      "step": 1233
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 2.3615942557898485e-06,
+      "loss": 0.268,
+      "step": 1234
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 2.3606320325349683e-06,
+      "loss": 0.311,
+      "step": 1235
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 2.3596692810434933e-06,
+      "loss": 0.2641,
+      "step": 1236
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 2.3587060019063393e-06,
+      "loss": 0.2861,
+      "step": 1237
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 2.3577421957147456e-06,
+      "loss": 0.2966,
+      "step": 1238
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 2.356777863060274e-06,
+      "loss": 0.2546,
+      "step": 1239
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 2.355813004534811e-06,
+      "loss": 0.2991,
+      "step": 1240
+    },
+    {
+      "epoch": 9.2,
+      "eval_loss": 1.7112075090408325,
+      "eval_runtime": 1.8172,
+      "eval_samples_per_second": 36.32,
+      "eval_steps_per_second": 3.302,
+      "step": 1240
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 2.354847620730564e-06,
+      "loss": 0.2842,
+      "step": 1241
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 2.353881712240065e-06,
+      "loss": 0.2712,
+      "step": 1242
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 2.3529152796561664e-06,
+      "loss": 0.2786,
+      "step": 1243
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 2.351948323572043e-06,
+      "loss": 0.2665,
+      "step": 1244
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 2.35098084458119e-06,
+      "loss": 0.2776,
+      "step": 1245
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 2.3500128432774254e-06,
+      "loss": 0.265,
+      "step": 1246
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 2.349044320254885e-06,
+      "loss": 0.3111,
+      "step": 1247
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 2.348075276108029e-06,
+      "loss": 0.2673,
+      "step": 1248
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 2.3471057114316323e-06,
+      "loss": 0.2899,
+      "step": 1249
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 2.3461356268207937e-06,
+      "loss": 0.2939,
+      "step": 1250
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 2.3451650228709286e-06,
+      "loss": 0.2757,
+      "step": 1251
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 2.3441939001777713e-06,
+      "loss": 0.2711,
+      "step": 1252
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 2.3432222593373764e-06,
+      "loss": 0.2794,
+      "step": 1253
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 2.3422501009461136e-06,
+      "loss": 0.2897,
+      "step": 1254
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 2.3412774256006735e-06,
+      "loss": 0.2889,
+      "step": 1255
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 2.3403042338980615e-06,
+      "loss": 0.2834,
+      "step": 1256
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 2.3393305264356013e-06,
+      "loss": 0.2672,
+      "step": 1257
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 2.338356303810932e-06,
+      "loss": 0.3021,
+      "step": 1258
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 2.33738156662201e-06,
+      "loss": 0.2925,
+      "step": 1259
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 2.3364063154671064e-06,
+      "loss": 0.2876,
+      "step": 1260
+    },
+    {
+      "epoch": 9.35,
+      "eval_loss": 1.7166179418563843,
+      "eval_runtime": 1.8214,
+      "eval_samples_per_second": 36.235,
+      "eval_steps_per_second": 3.294,
+      "step": 1260
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 2.335430550944809e-06,
+      "loss": 0.3093,
+      "step": 1261
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 2.3344542736540203e-06,
+      "loss": 0.2567,
+      "step": 1262
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 2.3334774841939572e-06,
+      "loss": 0.32,
+      "step": 1263
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 2.3325001831641508e-06,
+      "loss": 0.29,
+      "step": 1264
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 2.331522371164447e-06,
+      "loss": 0.3077,
+      "step": 1265
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 2.330544048795005e-06,
+      "loss": 0.2721,
+      "step": 1266
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 2.329565216656296e-06,
+      "loss": 0.3055,
+      "step": 1267
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 2.3285858753491056e-06,
+      "loss": 0.2839,
+      "step": 1268
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 2.327606025474532e-06,
+      "loss": 0.3001,
+      "step": 1269
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 2.3266256676339846e-06,
+      "loss": 0.2935,
+      "step": 1270
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 2.3256448024291856e-06,
+      "loss": 0.2798,
+      "step": 1271
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 2.324663430462167e-06,
+      "loss": 0.283,
+      "step": 1272
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 2.3236815523352733e-06,
+      "loss": 0.2783,
+      "step": 1273
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 2.3226991686511597e-06,
+      "loss": 0.2943,
+      "step": 1274
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 2.3217162800127905e-06,
+      "loss": 0.2805,
+      "step": 1275
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 2.3207328870234413e-06,
+      "loss": 0.2742,
+      "step": 1276
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 2.3197489902866957e-06,
+      "loss": 0.276,
+      "step": 1277
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 2.3187645904064486e-06,
+      "loss": 0.2689,
+      "step": 1278
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 2.3177796879869016e-06,
+      "loss": 0.2888,
+      "step": 1279
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 2.316794283632566e-06,
+      "loss": 0.2767,
+      "step": 1280
+    },
+    {
+      "epoch": 9.5,
+      "eval_loss": 1.7152905464172363,
+      "eval_runtime": 1.8203,
+      "eval_samples_per_second": 36.259,
+      "eval_steps_per_second": 3.296,
+      "step": 1280
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 2.3158083779482607e-06,
+      "loss": 0.3056,
+      "step": 1281
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 2.3148219715391128e-06,
+      "loss": 0.2717,
+      "step": 1282
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 2.3138350650105555e-06,
+      "loss": 0.277,
+      "step": 1283
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 2.312847658968331e-06,
+      "loss": 0.2735,
+      "step": 1284
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 2.3118597540184867e-06,
+      "loss": 0.2912,
+      "step": 1285
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 2.3108713507673757e-06,
+      "loss": 0.2888,
+      "step": 1286
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 2.3098824498216582e-06,
+      "loss": 0.3083,
+      "step": 1287
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 2.308893051788299e-06,
+      "loss": 0.3002,
+      "step": 1288
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 2.307903157274569e-06,
+      "loss": 0.3037,
+      "step": 1289
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 2.306912766888043e-06,
+      "loss": 0.3014,
+      "step": 1290
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 2.3059218812366003e-06,
+      "loss": 0.3027,
+      "step": 1291
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 2.3049305009284238e-06,
+      "loss": 0.2837,
+      "step": 1292
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 2.303938626572001e-06,
+      "loss": 0.2908,
+      "step": 1293
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 2.3029462587761218e-06,
+      "loss": 0.2952,
+      "step": 1294
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 2.3019533981498794e-06,
+      "loss": 0.2899,
+      "step": 1295
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 2.3009600453026693e-06,
+      "loss": 0.2828,
+      "step": 1296
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 2.2999662008441892e-06,
+      "loss": 0.2517,
+      "step": 1297
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 2.2989718653844384e-06,
+      "loss": 0.2961,
+      "step": 1298
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 2.2979770395337177e-06,
+      "loss": 0.2963,
+      "step": 1299
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 2.296981723902629e-06,
+      "loss": 0.3168,
+      "step": 1300
+    },
+    {
+      "epoch": 9.65,
+      "eval_loss": 1.7249263525009155,
+      "eval_runtime": 1.8172,
+      "eval_samples_per_second": 36.319,
+      "eval_steps_per_second": 3.302,
+      "step": 1300
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 2.295985919102074e-06,
+      "loss": 0.2841,
+      "step": 1301
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 2.294989625743256e-06,
+      "loss": 0.2698,
+      "step": 1302
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 2.293992844437677e-06,
+      "loss": 0.3091,
+      "step": 1303
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 2.29299557579714e-06,
+      "loss": 0.2933,
+      "step": 1304
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 2.2919978204337448e-06,
+      "loss": 0.3049,
+      "step": 1305
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 2.290999578959891e-06,
+      "loss": 0.285,
+      "step": 1306
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 2.290000851988278e-06,
+      "loss": 0.2676,
+      "step": 1307
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 2.2890016401319013e-06,
+      "loss": 0.3126,
+      "step": 1308
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 2.2880019440040554e-06,
+      "loss": 0.2801,
+      "step": 1309
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 2.28700176421833e-06,
+      "loss": 0.2507,
+      "step": 1310
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 2.2860011013886144e-06,
+      "loss": 0.272,
+      "step": 1311
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 2.284999956129091e-06,
+      "loss": 0.2971,
+      "step": 1312
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 2.2839983290542426e-06,
+      "loss": 0.3029,
+      "step": 1313
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 2.2829962207788437e-06,
+      "loss": 0.3039,
+      "step": 1314
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 2.2819936319179667e-06,
+      "loss": 0.2651,
+      "step": 1315
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 2.2809905630869776e-06,
+      "loss": 0.2748,
+      "step": 1316
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 2.279987014901538e-06,
+      "loss": 0.3025,
+      "step": 1317
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 2.2789829879776026e-06,
+      "loss": 0.2975,
+      "step": 1318
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 2.2779784829314213e-06,
+      "loss": 0.2807,
+      "step": 1319
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 2.276973500379536e-06,
+      "loss": 0.3227,
+      "step": 1320
+    },
+    {
+      "epoch": 9.8,
+      "eval_loss": 1.721448302268982,
+      "eval_runtime": 1.8175,
+      "eval_samples_per_second": 36.314,
+      "eval_steps_per_second": 3.301,
+      "step": 1320
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 2.2759680409387823e-06,
+      "loss": 0.276,
+      "step": 1321
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 2.27496210522629e-06,
+      "loss": 0.2659,
+      "step": 1322
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 2.2739556938594774e-06,
+      "loss": 0.2745,
+      "step": 1323
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 2.2729488074560595e-06,
+      "loss": 0.2996,
+      "step": 1324
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 2.2719414466340393e-06,
+      "loss": 0.2875,
+      "step": 1325
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 2.270933612011713e-06,
+      "loss": 0.2841,
+      "step": 1326
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 2.2699253042076658e-06,
+      "loss": 0.2791,
+      "step": 1327
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 2.268916523840775e-06,
+      "loss": 0.3189,
+      "step": 1328
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 2.267907271530207e-06,
+      "loss": 0.3102,
+      "step": 1329
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 2.2668975478954197e-06,
+      "loss": 0.294,
+      "step": 1330
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 2.2658873535561563e-06,
+      "loss": 0.2973,
+      "step": 1331
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 2.264876689132453e-06,
+      "loss": 0.289,
+      "step": 1332
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 2.263865555244632e-06,
+      "loss": 0.258,
+      "step": 1333
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 2.262853952513306e-06,
+      "loss": 0.2899,
+      "step": 1334
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 2.2618418815593724e-06,
+      "loss": 0.2776,
+      "step": 1335
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 2.2608293430040186e-06,
+      "loss": 0.282,
+      "step": 1336
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 2.259816337468717e-06,
+      "loss": 0.2994,
+      "step": 1337
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 2.258802865575229e-06,
+      "loss": 0.2983,
+      "step": 1338
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 2.2577889279455996e-06,
+      "loss": 0.2987,
+      "step": 1339
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 2.2567745252021613e-06,
+      "loss": 0.2909,
+      "step": 1340
+    },
+    {
+      "epoch": 9.94,
+      "eval_loss": 1.7276170253753662,
+      "eval_runtime": 1.8195,
+      "eval_samples_per_second": 36.273,
+      "eval_steps_per_second": 3.298,
+      "step": 1340
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 2.255759657967533e-06,
+      "loss": 0.2975,
+      "step": 1341
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 2.254744326864615e-06,
+      "loss": 0.264,
+      "step": 1342
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 2.2537285325165968e-06,
+      "loss": 0.2753,
+      "step": 1343
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 2.252712275546949e-06,
+      "loss": 0.2682,
+      "step": 1344
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 2.2516955565794272e-06,
+      "loss": 0.302,
+      "step": 1345
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 2.2506783762380715e-06,
+      "loss": 0.3355,
+      "step": 1346
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 2.2496607351472044e-06,
+      "loss": 0.2623,
+      "step": 1347
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 2.2486426339314295e-06,
+      "loss": 0.2521,
+      "step": 1348
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 2.2476240732156362e-06,
+      "loss": 0.2393,
+      "step": 1349
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 2.246605053624993e-06,
+      "loss": 0.2203,
+      "step": 1350
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 2.2455855757849522e-06,
+      "loss": 0.2317,
+      "step": 1351
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 2.2445656403212457e-06,
+      "loss": 0.2166,
+      "step": 1352
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 2.2435452478598872e-06,
+      "loss": 0.2292,
+      "step": 1353
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 2.24252439902717e-06,
+      "loss": 0.2103,
+      "step": 1354
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 2.24150309444967e-06,
+      "loss": 0.1997,
+      "step": 1355
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 2.2404813347542386e-06,
+      "loss": 0.2197,
+      "step": 1356
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 2.239459120568011e-06,
+      "loss": 0.2374,
+      "step": 1357
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 2.238436452518398e-06,
+      "loss": 0.2008,
+      "step": 1358
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 2.23741333123309e-06,
+      "loss": 0.2161,
+      "step": 1359
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 2.2363897573400574e-06,
+      "loss": 0.2181,
+      "step": 1360
+    },
+    {
+      "epoch": 10.09,
+      "eval_loss": 1.8039973974227905,
+      "eval_runtime": 1.8195,
+      "eval_samples_per_second": 36.275,
+      "eval_steps_per_second": 3.298,
+      "step": 1360
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 2.2353657314675458e-06,
+      "loss": 0.1998,
+      "step": 1361
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 2.2343412542440788e-06,
+      "loss": 0.2117,
+      "step": 1362
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 2.2333163262984584e-06,
+      "loss": 0.2051,
+      "step": 1363
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 2.232290948259762e-06,
+      "loss": 0.2081,
+      "step": 1364
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 2.231265120757344e-06,
+      "loss": 0.2213,
+      "step": 1365
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 2.2302388444208332e-06,
+      "loss": 0.1996,
+      "step": 1366
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 2.229212119880136e-06,
+      "loss": 0.2154,
+      "step": 1367
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 2.228184947765432e-06,
+      "loss": 0.2323,
+      "step": 1368
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 2.2271573287071775e-06,
+      "loss": 0.2286,
+      "step": 1369
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 2.226129263336101e-06,
+      "loss": 0.1951,
+      "step": 1370
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 2.225100752283206e-06,
+      "loss": 0.2081,
+      "step": 1371
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 2.22407179617977e-06,
+      "loss": 0.209,
+      "step": 1372
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 2.2230423956573423e-06,
+      "loss": 0.2276,
+      "step": 1373
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 2.222012551347747e-06,
+      "loss": 0.2093,
+      "step": 1374
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 2.2209822638830777e-06,
+      "loss": 0.2482,
+      "step": 1375
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 2.2199515338957036e-06,
+      "loss": 0.2034,
+      "step": 1376
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 2.2189203620182616e-06,
+      "loss": 0.1859,
+      "step": 1377
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 2.2178887488836635e-06,
+      "loss": 0.2161,
+      "step": 1378
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 2.21685669512509e-06,
+      "loss": 0.2004,
+      "step": 1379
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 2.2158242013759916e-06,
+      "loss": 0.2136,
+      "step": 1380
+    },
+    {
+      "epoch": 10.24,
+      "eval_loss": 1.839149832725525,
+      "eval_runtime": 1.8195,
+      "eval_samples_per_second": 36.274,
+      "eval_steps_per_second": 3.298,
+      "step": 1380
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 2.21479126827009e-06,
+      "loss": 0.2017,
+      "step": 1381
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 2.213757896441377e-06,
+      "loss": 0.2126,
+      "step": 1382
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 2.2127240865241126e-06,
+      "loss": 0.2035,
+      "step": 1383
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 2.2116898391528255e-06,
+      "loss": 0.1671,
+      "step": 1384
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 2.2106551549623146e-06,
+      "loss": 0.2148,
+      "step": 1385
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 2.2096200345876452e-06,
+      "loss": 0.2163,
+      "step": 1386
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 2.208584478664151e-06,
+      "loss": 0.2498,
+      "step": 1387
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 2.2075484878274325e-06,
+      "loss": 0.2238,
+      "step": 1388
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 2.2065120627133588e-06,
+      "loss": 0.2106,
+      "step": 1389
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 2.2054752039580627e-06,
+      "loss": 0.1946,
+      "step": 1390
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 2.204437912197946e-06,
+      "loss": 0.2189,
+      "step": 1391
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 2.2034001880696746e-06,
+      "loss": 0.2046,
+      "step": 1392
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 2.2023620322101804e-06,
+      "loss": 0.2032,
+      "step": 1393
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 2.20132344525666e-06,
+      "loss": 0.2211,
+      "step": 1394
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 2.200284427846575e-06,
+      "loss": 0.2201,
+      "step": 1395
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 2.1992449806176507e-06,
+      "loss": 0.1968,
+      "step": 1396
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 2.1982051042078766e-06,
+      "loss": 0.1801,
+      "step": 1397
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 2.1971647992555047e-06,
+      "loss": 0.206,
+      "step": 1398
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 2.196124066399052e-06,
+      "loss": 0.2006,
+      "step": 1399
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 2.1950829062772957e-06,
+      "loss": 0.212,
+      "step": 1400
+    },
+    {
+      "epoch": 10.39,
+      "eval_loss": 1.8139111995697021,
+      "eval_runtime": 1.8187,
+      "eval_samples_per_second": 36.289,
+      "eval_steps_per_second": 3.299,
+      "step": 1400
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 2.1940413195292776e-06,
+      "loss": 0.2221,
+      "step": 1401
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 2.1929993067943e-06,
+      "loss": 0.2081,
+      "step": 1402
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 2.191956868711926e-06,
+      "loss": 0.2236,
+      "step": 1403
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 2.1909140059219817e-06,
+      "loss": 0.2409,
+      "step": 1404
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 2.1898707190645526e-06,
+      "loss": 0.2113,
+      "step": 1405
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 2.188827008779984e-06,
+      "loss": 0.2179,
+      "step": 1406
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 2.187782875708883e-06,
+      "loss": 0.2151,
+      "step": 1407
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 2.1867383204921137e-06,
+      "loss": 0.237,
+      "step": 1408
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 2.185693343770802e-06,
+      "loss": 0.2382,
+      "step": 1409
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 2.18464794618633e-06,
+      "loss": 0.2045,
+      "step": 1410
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 2.18360212838034e-06,
+      "loss": 0.225,
+      "step": 1411
+    },
+    {
+      "epoch": 10.48,
+      "learning_rate": 2.1825558909947305e-06,
+      "loss": 0.191,
+      "step": 1412
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 2.1815092346716595e-06,
+      "loss": 0.2124,
+      "step": 1413
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 2.1804621600535406e-06,
+      "loss": 0.2198,
+      "step": 1414
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 2.1794146677830453e-06,
+      "loss": 0.2184,
+      "step": 1415
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 2.1783667585030995e-06,
+      "loss": 0.215,
+      "step": 1416
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 2.1773184328568876e-06,
+      "loss": 0.2291,
+      "step": 1417
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 2.1762696914878474e-06,
+      "loss": 0.2462,
+      "step": 1418
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 2.175220535039674e-06,
+      "loss": 0.2192,
+      "step": 1419
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 2.174170964156315e-06,
+      "loss": 0.2165,
+      "step": 1420
+    },
+    {
+      "epoch": 10.54,
+      "eval_loss": 1.8249884843826294,
+      "eval_runtime": 1.8177,
+      "eval_samples_per_second": 36.309,
+      "eval_steps_per_second": 3.301,
+      "step": 1420
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 2.1731209794819736e-06,
+      "loss": 0.218,
+      "step": 1421
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 2.172070581661107e-06,
+      "loss": 0.2272,
+      "step": 1422
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 2.171019771338425e-06,
+      "loss": 0.2187,
+      "step": 1423
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 2.169968549158893e-06,
+      "loss": 0.2029,
+      "step": 1424
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 2.1689169157677254e-06,
+      "loss": 0.2272,
+      "step": 1425
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 2.1678648718103935e-06,
+      "loss": 0.2367,
+      "step": 1426
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 2.166812417932616e-06,
+      "loss": 0.2237,
+      "step": 1427
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 2.165759554780367e-06,
+      "loss": 0.2365,
+      "step": 1428
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 2.1647062829998694e-06,
+      "loss": 0.2041,
+      "step": 1429
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 2.163652603237598e-06,
+      "loss": 0.2155,
+      "step": 1430
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 2.162598516140278e-06,
+      "loss": 0.1998,
+      "step": 1431
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 2.1615440223548837e-06,
+      "loss": 0.2312,
+      "step": 1432
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 2.160489122528641e-06,
+      "loss": 0.2055,
+      "step": 1433
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 2.159433817309022e-06,
+      "loss": 0.2103,
+      "step": 1434
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 2.1583781073437504e-06,
+      "loss": 0.194,
+      "step": 1435
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 2.157321993280797e-06,
+      "loss": 0.2173,
+      "step": 1436
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 2.1562654757683807e-06,
+      "loss": 0.2194,
+      "step": 1437
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 2.155208555454969e-06,
+      "loss": 0.2268,
+      "step": 1438
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 2.154151232989274e-06,
+      "loss": 0.236,
+      "step": 1439
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 2.153093509020259e-06,
+      "loss": 0.2087,
+      "step": 1440
+    },
+    {
+      "epoch": 10.69,
+      "eval_loss": 1.8216975927352905,
+      "eval_runtime": 1.8191,
+      "eval_samples_per_second": 36.282,
+      "eval_steps_per_second": 3.298,
+      "step": 1440
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 2.152035384197129e-06,
+      "loss": 0.2014,
+      "step": 1441
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 2.1509768591693382e-06,
+      "loss": 0.2169,
+      "step": 1442
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 2.149917934586585e-06,
+      "loss": 0.2137,
+      "step": 1443
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 2.1488586110988146e-06,
+      "loss": 0.2173,
+      "step": 1444
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 2.1477988893562143e-06,
+      "loss": 0.2361,
+      "step": 1445
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 2.146738770009219e-06,
+      "loss": 0.2135,
+      "step": 1446
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 2.145678253708505e-06,
+      "loss": 0.2011,
+      "step": 1447
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 2.1446173411049943e-06,
+      "loss": 0.2009,
+      "step": 1448
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 2.1435560328498513e-06,
+      "loss": 0.2214,
+      "step": 1449
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 2.1424943295944815e-06,
+      "loss": 0.2059,
+      "step": 1450
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 2.141432231990536e-06,
+      "loss": 0.2252,
+      "step": 1451
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 2.140369740689906e-06,
+      "loss": 0.2036,
+      "step": 1452
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 2.139306856344725e-06,
+      "loss": 0.1987,
+      "step": 1453
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 2.1382435796073666e-06,
+      "loss": 0.2187,
+      "step": 1454
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 2.1371799111304474e-06,
+      "loss": 0.2359,
+      "step": 1455
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 2.1361158515668228e-06,
+      "loss": 0.2242,
+      "step": 1456
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 2.1350514015695885e-06,
+      "loss": 0.2435,
+      "step": 1457
+    },
+    {
+      "epoch": 10.82,
+      "learning_rate": 2.1339865617920797e-06,
+      "loss": 0.2385,
+      "step": 1458
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 2.1329213328878716e-06,
+      "loss": 0.2099,
+      "step": 1459
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 2.1318557155107772e-06,
+      "loss": 0.2228,
+      "step": 1460
+    },
+    {
+      "epoch": 10.83,
+      "eval_loss": 1.8340202569961548,
+      "eval_runtime": 1.8198,
+      "eval_samples_per_second": 36.268,
+      "eval_steps_per_second": 3.297,
+      "step": 1460
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 2.1307897103148495e-06,
+      "loss": 0.2346,
+      "step": 1461
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 2.1297233179543773e-06,
+      "loss": 0.2095,
+      "step": 1462
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 2.1286565390838893e-06,
+      "loss": 0.2186,
+      "step": 1463
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 2.1275893743581496e-06,
+      "loss": 0.2131,
+      "step": 1464
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 2.1265218244321603e-06,
+      "loss": 0.2194,
+      "step": 1465
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 2.1254538899611598e-06,
+      "loss": 0.2459,
+      "step": 1466
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 2.124385571600622e-06,
+      "loss": 0.2182,
+      "step": 1467
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 2.123316870006256e-06,
+      "loss": 0.2315,
+      "step": 1468
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 2.122247785834008e-06,
+      "loss": 0.2606,
+      "step": 1469
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 2.1211783197400573e-06,
+      "loss": 0.2385,
+      "step": 1470
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 2.120108472380818e-06,
+      "loss": 0.2026,
+      "step": 1471
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 2.1190382444129385e-06,
+      "loss": 0.245,
+      "step": 1472
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 2.1179676364933008e-06,
+      "loss": 0.2323,
+      "step": 1473
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 2.11689664927902e-06,
+      "loss": 0.2144,
+      "step": 1474
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 2.1158252834274434e-06,
+      "loss": 0.2287,
+      "step": 1475
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 2.1147535395961517e-06,
+      "loss": 0.2197,
+      "step": 1476
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 2.1136814184429575e-06,
+      "loss": 0.2203,
+      "step": 1477
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 2.112608920625904e-06,
+      "loss": 0.204,
+      "step": 1478
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 2.111536046803267e-06,
+      "loss": 0.1965,
+      "step": 1479
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 2.110462797633552e-06,
+      "loss": 0.2086,
+      "step": 1480
+    },
+    {
+      "epoch": 10.98,
+      "eval_loss": 1.8305644989013672,
+      "eval_runtime": 1.8165,
+      "eval_samples_per_second": 36.334,
+      "eval_steps_per_second": 3.303,
+      "step": 1480
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 2.1093891737754943e-06,
+      "loss": 0.2275,
+      "step": 1481
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 2.1083151758880614e-06,
+      "loss": 0.2161,
+      "step": 1482
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 2.107240804630449e-06,
+      "loss": 0.1609,
+      "step": 1483
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 2.106166060662081e-06,
+      "loss": 0.1444,
+      "step": 1484
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 2.1050909446426123e-06,
+      "loss": 0.1637,
+      "step": 1485
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 2.1040154572319237e-06,
+      "loss": 0.1486,
+      "step": 1486
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 2.102939599090126e-06,
+      "loss": 0.1829,
+      "step": 1487
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 2.1018633708775567e-06,
+      "loss": 0.165,
+      "step": 1488
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 2.1007867732547804e-06,
+      "loss": 0.1662,
+      "step": 1489
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 2.0997098068825886e-06,
+      "loss": 0.1608,
+      "step": 1490
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 2.0986324724219988e-06,
+      "loss": 0.1672,
+      "step": 1491
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 2.0975547705342547e-06,
+      "loss": 0.1561,
+      "step": 1492
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 2.0964767018808256e-06,
+      "loss": 0.1429,
+      "step": 1493
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 2.095398267123406e-06,
+      "loss": 0.1675,
+      "step": 1494
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 2.0943194669239144e-06,
+      "loss": 0.1508,
+      "step": 1495
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 2.0932403019444948e-06,
+      "loss": 0.1641,
+      "step": 1496
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 2.0921607728475137e-06,
+      "loss": 0.157,
+      "step": 1497
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 2.0910808802955623e-06,
+      "loss": 0.1745,
+      "step": 1498
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 2.090000624951454e-06,
+      "loss": 0.1635,
+      "step": 1499
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 2.0889200074782253e-06,
+      "loss": 0.1695,
+      "step": 1500
+    },
+    {
+      "epoch": 11.13,
+      "eval_loss": 1.9333045482635498,
+      "eval_runtime": 1.8228,
+      "eval_samples_per_second": 36.208,
+      "eval_steps_per_second": 3.292,
+      "step": 1500
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 2.0878390285391355e-06,
+      "loss": 0.1646,
+      "step": 1501
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 2.0867576887976644e-06,
+      "loss": 0.144,
+      "step": 1502
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 2.0856759889175145e-06,
+      "loss": 0.1902,
+      "step": 1503
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 2.0845939295626084e-06,
+      "loss": 0.1385,
+      "step": 1504
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 2.0835115113970907e-06,
+      "loss": 0.1526,
+      "step": 1505
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 2.0824287350853245e-06,
+      "loss": 0.1866,
+      "step": 1506
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 2.0813456012918935e-06,
+      "loss": 0.1546,
+      "step": 1507
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 2.0802621106816013e-06,
+      "loss": 0.159,
+      "step": 1508
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 2.0791782639194703e-06,
+      "loss": 0.1543,
+      "step": 1509
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 2.078094061670741e-06,
+      "loss": 0.1559,
+      "step": 1510
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 2.077009504600872e-06,
+      "loss": 0.1456,
+      "step": 1511
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 2.07592459337554e-06,
+      "loss": 0.1616,
+      "step": 1512
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 2.0748393286606403e-06,
+      "loss": 0.1597,
+      "step": 1513
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 2.073753711122282e-06,
+      "loss": 0.1395,
+      "step": 1514
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 2.0726677414267948e-06,
+      "loss": 0.1543,
+      "step": 1515
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 2.0715814202407214e-06,
+      "loss": 0.1746,
+      "step": 1516
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 2.070494748230822e-06,
+      "loss": 0.1458,
+      "step": 1517
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 2.0694077260640703e-06,
+      "loss": 0.149,
+      "step": 1518
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 2.068320354407657e-06,
+      "loss": 0.1455,
+      "step": 1519
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 2.0672326339289863e-06,
+      "loss": 0.1474,
+      "step": 1520
+    },
+    {
+      "epoch": 11.28,
+      "eval_loss": 1.9166902303695679,
+      "eval_runtime": 1.8203,
+      "eval_samples_per_second": 36.257,
+      "eval_steps_per_second": 3.296,
+      "step": 1520
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 2.066144565295676e-06,
+      "loss": 0.1475,
+      "step": 1521
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 2.0650561491755586e-06,
+      "loss": 0.1547,
+      "step": 1522
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 2.0639673862366798e-06,
+      "loss": 0.1667,
+      "step": 1523
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 2.0628782771472965e-06,
+      "loss": 0.1537,
+      "step": 1524
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 2.0617888225758814e-06,
+      "loss": 0.1564,
+      "step": 1525
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 2.060699023191116e-06,
+      "loss": 0.1466,
+      "step": 1526
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 2.059608879661894e-06,
+      "loss": 0.1637,
+      "step": 1527
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 2.0585183926573222e-06,
+      "loss": 0.1859,
+      "step": 1528
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 2.057427562846717e-06,
+      "loss": 0.17,
+      "step": 1529
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 2.0563363908996042e-06,
+      "loss": 0.1468,
+      "step": 1530
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 2.055244877485721e-06,
+      "loss": 0.1462,
+      "step": 1531
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 2.0541530232750148e-06,
+      "loss": 0.1587,
+      "step": 1532
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 2.05306082893764e-06,
+      "loss": 0.1647,
+      "step": 1533
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 2.051968295143961e-06,
+      "loss": 0.1618,
+      "step": 1534
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 2.0508754225645514e-06,
+      "loss": 0.1623,
+      "step": 1535
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 2.049782211870191e-06,
+      "loss": 0.1508,
+      "step": 1536
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 2.048688663731868e-06,
+      "loss": 0.1568,
+      "step": 1537
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 2.047594778820778e-06,
+      "loss": 0.1691,
+      "step": 1538
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 2.0465005578083227e-06,
+      "loss": 0.1516,
+      "step": 1539
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 2.0454060013661105e-06,
+      "loss": 0.1587,
+      "step": 1540
+    },
+    {
+      "epoch": 11.43,
+      "eval_loss": 1.935088872909546,
+      "eval_runtime": 1.8221,
+      "eval_samples_per_second": 36.223,
+      "eval_steps_per_second": 3.293,
+      "step": 1540
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 2.0443111101659557e-06,
+      "loss": 0.1629,
+      "step": 1541
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 2.0432158848798777e-06,
+      "loss": 0.1838,
+      "step": 1542
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 2.0421203261801005e-06,
+      "loss": 0.1741,
+      "step": 1543
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 2.041024434739055e-06,
+      "loss": 0.1582,
+      "step": 1544
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 2.0399282112293732e-06,
+      "loss": 0.1503,
+      "step": 1545
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 2.038831656323893e-06,
+      "loss": 0.1648,
+      "step": 1546
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 2.0377347706956556e-06,
+      "loss": 0.1585,
+      "step": 1547
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 2.036637555017904e-06,
+      "loss": 0.1608,
+      "step": 1548
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 2.035540009964085e-06,
+      "loss": 0.1648,
+      "step": 1549
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 2.0344421362078466e-06,
+      "loss": 0.1427,
+      "step": 1550
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 2.0333439344230396e-06,
+      "loss": 0.1476,
+      "step": 1551
+    },
+    {
+      "epoch": 11.52,
+      "learning_rate": 2.0322454052837154e-06,
+      "loss": 0.1902,
+      "step": 1552
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 2.0311465494641263e-06,
+      "loss": 0.1691,
+      "step": 1553
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 2.030047367638725e-06,
+      "loss": 0.1442,
+      "step": 1554
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 2.028947860482166e-06,
+      "loss": 0.1557,
+      "step": 1555
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 2.0278480286693e-06,
+      "loss": 0.1517,
+      "step": 1556
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 2.026747872875181e-06,
+      "loss": 0.1599,
+      "step": 1557
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 2.0256473937750584e-06,
+      "loss": 0.1677,
+      "step": 1558
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 2.0245465920443827e-06,
+      "loss": 0.1647,
+      "step": 1559
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 2.0234454683588004e-06,
+      "loss": 0.1649,
+      "step": 1560
+    },
+    {
+      "epoch": 11.58,
+      "eval_loss": 1.934259295463562,
+      "eval_runtime": 1.816,
+      "eval_samples_per_second": 36.344,
+      "eval_steps_per_second": 3.304,
+      "step": 1560
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 2.022344023394157e-06,
+      "loss": 0.1645,
+      "step": 1561
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 2.0212422578264938e-06,
+      "loss": 0.1557,
+      "step": 1562
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 2.020140172332051e-06,
+      "loss": 0.1676,
+      "step": 1563
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 2.019037767587263e-06,
+      "loss": 0.1503,
+      "step": 1564
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 2.0179350442687617e-06,
+      "loss": 0.164,
+      "step": 1565
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 2.0168320030533733e-06,
+      "loss": 0.1703,
+      "step": 1566
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 2.0157286446181195e-06,
+      "loss": 0.1781,
+      "step": 1567
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 2.0146249696402176e-06,
+      "loss": 0.1759,
+      "step": 1568
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 2.0135209787970783e-06,
+      "loss": 0.1765,
+      "step": 1569
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 2.012416672766306e-06,
+      "loss": 0.1566,
+      "step": 1570
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 2.0113120522256998e-06,
+      "loss": 0.1542,
+      "step": 1571
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 2.0102071178532495e-06,
+      "loss": 0.1532,
+      "step": 1572
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 2.0091018703271404e-06,
+      "loss": 0.1582,
+      "step": 1573
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 2.0079963103257477e-06,
+      "loss": 0.1855,
+      "step": 1574
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 2.0068904385276397e-06,
+      "loss": 0.1666,
+      "step": 1575
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 2.0057842556115757e-06,
+      "loss": 0.1824,
+      "step": 1576
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 2.0046777622565054e-06,
+      "loss": 0.1739,
+      "step": 1577
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 2.00357095914157e-06,
+      "loss": 0.1748,
+      "step": 1578
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 2.0024638469461005e-06,
+      "loss": 0.1607,
+      "step": 1579
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 2.001356426349617e-06,
+      "loss": 0.1258,
+      "step": 1580
+    },
+    {
+      "epoch": 11.73,
+      "eval_loss": 1.9253904819488525,
+      "eval_runtime": 1.8198,
+      "eval_samples_per_second": 36.268,
+      "eval_steps_per_second": 3.297,
+      "step": 1580
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 2.0002486980318296e-06,
+      "loss": 0.1719,
+      "step": 1581
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 1.999140662672637e-06,
+      "loss": 0.1743,
+      "step": 1582
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 1.9980323209521266e-06,
+      "loss": 0.1647,
+      "step": 1583
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 1.9969236735505736e-06,
+      "loss": 0.1865,
+      "step": 1584
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 1.9958147211484404e-06,
+      "loss": 0.1557,
+      "step": 1585
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 1.9947054644263774e-06,
+      "loss": 0.175,
+      "step": 1586
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 1.9935959040652214e-06,
+      "loss": 0.1589,
+      "step": 1587
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 1.9924860407459956e-06,
+      "loss": 0.1588,
+      "step": 1588
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 1.991375875149909e-06,
+      "loss": 0.1453,
+      "step": 1589
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 1.990265407958357e-06,
+      "loss": 0.1665,
+      "step": 1590
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 1.9891546398529184e-06,
+      "loss": 0.1656,
+      "step": 1591
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 1.988043571515358e-06,
+      "loss": 0.1629,
+      "step": 1592
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 1.986932203627625e-06,
+      "loss": 0.159,
+      "step": 1593
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 1.9858205368718524e-06,
+      "loss": 0.1812,
+      "step": 1594
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 1.984708571930355e-06,
+      "loss": 0.1575,
+      "step": 1595
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 1.983596309485632e-06,
+      "loss": 0.1724,
+      "step": 1596
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 1.9824837502203664e-06,
+      "loss": 0.1631,
+      "step": 1597
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 1.9813708948174205e-06,
+      "loss": 0.1811,
+      "step": 1598
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 1.9802577439598417e-06,
+      "loss": 0.1763,
+      "step": 1599
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 1.9791442983308547e-06,
+      "loss": 0.1622,
+      "step": 1600
+    },
+    {
+      "epoch": 11.87,
+      "eval_loss": 1.931401014328003,
+      "eval_runtime": 1.8163,
+      "eval_samples_per_second": 36.337,
+      "eval_steps_per_second": 3.303,
+      "step": 1600
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 1.9780305586138694e-06,
+      "loss": 0.1736,
+      "step": 1601
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 1.976916525492473e-06,
+      "loss": 0.1576,
+      "step": 1602
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 1.9758021996504347e-06,
+      "loss": 0.158,
+      "step": 1603
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 1.9746875817717026e-06,
+      "loss": 0.179,
+      "step": 1604
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 1.9735726725404036e-06,
+      "loss": 0.1797,
+      "step": 1605
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 1.972457472640844e-06,
+      "loss": 0.1574,
+      "step": 1606
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 1.9713419827575093e-06,
+      "loss": 0.1638,
+      "step": 1607
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 1.9702262035750605e-06,
+      "loss": 0.1769,
+      "step": 1608
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 1.9691101357783398e-06,
+      "loss": 0.155,
+      "step": 1609
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 1.9679937800523627e-06,
+      "loss": 0.1684,
+      "step": 1610
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 1.966877137082324e-06,
+      "loss": 0.1637,
+      "step": 1611
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 1.965760207553594e-06,
+      "loss": 0.1859,
+      "step": 1612
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 1.9646429921517196e-06,
+      "loss": 0.1746,
+      "step": 1613
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 1.963525491562421e-06,
+      "loss": 0.1665,
+      "step": 1614
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 1.9624077064715964e-06,
+      "loss": 0.1608,
+      "step": 1615
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 1.9612896375653167e-06,
+      "loss": 0.1609,
+      "step": 1616
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 1.960171285529827e-06,
+      "loss": 0.1798,
+      "step": 1617
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 1.959052651051547e-06,
+      "loss": 0.1413,
+      "step": 1618
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 1.9579337348170693e-06,
+      "loss": 0.1299,
+      "step": 1619
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 1.9568145375131603e-06,
+      "loss": 0.1116,
+      "step": 1620
+    },
+    {
+      "epoch": 12.02,
+      "eval_loss": 2.004352569580078,
+      "eval_runtime": 1.8082,
+      "eval_samples_per_second": 36.5,
+      "eval_steps_per_second": 3.318,
+      "step": 1620
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 1.955695059826757e-06,
+      "loss": 0.1381,
+      "step": 1621
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 1.95457530244497e-06,
+      "loss": 0.1204,
+      "step": 1622
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 1.953455266055081e-06,
+      "loss": 0.1353,
+      "step": 1623
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 1.952334951344544e-06,
+      "loss": 0.1335,
+      "step": 1624
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 1.951214359000982e-06,
+      "loss": 0.1187,
+      "step": 1625
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 1.95009348971219e-06,
+      "loss": 0.1013,
+      "step": 1626
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 1.9489723441661323e-06,
+      "loss": 0.1077,
+      "step": 1627
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 1.9478509230509427e-06,
+      "loss": 0.1112,
+      "step": 1628
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 1.9467292270549243e-06,
+      "loss": 0.1218,
+      "step": 1629
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 1.945607256866549e-06,
+      "loss": 0.0985,
+      "step": 1630
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 1.9444850131744565e-06,
+      "loss": 0.0901,
+      "step": 1631
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 1.9433624966674554e-06,
+      "loss": 0.1046,
+      "step": 1632
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 1.9422397080345203e-06,
+      "loss": 0.1086,
+      "step": 1633
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 1.941116647964794e-06,
+      "loss": 0.1245,
+      "step": 1634
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 1.939993317147586e-06,
+      "loss": 0.1156,
+      "step": 1635
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 1.938869716272371e-06,
+      "loss": 0.1297,
+      "step": 1636
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 1.93774584602879e-06,
+      "loss": 0.1267,
+      "step": 1637
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 1.936621707106649e-06,
+      "loss": 0.129,
+      "step": 1638
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 1.93549730019592e-06,
+      "loss": 0.1012,
+      "step": 1639
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 1.934372625986738e-06,
+      "loss": 0.1107,
+      "step": 1640
+    },
+    {
+      "epoch": 12.17,
+      "eval_loss": 2.0484869480133057,
+      "eval_runtime": 1.8159,
+      "eval_samples_per_second": 36.345,
+      "eval_steps_per_second": 3.304,
+      "step": 1640
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 1.933247685169403e-06,
+      "loss": 0.1188,
+      "step": 1641
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 1.932122478434378e-06,
+      "loss": 0.1115,
+      "step": 1642
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 1.93099700647229e-06,
+      "loss": 0.1198,
+      "step": 1643
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 1.9298712699739284e-06,
+      "loss": 0.1214,
+      "step": 1644
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 1.928745269630245e-06,
+      "loss": 0.1241,
+      "step": 1645
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 1.9276190061323537e-06,
+      "loss": 0.1204,
+      "step": 1646
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 1.926492480171529e-06,
+      "loss": 0.1099,
+      "step": 1647
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 1.9253656924392076e-06,
+      "loss": 0.1295,
+      "step": 1648
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 1.9242386436269867e-06,
+      "loss": 0.1321,
+      "step": 1649
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 1.9231113344266226e-06,
+      "loss": 0.1302,
+      "step": 1650
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 1.9219837655300337e-06,
+      "loss": 0.1236,
+      "step": 1651
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 1.9208559376292958e-06,
+      "loss": 0.1282,
+      "step": 1652
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 1.919727851416644e-06,
+      "loss": 0.1214,
+      "step": 1653
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 1.918599507584473e-06,
+      "loss": 0.1217,
+      "step": 1654
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 1.917470906825334e-06,
+      "loss": 0.1156,
+      "step": 1655
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 1.9163420498319373e-06,
+      "loss": 0.1148,
+      "step": 1656
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 1.9152129372971503e-06,
+      "loss": 0.116,
+      "step": 1657
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 1.914083569913996e-06,
+      "loss": 0.1223,
+      "step": 1658
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 1.9129539483756553e-06,
+      "loss": 0.1203,
+      "step": 1659
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 1.9118240733754643e-06,
+      "loss": 0.1063,
+      "step": 1660
+    },
+    {
+      "epoch": 12.32,
+      "eval_loss": 2.0215635299682617,
+      "eval_runtime": 1.8069,
+      "eval_samples_per_second": 36.527,
+      "eval_steps_per_second": 3.321,
+      "step": 1660
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 1.9106939456069153e-06,
+      "loss": 0.1223,
+      "step": 1661
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 1.909563565763654e-06,
+      "loss": 0.1032,
+      "step": 1662
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 1.908432934539484e-06,
+      "loss": 0.1281,
+      "step": 1663
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 1.9073020526283601e-06,
+      "loss": 0.1324,
+      "step": 1664
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 1.9061709207243926e-06,
+      "loss": 0.1365,
+      "step": 1665
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 1.9050395395218447e-06,
+      "loss": 0.1324,
+      "step": 1666
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 1.903907909715133e-06,
+      "loss": 0.118,
+      "step": 1667
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 1.9027760319988261e-06,
+      "loss": 0.1087,
+      "step": 1668
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 1.9016439070676452e-06,
+      "loss": 0.1374,
+      "step": 1669
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 1.9005115356164632e-06,
+      "loss": 0.0969,
+      "step": 1670
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 1.8993789183403041e-06,
+      "loss": 0.1006,
+      "step": 1671
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 1.8982460559343434e-06,
+      "loss": 0.1123,
+      "step": 1672
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 1.8971129490939062e-06,
+      "loss": 0.1188,
+      "step": 1673
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 1.8959795985144677e-06,
+      "loss": 0.1091,
+      "step": 1674
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 1.8948460048916534e-06,
+      "loss": 0.1197,
+      "step": 1675
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 1.8937121689212378e-06,
+      "loss": 0.109,
+      "step": 1676
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 1.8925780912991434e-06,
+      "loss": 0.0979,
+      "step": 1677
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 1.8914437727214417e-06,
+      "loss": 0.1192,
+      "step": 1678
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 1.890309213884352e-06,
+      "loss": 0.1263,
+      "step": 1679
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 1.8891744154842412e-06,
+      "loss": 0.1291,
+      "step": 1680
+    },
+    {
+      "epoch": 12.47,
+      "eval_loss": 2.0126023292541504,
+      "eval_runtime": 1.818,
+      "eval_samples_per_second": 36.304,
+      "eval_steps_per_second": 3.3,
+      "step": 1680
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 1.8880393782176225e-06,
+      "loss": 0.1119,
+      "step": 1681
+    },
+    {
+      "epoch": 12.48,
+      "learning_rate": 1.8869041027811577e-06,
+      "loss": 0.1217,
+      "step": 1682
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 1.8857685898716512e-06,
+      "loss": 0.126,
+      "step": 1683
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 1.8846328401860568e-06,
+      "loss": 0.1345,
+      "step": 1684
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 1.8834968544214718e-06,
+      "loss": 0.113,
+      "step": 1685
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 1.8823606332751385e-06,
+      "loss": 0.1177,
+      "step": 1686
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 1.881224177444444e-06,
+      "loss": 0.1279,
+      "step": 1687
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 1.8800874876269196e-06,
+      "loss": 0.1327,
+      "step": 1688
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 1.8789505645202395e-06,
+      "loss": 0.1262,
+      "step": 1689
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 1.877813408822222e-06,
+      "loss": 0.1012,
+      "step": 1690
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 1.8766760212308272e-06,
+      "loss": 0.1349,
+      "step": 1691
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 1.8755384024441587e-06,
+      "loss": 0.1418,
+      "step": 1692
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 1.8744005531604606e-06,
+      "loss": 0.1262,
+      "step": 1693
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 1.8732624740781197e-06,
+      "loss": 0.1236,
+      "step": 1694
+    },
+    {
+      "epoch": 12.58,
+      "learning_rate": 1.8721241658956636e-06,
+      "loss": 0.1181,
+      "step": 1695
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 1.8709856293117592e-06,
+      "loss": 0.1179,
+      "step": 1696
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 1.869846865025216e-06,
+      "loss": 0.1113,
+      "step": 1697
+    },
+    {
+      "epoch": 12.6,
+      "learning_rate": 1.8687078737349807e-06,
+      "loss": 0.1127,
+      "step": 1698
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 1.8675686561401413e-06,
+      "loss": 0.128,
+      "step": 1699
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 1.8664292129399238e-06,
+      "loss": 0.1313,
+      "step": 1700
+    },
+    {
+      "epoch": 12.62,
+      "eval_loss": 2.034505605697632,
+      "eval_runtime": 1.8189,
+      "eval_samples_per_second": 36.285,
+      "eval_steps_per_second": 3.299,
+      "step": 1700
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 1.8652895448336932e-06,
+      "loss": 0.1505,
+      "step": 1701
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 1.864149652520951e-06,
+      "loss": 0.1269,
+      "step": 1702
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 1.8630095367013387e-06,
+      "loss": 0.131,
+      "step": 1703
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 1.8618691980746331e-06,
+      "loss": 0.089,
+      "step": 1704
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 1.8607286373407491e-06,
+      "loss": 0.1301,
+      "step": 1705
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 1.8595878551997363e-06,
+      "loss": 0.1204,
+      "step": 1706
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 1.8584468523517819e-06,
+      "loss": 0.1375,
+      "step": 1707
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 1.8573056294972068e-06,
+      "loss": 0.122,
+      "step": 1708
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 1.8561641873364693e-06,
+      "loss": 0.1287,
+      "step": 1709
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 1.85502252657016e-06,
+      "loss": 0.1342,
+      "step": 1710
+    },
+    {
+      "epoch": 12.7,
+      "learning_rate": 1.8538806478990047e-06,
+      "loss": 0.124,
+      "step": 1711
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 1.8527385520238632e-06,
+      "loss": 0.1155,
+      "step": 1712
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 1.851596239645728e-06,
+      "loss": 0.1144,
+      "step": 1713
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 1.8504537114657252e-06,
+      "loss": 0.112,
+      "step": 1714
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 1.849310968185112e-06,
+      "loss": 0.0993,
+      "step": 1715
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 1.8481680105052799e-06,
+      "loss": 0.128,
+      "step": 1716
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 1.8470248391277491e-06,
+      "loss": 0.13,
+      "step": 1717
+    },
+    {
+      "epoch": 12.75,
+      "learning_rate": 1.8458814547541736e-06,
+      "loss": 0.1365,
+      "step": 1718
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 1.8447378580863366e-06,
+      "loss": 0.1158,
+      "step": 1719
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 1.8435940498261518e-06,
+      "loss": 0.1273,
+      "step": 1720
+    },
+    {
+      "epoch": 12.76,
+      "eval_loss": 2.0201988220214844,
+      "eval_runtime": 1.8203,
+      "eval_samples_per_second": 36.258,
+      "eval_steps_per_second": 3.296,
+      "step": 1720
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 1.8424500306756637e-06,
+      "loss": 0.1384,
+      "step": 1721
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 1.8413058013370448e-06,
+      "loss": 0.1136,
+      "step": 1722
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 1.8401613625125977e-06,
+      "loss": 0.1223,
+      "step": 1723
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 1.8390167149047533e-06,
+      "loss": 0.1209,
+      "step": 1724
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 1.8378718592160704e-06,
+      "loss": 0.121,
+      "step": 1725
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 1.8367267961492357e-06,
+      "loss": 0.1232,
+      "step": 1726
+    },
+    {
+      "epoch": 12.82,
+      "learning_rate": 1.8355815264070632e-06,
+      "loss": 0.1343,
+      "step": 1727
+    },
+    {
+      "epoch": 12.82,
+      "learning_rate": 1.8344360506924942e-06,
+      "loss": 0.1193,
+      "step": 1728
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 1.8332903697085947e-06,
+      "loss": 0.1058,
+      "step": 1729
+    },
+    {
+      "epoch": 12.84,
+      "learning_rate": 1.832144484158559e-06,
+      "loss": 0.1171,
+      "step": 1730
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 1.8309983947457058e-06,
+      "loss": 0.1157,
+      "step": 1731
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 1.8298521021734786e-06,
+      "loss": 0.1541,
+      "step": 1732
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 1.8287056071454464e-06,
+      "loss": 0.1352,
+      "step": 1733
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 1.827558910365302e-06,
+      "loss": 0.1392,
+      "step": 1734
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 1.8264120125368617e-06,
+      "loss": 0.1171,
+      "step": 1735
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 1.8252649143640659e-06,
+      "loss": 0.1345,
+      "step": 1736
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 1.8241176165509776e-06,
+      "loss": 0.122,
+      "step": 1737
+    },
+    {
+      "epoch": 12.9,
+      "learning_rate": 1.822970119801783e-06,
+      "loss": 0.1335,
+      "step": 1738
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 1.8218224248207888e-06,
+      "loss": 0.1212,
+      "step": 1739
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 1.8206745323124253e-06,
+      "loss": 0.106,
+      "step": 1740
+    },
+    {
+      "epoch": 12.91,
+      "eval_loss": 2.01716947555542,
+      "eval_runtime": 1.818,
+      "eval_samples_per_second": 36.303,
+      "eval_steps_per_second": 3.3,
+      "step": 1740
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 1.8195264429812422e-06,
+      "loss": 0.1142,
+      "step": 1741
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 1.8183781575319118e-06,
+      "loss": 0.1236,
+      "step": 1742
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 1.8172296766692255e-06,
+      "loss": 0.1247,
+      "step": 1743
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 1.8160810010980948e-06,
+      "loss": 0.1192,
+      "step": 1744
+    },
+    {
+      "epoch": 12.95,
+      "learning_rate": 1.8149321315235514e-06,
+      "loss": 0.1112,
+      "step": 1745
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 1.8137830686507452e-06,
+      "loss": 0.1532,
+      "step": 1746
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 1.8126338131849456e-06,
+      "loss": 0.1058,
+      "step": 1747
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 1.811484365831539e-06,
+      "loss": 0.1335,
+      "step": 1748
+    },
+    {
+      "epoch": 12.98,
+      "learning_rate": 1.8103347272960316e-06,
+      "loss": 0.1156,
+      "step": 1749
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 1.8091848982840445e-06,
+      "loss": 0.1402,
+      "step": 1750
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 1.8080348795013176e-06,
+      "loss": 0.1175,
+      "step": 1751
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 1.8068846716537064e-06,
+      "loss": 0.1127,
+      "step": 1752
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 1.8057342754471824e-06,
+      "loss": 0.0888,
+      "step": 1753
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 1.8045836915878331e-06,
+      "loss": 0.0856,
+      "step": 1754
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 1.8034329207818611e-06,
+      "loss": 0.0969,
+      "step": 1755
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 1.8022819637355833e-06,
+      "loss": 0.1029,
+      "step": 1756
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 1.8011308211554313e-06,
+      "loss": 0.0898,
+      "step": 1757
+    },
+    {
+      "epoch": 13.05,
+      "learning_rate": 1.7999794937479517e-06,
+      "loss": 0.0848,
+      "step": 1758
+    },
+    {
+      "epoch": 13.05,
+      "learning_rate": 1.7988279822198018e-06,
+      "loss": 0.0915,
+      "step": 1759
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 1.7976762872777542e-06,
+      "loss": 0.0967,
+      "step": 1760
+    },
+    {
+      "epoch": 13.06,
+      "eval_loss": 2.1586484909057617,
+      "eval_runtime": 1.8105,
+      "eval_samples_per_second": 36.455,
+      "eval_steps_per_second": 3.314,
+      "step": 1760
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 1.796524409628693e-06,
+      "loss": 0.0848,
+      "step": 1761
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 1.7953723499796156e-06,
+      "loss": 0.1068,
+      "step": 1762
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 1.7942201090376293e-06,
+      "loss": 0.0799,
+      "step": 1763
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 1.7930676875099545e-06,
+      "loss": 0.0811,
+      "step": 1764
+    },
+    {
+      "epoch": 13.1,
+      "learning_rate": 1.7919150861039212e-06,
+      "loss": 0.0851,
+      "step": 1765
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 1.7907623055269704e-06,
+      "loss": 0.0774,
+      "step": 1766
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 1.7896093464866535e-06,
+      "loss": 0.0971,
+      "step": 1767
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 1.7884562096906296e-06,
+      "loss": 0.0906,
+      "step": 1768
+    },
+    {
+      "epoch": 13.13,
+      "learning_rate": 1.7873028958466686e-06,
+      "loss": 0.0854,
+      "step": 1769
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 1.7861494056626492e-06,
+      "loss": 0.0786,
+      "step": 1770
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 1.7849957398465567e-06,
+      "loss": 0.0875,
+      "step": 1771
+    },
+    {
+      "epoch": 13.15,
+      "learning_rate": 1.7838418991064865e-06,
+      "loss": 0.0809,
+      "step": 1772
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 1.7826878841506389e-06,
+      "loss": 0.091,
+      "step": 1773
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 1.7815336956873234e-06,
+      "loss": 0.0774,
+      "step": 1774
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 1.780379334424954e-06,
+      "loss": 0.1091,
+      "step": 1775
+    },
+    {
+      "epoch": 13.18,
+      "learning_rate": 1.7792248010720523e-06,
+      "loss": 0.0855,
+      "step": 1776
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 1.7780700963372444e-06,
+      "loss": 0.0785,
+      "step": 1777
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 1.776915220929263e-06,
+      "loss": 0.0921,
+      "step": 1778
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 1.7757601755569433e-06,
+      "loss": 0.0837,
+      "step": 1779
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 1.7746049609292279e-06,
+      "loss": 0.0916,
+      "step": 1780
+    },
+    {
+      "epoch": 13.21,
+      "eval_loss": 2.1249094009399414,
+      "eval_runtime": 1.8169,
+      "eval_samples_per_second": 36.326,
+      "eval_steps_per_second": 3.302,
+      "step": 1780
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 1.7734495777551601e-06,
+      "loss": 0.0914,
+      "step": 1781
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 1.772294026743889e-06,
+      "loss": 0.0817,
+      "step": 1782
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 1.771138308604665e-06,
+      "loss": 0.0904,
+      "step": 1783
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 1.769982424046843e-06,
+      "loss": 0.0899,
+      "step": 1784
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 1.7688263737798778e-06,
+      "loss": 0.0865,
+      "step": 1785
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 1.7676701585133282e-06,
+      "loss": 0.0937,
+      "step": 1786
+    },
+    {
+      "epoch": 13.26,
+      "learning_rate": 1.7665137789568522e-06,
+      "loss": 0.0824,
+      "step": 1787
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 1.7653572358202102e-06,
+      "loss": 0.087,
+      "step": 1788
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 1.7642005298132618e-06,
+      "loss": 0.091,
+      "step": 1789
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 1.763043661645968e-06,
+      "loss": 0.0876,
+      "step": 1790
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 1.7618866320283875e-06,
+      "loss": 0.0885,
+      "step": 1791
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 1.7607294416706797e-06,
+      "loss": 0.0832,
+      "step": 1792
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 1.7595720912831023e-06,
+      "loss": 0.1049,
+      "step": 1793
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 1.75841458157601e-06,
+      "loss": 0.09,
+      "step": 1794
+    },
+    {
+      "epoch": 13.32,
+      "learning_rate": 1.7572569132598566e-06,
+      "loss": 0.0875,
+      "step": 1795
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 1.756099087045193e-06,
+      "loss": 0.0861,
+      "step": 1796
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 1.7549411036426674e-06,
+      "loss": 0.0843,
+      "step": 1797
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 1.753782963763023e-06,
+      "loss": 0.0757,
+      "step": 1798
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 1.7526246681171011e-06,
+      "loss": 0.0941,
+      "step": 1799
+    },
+    {
+      "epoch": 13.36,
+      "learning_rate": 1.7514662174158363e-06,
+      "loss": 0.0876,
+      "step": 1800
+    },
+    {
+      "epoch": 13.36,
+      "eval_loss": 2.1199069023132324,
+      "eval_runtime": 1.8232,
+      "eval_samples_per_second": 36.2,
+      "eval_steps_per_second": 3.291,
+      "step": 1800
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 1.7503076123702608e-06,
+      "loss": 0.0911,
+      "step": 1801
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 1.7491488536914995e-06,
+      "loss": 0.0936,
+      "step": 1802
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 1.7479899420907725e-06,
+      "loss": 0.091,
+      "step": 1803
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 1.7468308782793946e-06,
+      "loss": 0.0985,
+      "step": 1804
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 1.7456716629687719e-06,
+      "loss": 0.0893,
+      "step": 1805
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 1.7445122968704055e-06,
+      "loss": 0.0918,
+      "step": 1806
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 1.7433527806958875e-06,
+      "loss": 0.1047,
+      "step": 1807
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 1.7421931151569037e-06,
+      "loss": 0.0938,
+      "step": 1808
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 1.74103330096523e-06,
+      "loss": 0.0886,
+      "step": 1809
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 1.739873338832735e-06,
+      "loss": 0.0693,
+      "step": 1810
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 1.7387132294713772e-06,
+      "loss": 0.0915,
+      "step": 1811
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 1.7375529735932048e-06,
+      "loss": 0.0886,
+      "step": 1812
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 1.7363925719103583e-06,
+      "loss": 0.081,
+      "step": 1813
+    },
+    {
+      "epoch": 13.46,
+      "learning_rate": 1.7352320251350646e-06,
+      "loss": 0.0884,
+      "step": 1814
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 1.7340713339796424e-06,
+      "loss": 0.085,
+      "step": 1815
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 1.732910499156497e-06,
+      "loss": 0.0953,
+      "step": 1816
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 1.7317495213781234e-06,
+      "loss": 0.091,
+      "step": 1817
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 1.7305884013571035e-06,
+      "loss": 0.0873,
+      "step": 1818
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 1.7294271398061058e-06,
+      "loss": 0.0832,
+      "step": 1819
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 1.728265737437888e-06,
+      "loss": 0.0978,
+      "step": 1820
+    },
+    {
+      "epoch": 13.51,
+      "eval_loss": 2.1126821041107178,
+      "eval_runtime": 1.8185,
+      "eval_samples_per_second": 36.294,
+      "eval_steps_per_second": 3.299,
+      "step": 1820
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 1.7271041949652916e-06,
+      "loss": 0.0939,
+      "step": 1821
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 1.7259425131012459e-06,
+      "loss": 0.0907,
+      "step": 1822
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 1.7247806925587648e-06,
+      "loss": 0.0805,
+      "step": 1823
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 1.7236187340509476e-06,
+      "loss": 0.094,
+      "step": 1824
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 1.7224566382909785e-06,
+      "loss": 0.0834,
+      "step": 1825
+    },
+    {
+      "epoch": 13.55,
+      "learning_rate": 1.7212944059921262e-06,
+      "loss": 0.0942,
+      "step": 1826
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 1.7201320378677411e-06,
+      "loss": 0.09,
+      "step": 1827
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 1.7189695346312608e-06,
+      "loss": 0.0887,
+      "step": 1828
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 1.7178068969962021e-06,
+      "loss": 0.0785,
+      "step": 1829
+    },
+    {
+      "epoch": 13.58,
+      "learning_rate": 1.7166441256761668e-06,
+      "loss": 0.0734,
+      "step": 1830
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 1.7154812213848368e-06,
+      "loss": 0.0905,
+      "step": 1831
+    },
+    {
+      "epoch": 13.6,
+      "learning_rate": 1.7143181848359774e-06,
+      "loss": 0.1071,
+      "step": 1832
+    },
+    {
+      "epoch": 13.6,
+      "learning_rate": 1.7131550167434343e-06,
+      "loss": 0.092,
+      "step": 1833
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 1.7119917178211336e-06,
+      "loss": 0.0869,
+      "step": 1834
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 1.7108282887830819e-06,
+      "loss": 0.0902,
+      "step": 1835
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 1.709664730343366e-06,
+      "loss": 0.0908,
+      "step": 1836
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 1.708501043216152e-06,
+      "loss": 0.0758,
+      "step": 1837
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 1.707337228115685e-06,
+      "loss": 0.0931,
+      "step": 1838
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 1.7061732857562882e-06,
+      "loss": 0.0855,
+      "step": 1839
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 1.7050092168523635e-06,
+      "loss": 0.0908,
+      "step": 1840
+    },
+    {
+      "epoch": 13.65,
+      "eval_loss": 2.1038267612457275,
+      "eval_runtime": 1.8149,
+      "eval_samples_per_second": 36.365,
+      "eval_steps_per_second": 3.306,
+      "step": 1840
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 1.70384502211839e-06,
+      "loss": 0.076,
+      "step": 1841
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 1.7026807022689246e-06,
+      "loss": 0.0875,
+      "step": 1842
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 1.7015162580186005e-06,
+      "loss": 0.094,
+      "step": 1843
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 1.7003516900821268e-06,
+      "loss": 0.0931,
+      "step": 1844
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 1.6991869991742905e-06,
+      "loss": 0.0813,
+      "step": 1845
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 1.698022186009951e-06,
+      "loss": 0.085,
+      "step": 1846
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 1.696857251304046e-06,
+      "loss": 0.0933,
+      "step": 1847
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 1.695692195771585e-06,
+      "loss": 0.0784,
+      "step": 1848
+    },
+    {
+      "epoch": 13.72,
+      "learning_rate": 1.6945270201276536e-06,
+      "loss": 0.1019,
+      "step": 1849
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 1.6933617250874103e-06,
+      "loss": 0.0883,
+      "step": 1850
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 1.6921963113660871e-06,
+      "loss": 0.0889,
+      "step": 1851
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 1.6910307796789885e-06,
+      "loss": 0.0937,
+      "step": 1852
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 1.6898651307414914e-06,
+      "loss": 0.0948,
+      "step": 1853
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 1.6886993652690462e-06,
+      "loss": 0.0949,
+      "step": 1854
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 1.6875334839771717e-06,
+      "loss": 0.0946,
+      "step": 1855
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 1.6863674875814603e-06,
+      "loss": 0.0848,
+      "step": 1856
+    },
+    {
+      "epoch": 13.78,
+      "learning_rate": 1.6852013767975756e-06,
+      "loss": 0.0935,
+      "step": 1857
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 1.6840351523412483e-06,
+      "loss": 0.102,
+      "step": 1858
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 1.6828688149282822e-06,
+      "loss": 0.1031,
+      "step": 1859
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 1.6817023652745483e-06,
+      "loss": 0.0918,
+      "step": 1860
+    },
+    {
+      "epoch": 13.8,
+      "eval_loss": 2.1295578479766846,
+      "eval_runtime": 1.8192,
+      "eval_samples_per_second": 36.28,
+      "eval_steps_per_second": 3.298,
+      "step": 1860
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 1.6805358040959872e-06,
+      "loss": 0.0998,
+      "step": 1861
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 1.679369132108608e-06,
+      "loss": 0.1061,
+      "step": 1862
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 1.678202350028488e-06,
+      "loss": 0.0875,
+      "step": 1863
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 1.6770354585717712e-06,
+      "loss": 0.0997,
+      "step": 1864
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 1.67586845845467e-06,
+      "loss": 0.1019,
+      "step": 1865
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 1.6747013503934625e-06,
+      "loss": 0.0863,
+      "step": 1866
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 1.6735341351044936e-06,
+      "loss": 0.0877,
+      "step": 1867
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 1.6723668133041736e-06,
+      "loss": 0.0963,
+      "step": 1868
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 1.6711993857089782e-06,
+      "loss": 0.0919,
+      "step": 1869
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 1.6700318530354487e-06,
+      "loss": 0.0863,
+      "step": 1870
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 1.6688642160001898e-06,
+      "loss": 0.0937,
+      "step": 1871
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 1.6676964753198718e-06,
+      "loss": 0.1014,
+      "step": 1872
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 1.6665286317112265e-06,
+      "loss": 0.0833,
+      "step": 1873
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 1.6653606858910511e-06,
+      "loss": 0.0799,
+      "step": 1874
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 1.6641926385762038e-06,
+      "loss": 0.085,
+      "step": 1875
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 1.6630244904836062e-06,
+      "loss": 0.1027,
+      "step": 1876
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 1.6618562423302406e-06,
+      "loss": 0.088,
+      "step": 1877
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 1.6606878948331528e-06,
+      "loss": 0.0915,
+      "step": 1878
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 1.6595194487094466e-06,
+      "loss": 0.0905,
+      "step": 1879
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 1.6583509046762894e-06,
+      "loss": 0.0861,
+      "step": 1880
+    },
+    {
+      "epoch": 13.95,
+      "eval_loss": 2.113924503326416,
+      "eval_runtime": 1.8213,
+      "eval_samples_per_second": 36.239,
+      "eval_steps_per_second": 3.294,
+      "step": 1880
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 1.6571822634509067e-06,
+      "loss": 0.0947,
+      "step": 1881
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 1.656013525750584e-06,
+      "loss": 0.0922,
+      "step": 1882
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 1.6548446922926663e-06,
+      "loss": 0.0923,
+      "step": 1883
+    },
+    {
+      "epoch": 13.98,
+      "learning_rate": 1.6536757637945578e-06,
+      "loss": 0.0757,
+      "step": 1884
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 1.6525067409737198e-06,
+      "loss": 0.097,
+      "step": 1885
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 1.6513376245476727e-06,
+      "loss": 0.0947,
+      "step": 1886
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 1.6501684152339935e-06,
+      "loss": 0.0733,
+      "step": 1887
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 1.6489991137503173e-06,
+      "loss": 0.0633,
+      "step": 1888
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 1.6478297208143345e-06,
+      "loss": 0.0643,
+      "step": 1889
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 1.6466602371437922e-06,
+      "loss": 0.0635,
+      "step": 1890
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 1.645490663456493e-06,
+      "loss": 0.0711,
+      "step": 1891
+    },
+    {
+      "epoch": 14.04,
+      "learning_rate": 1.6443210004702958e-06,
+      "loss": 0.0588,
+      "step": 1892
+    },
+    {
+      "epoch": 14.05,
+      "learning_rate": 1.6431512489031124e-06,
+      "loss": 0.0644,
+      "step": 1893
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 1.6419814094729107e-06,
+      "loss": 0.0557,
+      "step": 1894
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 1.6408114828977115e-06,
+      "loss": 0.0752,
+      "step": 1895
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 1.6396414698955899e-06,
+      "loss": 0.0632,
+      "step": 1896
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 1.6384713711846727e-06,
+      "loss": 0.0556,
+      "step": 1897
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 1.6373011874831414e-06,
+      "loss": 0.0696,
+      "step": 1898
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 1.6361309195092272e-06,
+      "loss": 0.0665,
+      "step": 1899
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 1.6349605679812153e-06,
+      "loss": 0.0555,
+      "step": 1900
+    },
+    {
+      "epoch": 14.1,
+      "eval_loss": 2.2332956790924072,
+      "eval_runtime": 1.8151,
+      "eval_samples_per_second": 36.361,
+      "eval_steps_per_second": 3.306,
+      "step": 1900
+    },
+    {
+      "epoch": 14.11,
+      "learning_rate": 1.6337901336174402e-06,
+      "loss": 0.0621,
+      "step": 1901
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 1.6326196171362888e-06,
+      "loss": 0.0683,
+      "step": 1902
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 1.631449019256198e-06,
+      "loss": 0.0784,
+      "step": 1903
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 1.630278340695654e-06,
+      "loss": 0.0668,
+      "step": 1904
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 1.6291075821731934e-06,
+      "loss": 0.0694,
+      "step": 1905
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 1.627936744407401e-06,
+      "loss": 0.0668,
+      "step": 1906
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 1.6267658281169113e-06,
+      "loss": 0.0578,
+      "step": 1907
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 1.6255948340204055e-06,
+      "loss": 0.0598,
+      "step": 1908
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 1.6244237628366145e-06,
+      "loss": 0.0745,
+      "step": 1909
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 1.6232526152843142e-06,
+      "loss": 0.0568,
+      "step": 1910
+    },
+    {
+      "epoch": 14.18,
+      "learning_rate": 1.6220813920823303e-06,
+      "loss": 0.0565,
+      "step": 1911
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 1.6209100939495317e-06,
+      "loss": 0.0666,
+      "step": 1912
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 1.6197387216048356e-06,
+      "loss": 0.0728,
+      "step": 1913
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 1.6185672757672042e-06,
+      "loss": 0.0724,
+      "step": 1914
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 1.617395757155644e-06,
+      "loss": 0.0738,
+      "step": 1915
+    },
+    {
+      "epoch": 14.22,
+      "learning_rate": 1.6162241664892078e-06,
+      "loss": 0.0701,
+      "step": 1916
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 1.6150525044869902e-06,
+      "loss": 0.0699,
+      "step": 1917
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 1.613880771868132e-06,
+      "loss": 0.0668,
+      "step": 1918
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 1.6127089693518158e-06,
+      "loss": 0.0677,
+      "step": 1919
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 1.6115370976572685e-06,
+      "loss": 0.0746,
+      "step": 1920
+    },
+    {
+      "epoch": 14.25,
+      "eval_loss": 2.1848902702331543,
+      "eval_runtime": 1.8194,
+      "eval_samples_per_second": 36.276,
+      "eval_steps_per_second": 3.298,
+      "step": 1920
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 1.6103651575037576e-06,
+      "loss": 0.0507,
+      "step": 1921
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 1.6091931496105945e-06,
+      "loss": 0.0767,
+      "step": 1922
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 1.6080210746971302e-06,
+      "loss": 0.0666,
+      "step": 1923
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 1.6068489334827594e-06,
+      "loss": 0.0505,
+      "step": 1924
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 1.6056767266869143e-06,
+      "loss": 0.0708,
+      "step": 1925
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 1.6045044550290708e-06,
+      "loss": 0.0606,
+      "step": 1926
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 1.6033321192287416e-06,
+      "loss": 0.0664,
+      "step": 1927
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 1.602159720005481e-06,
+      "loss": 0.059,
+      "step": 1928
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 1.60098725807888e-06,
+      "loss": 0.0587,
+      "step": 1929
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 1.599814734168571e-06,
+      "loss": 0.0695,
+      "step": 1930
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 1.5986421489942217e-06,
+      "loss": 0.0591,
+      "step": 1931
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 1.597469503275539e-06,
+      "loss": 0.0541,
+      "step": 1932
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 1.5962967977322661e-06,
+      "loss": 0.0668,
+      "step": 1933
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 1.5951240330841837e-06,
+      "loss": 0.067,
+      "step": 1934
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 1.593951210051108e-06,
+      "loss": 0.0701,
+      "step": 1935
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 1.592778329352892e-06,
+      "loss": 0.0807,
+      "step": 1936
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 1.5916053917094226e-06,
+      "loss": 0.0645,
+      "step": 1937
+    },
+    {
+      "epoch": 14.38,
+      "learning_rate": 1.5904323978406236e-06,
+      "loss": 0.0626,
+      "step": 1938
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 1.5892593484664512e-06,
+      "loss": 0.0651,
+      "step": 1939
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 1.5880862443068978e-06,
+      "loss": 0.0659,
+      "step": 1940
+    },
+    {
+      "epoch": 14.4,
+      "eval_loss": 2.202566385269165,
+      "eval_runtime": 1.8174,
+      "eval_samples_per_second": 36.315,
+      "eval_steps_per_second": 3.301,
+      "step": 1940
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 1.5869130860819875e-06,
+      "loss": 0.0634,
+      "step": 1941
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 1.5857398745117791e-06,
+      "loss": 0.0602,
+      "step": 1942
+    },
+    {
+      "epoch": 14.42,
+      "learning_rate": 1.5845666103163628e-06,
+      "loss": 0.0618,
+      "step": 1943
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 1.5833932942158623e-06,
+      "loss": 0.0659,
+      "step": 1944
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 1.5822199269304324e-06,
+      "loss": 0.067,
+      "step": 1945
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 1.5810465091802599e-06,
+      "loss": 0.0666,
+      "step": 1946
+    },
+    {
+      "epoch": 14.45,
+      "learning_rate": 1.5798730416855618e-06,
+      "loss": 0.0594,
+      "step": 1947
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 1.578699525166586e-06,
+      "loss": 0.0661,
+      "step": 1948
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 1.5775259603436117e-06,
+      "loss": 0.065,
+      "step": 1949
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 1.5763523479369447e-06,
+      "loss": 0.0677,
+      "step": 1950
+    },
+    {
+      "epoch": 14.48,
+      "learning_rate": 1.5751786886669238e-06,
+      "loss": 0.0761,
+      "step": 1951
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 1.5740049832539134e-06,
+      "loss": 0.078,
+      "step": 1952
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 1.5728312324183083e-06,
+      "loss": 0.067,
+      "step": 1953
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 1.5716574368805296e-06,
+      "loss": 0.0738,
+      "step": 1954
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 1.5704835973610272e-06,
+      "loss": 0.07,
+      "step": 1955
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 1.569309714580277e-06,
+      "loss": 0.0701,
+      "step": 1956
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 1.5681357892587825e-06,
+      "loss": 0.0658,
+      "step": 1957
+    },
+    {
+      "epoch": 14.53,
+      "learning_rate": 1.566961822117072e-06,
+      "loss": 0.0632,
+      "step": 1958
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 1.5657878138757004e-06,
+      "loss": 0.0654,
+      "step": 1959
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 1.5646137652552475e-06,
+      "loss": 0.0677,
+      "step": 1960
+    },
+    {
+      "epoch": 14.55,
+      "eval_loss": 2.200098752975464,
+      "eval_runtime": 1.8176,
+      "eval_samples_per_second": 36.312,
+      "eval_steps_per_second": 3.301,
+      "step": 1960
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 1.5634396769763182e-06,
+      "loss": 0.0506,
+      "step": 1961
+    },
+    {
+      "epoch": 14.56,
+      "learning_rate": 1.5622655497595411e-06,
+      "loss": 0.0712,
+      "step": 1962
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 1.5610913843255697e-06,
+      "loss": 0.0787,
+      "step": 1963
+    },
+    {
+      "epoch": 14.58,
+      "learning_rate": 1.5599171813950795e-06,
+      "loss": 0.0617,
+      "step": 1964
+    },
+    {
+      "epoch": 14.58,
+      "learning_rate": 1.5587429416887708e-06,
+      "loss": 0.0766,
+      "step": 1965
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 1.5575686659273648e-06,
+      "loss": 0.0756,
+      "step": 1966
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 1.5563943548316054e-06,
+      "loss": 0.0654,
+      "step": 1967
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 1.5552200091222595e-06,
+      "loss": 0.0734,
+      "step": 1968
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 1.5540456295201128e-06,
+      "loss": 0.0646,
+      "step": 1969
+    },
+    {
+      "epoch": 14.62,
+      "learning_rate": 1.5528712167459736e-06,
+      "loss": 0.0621,
+      "step": 1970
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 1.5516967715206705e-06,
+      "loss": 0.0736,
+      "step": 1971
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 1.5505222945650508e-06,
+      "loss": 0.0682,
+      "step": 1972
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 1.5493477865999827e-06,
+      "loss": 0.0743,
+      "step": 1973
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 1.5481732483463528e-06,
+      "loss": 0.0631,
+      "step": 1974
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 1.5469986805250656e-06,
+      "loss": 0.0619,
+      "step": 1975
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 1.545824083857045e-06,
+      "loss": 0.0674,
+      "step": 1976
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 1.5446494590632318e-06,
+      "loss": 0.0561,
+      "step": 1977
+    },
+    {
+      "epoch": 14.68,
+      "learning_rate": 1.5434748068645843e-06,
+      "loss": 0.0611,
+      "step": 1978
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 1.5423001279820777e-06,
+      "loss": 0.064,
+      "step": 1979
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 1.5411254231367037e-06,
+      "loss": 0.0611,
+      "step": 1980
+    },
+    {
+      "epoch": 14.69,
+      "eval_loss": 2.180213689804077,
+      "eval_runtime": 1.8174,
+      "eval_samples_per_second": 36.316,
+      "eval_steps_per_second": 3.301,
+      "step": 1980
+    },
+    {
+      "epoch": 14.7,
+      "learning_rate": 1.5399506930494687e-06,
+      "loss": 0.071,
+      "step": 1981
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 1.5387759384413975e-06,
+      "loss": 0.0817,
+      "step": 1982
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 1.537601160033526e-06,
+      "loss": 0.0606,
+      "step": 1983
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 1.5364263585469086e-06,
+      "loss": 0.0597,
+      "step": 1984
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 1.5352515347026105e-06,
+      "loss": 0.0718,
+      "step": 1985
+    },
+    {
+      "epoch": 14.74,
+      "learning_rate": 1.5340766892217133e-06,
+      "loss": 0.0581,
+      "step": 1986
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 1.5329018228253102e-06,
+      "loss": 0.0711,
+      "step": 1987
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 1.531726936234508e-06,
+      "loss": 0.067,
+      "step": 1988
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 1.5305520301704252e-06,
+      "loss": 0.0731,
+      "step": 1989
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 1.5293771053541936e-06,
+      "loss": 0.0605,
+      "step": 1990
+    },
+    {
+      "epoch": 14.78,
+      "learning_rate": 1.5282021625069543e-06,
+      "loss": 0.0577,
+      "step": 1991
+    },
+    {
+      "epoch": 14.78,
+      "learning_rate": 1.5270272023498627e-06,
+      "loss": 0.0659,
+      "step": 1992
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 1.5258522256040811e-06,
+      "loss": 0.0679,
+      "step": 1993
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 1.5246772329907848e-06,
+      "loss": 0.0631,
+      "step": 1994
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 1.5235022252311579e-06,
+      "loss": 0.065,
+      "step": 1995
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 1.522327203046393e-06,
+      "loss": 0.0592,
+      "step": 1996
+    },
+    {
+      "epoch": 14.82,
+      "learning_rate": 1.5211521671576934e-06,
+      "loss": 0.0689,
+      "step": 1997
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 1.5199771182862684e-06,
+      "loss": 0.0703,
+      "step": 1998
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 1.518802057153338e-06,
+      "loss": 0.0665,
+      "step": 1999
+    },
+    {
+      "epoch": 14.84,
+      "learning_rate": 1.5176269844801272e-06,
+      "loss": 0.0637,
+      "step": 2000
+    },
+    {
+      "epoch": 14.84,
+      "eval_loss": 2.2041208744049072,
+      "eval_runtime": 1.8185,
+      "eval_samples_per_second": 36.294,
+      "eval_steps_per_second": 3.299,
+      "step": 2000
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 1.5164519009878698e-06,
+      "loss": 0.0729,
+      "step": 2001
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 1.515276807397805e-06,
+      "loss": 0.0665,
+      "step": 2002
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 1.5141017044311797e-06,
+      "loss": 0.0606,
+      "step": 2003
+    },
+    {
+      "epoch": 14.87,
+      "learning_rate": 1.5129265928092445e-06,
+      "loss": 0.0707,
+      "step": 2004
+    },
+    {
+      "epoch": 14.88,
+      "learning_rate": 1.511751473253258e-06,
+      "loss": 0.0666,
+      "step": 2005
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 1.5105763464844812e-06,
+      "loss": 0.0725,
+      "step": 2006
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 1.5094012132241808e-06,
+      "loss": 0.0654,
+      "step": 2007
+    },
+    {
+      "epoch": 14.9,
+      "learning_rate": 1.5082260741936266e-06,
+      "loss": 0.0683,
+      "step": 2008
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 1.5070509301140937e-06,
+      "loss": 0.0709,
+      "step": 2009
+    },
+    {
+      "epoch": 14.92,
+      "learning_rate": 1.5058757817068579e-06,
+      "loss": 0.0658,
+      "step": 2010
+    },
+    {
+      "epoch": 14.92,
+      "learning_rate": 1.5047006296931995e-06,
+      "loss": 0.0672,
+      "step": 2011
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 1.5035254747944e-06,
+      "loss": 0.0641,
+      "step": 2012
+    },
+    {
+      "epoch": 14.94,
+      "learning_rate": 1.5023503177317436e-06,
+      "loss": 0.0628,
+      "step": 2013
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 1.5011751592265152e-06,
+      "loss": 0.0642,
+      "step": 2014
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 1.5e-06,
+      "loss": 0.0661,
+      "step": 2015
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 1.4988248407734853e-06,
+      "loss": 0.0744,
+      "step": 2016
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 1.497649682268257e-06,
+      "loss": 0.0674,
+      "step": 2017
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 1.4964745252055999e-06,
+      "loss": 0.0783,
+      "step": 2018
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 1.4952993703068008e-06,
+      "loss": 0.0711,
+      "step": 2019
+    },
+    {
+      "epoch": 14.99,
+      "learning_rate": 1.4941242182931426e-06,
+      "loss": 0.0634,
+      "step": 2020
+    },
+    {
+      "epoch": 14.99,
+      "eval_loss": 2.1904079914093018,
+      "eval_runtime": 1.8176,
+      "eval_samples_per_second": 36.311,
+      "eval_steps_per_second": 3.301,
+      "step": 2020
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 1.492949069885907e-06,
+      "loss": 0.0725,
+      "step": 2021
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4020,
+  "num_train_epochs": 30,
+  "save_steps": 500,
+  "total_flos": 3.5733940772672635e+18,
+  "trial_name": null,
+  "trial_params": null
+}