diff --git "a/checkpoint-3954/trainer_state.json" "b/checkpoint-3954/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-3954/trainer_state.json"
@@ -0,0 +1,23895 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 200,
+  "global_step": 3954,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 0.8854,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 1.2363,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8e-05,
+      "loss": 0.57,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4e-05,
+      "loss": 0.7468,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 1.1174,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.3413,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.9894,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.9246,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 0.8455,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-05,
+      "loss": 0.7254,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9999999407419926e-05,
+      "loss": 0.6649,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9999997629679705e-05,
+      "loss": 0.8749,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.999999466677943e-05,
+      "loss": 1.1179,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.99999905187192e-05,
+      "loss": 1.0552,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9999985185499186e-05,
+      "loss": 0.9388,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9999978667119607e-05,
+      "loss": 0.5965,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.99999709635807e-05,
+      "loss": 0.7575,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9999962074882785e-05,
+      "loss": 1.3591,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.999995200102622e-05,
+      "loss": 0.5293,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999940742011376e-05,
+      "loss": 1.118,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999928297838724e-05,
+      "loss": 0.5806,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999991466850874e-05,
+      "loss": 0.8742,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999989985402198e-05,
+      "loss": 0.8519,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999988385437901e-05,
+      "loss": 1.0442,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999986666958047e-05,
+      "loss": 0.5535,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999984829962703e-05,
+      "loss": 0.8342,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999982874451944e-05,
+      "loss": 0.5704,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999808004258444e-05,
+      "loss": 1.6309,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999978607884488e-05,
+      "loss": 0.8645,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999976296827961e-05,
+      "loss": 1.0076,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999973867256354e-05,
+      "loss": 0.2864,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999971319169764e-05,
+      "loss": 0.2028,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999686525682914e-05,
+      "loss": 0.542,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.99996586745204e-05,
+      "loss": 0.6453,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999629638211225e-05,
+      "loss": 0.3074,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999959941675652e-05,
+      "loss": 0.5244,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999568010157477e-05,
+      "loss": 0.6479,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999535418415346e-05,
+      "loss": 0.7683,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.99995016415314e-05,
+      "loss": 0.7861,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999466679506995e-05,
+      "loss": 0.891,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999943053234349e-05,
+      "loss": 0.7046,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999393200042336e-05,
+      "loss": 0.8554,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999935468260499e-05,
+      "loss": 0.7183,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999931498003297e-05,
+      "loss": 0.4414,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999927409232786e-05,
+      "loss": 0.831,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999232019491266e-05,
+      "loss": 0.8325,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999188761524856e-05,
+      "loss": 1.0734,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999144318430334e-05,
+      "loss": 0.835,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999909869020947e-05,
+      "loss": 0.9724,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999905187686403e-05,
+      "loss": 0.4949,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999900387839591e-05,
+      "loss": 0.9897,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9998954694806964e-05,
+      "loss": 0.5754,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9998904326099164e-05,
+      "loss": 1.0836,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999885277227449e-05,
+      "loss": 0.7409,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9998800033334974e-05,
+      "loss": 0.6784,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9998746109282705e-05,
+      "loss": 0.5594,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999869100011982e-05,
+      "loss": 0.567,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9998634705848484e-05,
+      "loss": 0.7662,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999857722647093e-05,
+      "loss": 0.6745,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9998518561989414e-05,
+      "loss": 0.9293,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999845871240627e-05,
+      "loss": 0.5923,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999839767772386e-05,
+      "loss": 0.5522,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999833545794458e-05,
+      "loss": 0.9283,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999827205307091e-05,
+      "loss": 1.4208,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999820746310535e-05,
+      "loss": 0.6536,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999814168805044e-05,
+      "loss": 0.5965,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999807472790878e-05,
+      "loss": 0.5124,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9998006582683025e-05,
+      "loss": 0.8341,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999793725237586e-05,
+      "loss": 0.7836,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999786673699002e-05,
+      "loss": 0.8772,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.99977950365283e-05,
+      "loss": 0.6505,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999772215099353e-05,
+      "loss": 0.4268,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.99976480803886e-05,
+      "loss": 0.8544,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999757282471641e-05,
+      "loss": 1.1559,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9997496383979956e-05,
+      "loss": 1.4848,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999741875818224e-05,
+      "loss": 0.9169,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9997339947326347e-05,
+      "loss": 0.9626,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9997259951415376e-05,
+      "loss": 0.8647,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999717877045249e-05,
+      "loss": 0.7588,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.99970964044409e-05,
+      "loss": 0.2353,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999701285338387e-05,
+      "loss": 1.0683,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999692811728468e-05,
+      "loss": 0.8542,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999684219614669e-05,
+      "loss": 0.5512,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999675508997329e-05,
+      "loss": 0.866,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999666679876792e-05,
+      "loss": 0.6513,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999657732253407e-05,
+      "loss": 0.9098,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9996486661275284e-05,
+      "loss": 1.0667,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999639481499513e-05,
+      "loss": 0.6993,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999630178369724e-05,
+      "loss": 0.8695,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.99962075673853e-05,
+      "loss": 0.784,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9996112166063004e-05,
+      "loss": 0.4206,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999601557973415e-05,
+      "loss": 0.9655,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999591780840254e-05,
+      "loss": 0.5169,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999581885207205e-05,
+      "loss": 0.6075,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999571871074657e-05,
+      "loss": 0.2471,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999561738443007e-05,
+      "loss": 1.0324,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999551487312654e-05,
+      "loss": 0.924,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999541117684004e-05,
+      "loss": 0.7946,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995306295574676e-05,
+      "loss": 0.9408,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999520022933456e-05,
+      "loss": 0.6059,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995092978123916e-05,
+      "loss": 0.7383,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999498454194696e-05,
+      "loss": 0.2429,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9994874920807985e-05,
+      "loss": 0.5436,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999476411471132e-05,
+      "loss": 0.2953,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999465212366134e-05,
+      "loss": 0.6523,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9994538947662474e-05,
+      "loss": 0.572,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9994424586719185e-05,
+      "loss": 0.7193,
+      "step": 107
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9994309040836e-05,
+      "loss": 0.8731,
+      "step": 108
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999419231001748e-05,
+      "loss": 0.5843,
+      "step": 109
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9994074394268234e-05,
+      "loss": 0.6412,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999395529359292e-05,
+      "loss": 0.9748,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999383500799625e-05,
+      "loss": 0.8403,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999371353748298e-05,
+      "loss": 0.2781,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9993590882057875e-05,
+      "loss": 0.5298,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999346704172582e-05,
+      "loss": 0.7843,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.99933420164917e-05,
+      "loss": 0.9732,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9993215806360446e-05,
+      "loss": 0.9223,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9993088411337035e-05,
+      "loss": 0.8888,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9992959831426515e-05,
+      "loss": 0.8844,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9992830066633956e-05,
+      "loss": 0.5614,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999269911696449e-05,
+      "loss": 0.7013,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9992566982423296e-05,
+      "loss": 0.4652,
+      "step": 122
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999243366301558e-05,
+      "loss": 0.7724,
+      "step": 123
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999229915874662e-05,
+      "loss": 0.9672,
+      "step": 124
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999216346962173e-05,
+      "loss": 1.1158,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999202659564626e-05,
+      "loss": 0.3481,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999188853682562e-05,
+      "loss": 1.1717,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999174929316527e-05,
+      "loss": 0.7328,
+      "step": 128
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999160886467072e-05,
+      "loss": 1.3884,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999146725134749e-05,
+      "loss": 0.9068,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999132445320119e-05,
+      "loss": 0.9644,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999118047023747e-05,
+      "loss": 0.285,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999103530246202e-05,
+      "loss": 0.5374,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9990888949880544e-05,
+      "loss": 1.4238,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999074141249885e-05,
+      "loss": 0.354,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999059269032277e-05,
+      "loss": 0.6119,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999044278335816e-05,
+      "loss": 0.816,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9990291691610956e-05,
+      "loss": 0.6969,
+      "step": 138
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999013941508713e-05,
+      "loss": 0.795,
+      "step": 139
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998998595379268e-05,
+      "loss": 0.9635,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998983130773368e-05,
+      "loss": 0.6558,
+      "step": 141
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998967547691624e-05,
+      "loss": 0.4313,
+      "step": 142
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998951846134652e-05,
+      "loss": 0.53,
+      "step": 143
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9989360261030715e-05,
+      "loss": 0.974,
+      "step": 144
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9989200875975085e-05,
+      "loss": 0.8632,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9989040306185904e-05,
+      "loss": 0.7214,
+      "step": 146
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998887855166955e-05,
+      "loss": 0.6944,
+      "step": 147
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998871561243238e-05,
+      "loss": 0.5707,
+      "step": 148
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998855148848085e-05,
+      "loss": 0.7763,
+      "step": 149
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9988386179821434e-05,
+      "loss": 1.0039,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998821968646067e-05,
+      "loss": 0.8207,
+      "step": 151
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998805200840513e-05,
+      "loss": 0.3861,
+      "step": 152
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998788314566145e-05,
+      "loss": 0.496,
+      "step": 153
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998771309823628e-05,
+      "loss": 0.5224,
+      "step": 154
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998754186613636e-05,
+      "loss": 0.5252,
+      "step": 155
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998736944936844e-05,
+      "loss": 0.961,
+      "step": 156
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998719584793933e-05,
+      "loss": 0.6731,
+      "step": 157
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998702106185589e-05,
+      "loss": 1.2646,
+      "step": 158
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998684509112504e-05,
+      "loss": 1.5771,
+      "step": 159
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998666793575372e-05,
+      "loss": 0.4212,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998648959574892e-05,
+      "loss": 0.5691,
+      "step": 161
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9986310071117705e-05,
+      "loss": 0.5723,
+      "step": 162
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998612936186716e-05,
+      "loss": 0.6776,
+      "step": 163
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9985947468004405e-05,
+      "loss": 0.8139,
+      "step": 164
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998576438953665e-05,
+      "loss": 0.7897,
+      "step": 165
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998558012647111e-05,
+      "loss": 0.5229,
+      "step": 166
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998539467881508e-05,
+      "loss": 0.9778,
+      "step": 167
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998520804657588e-05,
+      "loss": 1.0174,
+      "step": 168
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998502022976089e-05,
+      "loss": 0.3036,
+      "step": 169
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998483122837751e-05,
+      "loss": 0.771,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998464104243323e-05,
+      "loss": 0.6854,
+      "step": 171
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9984449671935545e-05,
+      "loss": 0.434,
+      "step": 172
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998425711689202e-05,
+      "loss": 1.0906,
+      "step": 173
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998406337731026e-05,
+      "loss": 0.8902,
+      "step": 174
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998386845319793e-05,
+      "loss": 0.541,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9983672344562726e-05,
+      "loss": 0.8334,
+      "step": 176
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998347505141239e-05,
+      "loss": 0.5004,
+      "step": 177
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998327657375472e-05,
+      "loss": 0.7286,
+      "step": 178
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998307691159755e-05,
+      "loss": 0.5692,
+      "step": 179
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9982876064948784e-05,
+      "loss": 0.4765,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998267403381634e-05,
+      "loss": 0.9517,
+      "step": 181
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998247081820821e-05,
+      "loss": 0.6937,
+      "step": 182
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998226641813241e-05,
+      "loss": 0.7616,
+      "step": 183
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998206083359703e-05,
+      "loss": 0.4327,
+      "step": 184
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998185406461018e-05,
+      "loss": 0.9744,
+      "step": 185
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998164611118003e-05,
+      "loss": 0.472,
+      "step": 186
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.99814369733148e-05,
+      "loss": 1.0047,
+      "step": 187
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9981226651022746e-05,
+      "loss": 0.6024,
+      "step": 188
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998101514431219e-05,
+      "loss": 1.1821,
+      "step": 189
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998080245319147e-05,
+      "loss": 0.5601,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998058857766901e-05,
+      "loss": 0.8611,
+      "step": 191
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9980373517753224e-05,
+      "loss": 0.6361,
+      "step": 192
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998015727345265e-05,
+      "loss": 0.4384,
+      "step": 193
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.99799398447758e-05,
+      "loss": 0.8944,
+      "step": 194
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997972123173128e-05,
+      "loss": 0.514,
+      "step": 195
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997950143432772e-05,
+      "loss": 0.4268,
+      "step": 196
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.99792804525738e-05,
+      "loss": 0.5087,
+      "step": 197
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9979058286478254e-05,
+      "loss": 0.8719,
+      "step": 198
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9978834936049864e-05,
+      "loss": 0.7477,
+      "step": 199
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997861040129745e-05,
+      "loss": 1.312,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.7646743059158325,
+      "eval_runtime": 39.1622,
+      "eval_samples_per_second": 1.021,
+      "eval_steps_per_second": 1.021,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997838468222987e-05,
+      "loss": 0.2435,
+      "step": 201
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997815777885606e-05,
+      "loss": 1.3608,
+      "step": 202
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9977929691184983e-05,
+      "loss": 0.4042,
+      "step": 203
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997770041922563e-05,
+      "loss": 1.0007,
+      "step": 204
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997746996298707e-05,
+      "loss": 0.8802,
+      "step": 205
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9977238322478416e-05,
+      "loss": 0.8522,
+      "step": 206
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997700549770881e-05,
+      "loss": 1.3858,
+      "step": 207
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9976771488687445e-05,
+      "loss": 0.6547,
+      "step": 208
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9976536295423566e-05,
+      "loss": 0.6354,
+      "step": 209
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997629991792648e-05,
+      "loss": 0.9353,
+      "step": 210
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9976062356205516e-05,
+      "loss": 2.3016,
+      "step": 211
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997582361027005e-05,
+      "loss": 0.7886,
+      "step": 212
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997558368012953e-05,
+      "loss": 0.4092,
+      "step": 213
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997534256579342e-05,
+      "loss": 0.216,
+      "step": 214
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997510026727126e-05,
+      "loss": 0.5512,
+      "step": 215
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9974856784572616e-05,
+      "loss": 0.5589,
+      "step": 216
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9974612117707086e-05,
+      "loss": 0.2174,
+      "step": 217
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997436626668437e-05,
+      "loss": 0.7967,
+      "step": 218
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997411923151416e-05,
+      "loss": 0.4445,
+      "step": 219
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997387101220622e-05,
+      "loss": 1.0046,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9973621608770354e-05,
+      "loss": 1.1762,
+      "step": 221
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997337102121641e-05,
+      "loss": 0.9577,
+      "step": 222
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9973119249554306e-05,
+      "loss": 0.456,
+      "step": 223
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997286629379397e-05,
+      "loss": 1.0601,
+      "step": 224
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997261215394541e-05,
+      "loss": 0.816,
+      "step": 225
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997235683001865e-05,
+      "loss": 0.6969,
+      "step": 226
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997210032202379e-05,
+      "loss": 0.8252,
+      "step": 227
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997184262997095e-05,
+      "loss": 0.5744,
+      "step": 228
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997158375387032e-05,
+      "loss": 0.529,
+      "step": 229
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997132369373212e-05,
+      "loss": 0.8767,
+      "step": 230
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9971062449566635e-05,
+      "loss": 0.6095,
+      "step": 231
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9970800021384185e-05,
+      "loss": 0.4393,
+      "step": 232
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997053640919512e-05,
+      "loss": 0.585,
+      "step": 233
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997027161300987e-05,
+      "loss": 1.0239,
+      "step": 234
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997000563283889e-05,
+      "loss": 0.7409,
+      "step": 235
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9969738468692685e-05,
+      "loss": 0.449,
+      "step": 236
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996947012058182e-05,
+      "loss": 0.5384,
+      "step": 237
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996920058851689e-05,
+      "loss": 0.7052,
+      "step": 238
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996892987250854e-05,
+      "loss": 0.3529,
+      "step": 239
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996865797256746e-05,
+      "loss": 0.4189,
+      "step": 240
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996838488870441e-05,
+      "loss": 0.6253,
+      "step": 241
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996811062093017e-05,
+      "loss": 0.6808,
+      "step": 242
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996783516925556e-05,
+      "loss": 0.8701,
+      "step": 243
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9967558533691476e-05,
+      "loss": 0.4946,
+      "step": 244
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996728071424885e-05,
+      "loss": 0.7215,
+      "step": 245
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9967001710938645e-05,
+      "loss": 0.6763,
+      "step": 246
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.99667215237719e-05,
+      "loss": 0.8891,
+      "step": 247
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996644015275966e-05,
+      "loss": 0.4873,
+      "step": 248
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996615759791306e-05,
+      "loss": 0.8455,
+      "step": 249
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996587385924326e-05,
+      "loss": 0.6652,
+      "step": 250
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9965588936761466e-05,
+      "loss": 0.7879,
+      "step": 251
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9965302830478934e-05,
+      "loss": 0.8195,
+      "step": 252
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996501554040696e-05,
+      "loss": 0.9725,
+      "step": 253
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996472706655691e-05,
+      "loss": 0.7879,
+      "step": 254
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996443740894017e-05,
+      "loss": 0.5695,
+      "step": 255
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9964146567568165e-05,
+      "loss": 0.5787,
+      "step": 256
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9963854542452426e-05,
+      "loss": 0.8464,
+      "step": 257
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996356133360446e-05,
+      "loss": 0.966,
+      "step": 258
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9963266941035847e-05,
+      "loss": 0.4157,
+      "step": 259
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996297136475824e-05,
+      "loss": 0.4295,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9962674604783296e-05,
+      "loss": 0.8546,
+      "step": 261
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996237666112275e-05,
+      "loss": 1.0376,
+      "step": 262
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996207753378837e-05,
+      "loss": 0.7363,
+      "step": 263
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996177722279196e-05,
+      "loss": 0.2626,
+      "step": 264
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996147572814541e-05,
+      "loss": 0.6511,
+      "step": 265
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.99611730498606e-05,
+      "loss": 0.7874,
+      "step": 266
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9960869187949515e-05,
+      "loss": 0.5002,
+      "step": 267
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996056414242414e-05,
+      "loss": 0.6879,
+      "step": 268
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9960257913296536e-05,
+      "loss": 1.0779,
+      "step": 269
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9959950500578805e-05,
+      "loss": 1.4456,
+      "step": 270
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995964190428307e-05,
+      "loss": 0.7683,
+      "step": 271
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995933212442155e-05,
+      "loss": 0.7847,
+      "step": 272
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9959021161006466e-05,
+      "loss": 0.554,
+      "step": 273
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9958709014050105e-05,
+      "loss": 0.535,
+      "step": 274
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9958395683564795e-05,
+      "loss": 0.6688,
+      "step": 275
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995808116956292e-05,
+      "loss": 1.0189,
+      "step": 276
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9957765472056904e-05,
+      "loss": 1.3465,
+      "step": 277
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9957448591059226e-05,
+      "loss": 0.5812,
+      "step": 278
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995713052658239e-05,
+      "loss": 0.2582,
+      "step": 279
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995681127863897e-05,
+      "loss": 0.7336,
+      "step": 280
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9956490847241575e-05,
+      "loss": 1.0346,
+      "step": 281
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995616923240287e-05,
+      "loss": 0.863,
+      "step": 282
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9955846434135555e-05,
+      "loss": 0.6711,
+      "step": 283
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995552245245237e-05,
+      "loss": 0.5115,
+      "step": 284
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9955197287366136e-05,
+      "loss": 0.8629,
+      "step": 285
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9954870938889695e-05,
+      "loss": 0.9312,
+      "step": 286
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995454340703594e-05,
+      "loss": 0.6739,
+      "step": 287
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9954214691817786e-05,
+      "loss": 0.9671,
+      "step": 288
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9953884793248244e-05,
+      "loss": 0.9489,
+      "step": 289
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9953553711340335e-05,
+      "loss": 0.5664,
+      "step": 290
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995322144610716e-05,
+      "loss": 0.5009,
+      "step": 291
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995288799756181e-05,
+      "loss": 0.651,
+      "step": 292
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.99525533657175e-05,
+      "loss": 0.8262,
+      "step": 293
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9952217550587406e-05,
+      "loss": 0.4523,
+      "step": 294
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995188055218482e-05,
+      "loss": 0.6064,
+      "step": 295
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9951542370523055e-05,
+      "loss": 1.1455,
+      "step": 296
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995120300561546e-05,
+      "loss": 0.7861,
+      "step": 297
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995086245747545e-05,
+      "loss": 0.6733,
+      "step": 298
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9950520726116476e-05,
+      "loss": 0.7927,
+      "step": 299
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995017781155204e-05,
+      "loss": 0.4693,
+      "step": 300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994983371379569e-05,
+      "loss": 0.6307,
+      "step": 301
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9949488432861015e-05,
+      "loss": 0.8713,
+      "step": 302
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994914196876166e-05,
+      "loss": 0.6627,
+      "step": 303
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9948794321511306e-05,
+      "loss": 0.8672,
+      "step": 304
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994844549112369e-05,
+      "loss": 0.7639,
+      "step": 305
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.99480954776126e-05,
+      "loss": 0.348,
+      "step": 306
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994774428099185e-05,
+      "loss": 0.7799,
+      "step": 307
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9947391901275335e-05,
+      "loss": 0.5178,
+      "step": 308
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994703833847695e-05,
+      "loss": 0.4215,
+      "step": 309
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994668359261067e-05,
+      "loss": 0.4868,
+      "step": 310
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994632766369053e-05,
+      "loss": 0.487,
+      "step": 311
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994597055173056e-05,
+      "loss": 0.805,
+      "step": 312
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994561225674489e-05,
+      "loss": 0.7643,
+      "step": 313
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994525277874767e-05,
+      "loss": 0.7875,
+      "step": 314
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994489211775309e-05,
+      "loss": 0.6267,
+      "step": 315
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994453027377542e-05,
+      "loss": 0.5351,
+      "step": 316
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994416724682893e-05,
+      "loss": 0.754,
+      "step": 317
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9943803036927975e-05,
+      "loss": 0.6203,
+      "step": 318
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9943437644086944e-05,
+      "loss": 0.5753,
+      "step": 319
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994307106832027e-05,
+      "loss": 0.8877,
+      "step": 320
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994270330964243e-05,
+      "loss": 0.5502,
+      "step": 321
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994233436806796e-05,
+      "loss": 1.2049,
+      "step": 322
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9941964243611424e-05,
+      "loss": 0.7489,
+      "step": 323
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994159293628746e-05,
+      "loss": 0.7258,
+      "step": 324
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994122044611073e-05,
+      "loss": 0.4458,
+      "step": 325
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994084677309594e-05,
+      "loss": 0.6937,
+      "step": 326
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994047191725787e-05,
+      "loss": 0.9687,
+      "step": 327
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9940095878611296e-05,
+      "loss": 0.4605,
+      "step": 328
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.993971865717112e-05,
+      "loss": 0.805,
+      "step": 329
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.993934025295221e-05,
+      "loss": 0.4787,
+      "step": 330
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.993896066596953e-05,
+      "loss": 0.12,
+      "step": 331
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9938579896238065e-05,
+      "loss": 0.3086,
+      "step": 332
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.993819794377286e-05,
+      "loss": 0.9938,
+      "step": 333
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9937814808589014e-05,
+      "loss": 1.2353,
+      "step": 334
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.993743049070165e-05,
+      "loss": 0.8871,
+      "step": 335
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.993704499012597e-05,
+      "loss": 0.5853,
+      "step": 336
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993665830687718e-05,
+      "loss": 0.9336,
+      "step": 337
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993627044097057e-05,
+      "loss": 0.7966,
+      "step": 338
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993588139242146e-05,
+      "loss": 0.4641,
+      "step": 339
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993549116124521e-05,
+      "loss": 0.8981,
+      "step": 340
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993509974745726e-05,
+      "loss": 0.7578,
+      "step": 341
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993470715107304e-05,
+      "loss": 0.9565,
+      "step": 342
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993431337210809e-05,
+      "loss": 0.56,
+      "step": 343
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993391841057795e-05,
+      "loss": 0.5455,
+      "step": 344
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993352226649823e-05,
+      "loss": 0.5952,
+      "step": 345
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9933124939884566e-05,
+      "loss": 0.8447,
+      "step": 346
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993272643075267e-05,
+      "loss": 1.1634,
+      "step": 347
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993232673911829e-05,
+      "loss": 0.8084,
+      "step": 348
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993192586499719e-05,
+      "loss": 0.9682,
+      "step": 349
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993152380840524e-05,
+      "loss": 0.4549,
+      "step": 350
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.99311205693583e-05,
+      "loss": 0.6784,
+      "step": 351
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.99307161478723e-05,
+      "loss": 0.1557,
+      "step": 352
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9930310543963236e-05,
+      "loss": 0.724,
+      "step": 353
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992990375764711e-05,
+      "loss": 1.0012,
+      "step": 354
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992949578894e-05,
+      "loss": 0.3667,
+      "step": 355
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992908663785802e-05,
+      "loss": 0.6989,
+      "step": 356
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992867630441734e-05,
+      "loss": 0.8402,
+      "step": 357
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9928264788634174e-05,
+      "loss": 0.8077,
+      "step": 358
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992785209052478e-05,
+      "loss": 0.7501,
+      "step": 359
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992743821010545e-05,
+      "loss": 0.3213,
+      "step": 360
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992702314739253e-05,
+      "loss": 0.489,
+      "step": 361
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992660690240242e-05,
+      "loss": 0.6875,
+      "step": 362
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9926189475151594e-05,
+      "loss": 0.3951,
+      "step": 363
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9925770865656504e-05,
+      "loss": 0.8949,
+      "step": 364
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.99253510739337e-05,
+      "loss": 0.428,
+      "step": 365
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9924930099999776e-05,
+      "loss": 0.59,
+      "step": 366
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992450794387136e-05,
+      "loss": 0.9964,
+      "step": 367
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992408460556511e-05,
+      "loss": 0.8316,
+      "step": 368
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992366008509777e-05,
+      "loss": 0.9497,
+      "step": 369
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992323438248612e-05,
+      "loss": 1.0118,
+      "step": 370
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9922807497746945e-05,
+      "loss": 0.5373,
+      "step": 371
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992237943089713e-05,
+      "loss": 0.8943,
+      "step": 372
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992195018195359e-05,
+      "loss": 1.0365,
+      "step": 373
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992151975093327e-05,
+      "loss": 0.9433,
+      "step": 374
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.992108813785318e-05,
+      "loss": 0.5963,
+      "step": 375
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9920655342730375e-05,
+      "loss": 0.4593,
+      "step": 376
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.992022136558194e-05,
+      "loss": 0.803,
+      "step": 377
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991978620642504e-05,
+      "loss": 0.8821,
+      "step": 378
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9919349865276846e-05,
+      "loss": 0.3701,
+      "step": 379
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991891234215461e-05,
+      "loss": 0.8142,
+      "step": 380
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991847363707561e-05,
+      "loss": 0.5424,
+      "step": 381
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991803375005717e-05,
+      "loss": 0.1847,
+      "step": 382
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9917592681116684e-05,
+      "loss": 1.0403,
+      "step": 383
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991715043027157e-05,
+      "loss": 0.9426,
+      "step": 384
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9916706997539296e-05,
+      "loss": 0.9148,
+      "step": 385
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991626238293737e-05,
+      "loss": 0.1032,
+      "step": 386
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991581658648338e-05,
+      "loss": 0.5521,
+      "step": 387
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991536960819492e-05,
+      "loss": 0.4574,
+      "step": 388
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991492144808966e-05,
+      "loss": 0.5007,
+      "step": 389
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.99144721061853e-05,
+      "loss": 0.644,
+      "step": 390
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991402158249958e-05,
+      "loss": 0.3265,
+      "step": 391
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9913569877050315e-05,
+      "loss": 1.1615,
+      "step": 392
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991311698985534e-05,
+      "loss": 0.4604,
+      "step": 393
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991266292093255e-05,
+      "loss": 0.4394,
+      "step": 394
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991220767029987e-05,
+      "loss": 0.7135,
+      "step": 395
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991175123797531e-05,
+      "loss": 0.4881,
+      "step": 396
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991129362397689e-05,
+      "loss": 0.4556,
+      "step": 397
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991083482832268e-05,
+      "loss": 0.9216,
+      "step": 398
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991037485103081e-05,
+      "loss": 0.5154,
+      "step": 399
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990991369211946e-05,
+      "loss": 0.4344,
+      "step": 400
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.7615162134170532,
+      "eval_runtime": 38.4269,
+      "eval_samples_per_second": 1.041,
+      "eval_steps_per_second": 1.041,
+      "step": 400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990945135160684e-05,
+      "loss": 1.0702,
+      "step": 401
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990898782951121e-05,
+      "loss": 0.5102,
+      "step": 402
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.99085231258509e-05,
+      "loss": 1.3747,
+      "step": 403
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990805724064425e-05,
+      "loss": 0.7087,
+      "step": 404
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990759017390967e-05,
+      "loss": 0.6781,
+      "step": 405
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9907121925665616e-05,
+      "loss": 0.7867,
+      "step": 406
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9906652495930576e-05,
+      "loss": 1.2853,
+      "step": 407
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990618188472311e-05,
+      "loss": 0.884,
+      "step": 408
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9905710092061794e-05,
+      "loss": 0.3925,
+      "step": 409
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990523711796528e-05,
+      "loss": 0.7992,
+      "step": 410
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9904762962452244e-05,
+      "loss": 0.4367,
+      "step": 411
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990428762554142e-05,
+      "loss": 0.6135,
+      "step": 412
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9903811107251594e-05,
+      "loss": 0.4931,
+      "step": 413
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9903333407601575e-05,
+      "loss": 0.3081,
+      "step": 414
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.990285452661024e-05,
+      "loss": 0.5781,
+      "step": 415
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.990237446429653e-05,
+      "loss": 1.0182,
+      "step": 416
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9901893220679375e-05,
+      "loss": 0.6112,
+      "step": 417
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9901410795777814e-05,
+      "loss": 0.635,
+      "step": 418
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.990092718961089e-05,
+      "loss": 0.5351,
+      "step": 419
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.99004424021977e-05,
+      "loss": 0.8349,
+      "step": 420
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989995643355742e-05,
+      "loss": 0.3452,
+      "step": 421
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9899469283709234e-05,
+      "loss": 0.617,
+      "step": 422
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989898095267239e-05,
+      "loss": 0.7995,
+      "step": 423
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989849144046618e-05,
+      "loss": 0.9127,
+      "step": 424
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9898000747109936e-05,
+      "loss": 0.6608,
+      "step": 425
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9897508872623055e-05,
+      "loss": 0.3631,
+      "step": 426
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9897015817024956e-05,
+      "loss": 0.6584,
+      "step": 427
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989652158033512e-05,
+      "loss": 0.8233,
+      "step": 428
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989602616257309e-05,
+      "loss": 0.5253,
+      "step": 429
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989552956375841e-05,
+      "loss": 0.6646,
+      "step": 430
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989503178391072e-05,
+      "loss": 0.6089,
+      "step": 431
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9894532823049665e-05,
+      "loss": 1.0536,
+      "step": 432
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9894032681194976e-05,
+      "loss": 1.7934,
+      "step": 433
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989353135836641e-05,
+      "loss": 0.5623,
+      "step": 434
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989302885458375e-05,
+      "loss": 0.6573,
+      "step": 435
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9892525169866865e-05,
+      "loss": 0.889,
+      "step": 436
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989202030423565e-05,
+      "loss": 0.4635,
+      "step": 437
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9891514257710065e-05,
+      "loss": 0.6891,
+      "step": 438
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989100703031008e-05,
+      "loss": 0.4895,
+      "step": 439
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989049862205574e-05,
+      "loss": 0.5198,
+      "step": 440
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988998903296712e-05,
+      "loss": 0.5811,
+      "step": 441
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9889478263064374e-05,
+      "loss": 1.0391,
+      "step": 442
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988896631236766e-05,
+      "loss": 1.3025,
+      "step": 443
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988845318089722e-05,
+      "loss": 0.6871,
+      "step": 444
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988793886867331e-05,
+      "loss": 0.5677,
+      "step": 445
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988742337571625e-05,
+      "loss": 0.8204,
+      "step": 446
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988690670204642e-05,
+      "loss": 0.477,
+      "step": 447
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988638884768421e-05,
+      "loss": 1.2118,
+      "step": 448
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.98858698126501e-05,
+      "loss": 1.2564,
+      "step": 449
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988534959696457e-05,
+      "loss": 0.6917,
+      "step": 450
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988482820064819e-05,
+      "loss": 0.6447,
+      "step": 451
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988430562372156e-05,
+      "loss": 0.9222,
+      "step": 452
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9883781866205306e-05,
+      "loss": 0.98,
+      "step": 453
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.988325692812013e-05,
+      "loss": 0.435,
+      "step": 454
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9882730809486766e-05,
+      "loss": 0.6581,
+      "step": 455
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9882203510326016e-05,
+      "loss": 0.585,
+      "step": 456
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9881675030658694e-05,
+      "loss": 1.0423,
+      "step": 457
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.988114537050568e-05,
+      "loss": 0.5581,
+      "step": 458
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.988061452988789e-05,
+      "loss": 0.589,
+      "step": 459
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.988008250882631e-05,
+      "loss": 0.6804,
+      "step": 460
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987954930734196e-05,
+      "loss": 0.2107,
+      "step": 461
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987901492545589e-05,
+      "loss": 0.4018,
+      "step": 462
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987847936318922e-05,
+      "loss": 0.9552,
+      "step": 463
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98779426205631e-05,
+      "loss": 0.1783,
+      "step": 464
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9877404697598745e-05,
+      "loss": 0.7746,
+      "step": 465
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98768655943174e-05,
+      "loss": 0.6824,
+      "step": 466
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987632531074036e-05,
+      "loss": 0.4199,
+      "step": 467
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987578384688898e-05,
+      "loss": 0.7741,
+      "step": 468
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9875241202784625e-05,
+      "loss": 0.3413,
+      "step": 469
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987469737844876e-05,
+      "loss": 0.2451,
+      "step": 470
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9874152373902866e-05,
+      "loss": 0.9966,
+      "step": 471
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9873606189168456e-05,
+      "loss": 0.963,
+      "step": 472
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987305882426712e-05,
+      "loss": 0.5658,
+      "step": 473
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987251027922048e-05,
+      "loss": 0.6155,
+      "step": 474
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987196055405021e-05,
+      "loss": 1.0087,
+      "step": 475
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987140964877802e-05,
+      "loss": 1.1266,
+      "step": 476
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9870857563425675e-05,
+      "loss": 1.0651,
+      "step": 477
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987030429801499e-05,
+      "loss": 0.8072,
+      "step": 478
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986974985256781e-05,
+      "loss": 0.6552,
+      "step": 479
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9869194227106063e-05,
+      "loss": 0.7139,
+      "step": 480
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9868637421651665e-05,
+      "loss": 0.4475,
+      "step": 481
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986807943622664e-05,
+      "loss": 1.1717,
+      "step": 482
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9867520270853026e-05,
+      "loss": 0.3493,
+      "step": 483
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986695992555291e-05,
+      "loss": 0.5741,
+      "step": 484
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986639840034843e-05,
+      "loss": 1.046,
+      "step": 485
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986583569526176e-05,
+      "loss": 1.0905,
+      "step": 486
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9865271810315145e-05,
+      "loss": 0.5712,
+      "step": 487
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986470674553085e-05,
+      "loss": 0.3063,
+      "step": 488
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98641405009312e-05,
+      "loss": 0.3301,
+      "step": 489
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986357307653857e-05,
+      "loss": 0.5914,
+      "step": 490
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986300447237537e-05,
+      "loss": 0.6798,
+      "step": 491
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986243468846407e-05,
+      "loss": 1.7471,
+      "step": 492
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9861863724827176e-05,
+      "loss": 0.9841,
+      "step": 493
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.986129158148724e-05,
+      "loss": 0.6689,
+      "step": 494
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.986071825846687e-05,
+      "loss": 0.3661,
+      "step": 495
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.986014375578872e-05,
+      "loss": 0.6552,
+      "step": 496
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985956807347547e-05,
+      "loss": 0.8427,
+      "step": 497
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9858991211549866e-05,
+      "loss": 0.5668,
+      "step": 498
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9858413170034713e-05,
+      "loss": 0.8241,
+      "step": 499
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9857833948952835e-05,
+      "loss": 0.5546,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9857253548327116e-05,
+      "loss": 0.6141,
+      "step": 501
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985667196818049e-05,
+      "loss": 1.3933,
+      "step": 502
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985608920853593e-05,
+      "loss": 0.7597,
+      "step": 503
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985550526941645e-05,
+      "loss": 1.0044,
+      "step": 504
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9854920150845126e-05,
+      "loss": 1.0349,
+      "step": 505
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985433385284507e-05,
+      "loss": 0.8327,
+      "step": 506
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9853746375439444e-05,
+      "loss": 0.7825,
+      "step": 507
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985315771865146e-05,
+      "loss": 0.1198,
+      "step": 508
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985256788250437e-05,
+      "loss": 0.9072,
+      "step": 509
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9851976867021485e-05,
+      "loss": 1.2163,
+      "step": 510
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985138467222614e-05,
+      "loss": 0.7235,
+      "step": 511
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985079129814173e-05,
+      "loss": 0.8765,
+      "step": 512
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.98501967447917e-05,
+      "loss": 1.1275,
+      "step": 513
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984960101219955e-05,
+      "loss": 0.7635,
+      "step": 514
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.98490041003888e-05,
+      "loss": 0.6363,
+      "step": 515
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9848406009383035e-05,
+      "loss": 0.7161,
+      "step": 516
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9847806739205886e-05,
+      "loss": 0.4835,
+      "step": 517
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9847206289881016e-05,
+      "loss": 1.052,
+      "step": 518
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984660466143216e-05,
+      "loss": 0.9878,
+      "step": 519
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984600185388308e-05,
+      "loss": 0.8029,
+      "step": 520
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984539786725759e-05,
+      "loss": 0.4118,
+      "step": 521
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984479270157955e-05,
+      "loss": 0.7589,
+      "step": 522
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984418635687287e-05,
+      "loss": 0.9337,
+      "step": 523
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9843578833161506e-05,
+      "loss": 0.4736,
+      "step": 524
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984297013046945e-05,
+      "loss": 0.6372,
+      "step": 525
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984236024882075e-05,
+      "loss": 0.8303,
+      "step": 526
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9841749188239514e-05,
+      "loss": 0.9619,
+      "step": 527
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984113694874986e-05,
+      "loss": 0.3287,
+      "step": 528
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.984052353037599e-05,
+      "loss": 0.3507,
+      "step": 529
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.983990893314213e-05,
+      "loss": 0.8787,
+      "step": 530
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9839293157072566e-05,
+      "loss": 1.3396,
+      "step": 531
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.983867620219162e-05,
+      "loss": 0.6647,
+      "step": 532
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.983805806852367e-05,
+      "loss": 0.6827,
+      "step": 533
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983743875609313e-05,
+      "loss": 0.3623,
+      "step": 534
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983681826492447e-05,
+      "loss": 0.7841,
+      "step": 535
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.98361965950422e-05,
+      "loss": 0.0,
+      "step": 536
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983557374647088e-05,
+      "loss": 1.0494,
+      "step": 537
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9834949719235116e-05,
+      "loss": 0.7345,
+      "step": 538
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9834324513359554e-05,
+      "loss": 0.7587,
+      "step": 539
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9833698128868904e-05,
+      "loss": 0.6899,
+      "step": 540
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983307056578791e-05,
+      "loss": 0.5664,
+      "step": 541
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983244182414136e-05,
+      "loss": 0.8609,
+      "step": 542
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983181190395409e-05,
+      "loss": 1.0576,
+      "step": 543
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9831180805250997e-05,
+      "loss": 0.7906,
+      "step": 544
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9830548528056996e-05,
+      "loss": 0.6971,
+      "step": 545
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982991507239708e-05,
+      "loss": 0.4068,
+      "step": 546
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982928043829627e-05,
+      "loss": 0.9058,
+      "step": 547
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982864462577963e-05,
+      "loss": 0.9144,
+      "step": 548
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9828007634872285e-05,
+      "loss": 0.5185,
+      "step": 549
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982736946559939e-05,
+      "loss": 0.7414,
+      "step": 550
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982673011798618e-05,
+      "loss": 0.7085,
+      "step": 551
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9826089592057886e-05,
+      "loss": 0.5191,
+      "step": 552
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9825447887839823e-05,
+      "loss": 1.109,
+      "step": 553
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982480500535734e-05,
+      "loss": 0.8237,
+      "step": 554
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982416094463584e-05,
+      "loss": 0.3588,
+      "step": 555
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9823515705700765e-05,
+      "loss": 1.0393,
+      "step": 556
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.98228692885776e-05,
+      "loss": 0.8576,
+      "step": 557
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982222169329189e-05,
+      "loss": 0.9884,
+      "step": 558
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.98215729198692e-05,
+      "loss": 0.4797,
+      "step": 559
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.982092296833519e-05,
+      "loss": 0.8401,
+      "step": 560
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9820271838715505e-05,
+      "loss": 0.3939,
+      "step": 561
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981961953103589e-05,
+      "loss": 0.4384,
+      "step": 562
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9818966045322115e-05,
+      "loss": 0.5004,
+      "step": 563
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981831138159998e-05,
+      "loss": 1.1769,
+      "step": 564
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981765553989536e-05,
+      "loss": 0.5407,
+      "step": 565
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981699852023416e-05,
+      "loss": 1.9469,
+      "step": 566
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981634032264233e-05,
+      "loss": 0.8303,
+      "step": 567
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981568094714589e-05,
+      "loss": 0.7636,
+      "step": 568
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9815020393770876e-05,
+      "loss": 1.0147,
+      "step": 569
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981435866254338e-05,
+      "loss": 0.4845,
+      "step": 570
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981369575348955e-05,
+      "loss": 0.7647,
+      "step": 571
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981303166663557e-05,
+      "loss": 0.5465,
+      "step": 572
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.981236640200769e-05,
+      "loss": 0.9299,
+      "step": 573
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9811699959632174e-05,
+      "loss": 1.1041,
+      "step": 574
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.981103233953536e-05,
+      "loss": 0.6253,
+      "step": 575
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.981036354174361e-05,
+      "loss": 0.5155,
+      "step": 576
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9809693566283354e-05,
+      "loss": 0.4748,
+      "step": 577
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980902241318107e-05,
+      "loss": 0.5962,
+      "step": 578
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9808350082463257e-05,
+      "loss": 0.4412,
+      "step": 579
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980767657415647e-05,
+      "loss": 1.0084,
+      "step": 580
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980700188828734e-05,
+      "loss": 0.6452,
+      "step": 581
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.98063260248825e-05,
+      "loss": 0.728,
+      "step": 582
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9805648983968664e-05,
+      "loss": 1.0156,
+      "step": 583
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980497076557256e-05,
+      "loss": 0.9314,
+      "step": 584
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9804291369721e-05,
+      "loss": 0.5395,
+      "step": 585
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980361079644082e-05,
+      "loss": 0.3717,
+      "step": 586
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.98029290457589e-05,
+      "loss": 0.9482,
+      "step": 587
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9802246117702174e-05,
+      "loss": 1.058,
+      "step": 588
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980156201229763e-05,
+      "loss": 0.2553,
+      "step": 589
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9800876729572286e-05,
+      "loss": 0.6815,
+      "step": 590
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980019026955321e-05,
+      "loss": 0.936,
+      "step": 591
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979950263226753e-05,
+      "loss": 1.2634,
+      "step": 592
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979881381774241e-05,
+      "loss": 0.8621,
+      "step": 593
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9798123826005045e-05,
+      "loss": 0.2491,
+      "step": 594
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9797432657082724e-05,
+      "loss": 1.2449,
+      "step": 595
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979674031100273e-05,
+      "loss": 1.0205,
+      "step": 596
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9796046787792426e-05,
+      "loss": 0.3124,
+      "step": 597
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979535208747919e-05,
+      "loss": 0.5239,
+      "step": 598
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979465621009049e-05,
+      "loss": 0.5491,
+      "step": 599
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9793959155653815e-05,
+      "loss": 0.5186,
+      "step": 600
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": NaN,
+      "eval_runtime": 38.4131,
+      "eval_samples_per_second": 1.041,
+      "eval_steps_per_second": 1.041,
+      "step": 600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979326092419668e-05,
+      "loss": 0.5517,
+      "step": 601
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9792561515746694e-05,
+      "loss": 0.1782,
+      "step": 602
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979186093033147e-05,
+      "loss": 0.2778,
+      "step": 603
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9791159167978694e-05,
+      "loss": 0.8858,
+      "step": 604
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979045622871608e-05,
+      "loss": 0.4152,
+      "step": 605
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9789752112571414e-05,
+      "loss": 0.323,
+      "step": 606
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9789046819572504e-05,
+      "loss": 0.3979,
+      "step": 607
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.978834034974721e-05,
+      "loss": 0.6477,
+      "step": 608
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.978763270312344e-05,
+      "loss": 0.5177,
+      "step": 609
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9786923879729156e-05,
+      "loss": 0.9117,
+      "step": 610
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9786213879592354e-05,
+      "loss": 0.8207,
+      "step": 611
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.978550270274109e-05,
+      "loss": 0.9046,
+      "step": 612
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9784790349203464e-05,
+      "loss": 0.3985,
+      "step": 613
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978407681900759e-05,
+      "loss": 0.6626,
+      "step": 614
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.97833621121817e-05,
+      "loss": 0.6015,
+      "step": 615
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9782646228753985e-05,
+      "loss": 0.6706,
+      "step": 616
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978192916875275e-05,
+      "loss": 0.4638,
+      "step": 617
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978121093220632e-05,
+      "loss": 0.8315,
+      "step": 618
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978049151914307e-05,
+      "loss": 0.3895,
+      "step": 619
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977977092959141e-05,
+      "loss": 0.7678,
+      "step": 620
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977904916357983e-05,
+      "loss": 0.4257,
+      "step": 621
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977832622113681e-05,
+      "loss": 0.3887,
+      "step": 622
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977760210229094e-05,
+      "loss": 0.9261,
+      "step": 623
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9776876807070815e-05,
+      "loss": 1.0082,
+      "step": 624
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977615033550509e-05,
+      "loss": 0.8356,
+      "step": 625
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9775422687622455e-05,
+      "loss": 0.7498,
+      "step": 626
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9774693863451674e-05,
+      "loss": 0.3833,
+      "step": 627
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977396386302152e-05,
+      "loss": 0.8762,
+      "step": 628
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977323268636085e-05,
+      "loss": 0.756,
+      "step": 629
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977250033349854e-05,
+      "loss": 0.8989,
+      "step": 630
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977176680446352e-05,
+      "loss": 0.9637,
+      "step": 631
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977103209928477e-05,
+      "loss": 0.2843,
+      "step": 632
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977029621799132e-05,
+      "loss": 0.5,
+      "step": 633
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976955916061223e-05,
+      "loss": 1.1824,
+      "step": 634
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9768820927176634e-05,
+      "loss": 0.5123,
+      "step": 635
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9768081517713675e-05,
+      "loss": 0.9832,
+      "step": 636
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976734093225258e-05,
+      "loss": 0.5566,
+      "step": 637
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976659917082261e-05,
+      "loss": 0.5623,
+      "step": 638
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976585623345305e-05,
+      "loss": 0.0,
+      "step": 639
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976511212017326e-05,
+      "loss": 0.431,
+      "step": 640
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976436683101264e-05,
+      "loss": 0.9735,
+      "step": 641
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976362036600063e-05,
+      "loss": 0.8463,
+      "step": 642
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976287272516672e-05,
+      "loss": 0.5398,
+      "step": 643
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9762123908540446e-05,
+      "loss": 0.3544,
+      "step": 644
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9761373916151384e-05,
+      "loss": 0.4609,
+      "step": 645
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.976062274802917e-05,
+      "loss": 1.0651,
+      "step": 646
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.975987040420347e-05,
+      "loss": 0.5683,
+      "step": 647
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.975911688470402e-05,
+      "loss": 0.5033,
+      "step": 648
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.975836218956057e-05,
+      "loss": 0.8687,
+      "step": 649
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.975760631880296e-05,
+      "loss": 0.38,
+      "step": 650
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.975684927246103e-05,
+      "loss": 0.5074,
+      "step": 651
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9756091050564686e-05,
+      "loss": 0.574,
+      "step": 652
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97553316531439e-05,
+      "loss": 0.2834,
+      "step": 653
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9754571080228656e-05,
+      "loss": 1.0199,
+      "step": 654
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9753809331849e-05,
+      "loss": 0.8549,
+      "step": 655
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.975304640803504e-05,
+      "loss": 0.6344,
+      "step": 656
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97522823088169e-05,
+      "loss": 0.2941,
+      "step": 657
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.975151703422478e-05,
+      "loss": 0.7182,
+      "step": 658
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97507505842889e-05,
+      "loss": 0.9642,
+      "step": 659
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974998295903955e-05,
+      "loss": 0.8228,
+      "step": 660
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974921415850703e-05,
+      "loss": 1.0469,
+      "step": 661
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974844418272175e-05,
+      "loss": 1.0153,
+      "step": 662
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97476730317141e-05,
+      "loss": 0.7459,
+      "step": 663
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974690070551456e-05,
+      "loss": 0.6995,
+      "step": 664
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9746127204153623e-05,
+      "loss": 0.6856,
+      "step": 665
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974535252766187e-05,
+      "loss": 1.3125,
+      "step": 666
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9744576676069896e-05,
+      "loss": 0.8592,
+      "step": 667
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974379964940834e-05,
+      "loss": 0.8525,
+      "step": 668
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9743021447707915e-05,
+      "loss": 0.1707,
+      "step": 669
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9742242070999345e-05,
+      "loss": 0.6415,
+      "step": 670
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974146151931343e-05,
+      "loss": 0.3453,
+      "step": 671
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9740679792681016e-05,
+      "loss": 0.4477,
+      "step": 672
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9739896891132975e-05,
+      "loss": 0.633,
+      "step": 673
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9739112814700234e-05,
+      "loss": 0.6456,
+      "step": 674
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973832756341376e-05,
+      "loss": 0.9755,
+      "step": 675
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9737541137304604e-05,
+      "loss": 0.4085,
+      "step": 676
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97367535364038e-05,
+      "loss": 0.4389,
+      "step": 677
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973596476074248e-05,
+      "loss": 1.1012,
+      "step": 678
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973517481035182e-05,
+      "loss": 0.0,
+      "step": 679
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973438368526299e-05,
+      "loss": 0.4893,
+      "step": 680
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973359138550727e-05,
+      "loss": 1.017,
+      "step": 681
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973279791111596e-05,
+      "loss": 0.8517,
+      "step": 682
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973200326212039e-05,
+      "loss": 1.3503,
+      "step": 683
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973120743855197e-05,
+      "loss": 0.7602,
+      "step": 684
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.973041044044213e-05,
+      "loss": 0.4444,
+      "step": 685
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.972961226782236e-05,
+      "loss": 0.6146,
+      "step": 686
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.972881292072419e-05,
+      "loss": 0.8157,
+      "step": 687
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97280123991792e-05,
+      "loss": 0.4871,
+      "step": 688
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.972721070321901e-05,
+      "loss": 1.0476,
+      "step": 689
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9726407832875304e-05,
+      "loss": 0.4555,
+      "step": 690
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9725603788179774e-05,
+      "loss": 0.6441,
+      "step": 691
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.972479856916421e-05,
+      "loss": 0.882,
+      "step": 692
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9723992175860416e-05,
+      "loss": 0.676,
+      "step": 693
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.972318460830024e-05,
+      "loss": 0.8777,
+      "step": 694
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.97223758665156e-05,
+      "loss": 0.7623,
+      "step": 695
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.972156595053843e-05,
+      "loss": 0.7604,
+      "step": 696
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9720754860400726e-05,
+      "loss": 0.7835,
+      "step": 697
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971994259613454e-05,
+      "loss": 1.3098,
+      "step": 698
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971912915777197e-05,
+      "loss": 0.8731,
+      "step": 699
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971831454534512e-05,
+      "loss": 0.84,
+      "step": 700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9717498758886206e-05,
+      "loss": 1.02,
+      "step": 701
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9716681798427444e-05,
+      "loss": 0.8852,
+      "step": 702
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971586366400109e-05,
+      "loss": 0.8711,
+      "step": 703
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971504435563948e-05,
+      "loss": 0.5718,
+      "step": 704
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971422387337499e-05,
+      "loss": 1.045,
+      "step": 705
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971340221724002e-05,
+      "loss": 0.7086,
+      "step": 706
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971257938726704e-05,
+      "loss": 0.6384,
+      "step": 707
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971175538348854e-05,
+      "loss": 0.4659,
+      "step": 708
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971093020593709e-05,
+      "loss": 0.7454,
+      "step": 709
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971010385464528e-05,
+      "loss": 0.7624,
+      "step": 710
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9709276329645743e-05,
+      "loss": 0.2785,
+      "step": 711
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.97084476309712e-05,
+      "loss": 0.9069,
+      "step": 712
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970761775865437e-05,
+      "loss": 1.0217,
+      "step": 713
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970678671272804e-05,
+      "loss": 0.2893,
+      "step": 714
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970595449322504e-05,
+      "loss": 0.5966,
+      "step": 715
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9705121100178244e-05,
+      "loss": 0.584,
+      "step": 716
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970428653362059e-05,
+      "loss": 0.5487,
+      "step": 717
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970345079358504e-05,
+      "loss": 0.5771,
+      "step": 718
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.97026138801046e-05,
+      "loss": 0.8909,
+      "step": 719
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970177579321234e-05,
+      "loss": 0.9557,
+      "step": 720
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970093653294137e-05,
+      "loss": 0.7932,
+      "step": 721
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.970009609932485e-05,
+      "loss": 0.1584,
+      "step": 722
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9699254492395973e-05,
+      "loss": 0.9521,
+      "step": 723
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.969841171218799e-05,
+      "loss": 0.5005,
+      "step": 724
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9697567758734186e-05,
+      "loss": 1.1867,
+      "step": 725
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9696722632067926e-05,
+      "loss": 0.4459,
+      "step": 726
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.969587633222258e-05,
+      "loss": 1.1169,
+      "step": 727
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.969502885923158e-05,
+      "loss": 0.7873,
+      "step": 728
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.969418021312842e-05,
+      "loss": 0.4391,
+      "step": 729
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.96933303939466e-05,
+      "loss": 0.4134,
+      "step": 730
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9692479401719715e-05,
+      "loss": 0.6766,
+      "step": 731
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.969162723648138e-05,
+      "loss": 0.6211,
+      "step": 732
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.969077389826525e-05,
+      "loss": 0.4241,
+      "step": 733
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9689919387105046e-05,
+      "loss": 1.0092,
+      "step": 734
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968906370303452e-05,
+      "loss": 1.092,
+      "step": 735
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968820684608748e-05,
+      "loss": 0.5606,
+      "step": 736
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968734881629778e-05,
+      "loss": 1.0255,
+      "step": 737
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968648961369931e-05,
+      "loss": 0.7812,
+      "step": 738
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968562923832601e-05,
+      "loss": 0.9258,
+      "step": 739
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968476769021188e-05,
+      "loss": 0.966,
+      "step": 740
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968390496939095e-05,
+      "loss": 0.6755,
+      "step": 741
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9683041075897306e-05,
+      "loss": 0.7615,
+      "step": 742
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968217600976507e-05,
+      "loss": 0.9228,
+      "step": 743
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9681309771028415e-05,
+      "loss": 0.6916,
+      "step": 744
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968044235972157e-05,
+      "loss": 0.2906,
+      "step": 745
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.96795737758788e-05,
+      "loss": 0.8763,
+      "step": 746
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.967870401953442e-05,
+      "loss": 0.8196,
+      "step": 747
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.967783309072278e-05,
+      "loss": 1.0438,
+      "step": 748
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9676960989478304e-05,
+      "loss": 0.9119,
+      "step": 749
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.967608771583543e-05,
+      "loss": 0.3498,
+      "step": 750
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9675213269828664e-05,
+      "loss": 0.9942,
+      "step": 751
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9674337651492545e-05,
+      "loss": 0.5793,
+      "step": 752
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.967346086086166e-05,
+      "loss": 1.018,
+      "step": 753
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9672582897970676e-05,
+      "loss": 0.4344,
+      "step": 754
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9671703762854245e-05,
+      "loss": 0.6866,
+      "step": 755
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.967082345554711e-05,
+      "loss": 0.2913,
+      "step": 756
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966994197608404e-05,
+      "loss": 0.5427,
+      "step": 757
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9669059324499876e-05,
+      "loss": 0.429,
+      "step": 758
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966817550082948e-05,
+      "loss": 0.8835,
+      "step": 759
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966729050510776e-05,
+      "loss": 0.8267,
+      "step": 760
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966640433736968e-05,
+      "loss": 0.8429,
+      "step": 761
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966551699765025e-05,
+      "loss": 0.753,
+      "step": 762
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9664628485984525e-05,
+      "loss": 0.997,
+      "step": 763
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966373880240761e-05,
+      "loss": 0.8655,
+      "step": 764
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966284794695465e-05,
+      "loss": 0.432,
+      "step": 765
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9661955919660834e-05,
+      "loss": 1.114,
+      "step": 766
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.96610627205614e-05,
+      "loss": 0.597,
+      "step": 767
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.966016834969165e-05,
+      "loss": 0.9548,
+      "step": 768
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.96592728070869e-05,
+      "loss": 0.8023,
+      "step": 769
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.965837609278254e-05,
+      "loss": 0.7178,
+      "step": 770
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.965747820681398e-05,
+      "loss": 0.5022,
+      "step": 771
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.96565791492167e-05,
+      "loss": 1.1367,
+      "step": 772
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9655678920026234e-05,
+      "loss": 1.0272,
+      "step": 773
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965477751927812e-05,
+      "loss": 0.9677,
+      "step": 774
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965387494700797e-05,
+      "loss": 0.2515,
+      "step": 775
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965297120325146e-05,
+      "loss": 0.798,
+      "step": 776
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965206628804428e-05,
+      "loss": 0.5355,
+      "step": 777
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965116020142218e-05,
+      "loss": 0.5037,
+      "step": 778
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9650252943420956e-05,
+      "loss": 1.148,
+      "step": 779
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9649344514076444e-05,
+      "loss": 0.8077,
+      "step": 780
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9648434913424544e-05,
+      "loss": 0.684,
+      "step": 781
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.964752414150118e-05,
+      "loss": 0.1205,
+      "step": 782
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9646612198342325e-05,
+      "loss": 0.5679,
+      "step": 783
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.964569908398403e-05,
+      "loss": 0.5582,
+      "step": 784
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.964478479846235e-05,
+      "loss": 0.2099,
+      "step": 785
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.964386934181341e-05,
+      "loss": 1.2486,
+      "step": 786
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9642952714073366e-05,
+      "loss": 0.5466,
+      "step": 787
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.964203491527844e-05,
+      "loss": 1.3289,
+      "step": 788
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.96411159454649e-05,
+      "loss": 0.5627,
+      "step": 789
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9640195804669026e-05,
+      "loss": 0.7939,
+      "step": 790
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9639274492927184e-05,
+      "loss": 0.6946,
+      "step": 791
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.963835201027575e-05,
+      "loss": 0.4024,
+      "step": 792
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9637428356751205e-05,
+      "loss": 0.6141,
+      "step": 793
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9636503532390015e-05,
+      "loss": 0.7694,
+      "step": 794
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.96355775372287e-05,
+      "loss": 0.813,
+      "step": 795
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.963465037130388e-05,
+      "loss": 0.6485,
+      "step": 796
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.963372203465215e-05,
+      "loss": 0.8902,
+      "step": 797
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.963279252731019e-05,
+      "loss": 1.215,
+      "step": 798
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9631861849314746e-05,
+      "loss": 0.6294,
+      "step": 799
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.963093000070255e-05,
+      "loss": 0.3333,
+      "step": 800
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.7615382075309753,
+      "eval_runtime": 39.0122,
+      "eval_samples_per_second": 1.025,
+      "eval_steps_per_second": 1.025,
+      "step": 800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962999698151044e-05,
+      "loss": 0.4977,
+      "step": 801
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962906279177525e-05,
+      "loss": 1.025,
+      "step": 802
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962812743153392e-05,
+      "loss": 0.9211,
+      "step": 803
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962719090082338e-05,
+      "loss": 0.9257,
+      "step": 804
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9626253199680625e-05,
+      "loss": 0.8042,
+      "step": 805
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962531432814271e-05,
+      "loss": 0.7421,
+      "step": 806
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9624374286246714e-05,
+      "loss": 0.562,
+      "step": 807
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962343307402978e-05,
+      "loss": 0.3951,
+      "step": 808
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.96224906915291e-05,
+      "loss": 0.5552,
+      "step": 809
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.962154713878189e-05,
+      "loss": 0.6422,
+      "step": 810
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.962060241582543e-05,
+      "loss": 0.1655,
+      "step": 811
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961965652269704e-05,
+      "loss": 0.8727,
+      "step": 812
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961870945943408e-05,
+      "loss": 0.6662,
+      "step": 813
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9617761226073985e-05,
+      "loss": 0.1901,
+      "step": 814
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9616811822654206e-05,
+      "loss": 1.6183,
+      "step": 815
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961586124921224e-05,
+      "loss": 0.5908,
+      "step": 816
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961490950578565e-05,
+      "loss": 0.6052,
+      "step": 817
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9613956592412026e-05,
+      "loss": 0.5041,
+      "step": 818
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9613002509129026e-05,
+      "loss": 0.6526,
+      "step": 819
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961204725597433e-05,
+      "loss": 0.8437,
+      "step": 820
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961109083298569e-05,
+      "loss": 0.8783,
+      "step": 821
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961013324020086e-05,
+      "loss": 0.5504,
+      "step": 822
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960917447765771e-05,
+      "loss": 0.897,
+      "step": 823
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960821454539408e-05,
+      "loss": 0.8128,
+      "step": 824
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960725344344792e-05,
+      "loss": 0.8118,
+      "step": 825
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960629117185718e-05,
+      "loss": 0.7126,
+      "step": 826
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960532773065988e-05,
+      "loss": 1.2627,
+      "step": 827
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960436311989409e-05,
+      "loss": 0.2971,
+      "step": 828
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.96033973395979e-05,
+      "loss": 0.8212,
+      "step": 829
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960243038980948e-05,
+      "loss": 0.7756,
+      "step": 830
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960146227056703e-05,
+      "loss": 0.4844,
+      "step": 831
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960049298190878e-05,
+      "loss": 0.7713,
+      "step": 832
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9599522523873037e-05,
+      "loss": 0.9457,
+      "step": 833
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.959855089649813e-05,
+      "loss": 0.5766,
+      "step": 834
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9597578099822446e-05,
+      "loss": 0.4516,
+      "step": 835
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9596604133884415e-05,
+      "loss": 0.2418,
+      "step": 836
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.959562899872252e-05,
+      "loss": 1.0081,
+      "step": 837
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.959465269437528e-05,
+      "loss": 0.6673,
+      "step": 838
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.959367522088126e-05,
+      "loss": 1.029,
+      "step": 839
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.959269657827908e-05,
+      "loss": 0.5733,
+      "step": 840
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9591716766607404e-05,
+      "loss": 0.8322,
+      "step": 841
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.959073578590494e-05,
+      "loss": 1.1657,
+      "step": 842
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958975363621043e-05,
+      "loss": 0.7532,
+      "step": 843
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958877031756269e-05,
+      "loss": 0.4601,
+      "step": 844
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958778583000055e-05,
+      "loss": 0.4183,
+      "step": 845
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958680017356292e-05,
+      "loss": 0.6131,
+      "step": 846
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958581334828874e-05,
+      "loss": 0.8911,
+      "step": 847
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958482535421696e-05,
+      "loss": 0.945,
+      "step": 848
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9583836191386656e-05,
+      "loss": 0.5979,
+      "step": 849
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.958284585983687e-05,
+      "loss": 0.9321,
+      "step": 850
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9581854359606754e-05,
+      "loss": 0.6028,
+      "step": 851
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.958086169073546e-05,
+      "loss": 0.4407,
+      "step": 852
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.95798678532622e-05,
+      "loss": 0.8736,
+      "step": 853
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957887284722626e-05,
+      "loss": 0.9917,
+      "step": 854
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9577876672666924e-05,
+      "loss": 0.7549,
+      "step": 855
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957687932962355e-05,
+      "loss": 0.4721,
+      "step": 856
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957588081813555e-05,
+      "loss": 1.0013,
+      "step": 857
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9574881138242356e-05,
+      "loss": 0.7994,
+      "step": 858
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957388028998348e-05,
+      "loss": 0.2531,
+      "step": 859
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957287827339844e-05,
+      "loss": 0.616,
+      "step": 860
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957187508852683e-05,
+      "loss": 0.5798,
+      "step": 861
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9570870735408276e-05,
+      "loss": 0.502,
+      "step": 862
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956986521408246e-05,
+      "loss": 0.5222,
+      "step": 863
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956885852458912e-05,
+      "loss": 0.7518,
+      "step": 864
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9567850666968e-05,
+      "loss": 0.4949,
+      "step": 865
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956684164125893e-05,
+      "loss": 0.4812,
+      "step": 866
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956583144750177e-05,
+      "loss": 0.7017,
+      "step": 867
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956482008573642e-05,
+      "loss": 0.8375,
+      "step": 868
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956380755600285e-05,
+      "loss": 0.6748,
+      "step": 869
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9562793858341046e-05,
+      "loss": 0.7547,
+      "step": 870
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9561778992791064e-05,
+      "loss": 0.4284,
+      "step": 871
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956076295939299e-05,
+      "loss": 0.5725,
+      "step": 872
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955974575818696e-05,
+      "loss": 0.1601,
+      "step": 873
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9558727389213174e-05,
+      "loss": 1.2319,
+      "step": 874
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955770785251185e-05,
+      "loss": 0.8134,
+      "step": 875
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955668714812326e-05,
+      "loss": 0.4017,
+      "step": 876
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955566527608774e-05,
+      "loss": 0.8638,
+      "step": 877
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9554642236445665e-05,
+      "loss": 1.174,
+      "step": 878
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955361802923743e-05,
+      "loss": 0.5909,
+      "step": 879
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955259265450351e-05,
+      "loss": 0.7237,
+      "step": 880
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955156611228442e-05,
+      "loss": 0.4511,
+      "step": 881
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.955053840262068e-05,
+      "loss": 0.9587,
+      "step": 882
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9549509525552937e-05,
+      "loss": 0.4167,
+      "step": 883
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9548479481121805e-05,
+      "loss": 1.163,
+      "step": 884
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9547448269367986e-05,
+      "loss": 0.0575,
+      "step": 885
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.954641589033221e-05,
+      "loss": 0.4612,
+      "step": 886
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.954538234405528e-05,
+      "loss": 1.046,
+      "step": 887
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9544347630578015e-05,
+      "loss": 0.3963,
+      "step": 888
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.954331174994129e-05,
+      "loss": 0.4283,
+      "step": 889
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.954227470218603e-05,
+      "loss": 0.5654,
+      "step": 890
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.95412364873532e-05,
+      "loss": 0.7174,
+      "step": 891
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9540197105483826e-05,
+      "loss": 0.6267,
+      "step": 892
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9539156556618954e-05,
+      "loss": 0.5234,
+      "step": 893
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953811484079971e-05,
+      "loss": 0.6742,
+      "step": 894
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9537071958067234e-05,
+      "loss": 1.2314,
+      "step": 895
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953602790846272e-05,
+      "loss": 0.8291,
+      "step": 896
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953498269202743e-05,
+      "loss": 0.6891,
+      "step": 897
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953393630880264e-05,
+      "loss": 0.5632,
+      "step": 898
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953288875882971e-05,
+      "loss": 0.5854,
+      "step": 899
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9531840042149993e-05,
+      "loss": 1.0753,
+      "step": 900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953079015880494e-05,
+      "loss": 0.9642,
+      "step": 901
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952973910883602e-05,
+      "loss": 1.0263,
+      "step": 902
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952868689228477e-05,
+      "loss": 1.8415,
+      "step": 903
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9527633509192724e-05,
+      "loss": 0.8278,
+      "step": 904
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952657895960153e-05,
+      "loss": 1.4415,
+      "step": 905
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952552324355284e-05,
+      "loss": 0.8477,
+      "step": 906
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9524466361088346e-05,
+      "loss": 0.5146,
+      "step": 907
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952340831224982e-05,
+      "loss": 0.5327,
+      "step": 908
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952234909707905e-05,
+      "loss": 0.5911,
+      "step": 909
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9521288715617875e-05,
+      "loss": 0.817,
+      "step": 910
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9520227167908195e-05,
+      "loss": 0.5918,
+      "step": 911
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.951916445399195e-05,
+      "loss": 0.7832,
+      "step": 912
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9518100573911106e-05,
+      "loss": 0.5615,
+      "step": 913
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9517035527707715e-05,
+      "loss": 1.215,
+      "step": 914
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.951596931542384e-05,
+      "loss": 0.7088,
+      "step": 915
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.951490193710159e-05,
+      "loss": 0.5522,
+      "step": 916
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.951383339278316e-05,
+      "loss": 0.8805,
+      "step": 917
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.951276368251074e-05,
+      "loss": 0.6197,
+      "step": 918
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9511692806326597e-05,
+      "loss": 0.8468,
+      "step": 919
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.951062076427303e-05,
+      "loss": 0.5901,
+      "step": 920
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.950954755639241e-05,
+      "loss": 0.4967,
+      "step": 921
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.950847318272711e-05,
+      "loss": 0.3619,
+      "step": 922
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.950739764331959e-05,
+      "loss": 0.7518,
+      "step": 923
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9506320938212335e-05,
+      "loss": 0.9904,
+      "step": 924
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.950524306744788e-05,
+      "loss": 0.5851,
+      "step": 925
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.95041640310688e-05,
+      "loss": 0.3547,
+      "step": 926
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9503083829117736e-05,
+      "loss": 0.8399,
+      "step": 927
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.950200246163735e-05,
+      "loss": 0.5277,
+      "step": 928
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.950091992867038e-05,
+      "loss": 0.5781,
+      "step": 929
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949983623025956e-05,
+      "loss": 0.2153,
+      "step": 930
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949875136644773e-05,
+      "loss": 0.7033,
+      "step": 931
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9497665337277734e-05,
+      "loss": 1.1275,
+      "step": 932
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9496578142792484e-05,
+      "loss": 0.5945,
+      "step": 933
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949548978303493e-05,
+      "loss": 1.0196,
+      "step": 934
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9494400258048056e-05,
+      "loss": 0.4635,
+      "step": 935
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949330956787491e-05,
+      "loss": 0.5602,
+      "step": 936
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9492217712558585e-05,
+      "loss": 0.9025,
+      "step": 937
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949112469214222e-05,
+      "loss": 0.9187,
+      "step": 938
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949003050666898e-05,
+      "loss": 1.1808,
+      "step": 939
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.94889351561821e-05,
+      "loss": 0.9822,
+      "step": 940
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948783864072486e-05,
+      "loss": 0.6586,
+      "step": 941
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9486740960340555e-05,
+      "loss": 1.2622,
+      "step": 942
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948564211507257e-05,
+      "loss": 0.9431,
+      "step": 943
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948454210496431e-05,
+      "loss": 1.0412,
+      "step": 944
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948344093005923e-05,
+      "loss": 1.0572,
+      "step": 945
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948233859040083e-05,
+      "loss": 0.4209,
+      "step": 946
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948123508603267e-05,
+      "loss": 1.551,
+      "step": 947
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.948013041699832e-05,
+      "loss": 1.1328,
+      "step": 948
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947902458334144e-05,
+      "loss": 0.9336,
+      "step": 949
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947791758510571e-05,
+      "loss": 0.9628,
+      "step": 950
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947680942233487e-05,
+      "loss": 1.0437,
+      "step": 951
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947570009507269e-05,
+      "loss": 0.4845,
+      "step": 952
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947458960336299e-05,
+      "loss": 1.0038,
+      "step": 953
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947347794724966e-05,
+      "loss": 0.3998,
+      "step": 954
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9472365126776594e-05,
+      "loss": 0.63,
+      "step": 955
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947125114198777e-05,
+      "loss": 0.9183,
+      "step": 956
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.947013599292718e-05,
+      "loss": 1.2634,
+      "step": 957
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946901967963889e-05,
+      "loss": 1.4595,
+      "step": 958
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946790220216701e-05,
+      "loss": 0.4982,
+      "step": 959
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946678356055567e-05,
+      "loss": 0.9711,
+      "step": 960
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946566375484906e-05,
+      "loss": 0.9508,
+      "step": 961
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946454278509143e-05,
+      "loss": 0.4829,
+      "step": 962
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946342065132706e-05,
+      "loss": 0.6701,
+      "step": 963
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946229735360028e-05,
+      "loss": 0.6656,
+      "step": 964
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9461172891955464e-05,
+      "loss": 0.8745,
+      "step": 965
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.946004726643704e-05,
+      "loss": 0.7693,
+      "step": 966
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.945892047708948e-05,
+      "loss": 0.7872,
+      "step": 967
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.945779252395728e-05,
+      "loss": 1.0007,
+      "step": 968
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9456663407085005e-05,
+      "loss": 1.129,
+      "step": 969
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.945553312651728e-05,
+      "loss": 0.8474,
+      "step": 970
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.945440168229875e-05,
+      "loss": 0.8179,
+      "step": 971
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.945326907447409e-05,
+      "loss": 0.4537,
+      "step": 972
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9452135303088076e-05,
+      "loss": 0.8081,
+      "step": 973
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.945100036818547e-05,
+      "loss": 0.836,
+      "step": 974
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9449864269811134e-05,
+      "loss": 0.8352,
+      "step": 975
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9448727008009936e-05,
+      "loss": 0.7871,
+      "step": 976
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.94475885828268e-05,
+      "loss": 0.5118,
+      "step": 977
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.944644899430671e-05,
+      "loss": 0.5388,
+      "step": 978
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.944530824249468e-05,
+      "loss": 0.9298,
+      "step": 979
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9444166327435774e-05,
+      "loss": 0.3085,
+      "step": 980
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.944302324917511e-05,
+      "loss": 0.4033,
+      "step": 981
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.944187900775784e-05,
+      "loss": 0.774,
+      "step": 982
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9440733603229176e-05,
+      "loss": 0.6587,
+      "step": 983
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.943958703563435e-05,
+      "loss": 0.7901,
+      "step": 984
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.943843930501868e-05,
+      "loss": 0.7482,
+      "step": 985
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.943729041142749e-05,
+      "loss": 0.5077,
+      "step": 986
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9436140354906176e-05,
+      "loss": 0.7052,
+      "step": 987
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.943498913550017e-05,
+      "loss": 1.0359,
+      "step": 988
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9433836753254953e-05,
+      "loss": 0.7171,
+      "step": 989
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9432683208216044e-05,
+      "loss": 0.3096,
+      "step": 990
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.943152850042902e-05,
+      "loss": 0.6341,
+      "step": 991
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9430372629939503e-05,
+      "loss": 0.7912,
+      "step": 992
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.942921559679314e-05,
+      "loss": 0.6694,
+      "step": 993
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9428057401035654e-05,
+      "loss": 0.8449,
+      "step": 994
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9426898042712795e-05,
+      "loss": 0.7973,
+      "step": 995
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.942573752187036e-05,
+      "loss": 0.8771,
+      "step": 996
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.942457583855421e-05,
+      "loss": 1.3826,
+      "step": 997
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9423412992810215e-05,
+      "loss": 0.7641,
+      "step": 998
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.942224898468433e-05,
+      "loss": 0.5744,
+      "step": 999
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.942108381422254e-05,
+      "loss": 0.3806,
+      "step": 1000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.74680495262146,
+      "eval_runtime": 38.0688,
+      "eval_samples_per_second": 1.051,
+      "eval_steps_per_second": 1.051,
+      "step": 1000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.941991748147087e-05,
+      "loss": 0.4738,
+      "step": 1001
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.941874998647539e-05,
+      "loss": 0.8427,
+      "step": 1002
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9417581329282235e-05,
+      "loss": 1.1159,
+      "step": 1003
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.941641150993758e-05,
+      "loss": 1.3073,
+      "step": 1004
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9415240528487616e-05,
+      "loss": 0.8298,
+      "step": 1005
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.941406838497861e-05,
+      "loss": 1.2294,
+      "step": 1006
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9412895079456876e-05,
+      "loss": 0.4976,
+      "step": 1007
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.941172061196876e-05,
+      "loss": 0.6595,
+      "step": 1008
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.941054498256067e-05,
+      "loss": 0.8598,
+      "step": 1009
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940936819127903e-05,
+      "loss": 0.7398,
+      "step": 1010
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940819023817035e-05,
+      "loss": 1.0836,
+      "step": 1011
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940701112328116e-05,
+      "loss": 0.5301,
+      "step": 1012
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9405830846658025e-05,
+      "loss": 0.5051,
+      "step": 1013
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940464940834759e-05,
+      "loss": 1.346,
+      "step": 1014
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9403466808396524e-05,
+      "loss": 0.4535,
+      "step": 1015
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9402283046851546e-05,
+      "loss": 0.9299,
+      "step": 1016
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940109812375942e-05,
+      "loss": 0.6718,
+      "step": 1017
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939991203916696e-05,
+      "loss": 0.5103,
+      "step": 1018
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939872479312101e-05,
+      "loss": 0.6974,
+      "step": 1019
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939753638566849e-05,
+      "loss": 0.4595,
+      "step": 1020
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9396346816856326e-05,
+      "loss": 1.1487,
+      "step": 1021
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939515608673155e-05,
+      "loss": 1.2432,
+      "step": 1022
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939396419534116e-05,
+      "loss": 1.1913,
+      "step": 1023
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9392771142732274e-05,
+      "loss": 0.6698,
+      "step": 1024
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9391576928951994e-05,
+      "loss": 0.4274,
+      "step": 1025
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939038155404753e-05,
+      "loss": 0.5334,
+      "step": 1026
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9389185018066084e-05,
+      "loss": 0.4617,
+      "step": 1027
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9387987321054936e-05,
+      "loss": 0.2429,
+      "step": 1028
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.938678846306139e-05,
+      "loss": 0.5374,
+      "step": 1029
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.938558844413282e-05,
+      "loss": 1.2736,
+      "step": 1030
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.938438726431663e-05,
+      "loss": 0.9455,
+      "step": 1031
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9383184923660264e-05,
+      "loss": 0.2966,
+      "step": 1032
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.938198142221123e-05,
+      "loss": 0.6437,
+      "step": 1033
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9380776760017076e-05,
+      "loss": 0.6301,
+      "step": 1034
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9379570937125385e-05,
+      "loss": 0.719,
+      "step": 1035
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9378363953583796e-05,
+      "loss": 0.5496,
+      "step": 1036
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.937715580943999e-05,
+      "loss": 0.4915,
+      "step": 1037
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.937594650474169e-05,
+      "loss": 0.6062,
+      "step": 1038
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9374736039536684e-05,
+      "loss": 0.6607,
+      "step": 1039
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9373524413872775e-05,
+      "loss": 0.8415,
+      "step": 1040
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9372311627797846e-05,
+      "loss": 1.1888,
+      "step": 1041
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.93710976813598e-05,
+      "loss": 0.5568,
+      "step": 1042
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9369882574606585e-05,
+      "loss": 0.8896,
+      "step": 1043
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.936866630758622e-05,
+      "loss": 0.2924,
+      "step": 1044
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.936744888034674e-05,
+      "loss": 0.1894,
+      "step": 1045
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.936623029293626e-05,
+      "loss": 0.8082,
+      "step": 1046
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.936501054540291e-05,
+      "loss": 0.8812,
+      "step": 1047
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.936378963779486e-05,
+      "loss": 0.7861,
+      "step": 1048
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.936256757016037e-05,
+      "loss": 0.5843,
+      "step": 1049
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9361344342547704e-05,
+      "loss": 0.1518,
+      "step": 1050
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.936011995500517e-05,
+      "loss": 0.7739,
+      "step": 1051
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.935889440758118e-05,
+      "loss": 0.4484,
+      "step": 1052
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.935766770032411e-05,
+      "loss": 0.7234,
+      "step": 1053
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.935643983328243e-05,
+      "loss": 0.939,
+      "step": 1054
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.935521080650467e-05,
+      "loss": 0.8332,
+      "step": 1055
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.935398062003936e-05,
+      "loss": 0.4331,
+      "step": 1056
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93527492739351e-05,
+      "loss": 1.1111,
+      "step": 1057
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9351516768240546e-05,
+      "loss": 1.2546,
+      "step": 1058
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9350283103004384e-05,
+      "loss": 0.958,
+      "step": 1059
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934904827827535e-05,
+      "loss": 0.2438,
+      "step": 1060
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934781229410222e-05,
+      "loss": 0.6399,
+      "step": 1061
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934657515053383e-05,
+      "loss": 1.6678,
+      "step": 1062
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9345336847619054e-05,
+      "loss": 0.689,
+      "step": 1063
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9344097385406806e-05,
+      "loss": 0.9553,
+      "step": 1064
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934285676394606e-05,
+      "loss": 0.5002,
+      "step": 1065
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934161498328581e-05,
+      "loss": 1.3249,
+      "step": 1066
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934037204347513e-05,
+      "loss": 0.7522,
+      "step": 1067
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9339127944563116e-05,
+      "loss": 0.6611,
+      "step": 1068
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.933788268659892e-05,
+      "loss": 0.9879,
+      "step": 1069
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.933663626963174e-05,
+      "loss": 0.9013,
+      "step": 1070
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93353886937108e-05,
+      "loss": 0.7113,
+      "step": 1071
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9334139958885395e-05,
+      "loss": 0.4651,
+      "step": 1072
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.933289006520487e-05,
+      "loss": 0.7993,
+      "step": 1073
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.933163901271857e-05,
+      "loss": 0.7388,
+      "step": 1074
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.933038680147595e-05,
+      "loss": 0.4926,
+      "step": 1075
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.932913343152647e-05,
+      "loss": 0.4414,
+      "step": 1076
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9327878902919644e-05,
+      "loss": 0.6463,
+      "step": 1077
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.932662321570502e-05,
+      "loss": 0.6136,
+      "step": 1078
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.932536636993223e-05,
+      "loss": 1.3414,
+      "step": 1079
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93241083656509e-05,
+      "loss": 0.407,
+      "step": 1080
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.932284920291075e-05,
+      "loss": 1.0213,
+      "step": 1081
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93215888817615e-05,
+      "loss": 0.2336,
+      "step": 1082
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.932032740225296e-05,
+      "loss": 0.4667,
+      "step": 1083
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.931906476443495e-05,
+      "loss": 0.3041,
+      "step": 1084
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.931780096835736e-05,
+      "loss": 0.883,
+      "step": 1085
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.931653601407012e-05,
+      "loss": 0.5688,
+      "step": 1086
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93152699016232e-05,
+      "loss": 0.4083,
+      "step": 1087
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.931400263106661e-05,
+      "loss": 1.8275,
+      "step": 1088
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.931273420245042e-05,
+      "loss": 0.7335,
+      "step": 1089
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.931146461582474e-05,
+      "loss": 0.6594,
+      "step": 1090
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.931019387123973e-05,
+      "loss": 1.5216,
+      "step": 1091
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.930892196874557e-05,
+      "loss": 0.6286,
+      "step": 1092
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9307648908392535e-05,
+      "loss": 0.4882,
+      "step": 1093
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.930637469023091e-05,
+      "loss": 1.0298,
+      "step": 1094
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.930509931431102e-05,
+      "loss": 0.5089,
+      "step": 1095
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9303822780683264e-05,
+      "loss": 0.592,
+      "step": 1096
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.930254508939806e-05,
+      "loss": 0.8365,
+      "step": 1097
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.930126624050589e-05,
+      "loss": 0.4642,
+      "step": 1098
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929998623405728e-05,
+      "loss": 0.6549,
+      "step": 1099
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9298705070102794e-05,
+      "loss": 0.7576,
+      "step": 1100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929742274869304e-05,
+      "loss": 1.2278,
+      "step": 1101
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929613926987868e-05,
+      "loss": 0.8252,
+      "step": 1102
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929485463371041e-05,
+      "loss": 1.0208,
+      "step": 1103
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9293568840239e-05,
+      "loss": 1.1395,
+      "step": 1104
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9292281889515224e-05,
+      "loss": 0.8688,
+      "step": 1105
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929099378158994e-05,
+      "loss": 0.2268,
+      "step": 1106
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.928970451651403e-05,
+      "loss": 0.6066,
+      "step": 1107
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9288414094338414e-05,
+      "loss": 0.4053,
+      "step": 1108
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9287122515114085e-05,
+      "loss": 0.5469,
+      "step": 1109
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9285829778892074e-05,
+      "loss": 0.7251,
+      "step": 1110
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.928453588572342e-05,
+      "loss": 1.508,
+      "step": 1111
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.928324083565928e-05,
+      "loss": 1.2747,
+      "step": 1112
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.928194462875079e-05,
+      "loss": 0.4701,
+      "step": 1113
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.928064726504915e-05,
+      "loss": 0.8879,
+      "step": 1114
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9279348744605635e-05,
+      "loss": 0.2238,
+      "step": 1115
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9278049067471525e-05,
+      "loss": 0.4867,
+      "step": 1116
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9276748233698185e-05,
+      "loss": 0.8134,
+      "step": 1117
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9275446243336985e-05,
+      "loss": 0.4936,
+      "step": 1118
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.927414309643936e-05,
+      "loss": 0.8675,
+      "step": 1119
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.927283879305681e-05,
+      "loss": 0.7939,
+      "step": 1120
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.927153333324086e-05,
+      "loss": 0.8048,
+      "step": 1121
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.927022671704306e-05,
+      "loss": 0.8395,
+      "step": 1122
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.926891894451504e-05,
+      "loss": 0.9218,
+      "step": 1123
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.926761001570847e-05,
+      "loss": 0.7637,
+      "step": 1124
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.926629993067506e-05,
+      "loss": 0.6744,
+      "step": 1125
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9264988689466553e-05,
+      "loss": 0.2639,
+      "step": 1126
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.926367629213476e-05,
+      "loss": 0.208,
+      "step": 1127
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.926236273873153e-05,
+      "loss": 0.8506,
+      "step": 1128
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.926104802930875e-05,
+      "loss": 0.1442,
+      "step": 1129
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925973216391836e-05,
+      "loss": 1.0569,
+      "step": 1130
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925841514261233e-05,
+      "loss": 0.4853,
+      "step": 1131
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925709696544271e-05,
+      "loss": 0.8939,
+      "step": 1132
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925577763246158e-05,
+      "loss": 0.8361,
+      "step": 1133
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9254457143721036e-05,
+      "loss": 0.4336,
+      "step": 1134
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925313549927325e-05,
+      "loss": 1.0787,
+      "step": 1135
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925181269917045e-05,
+      "loss": 1.1907,
+      "step": 1136
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925048874346489e-05,
+      "loss": 1.3153,
+      "step": 1137
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.924916363220885e-05,
+      "loss": 0.999,
+      "step": 1138
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.924783736545471e-05,
+      "loss": 0.3904,
+      "step": 1139
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.924650994325485e-05,
+      "loss": 0.5448,
+      "step": 1140
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.92451813656617e-05,
+      "loss": 0.5804,
+      "step": 1141
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.924385163272777e-05,
+      "loss": 0.4875,
+      "step": 1142
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9242520744505574e-05,
+      "loss": 0.4363,
+      "step": 1143
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.924118870104769e-05,
+      "loss": 0.8629,
+      "step": 1144
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923985550240675e-05,
+      "loss": 0.4687,
+      "step": 1145
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923852114863541e-05,
+      "loss": 0.6639,
+      "step": 1146
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9237185639786396e-05,
+      "loss": 0.5089,
+      "step": 1147
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9235848975912464e-05,
+      "loss": 0.7795,
+      "step": 1148
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923451115706641e-05,
+      "loss": 0.2589,
+      "step": 1149
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923317218330111e-05,
+      "loss": 0.4454,
+      "step": 1150
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923183205466943e-05,
+      "loss": 0.6004,
+      "step": 1151
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923049077122433e-05,
+      "loss": 0.6238,
+      "step": 1152
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9229148333018786e-05,
+      "loss": 1.2659,
+      "step": 1153
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.922780474010585e-05,
+      "loss": 0.7571,
+      "step": 1154
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9226459992538585e-05,
+      "loss": 0.5154,
+      "step": 1155
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.922511409037012e-05,
+      "loss": 0.7589,
+      "step": 1156
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.922376703365364e-05,
+      "loss": 0.3998,
+      "step": 1157
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9222418822442334e-05,
+      "loss": 1.5131,
+      "step": 1158
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.922106945678948e-05,
+      "loss": 1.0097,
+      "step": 1159
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.921971893674839e-05,
+      "loss": 0.7705,
+      "step": 1160
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9218367262372404e-05,
+      "loss": 0.5361,
+      "step": 1161
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.921701443371492e-05,
+      "loss": 0.6961,
+      "step": 1162
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.921566045082939e-05,
+      "loss": 0.5952,
+      "step": 1163
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.92143053137693e-05,
+      "loss": 1.1405,
+      "step": 1164
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.921294902258819e-05,
+      "loss": 1.0314,
+      "step": 1165
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.921159157733963e-05,
+      "loss": 1.1863,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.921023297807725e-05,
+      "loss": 0.8352,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920887322485474e-05,
+      "loss": 0.8353,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920751231772579e-05,
+      "loss": 0.8257,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9206150256744176e-05,
+      "loss": 0.6089,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920478704196372e-05,
+      "loss": 0.4049,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9203422673438244e-05,
+      "loss": 0.526,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920205715122167e-05,
+      "loss": 1.0577,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920069047536795e-05,
+      "loss": 0.8253,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.919932264593106e-05,
+      "loss": 0.3562,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.919795366296504e-05,
+      "loss": 0.6914,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.919658352652397e-05,
+      "loss": 0.6368,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.919521223666198e-05,
+      "loss": 0.8265,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9193839793433254e-05,
+      "loss": 0.7449,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9192466196891995e-05,
+      "loss": 0.9174,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9191091447092475e-05,
+      "loss": 0.2088,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9189715544089006e-05,
+      "loss": 0.9509,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.918833848793593e-05,
+      "loss": 0.9134,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9186960278687675e-05,
+      "loss": 0.7577,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9185580916398654e-05,
+      "loss": 0.5887,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9184200401123384e-05,
+      "loss": 0.595,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9182818732916394e-05,
+      "loss": 0.5269,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.918143591183226e-05,
+      "loss": 0.5916,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9180051937925626e-05,
+      "loss": 0.5155,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.917866681125116e-05,
+      "loss": 0.9692,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.917728053186359e-05,
+      "loss": 0.5255,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9175893099817654e-05,
+      "loss": 0.4909,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.917450451516819e-05,
+      "loss": 0.8824,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9173114777970044e-05,
+      "loss": 1.03,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9171723888278116e-05,
+      "loss": 0.6719,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.917033184614737e-05,
+      "loss": 0.8666,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.916893865163278e-05,
+      "loss": 0.4383,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.916754430478939e-05,
+      "loss": 0.5486,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.916614880567228e-05,
+      "loss": 0.2647,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.91647521543366e-05,
+      "loss": 0.8339,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.7496468424797058,
+      "eval_runtime": 38.599,
+      "eval_samples_per_second": 1.036,
+      "eval_steps_per_second": 1.036,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9163354350837494e-05,
+      "loss": 0.7949,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.916195539523021e-05,
+      "loss": 0.6843,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.916055528757e-05,
+      "loss": 0.5797,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.915915402791218e-05,
+      "loss": 0.4891,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9157751616312107e-05,
+      "loss": 0.764,
+      "step": 1205
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9156348052825175e-05,
+      "loss": 1.1367,
+      "step": 1206
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.915494333750685e-05,
+      "loss": 0.8591,
+      "step": 1207
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.915353747041261e-05,
+      "loss": 0.6608,
+      "step": 1208
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9152130451598e-05,
+      "loss": 0.5273,
+      "step": 1209
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.915072228111861e-05,
+      "loss": 0.5034,
+      "step": 1210
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914931295903006e-05,
+      "loss": 0.382,
+      "step": 1211
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914790248538803e-05,
+      "loss": 0.9885,
+      "step": 1212
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914649086024825e-05,
+      "loss": 0.8375,
+      "step": 1213
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914507808366647e-05,
+      "loss": 0.4579,
+      "step": 1214
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9143664155698516e-05,
+      "loss": 1.2297,
+      "step": 1215
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914224907640024e-05,
+      "loss": 0.5103,
+      "step": 1216
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914083284582755e-05,
+      "loss": 1.0733,
+      "step": 1217
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9139415464036374e-05,
+      "loss": 0.4114,
+      "step": 1218
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.913799693108274e-05,
+      "loss": 0.3371,
+      "step": 1219
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.913657724702267e-05,
+      "loss": 0.5982,
+      "step": 1220
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.913515641191224e-05,
+      "loss": 0.9224,
+      "step": 1221
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9133734425807595e-05,
+      "loss": 0.8095,
+      "step": 1222
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.91323112887649e-05,
+      "loss": 1.1187,
+      "step": 1223
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.91308870008404e-05,
+      "loss": 0.3692,
+      "step": 1224
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.912946156209033e-05,
+      "loss": 0.2138,
+      "step": 1225
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9128034972571016e-05,
+      "loss": 0.0853,
+      "step": 1226
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9126607232338816e-05,
+      "loss": 0.3911,
+      "step": 1227
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.912517834145014e-05,
+      "loss": 0.786,
+      "step": 1228
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9123748299961433e-05,
+      "loss": 0.7656,
+      "step": 1229
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.912231710792919e-05,
+      "loss": 0.417,
+      "step": 1230
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.912088476540994e-05,
+      "loss": 0.7035,
+      "step": 1231
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.911945127246028e-05,
+      "loss": 0.9331,
+      "step": 1232
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9118016629136834e-05,
+      "loss": 0.4295,
+      "step": 1233
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9116580835496285e-05,
+      "loss": 0.308,
+      "step": 1234
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.911514389159535e-05,
+      "loss": 0.4518,
+      "step": 1235
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.91137057974908e-05,
+      "loss": 0.5122,
+      "step": 1236
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9112266553239434e-05,
+      "loss": 0.8486,
+      "step": 1237
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9110826158898124e-05,
+      "loss": 0.9193,
+      "step": 1238
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.910938461452377e-05,
+      "loss": 0.6371,
+      "step": 1239
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9107941920173316e-05,
+      "loss": 0.9696,
+      "step": 1240
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.910649807590377e-05,
+      "loss": 0.6701,
+      "step": 1241
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.910505308177215e-05,
+      "loss": 0.3971,
+      "step": 1242
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9103606937835554e-05,
+      "loss": 0.6503,
+      "step": 1243
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.910215964415112e-05,
+      "loss": 0.59,
+      "step": 1244
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9100711200776006e-05,
+      "loss": 1.1223,
+      "step": 1245
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.909926160776744e-05,
+      "loss": 0.7765,
+      "step": 1246
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.909781086518269e-05,
+      "loss": 0.5022,
+      "step": 1247
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9096358973079084e-05,
+      "loss": 0.9155,
+      "step": 1248
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.909490593151395e-05,
+      "loss": 0.2081,
+      "step": 1249
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.909345174054471e-05,
+      "loss": 0.3318,
+      "step": 1250
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.90919964002288e-05,
+      "loss": 0.7604,
+      "step": 1251
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.909053991062372e-05,
+      "loss": 0.548,
+      "step": 1252
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.908908227178702e-05,
+      "loss": 0.8028,
+      "step": 1253
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.908762348377628e-05,
+      "loss": 1.1675,
+      "step": 1254
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9086163546649113e-05,
+      "loss": 0.8905,
+      "step": 1255
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.908470246046321e-05,
+      "loss": 0.9825,
+      "step": 1256
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.908324022527628e-05,
+      "loss": 0.46,
+      "step": 1257
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.90817768411461e-05,
+      "loss": 0.7942,
+      "step": 1258
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9080312308130485e-05,
+      "loss": 1.1558,
+      "step": 1259
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.907884662628727e-05,
+      "loss": 0.8335,
+      "step": 1260
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.907737979567439e-05,
+      "loss": 0.5695,
+      "step": 1261
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.907591181634976e-05,
+      "loss": 0.6093,
+      "step": 1262
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.90744426883714e-05,
+      "loss": 0.9693,
+      "step": 1263
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.907297241179734e-05,
+      "loss": 0.6122,
+      "step": 1264
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9071500986685646e-05,
+      "loss": 0.9568,
+      "step": 1265
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9070028413094464e-05,
+      "loss": 0.9696,
+      "step": 1266
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.906855469108197e-05,
+      "loss": 0.5976,
+      "step": 1267
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9067079820706375e-05,
+      "loss": 0.5162,
+      "step": 1268
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9065603802025954e-05,
+      "loss": 1.4311,
+      "step": 1269
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.906412663509901e-05,
+      "loss": 0.7399,
+      "step": 1270
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9062648319983905e-05,
+      "loss": 0.7819,
+      "step": 1271
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.906116885673903e-05,
+      "loss": 0.5788,
+      "step": 1272
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.905968824542284e-05,
+      "loss": 0.1952,
+      "step": 1273
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.905820648609383e-05,
+      "loss": 0.2294,
+      "step": 1274
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9056723578810535e-05,
+      "loss": 0.5443,
+      "step": 1275
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.905523952363153e-05,
+      "loss": 0.9482,
+      "step": 1276
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.905375432061544e-05,
+      "loss": 0.9256,
+      "step": 1277
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.905226796982096e-05,
+      "loss": 1.1936,
+      "step": 1278
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.90507804713068e-05,
+      "loss": 0.6936,
+      "step": 1279
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9049291825131704e-05,
+      "loss": 0.8894,
+      "step": 1280
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.904780203135451e-05,
+      "loss": 0.2037,
+      "step": 1281
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.904631109003406e-05,
+      "loss": 0.5867,
+      "step": 1282
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.904481900122925e-05,
+      "loss": 0.8552,
+      "step": 1283
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.904332576499904e-05,
+      "loss": 0.4246,
+      "step": 1284
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9041831381402394e-05,
+      "loss": 1.0295,
+      "step": 1285
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9040335850498376e-05,
+      "loss": 0.774,
+      "step": 1286
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903883917234605e-05,
+      "loss": 0.55,
+      "step": 1287
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903734134700456e-05,
+      "loss": 0.8934,
+      "step": 1288
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903584237453306e-05,
+      "loss": 0.8684,
+      "step": 1289
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903434225499078e-05,
+      "loss": 0.6622,
+      "step": 1290
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903284098843697e-05,
+      "loss": 1.0629,
+      "step": 1291
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9031338574930954e-05,
+      "loss": 0.774,
+      "step": 1292
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.902983501453207e-05,
+      "loss": 0.7319,
+      "step": 1293
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9028330307299725e-05,
+      "loss": 0.3425,
+      "step": 1294
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9026824453293356e-05,
+      "loss": 0.1708,
+      "step": 1295
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.902531745257247e-05,
+      "loss": 0.4172,
+      "step": 1296
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.902380930519658e-05,
+      "loss": 0.5625,
+      "step": 1297
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.902230001122528e-05,
+      "loss": 0.4049,
+      "step": 1298
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.902078957071819e-05,
+      "loss": 0.3299,
+      "step": 1299
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9019277983734974e-05,
+      "loss": 0.0466,
+      "step": 1300
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.901776525033536e-05,
+      "loss": 0.7186,
+      "step": 1301
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.901625137057911e-05,
+      "loss": 0.582,
+      "step": 1302
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.901473634452602e-05,
+      "loss": 0.7563,
+      "step": 1303
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.901322017223594e-05,
+      "loss": 0.4953,
+      "step": 1304
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.901170285376878e-05,
+      "loss": 1.0195,
+      "step": 1305
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.901018438918446e-05,
+      "loss": 0.9867,
+      "step": 1306
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9008664778543e-05,
+      "loss": 0.5448,
+      "step": 1307
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.900714402190441e-05,
+      "loss": 0.0875,
+      "step": 1308
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.900562211932877e-05,
+      "loss": 0.9428,
+      "step": 1309
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9004099070876204e-05,
+      "loss": 0.5971,
+      "step": 1310
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9002574876606884e-05,
+      "loss": 0.551,
+      "step": 1311
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.900104953658102e-05,
+      "loss": 0.9879,
+      "step": 1312
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.899952305085888e-05,
+      "loss": 1.105,
+      "step": 1313
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.899799541950076e-05,
+      "loss": 0.8258,
+      "step": 1314
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.899646664256701e-05,
+      "loss": 0.535,
+      "step": 1315
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.899493672011802e-05,
+      "loss": 0.184,
+      "step": 1316
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.8993405652214255e-05,
+      "loss": 0.7485,
+      "step": 1317
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.899187343891617e-05,
+      "loss": 0.444,
+      "step": 1318
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.89903400802843e-05,
+      "loss": 0.6234,
+      "step": 1319
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.898880557637924e-05,
+      "loss": 0.9469,
+      "step": 1320
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.8987269927261596e-05,
+      "loss": 0.8352,
+      "step": 1321
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.8985733132992044e-05,
+      "loss": 0.8233,
+      "step": 1322
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.898419519363128e-05,
+      "loss": 0.3879,
+      "step": 1323
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.898265610924007e-05,
+      "loss": 0.3328,
+      "step": 1324
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.898111587987922e-05,
+      "loss": 0.7853,
+      "step": 1325
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8979574505609574e-05,
+      "loss": 0.7278,
+      "step": 1326
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8978031986492024e-05,
+      "loss": 0.4904,
+      "step": 1327
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8976488322587505e-05,
+      "loss": 0.9817,
+      "step": 1328
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8974943513957e-05,
+      "loss": 0.571,
+      "step": 1329
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.897339756066155e-05,
+      "loss": 0.7662,
+      "step": 1330
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.897185046276221e-05,
+      "loss": 0.2677,
+      "step": 1331
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.897030222032011e-05,
+      "loss": 1.075,
+      "step": 1332
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.896875283339641e-05,
+      "loss": 0.6429,
+      "step": 1333
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8967202302052314e-05,
+      "loss": 1.0514,
+      "step": 1334
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.896565062634909e-05,
+      "loss": 0.8698,
+      "step": 1335
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.896409780634803e-05,
+      "loss": 0.8626,
+      "step": 1336
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8962543842110474e-05,
+      "loss": 0.8138,
+      "step": 1337
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.896098873369782e-05,
+      "loss": 0.3084,
+      "step": 1338
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895943248117149e-05,
+      "loss": 0.5818,
+      "step": 1339
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8957875084592976e-05,
+      "loss": 0.4695,
+      "step": 1340
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895631654402381e-05,
+      "loss": 0.4678,
+      "step": 1341
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895475685952555e-05,
+      "loss": 0.5829,
+      "step": 1342
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8953196031159814e-05,
+      "loss": 1.3858,
+      "step": 1343
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895163405898826e-05,
+      "loss": 0.8824,
+      "step": 1344
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895007094307261e-05,
+      "loss": 0.6689,
+      "step": 1345
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8948506683474595e-05,
+      "loss": 0.6724,
+      "step": 1346
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.894694128025602e-05,
+      "loss": 0.3838,
+      "step": 1347
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.894537473347873e-05,
+      "loss": 0.5651,
+      "step": 1348
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.894380704320462e-05,
+      "loss": 0.2837,
+      "step": 1349
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.894223820949559e-05,
+      "loss": 0.6312,
+      "step": 1350
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.894066823241365e-05,
+      "loss": 0.5827,
+      "step": 1351
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.893909711202082e-05,
+      "loss": 0.4492,
+      "step": 1352
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.893752484837915e-05,
+      "loss": 0.396,
+      "step": 1353
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8935951441550766e-05,
+      "loss": 0.722,
+      "step": 1354
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8934376891597814e-05,
+      "loss": 0.5021,
+      "step": 1355
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.893280119858251e-05,
+      "loss": 0.8738,
+      "step": 1356
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.89312243625671e-05,
+      "loss": 0.5279,
+      "step": 1357
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8929646383613875e-05,
+      "loss": 0.6545,
+      "step": 1358
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.892806726178517e-05,
+      "loss": 0.4754,
+      "step": 1359
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8926486997143375e-05,
+      "loss": 0.6262,
+      "step": 1360
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.892490558975091e-05,
+      "loss": 0.9926,
+      "step": 1361
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.892332303967026e-05,
+      "loss": 0.4005,
+      "step": 1362
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.892173934696394e-05,
+      "loss": 0.7698,
+      "step": 1363
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.892015451169451e-05,
+      "loss": 0.9631,
+      "step": 1364
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.891856853392458e-05,
+      "loss": 0.5726,
+      "step": 1365
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.891698141371682e-05,
+      "loss": 0.6461,
+      "step": 1366
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.89153931511339e-05,
+      "loss": 1.0637,
+      "step": 1367
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.89138037462386e-05,
+      "loss": 1.5453,
+      "step": 1368
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8912213199093675e-05,
+      "loss": 0.7418,
+      "step": 1369
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.891062150976199e-05,
+      "loss": 0.5512,
+      "step": 1370
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.890902867830641e-05,
+      "loss": 1.1713,
+      "step": 1371
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8907434704789855e-05,
+      "loss": 0.6168,
+      "step": 1372
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.890583958927531e-05,
+      "loss": 0.5444,
+      "step": 1373
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.890424333182578e-05,
+      "loss": 0.3876,
+      "step": 1374
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.890264593250434e-05,
+      "loss": 0.4995,
+      "step": 1375
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.890104739137407e-05,
+      "loss": 0.7559,
+      "step": 1376
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.889944770849815e-05,
+      "loss": 0.831,
+      "step": 1377
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8897846883939755e-05,
+      "loss": 1.0407,
+      "step": 1378
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.889624491776213e-05,
+      "loss": 0.2451,
+      "step": 1379
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.889464181002857e-05,
+      "loss": 0.4188,
+      "step": 1380
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8893037560802396e-05,
+      "loss": 0.2398,
+      "step": 1381
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8891432170147e-05,
+      "loss": 1.2566,
+      "step": 1382
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8889825638125784e-05,
+      "loss": 0.8542,
+      "step": 1383
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.888821796480222e-05,
+      "loss": 0.6554,
+      "step": 1384
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.888660915023984e-05,
+      "loss": 0.9475,
+      "step": 1385
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.888499919450217e-05,
+      "loss": 0.4093,
+      "step": 1386
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.888338809765283e-05,
+      "loss": 1.0281,
+      "step": 1387
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.888177585975545e-05,
+      "loss": 0.6351,
+      "step": 1388
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8880162480873745e-05,
+      "loss": 0.1137,
+      "step": 1389
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.887854796107144e-05,
+      "loss": 0.6905,
+      "step": 1390
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.887693230041232e-05,
+      "loss": 0.771,
+      "step": 1391
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.887531549896022e-05,
+      "loss": 1.0703,
+      "step": 1392
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.887369755677899e-05,
+      "loss": 0.4057,
+      "step": 1393
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.887207847393257e-05,
+      "loss": 0.9359,
+      "step": 1394
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8870458250484905e-05,
+      "loss": 0.4064,
+      "step": 1395
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8868836886500013e-05,
+      "loss": 0.6552,
+      "step": 1396
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.886721438204194e-05,
+      "loss": 0.4984,
+      "step": 1397
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8865590737174794e-05,
+      "loss": 1.1186,
+      "step": 1398
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.886396595196271e-05,
+      "loss": 0.9223,
+      "step": 1399
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.886234002646988e-05,
+      "loss": 0.9304,
+      "step": 1400
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.7503051161766052,
+      "eval_runtime": 38.3562,
+      "eval_samples_per_second": 1.043,
+      "eval_steps_per_second": 1.043,
+      "step": 1400
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.886071296076053e-05,
+      "loss": 0.4758,
+      "step": 1401
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.885908475489894e-05,
+      "loss": 0.8806,
+      "step": 1402
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8857455408949445e-05,
+      "loss": 0.9602,
+      "step": 1403
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.885582492297639e-05,
+      "loss": 0.6884,
+      "step": 1404
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8854193297044206e-05,
+      "loss": 0.4663,
+      "step": 1405
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.885256053121734e-05,
+      "loss": 1.0772,
+      "step": 1406
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.885092662556031e-05,
+      "loss": 0.934,
+      "step": 1407
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8849291580137644e-05,
+      "loss": 0.9204,
+      "step": 1408
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.884765539501394e-05,
+      "loss": 1.1525,
+      "step": 1409
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8846018070253856e-05,
+      "loss": 0.6152,
+      "step": 1410
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.884437960592205e-05,
+      "loss": 0.8692,
+      "step": 1411
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.884274000208327e-05,
+      "loss": 0.8194,
+      "step": 1412
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8841099258802276e-05,
+      "loss": 0.7535,
+      "step": 1413
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.883945737614388e-05,
+      "loss": 0.84,
+      "step": 1414
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.883781435417296e-05,
+      "loss": 0.3645,
+      "step": 1415
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.883617019295442e-05,
+      "loss": 0.5824,
+      "step": 1416
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8834524892553216e-05,
+      "loss": 0.9631,
+      "step": 1417
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8832878453034335e-05,
+      "loss": 0.543,
+      "step": 1418
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.883123087446283e-05,
+      "loss": 0.3457,
+      "step": 1419
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.882958215690379e-05,
+      "loss": 0.3652,
+      "step": 1420
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.882793230042234e-05,
+      "loss": 0.2712,
+      "step": 1421
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.882628130508367e-05,
+      "loss": 0.8634,
+      "step": 1422
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.882462917095298e-05,
+      "loss": 0.3931,
+      "step": 1423
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8822975898095563e-05,
+      "loss": 0.9728,
+      "step": 1424
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8821321486576724e-05,
+      "loss": 0.8801,
+      "step": 1425
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.881966593646182e-05,
+      "loss": 0.9122,
+      "step": 1426
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8818009247816256e-05,
+      "loss": 0.6071,
+      "step": 1427
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.881635142070548e-05,
+      "loss": 1.0121,
+      "step": 1428
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8814692455194966e-05,
+      "loss": 0.8232,
+      "step": 1429
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.881303235135029e-05,
+      "loss": 0.7892,
+      "step": 1430
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8811371109237006e-05,
+      "loss": 0.9518,
+      "step": 1431
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.880970872892075e-05,
+      "loss": 0.9974,
+      "step": 1432
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8808045210467196e-05,
+      "loss": 0.3978,
+      "step": 1433
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8806380553942057e-05,
+      "loss": 1.228,
+      "step": 1434
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8804714759411104e-05,
+      "loss": 0.9577,
+      "step": 1435
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.880304782694014e-05,
+      "loss": 0.3137,
+      "step": 1436
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8801379756595014e-05,
+      "loss": 0.1755,
+      "step": 1437
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8799710548441636e-05,
+      "loss": 0.674,
+      "step": 1438
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.879804020254593e-05,
+      "loss": 0.5653,
+      "step": 1439
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.87963687189739e-05,
+      "loss": 0.9895,
+      "step": 1440
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.879469609779158e-05,
+      "loss": 1.0278,
+      "step": 1441
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.879302233906503e-05,
+      "loss": 0.9052,
+      "step": 1442
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8791347442860385e-05,
+      "loss": 0.4667,
+      "step": 1443
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878967140924381e-05,
+      "loss": 0.5531,
+      "step": 1444
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878799423828151e-05,
+      "loss": 0.9915,
+      "step": 1445
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878631593003976e-05,
+      "loss": 1.1672,
+      "step": 1446
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878463648458485e-05,
+      "loss": 0.3166,
+      "step": 1447
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878295590198313e-05,
+      "loss": 0.4339,
+      "step": 1448
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878127418230098e-05,
+      "loss": 0.2514,
+      "step": 1449
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8779591325604856e-05,
+      "loss": 0.7436,
+      "step": 1450
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.877790733196123e-05,
+      "loss": 0.7142,
+      "step": 1451
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.877622220143663e-05,
+      "loss": 0.5992,
+      "step": 1452
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.877453593409762e-05,
+      "loss": 0.8641,
+      "step": 1453
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.877284853001084e-05,
+      "loss": 0.9125,
+      "step": 1454
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.877115998924293e-05,
+      "loss": 0.4552,
+      "step": 1455
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8769470311860596e-05,
+      "loss": 0.9349,
+      "step": 1456
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.87677794979306e-05,
+      "loss": 0.8702,
+      "step": 1457
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.876608754751973e-05,
+      "loss": 0.8543,
+      "step": 1458
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.876439446069483e-05,
+      "loss": 0.4365,
+      "step": 1459
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.876270023752278e-05,
+      "loss": 0.2401,
+      "step": 1460
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.876100487807053e-05,
+      "loss": 0.8984,
+      "step": 1461
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.875930838240504e-05,
+      "loss": 0.9954,
+      "step": 1462
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8757610750593335e-05,
+      "loss": 0.6037,
+      "step": 1463
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.875591198270247e-05,
+      "loss": 0.6592,
+      "step": 1464
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8754212078799584e-05,
+      "loss": 1.0768,
+      "step": 1465
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.875251103895179e-05,
+      "loss": 1.0465,
+      "step": 1466
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.875080886322632e-05,
+      "loss": 0.5973,
+      "step": 1467
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.874910555169042e-05,
+      "loss": 0.5641,
+      "step": 1468
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.874740110441136e-05,
+      "loss": 0.8865,
+      "step": 1469
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.874569552145648e-05,
+      "loss": 0.758,
+      "step": 1470
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.874398880289318e-05,
+      "loss": 1.1136,
+      "step": 1471
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.874228094878885e-05,
+      "loss": 1.0929,
+      "step": 1472
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8740571959211e-05,
+      "loss": 0.7198,
+      "step": 1473
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8738861834227104e-05,
+      "loss": 1.478,
+      "step": 1474
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.873715057390475e-05,
+      "loss": 0.8479,
+      "step": 1475
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.873543817831153e-05,
+      "loss": 0.8811,
+      "step": 1476
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.873372464751509e-05,
+      "loss": 0.564,
+      "step": 1477
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8732009981583125e-05,
+      "loss": 0.5996,
+      "step": 1478
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.873029418058338e-05,
+      "loss": 0.8714,
+      "step": 1479
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.872857724458364e-05,
+      "loss": 0.6821,
+      "step": 1480
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8726859173651725e-05,
+      "loss": 0.7693,
+      "step": 1481
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.872513996785551e-05,
+      "loss": 0.0968,
+      "step": 1482
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.872341962726291e-05,
+      "loss": 0.5205,
+      "step": 1483
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.872169815194189e-05,
+      "loss": 0.4439,
+      "step": 1484
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.871997554196047e-05,
+      "loss": 0.5522,
+      "step": 1485
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.871825179738668e-05,
+      "loss": 0.4686,
+      "step": 1486
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.871652691828863e-05,
+      "loss": 0.9297,
+      "step": 1487
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.871480090473446e-05,
+      "loss": 0.0941,
+      "step": 1488
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.871307375679236e-05,
+      "loss": 0.6016,
+      "step": 1489
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.871134547453056e-05,
+      "loss": 1.0442,
+      "step": 1490
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8709616058017334e-05,
+      "loss": 0.694,
+      "step": 1491
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.870788550732099e-05,
+      "loss": 0.502,
+      "step": 1492
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.870615382250993e-05,
+      "loss": 0.8294,
+      "step": 1493
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8704421003652525e-05,
+      "loss": 0.857,
+      "step": 1494
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8702687050817254e-05,
+      "loss": 0.9896,
+      "step": 1495
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.870095196407262e-05,
+      "loss": 0.8676,
+      "step": 1496
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8699215743487143e-05,
+      "loss": 0.1524,
+      "step": 1497
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.869747838912945e-05,
+      "loss": 0.8866,
+      "step": 1498
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.869573990106814e-05,
+      "loss": 0.5936,
+      "step": 1499
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8694000279371914e-05,
+      "loss": 0.5552,
+      "step": 1500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.869225952410949e-05,
+      "loss": 1.365,
+      "step": 1501
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.869051763534964e-05,
+      "loss": 0.8452,
+      "step": 1502
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.868877461316117e-05,
+      "loss": 0.3458,
+      "step": 1503
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.868703045761294e-05,
+      "loss": 0.3244,
+      "step": 1504
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.868528516877387e-05,
+      "loss": 0.8587,
+      "step": 1505
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.868353874671289e-05,
+      "loss": 0.9946,
+      "step": 1506
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8681791191499e-05,
+      "loss": 0.5743,
+      "step": 1507
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.868004250320123e-05,
+      "loss": 1.0699,
+      "step": 1508
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.867829268188867e-05,
+      "loss": 1.068,
+      "step": 1509
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.867654172763046e-05,
+      "loss": 1.1351,
+      "step": 1510
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8674789640495746e-05,
+      "loss": 0.3463,
+      "step": 1511
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.867303642055376e-05,
+      "loss": 0.5759,
+      "step": 1512
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.867128206787376e-05,
+      "loss": 0.546,
+      "step": 1513
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.866952658252506e-05,
+      "loss": 0.865,
+      "step": 1514
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8667769964577e-05,
+      "loss": 0.645,
+      "step": 1515
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.866601221409897e-05,
+      "loss": 0.8338,
+      "step": 1516
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.866425333116044e-05,
+      "loss": 0.8253,
+      "step": 1517
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.866249331583086e-05,
+      "loss": 0.5439,
+      "step": 1518
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.866073216817978e-05,
+      "loss": 0.998,
+      "step": 1519
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8658969888276774e-05,
+      "loss": 0.4254,
+      "step": 1520
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8657206476191455e-05,
+      "loss": 0.518,
+      "step": 1521
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.865544193199349e-05,
+      "loss": 0.4443,
+      "step": 1522
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8653676255752584e-05,
+      "loss": 0.7454,
+      "step": 1523
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.86519094475385e-05,
+      "loss": 0.8426,
+      "step": 1524
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.865014150742103e-05,
+      "loss": 0.5469,
+      "step": 1525
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.864837243547002e-05,
+      "loss": 0.1662,
+      "step": 1526
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.864660223175536e-05,
+      "loss": 0.4678,
+      "step": 1527
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.864483089634697e-05,
+      "loss": 0.4738,
+      "step": 1528
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.864305842931484e-05,
+      "loss": 0.597,
+      "step": 1529
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.864128483072898e-05,
+      "loss": 0.5478,
+      "step": 1530
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863951010065947e-05,
+      "loss": 0.8952,
+      "step": 1531
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863773423917642e-05,
+      "loss": 0.9484,
+      "step": 1532
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863595724634997e-05,
+      "loss": 0.8669,
+      "step": 1533
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8634179122250345e-05,
+      "loss": 0.89,
+      "step": 1534
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863239986694777e-05,
+      "loss": 1.1242,
+      "step": 1535
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863061948051255e-05,
+      "loss": 0.4359,
+      "step": 1536
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8628837963015e-05,
+      "loss": 0.5395,
+      "step": 1537
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.862705531452552e-05,
+      "loss": 0.5443,
+      "step": 1538
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.862527153511452e-05,
+      "loss": 1.063,
+      "step": 1539
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.862348662485248e-05,
+      "loss": 0.9187,
+      "step": 1540
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8621700583809905e-05,
+      "loss": 0.3822,
+      "step": 1541
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8619913412057365e-05,
+      "loss": 0.8023,
+      "step": 1542
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.861812510966544e-05,
+      "loss": 0.5992,
+      "step": 1543
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.86163356767048e-05,
+      "loss": 1.1752,
+      "step": 1544
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.861454511324613e-05,
+      "loss": 1.0501,
+      "step": 1545
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.861275341936016e-05,
+      "loss": 0.3466,
+      "step": 1546
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.861096059511769e-05,
+      "loss": 1.3023,
+      "step": 1547
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8609166640589525e-05,
+      "loss": 0.4328,
+      "step": 1548
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8607371555846545e-05,
+      "loss": 0.5305,
+      "step": 1549
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.860557534095966e-05,
+      "loss": 0.9892,
+      "step": 1550
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.860377799599984e-05,
+      "loss": 0.6667,
+      "step": 1551
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.860197952103808e-05,
+      "loss": 0.8505,
+      "step": 1552
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.860017991614543e-05,
+      "loss": 0.5373,
+      "step": 1553
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8598379181393e-05,
+      "loss": 0.6126,
+      "step": 1554
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.859657731685191e-05,
+      "loss": 0.5509,
+      "step": 1555
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.859477432259335e-05,
+      "loss": 0.6943,
+      "step": 1556
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.859297019868854e-05,
+      "loss": 0.4788,
+      "step": 1557
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.859116494520877e-05,
+      "loss": 0.7418,
+      "step": 1558
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8589358562225346e-05,
+      "loss": 0.9025,
+      "step": 1559
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8587551049809624e-05,
+      "loss": 0.4907,
+      "step": 1560
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.858574240803303e-05,
+      "loss": 0.4993,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8583932636966995e-05,
+      "loss": 0.4125,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8582121736683027e-05,
+      "loss": 0.8577,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.858030970725265e-05,
+      "loss": 1.0045,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8578496548747466e-05,
+      "loss": 0.9697,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8576682261239105e-05,
+      "loss": 0.885,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.857486684479922e-05,
+      "loss": 0.8597,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.857305029949956e-05,
+      "loss": 1.1737,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.857123262541186e-05,
+      "loss": 1.1649,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.856941382260794e-05,
+      "loss": 0.5088,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.856759389115966e-05,
+      "loss": 0.7499,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8565772831138914e-05,
+      "loss": 0.428,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.856395064261763e-05,
+      "loss": 0.9811,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8562127325667804e-05,
+      "loss": 0.9255,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.856030288036147e-05,
+      "loss": 2.032,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.85584773067707e-05,
+      "loss": 0.767,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8556650604967605e-05,
+      "loss": 1.2836,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.855482277502436e-05,
+      "loss": 0.6337,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.855299381701318e-05,
+      "loss": 1.1379,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.855116373100631e-05,
+      "loss": 0.4784,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.854933251707604e-05,
+      "loss": 0.9148,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8547500175294723e-05,
+      "loss": 1.484,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.854566670573475e-05,
+      "loss": 0.5193,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.854383210846855e-05,
+      "loss": 0.7753,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8541996383568596e-05,
+      "loss": 0.8142,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.85401595311074e-05,
+      "loss": 0.8942,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.853832155115754e-05,
+      "loss": 0.4837,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.853648244379163e-05,
+      "loss": 0.9841,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.853464220908232e-05,
+      "loss": 0.8996,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.85328008471023e-05,
+      "loss": 0.1448,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.853095835792431e-05,
+      "loss": 0.2692,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8529114741621174e-05,
+      "loss": 0.2779,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8527269998265686e-05,
+      "loss": 0.7487,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.852542412793073e-05,
+      "loss": 1.012,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.852357713068925e-05,
+      "loss": 0.5556,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.852172900661419e-05,
+      "loss": 1.2211,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8519879755778555e-05,
+      "loss": 0.2863,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8518029378255435e-05,
+      "loss": 0.4193,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8516177874117894e-05,
+      "loss": 0.64,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.851432524343909e-05,
+      "loss": 0.5314,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.7522938251495361,
+      "eval_runtime": 37.8093,
+      "eval_samples_per_second": 1.058,
+      "eval_steps_per_second": 1.058,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.851247148629221e-05,
+      "loss": 0.9139,
+      "step": 1601
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8510616602750494e-05,
+      "loss": 0.1425,
+      "step": 1602
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.850876059288721e-05,
+      "loss": 0.2705,
+      "step": 1603
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.850690345677569e-05,
+      "loss": 0.535,
+      "step": 1604
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8505045194489295e-05,
+      "loss": 0.6015,
+      "step": 1605
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.850318580610144e-05,
+      "loss": 0.3521,
+      "step": 1606
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.850132529168558e-05,
+      "loss": 0.8435,
+      "step": 1607
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.84994636513152e-05,
+      "loss": 0.4457,
+      "step": 1608
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8497600885063866e-05,
+      "loss": 0.7521,
+      "step": 1609
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.849573699300516e-05,
+      "loss": 0.5643,
+      "step": 1610
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.849387197521271e-05,
+      "loss": 0.5936,
+      "step": 1611
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.84920058317602e-05,
+      "loss": 0.3199,
+      "step": 1612
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.849013856272136e-05,
+      "loss": 1.0658,
+      "step": 1613
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.848827016816994e-05,
+      "loss": 0.6742,
+      "step": 1614
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.848640064817976e-05,
+      "loss": 0.4235,
+      "step": 1615
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.848453000282468e-05,
+      "loss": 1.0451,
+      "step": 1616
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.848265823217859e-05,
+      "loss": 0.624,
+      "step": 1617
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8480785336315445e-05,
+      "loss": 0.9888,
+      "step": 1618
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.847891131530924e-05,
+      "loss": 0.6146,
+      "step": 1619
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8477036169233986e-05,
+      "loss": 0.377,
+      "step": 1620
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.847515989816378e-05,
+      "loss": 1.34,
+      "step": 1621
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8473282502172745e-05,
+      "loss": 0.6745,
+      "step": 1622
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.847140398133504e-05,
+      "loss": 0.7822,
+      "step": 1623
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8469524335724875e-05,
+      "loss": 0.5832,
+      "step": 1624
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8467643565416515e-05,
+      "loss": 1.0789,
+      "step": 1625
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.846576167048426e-05,
+      "loss": 0.8124,
+      "step": 1626
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.846387865100244e-05,
+      "loss": 0.7029,
+      "step": 1627
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.846199450704547e-05,
+      "loss": 0.7635,
+      "step": 1628
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.846010923868776e-05,
+      "loss": 0.7188,
+      "step": 1629
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.84582228460038e-05,
+      "loss": 1.0995,
+      "step": 1630
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8456335329068106e-05,
+      "loss": 0.4828,
+      "step": 1631
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.845444668795525e-05,
+      "loss": 0.2784,
+      "step": 1632
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.845255692273985e-05,
+      "loss": 0.9238,
+      "step": 1633
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.845066603349654e-05,
+      "loss": 0.1157,
+      "step": 1634
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.844877402030004e-05,
+      "loss": 0.3721,
+      "step": 1635
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.844688088322509e-05,
+      "loss": 0.8947,
+      "step": 1636
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8444986622346476e-05,
+      "loss": 0.0489,
+      "step": 1637
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.844309123773904e-05,
+      "loss": 0.5853,
+      "step": 1638
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8441194729477645e-05,
+      "loss": 0.6994,
+      "step": 1639
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.843929709763722e-05,
+      "loss": 0.8049,
+      "step": 1640
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.843739834229275e-05,
+      "loss": 0.7978,
+      "step": 1641
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.843549846351921e-05,
+      "loss": 1.0982,
+      "step": 1642
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.843359746139168e-05,
+      "loss": 0.998,
+      "step": 1643
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.843169533598526e-05,
+      "loss": 1.0879,
+      "step": 1644
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8429792087375086e-05,
+      "loss": 0.6441,
+      "step": 1645
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842788771563634e-05,
+      "loss": 0.8857,
+      "step": 1646
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842598222084427e-05,
+      "loss": 0.7556,
+      "step": 1647
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842407560307414e-05,
+      "loss": 0.886,
+      "step": 1648
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842216786240129e-05,
+      "loss": 0.9107,
+      "step": 1649
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842025899890106e-05,
+      "loss": 0.9593,
+      "step": 1650
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.841834901264889e-05,
+      "loss": 0.6385,
+      "step": 1651
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.841643790372021e-05,
+      "loss": 0.7848,
+      "step": 1652
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8414525672190523e-05,
+      "loss": 1.0194,
+      "step": 1653
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8412612318135394e-05,
+      "loss": 1.1359,
+      "step": 1654
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.841069784163038e-05,
+      "loss": 0.5437,
+      "step": 1655
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.840878224275114e-05,
+      "loss": 0.4581,
+      "step": 1656
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.840686552157332e-05,
+      "loss": 0.4615,
+      "step": 1657
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8404947678172676e-05,
+      "loss": 0.3436,
+      "step": 1658
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.840302871262494e-05,
+      "loss": 1.1757,
+      "step": 1659
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.840110862500595e-05,
+      "loss": 0.7846,
+      "step": 1660
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8399187415391536e-05,
+      "loss": 0.7608,
+      "step": 1661
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.839726508385761e-05,
+      "loss": 0.5541,
+      "step": 1662
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.839534163048011e-05,
+      "loss": 0.2802,
+      "step": 1663
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8393417055335025e-05,
+      "loss": 0.4383,
+      "step": 1664
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.839149135849838e-05,
+      "loss": 0.4731,
+      "step": 1665
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.838956454004626e-05,
+      "loss": 0.6038,
+      "step": 1666
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8387636600054784e-05,
+      "loss": 1.082,
+      "step": 1667
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.838570753860011e-05,
+      "loss": 0.4258,
+      "step": 1668
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8383777355758444e-05,
+      "loss": 0.9024,
+      "step": 1669
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.838184605160604e-05,
+      "loss": 0.9716,
+      "step": 1670
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.83799136262192e-05,
+      "loss": 0.8294,
+      "step": 1671
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.837798007967426e-05,
+      "loss": 0.8098,
+      "step": 1672
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.837604541204761e-05,
+      "loss": 0.7914,
+      "step": 1673
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.837410962341567e-05,
+      "loss": 0.5247,
+      "step": 1674
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8372172713854926e-05,
+      "loss": 0.7249,
+      "step": 1675
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.837023468344189e-05,
+      "loss": 0.3302,
+      "step": 1676
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.836829553225313e-05,
+      "loss": 0.5641,
+      "step": 1677
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.836635526036525e-05,
+      "loss": 1.0647,
+      "step": 1678
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8364413867854896e-05,
+      "loss": 1.1064,
+      "step": 1679
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.836247135479877e-05,
+      "loss": 0.7924,
+      "step": 1680
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8360527721273605e-05,
+      "loss": 0.5391,
+      "step": 1681
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8358582967356184e-05,
+      "loss": 0.7337,
+      "step": 1682
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8356637093123355e-05,
+      "loss": 0.8415,
+      "step": 1683
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.835469009865197e-05,
+      "loss": 0.5393,
+      "step": 1684
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8352741984018944e-05,
+      "loss": 0.3186,
+      "step": 1685
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8350792749301244e-05,
+      "loss": 1.5736,
+      "step": 1686
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8348842394575886e-05,
+      "loss": 0.9144,
+      "step": 1687
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.83468909199199e-05,
+      "loss": 1.0457,
+      "step": 1688
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8344938325410395e-05,
+      "loss": 0.6201,
+      "step": 1689
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.83429846111245e-05,
+      "loss": 0.9091,
+      "step": 1690
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.83410297771394e-05,
+      "loss": 1.0061,
+      "step": 1691
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8339073823532335e-05,
+      "loss": 0.6002,
+      "step": 1692
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.833711675038055e-05,
+      "loss": 0.5167,
+      "step": 1693
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.833515855776137e-05,
+      "loss": 0.8247,
+      "step": 1694
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.833319924575216e-05,
+      "loss": 0.7894,
+      "step": 1695
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8331238814430316e-05,
+      "loss": 0.3392,
+      "step": 1696
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8329277263873295e-05,
+      "loss": 0.1306,
+      "step": 1697
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8327314594158584e-05,
+      "loss": 0.7472,
+      "step": 1698
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.832535080536371e-05,
+      "loss": 0.245,
+      "step": 1699
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8323385897566266e-05,
+      "loss": 0.7792,
+      "step": 1700
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8321419870843876e-05,
+      "loss": 0.8057,
+      "step": 1701
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8319452725274195e-05,
+      "loss": 0.5514,
+      "step": 1702
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.831748446093494e-05,
+      "loss": 0.5866,
+      "step": 1703
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8315515077903885e-05,
+      "loss": 0.7583,
+      "step": 1704
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.831354457625881e-05,
+      "loss": 0.5951,
+      "step": 1705
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.831157295607758e-05,
+      "loss": 0.5789,
+      "step": 1706
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8309600217438055e-05,
+      "loss": 0.6454,
+      "step": 1707
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.83076263604182e-05,
+      "loss": 0.6393,
+      "step": 1708
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.830565138509598e-05,
+      "loss": 0.4661,
+      "step": 1709
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8303675291549405e-05,
+      "loss": 0.5302,
+      "step": 1710
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.830169807985657e-05,
+      "loss": 0.6863,
+      "step": 1711
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.829971975009556e-05,
+      "loss": 0.4464,
+      "step": 1712
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.829774030234454e-05,
+      "loss": 0.8027,
+      "step": 1713
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.829575973668171e-05,
+      "loss": 1.2325,
+      "step": 1714
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.82937780531853e-05,
+      "loss": 0.6929,
+      "step": 1715
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8291795251933615e-05,
+      "loss": 1.1131,
+      "step": 1716
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8289811333004975e-05,
+      "loss": 0.6946,
+      "step": 1717
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.828782629647776e-05,
+      "loss": 0.4321,
+      "step": 1718
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.828584014243039e-05,
+      "loss": 0.5451,
+      "step": 1719
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.828385287094132e-05,
+      "loss": 0.6217,
+      "step": 1720
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.828186448208907e-05,
+      "loss": 0.9565,
+      "step": 1721
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8279874975952195e-05,
+      "loss": 0.795,
+      "step": 1722
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8277884352609275e-05,
+      "loss": 0.5664,
+      "step": 1723
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8275892612138954e-05,
+      "loss": 0.4356,
+      "step": 1724
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8273899754619927e-05,
+      "loss": 0.7441,
+      "step": 1725
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.827190578013091e-05,
+      "loss": 0.668,
+      "step": 1726
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.826991068875069e-05,
+      "loss": 0.8504,
+      "step": 1727
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.826791448055807e-05,
+      "loss": 0.4182,
+      "step": 1728
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.826591715563191e-05,
+      "loss": 0.7289,
+      "step": 1729
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8263918714051125e-05,
+      "loss": 0.5726,
+      "step": 1730
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.826191915589466e-05,
+      "loss": 0.6866,
+      "step": 1731
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8259918481241515e-05,
+      "loss": 0.7236,
+      "step": 1732
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.825791669017072e-05,
+      "loss": 0.5972,
+      "step": 1733
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.825591378276135e-05,
+      "loss": 0.3553,
+      "step": 1734
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.825390975909255e-05,
+      "loss": 0.6118,
+      "step": 1735
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.825190461924347e-05,
+      "loss": 0.6552,
+      "step": 1736
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.824989836329333e-05,
+      "loss": 1.2788,
+      "step": 1737
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.824789099132139e-05,
+      "loss": 0.939,
+      "step": 1738
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8245882503406956e-05,
+      "loss": 1.2313,
+      "step": 1739
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.824387289962936e-05,
+      "loss": 0.7734,
+      "step": 1740
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.824186218006801e-05,
+      "loss": 0.8169,
+      "step": 1741
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.823985034480232e-05,
+      "loss": 0.665,
+      "step": 1742
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8237837393911786e-05,
+      "loss": 0.8071,
+      "step": 1743
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8235823327475926e-05,
+      "loss": 0.6053,
+      "step": 1744
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.82338081455743e-05,
+      "loss": 0.7949,
+      "step": 1745
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.823179184828653e-05,
+      "loss": 1.0375,
+      "step": 1746
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.822977443569226e-05,
+      "loss": 0.0429,
+      "step": 1747
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.822775590787118e-05,
+      "loss": 0.3411,
+      "step": 1748
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8225736264903055e-05,
+      "loss": 0.3489,
+      "step": 1749
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.822371550686766e-05,
+      "loss": 0.9987,
+      "step": 1750
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8221693633844823e-05,
+      "loss": 0.7195,
+      "step": 1751
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8219670645914426e-05,
+      "loss": 0.6597,
+      "step": 1752
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.821764654315638e-05,
+      "loss": 0.5846,
+      "step": 1753
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.821562132565066e-05,
+      "loss": 0.7987,
+      "step": 1754
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8213594993477264e-05,
+      "loss": 0.3888,
+      "step": 1755
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.821156754671624e-05,
+      "loss": 0.7814,
+      "step": 1756
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.820953898544769e-05,
+      "loss": 0.124,
+      "step": 1757
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.820750930975174e-05,
+      "loss": 1.1248,
+      "step": 1758
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8205478519708595e-05,
+      "loss": 0.8827,
+      "step": 1759
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.820344661539847e-05,
+      "loss": 0.5398,
+      "step": 1760
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.820141359690163e-05,
+      "loss": 0.9496,
+      "step": 1761
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.819937946429839e-05,
+      "loss": 0.7363,
+      "step": 1762
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.819734421766913e-05,
+      "loss": 0.702,
+      "step": 1763
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.819530785709422e-05,
+      "loss": 0.7408,
+      "step": 1764
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.819327038265415e-05,
+      "loss": 0.8225,
+      "step": 1765
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.819123179442937e-05,
+      "loss": 0.4366,
+      "step": 1766
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8189192092500436e-05,
+      "loss": 0.401,
+      "step": 1767
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.818715127694792e-05,
+      "loss": 0.4997,
+      "step": 1768
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.818510934785245e-05,
+      "loss": 1.3866,
+      "step": 1769
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.818306630529469e-05,
+      "loss": 0.9869,
+      "step": 1770
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.818102214935536e-05,
+      "loss": 0.9726,
+      "step": 1771
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8178976880115195e-05,
+      "loss": 0.8,
+      "step": 1772
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.817693049765501e-05,
+      "loss": 0.9229,
+      "step": 1773
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.817488300205565e-05,
+      "loss": 0.1507,
+      "step": 1774
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8172834393397994e-05,
+      "loss": 1.1575,
+      "step": 1775
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.817078467176297e-05,
+      "loss": 0.9995,
+      "step": 1776
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.816873383723156e-05,
+      "loss": 1.0564,
+      "step": 1777
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8166681889884786e-05,
+      "loss": 0.8103,
+      "step": 1778
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.81646288298037e-05,
+      "loss": 0.4355,
+      "step": 1779
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8162574657069424e-05,
+      "loss": 1.2688,
+      "step": 1780
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.81605193717631e-05,
+      "loss": 1.0622,
+      "step": 1781
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.81584629739659e-05,
+      "loss": 0.8895,
+      "step": 1782
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.815640546375911e-05,
+      "loss": 0.587,
+      "step": 1783
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8154346841223974e-05,
+      "loss": 0.7022,
+      "step": 1784
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.815228710644184e-05,
+      "loss": 0.7234,
+      "step": 1785
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.815022625949407e-05,
+      "loss": 0.544,
+      "step": 1786
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8148164300462077e-05,
+      "loss": 1.1852,
+      "step": 1787
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8146101229427315e-05,
+      "loss": 1.0103,
+      "step": 1788
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8144037046471296e-05,
+      "loss": 0.6101,
+      "step": 1789
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.814197175167557e-05,
+      "loss": 0.9482,
+      "step": 1790
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.813990534512173e-05,
+      "loss": 0.8899,
+      "step": 1791
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.813783782689138e-05,
+      "loss": 0.8409,
+      "step": 1792
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.813576919706623e-05,
+      "loss": 0.7573,
+      "step": 1793
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.813369945572798e-05,
+      "loss": 0.6502,
+      "step": 1794
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.813162860295841e-05,
+      "loss": 0.5777,
+      "step": 1795
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.812955663883933e-05,
+      "loss": 0.6417,
+      "step": 1796
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.812748356345259e-05,
+      "loss": 0.5904,
+      "step": 1797
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.812540937688009e-05,
+      "loss": 0.5773,
+      "step": 1798
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.812333407920375e-05,
+      "loss": 0.9753,
+      "step": 1799
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.81212576705056e-05,
+      "loss": 0.7599,
+      "step": 1800
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.7438657283782959,
+      "eval_runtime": 38.7936,
+      "eval_samples_per_second": 1.031,
+      "eval_steps_per_second": 1.031,
+      "step": 1800
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.811918015086762e-05,
+      "loss": 0.6057,
+      "step": 1801
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.811710152037192e-05,
+      "loss": 0.7025,
+      "step": 1802
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8115021779100606e-05,
+      "loss": 0.5629,
+      "step": 1803
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.811294092713583e-05,
+      "loss": 0.3163,
+      "step": 1804
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.811085896455981e-05,
+      "loss": 0.232,
+      "step": 1805
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.810877589145478e-05,
+      "loss": 0.6216,
+      "step": 1806
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.810669170790305e-05,
+      "loss": 0.7735,
+      "step": 1807
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.810460641398693e-05,
+      "loss": 0.4612,
+      "step": 1808
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8102520009788824e-05,
+      "loss": 0.744,
+      "step": 1809
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8100432495391154e-05,
+      "loss": 0.7511,
+      "step": 1810
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.809834387087637e-05,
+      "loss": 0.3821,
+      "step": 1811
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.809625413632701e-05,
+      "loss": 0.9962,
+      "step": 1812
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.809416329182561e-05,
+      "loss": 0.296,
+      "step": 1813
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.809207133745478e-05,
+      "loss": 0.8696,
+      "step": 1814
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.808997827329715e-05,
+      "loss": 0.442,
+      "step": 1815
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8087884099435427e-05,
+      "loss": 0.7275,
+      "step": 1816
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8085788815952326e-05,
+      "loss": 0.8694,
+      "step": 1817
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8083692422930624e-05,
+      "loss": 0.7416,
+      "step": 1818
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.808159492045315e-05,
+      "loss": 0.6889,
+      "step": 1819
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.807949630860276e-05,
+      "loss": 0.7356,
+      "step": 1820
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.807739658746236e-05,
+      "loss": 0.8245,
+      "step": 1821
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.80752957571149e-05,
+      "loss": 0.6664,
+      "step": 1822
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.807319381764337e-05,
+      "loss": 0.6907,
+      "step": 1823
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8071090769130805e-05,
+      "loss": 0.8252,
+      "step": 1824
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8068986611660304e-05,
+      "loss": 0.9356,
+      "step": 1825
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.806688134531498e-05,
+      "loss": 0.769,
+      "step": 1826
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.806477497017801e-05,
+      "loss": 0.6103,
+      "step": 1827
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.806266748633259e-05,
+      "loss": 0.4971,
+      "step": 1828
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.806055889386199e-05,
+      "loss": 1.065,
+      "step": 1829
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.805844919284952e-05,
+      "loss": 0.6006,
+      "step": 1830
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.805633838337851e-05,
+      "loss": 1.0196,
+      "step": 1831
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8054226465532345e-05,
+      "loss": 0.4579,
+      "step": 1832
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.805211343939446e-05,
+      "loss": 0.6711,
+      "step": 1833
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8049999305048344e-05,
+      "loss": 0.4856,
+      "step": 1834
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.80478840625775e-05,
+      "loss": 0.5098,
+      "step": 1835
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.80457677120655e-05,
+      "loss": 0.9515,
+      "step": 1836
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.804365025359595e-05,
+      "loss": 0.627,
+      "step": 1837
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.804153168725251e-05,
+      "loss": 0.8921,
+      "step": 1838
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.803941201311886e-05,
+      "loss": 0.8674,
+      "step": 1839
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.803729123127874e-05,
+      "loss": 0.8627,
+      "step": 1840
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8035169341815936e-05,
+      "loss": 0.4042,
+      "step": 1841
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.803304634481427e-05,
+      "loss": 0.6473,
+      "step": 1842
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.803092224035762e-05,
+      "loss": 0.8394,
+      "step": 1843
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.80287970285299e-05,
+      "loss": 0.3958,
+      "step": 1844
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.802667070941505e-05,
+      "loss": 0.7597,
+      "step": 1845
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.802454328309709e-05,
+      "loss": 1.5241,
+      "step": 1846
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.802241474966006e-05,
+      "loss": 0.933,
+      "step": 1847
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8020285109188046e-05,
+      "loss": 0.9731,
+      "step": 1848
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.801815436176518e-05,
+      "loss": 0.9398,
+      "step": 1849
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8016022507475635e-05,
+      "loss": 0.3452,
+      "step": 1850
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8013889546403634e-05,
+      "loss": 0.6816,
+      "step": 1851
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.801175547863344e-05,
+      "loss": 1.3444,
+      "step": 1852
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8009620304249365e-05,
+      "loss": 0.8643,
+      "step": 1853
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.800748402333575e-05,
+      "loss": 0.7694,
+      "step": 1854
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.800534663597701e-05,
+      "loss": 0.3208,
+      "step": 1855
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8003208142257546e-05,
+      "loss": 0.6183,
+      "step": 1856
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.800106854226187e-05,
+      "loss": 0.3698,
+      "step": 1857
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.79989278360745e-05,
+      "loss": 0.8582,
+      "step": 1858
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7996786023780005e-05,
+      "loss": 0.9578,
+      "step": 1859
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7994643105463e-05,
+      "loss": 0.4692,
+      "step": 1860
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.799249908120814e-05,
+      "loss": 0.9031,
+      "step": 1861
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.799035395110012e-05,
+      "loss": 0.8695,
+      "step": 1862
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7988207715223684e-05,
+      "loss": 0.3386,
+      "step": 1863
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.798606037366364e-05,
+      "loss": 0.8424,
+      "step": 1864
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.798391192650479e-05,
+      "loss": 0.2112,
+      "step": 1865
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7981762373832025e-05,
+      "loss": 0.9805,
+      "step": 1866
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.797961171573028e-05,
+      "loss": 0.9841,
+      "step": 1867
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7977459952284475e-05,
+      "loss": 0.2167,
+      "step": 1868
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7975307083579654e-05,
+      "loss": 0.6198,
+      "step": 1869
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.797315310970086e-05,
+      "loss": 0.5006,
+      "step": 1870
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.797099803073316e-05,
+      "loss": 1.0905,
+      "step": 1871
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7968841846761726e-05,
+      "loss": 0.37,
+      "step": 1872
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.796668455787172e-05,
+      "loss": 0.9461,
+      "step": 1873
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.796452616414837e-05,
+      "loss": 0.4529,
+      "step": 1874
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.796236666567694e-05,
+      "loss": 0.3319,
+      "step": 1875
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7960206062542755e-05,
+      "loss": 0.4971,
+      "step": 1876
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.795804435483115e-05,
+      "loss": 0.7524,
+      "step": 1877
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7955881542627554e-05,
+      "loss": 0.1503,
+      "step": 1878
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.795371762601738e-05,
+      "loss": 0.271,
+      "step": 1879
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.795155260508612e-05,
+      "loss": 0.9718,
+      "step": 1880
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.794938647991931e-05,
+      "loss": 0.4332,
+      "step": 1881
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7947219250602536e-05,
+      "loss": 0.3571,
+      "step": 1882
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7945050917221396e-05,
+      "loss": 0.8536,
+      "step": 1883
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.794288147986155e-05,
+      "loss": 0.2372,
+      "step": 1884
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.794071093860872e-05,
+      "loss": 0.9851,
+      "step": 1885
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7938539293548634e-05,
+      "loss": 0.7598,
+      "step": 1886
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.793636654476709e-05,
+      "loss": 0.3359,
+      "step": 1887
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7934192692349935e-05,
+      "loss": 0.0223,
+      "step": 1888
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7932017736383035e-05,
+      "loss": 0.5036,
+      "step": 1889
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.792984167695231e-05,
+      "loss": 0.1223,
+      "step": 1890
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.792766451414373e-05,
+      "loss": 0.6588,
+      "step": 1891
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.792548624804332e-05,
+      "loss": 0.2395,
+      "step": 1892
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.792330687873711e-05,
+      "loss": 0.5701,
+      "step": 1893
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.792112640631121e-05,
+      "loss": 0.8618,
+      "step": 1894
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.791894483085176e-05,
+      "loss": 0.7937,
+      "step": 1895
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7916762152444934e-05,
+      "loss": 0.7499,
+      "step": 1896
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.791457837117696e-05,
+      "loss": 1.0606,
+      "step": 1897
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.791239348713413e-05,
+      "loss": 1.009,
+      "step": 1898
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.791020750040273e-05,
+      "loss": 0.2752,
+      "step": 1899
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.790802041106914e-05,
+      "loss": 0.5265,
+      "step": 1900
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.790583221921975e-05,
+      "loss": 0.9469,
+      "step": 1901
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7903642924941005e-05,
+      "loss": 0.856,
+      "step": 1902
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.79014525283194e-05,
+      "loss": 1.4056,
+      "step": 1903
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.789926102944147e-05,
+      "loss": 0.6447,
+      "step": 1904
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.789706842839378e-05,
+      "loss": 0.4479,
+      "step": 1905
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7894874725262965e-05,
+      "loss": 0.5505,
+      "step": 1906
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.789267992013567e-05,
+      "loss": 0.4903,
+      "step": 1907
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7890484013098606e-05,
+      "loss": 0.4528,
+      "step": 1908
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.788828700423853e-05,
+      "loss": 0.6,
+      "step": 1909
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.788608889364224e-05,
+      "loss": 0.8995,
+      "step": 1910
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7883889681396555e-05,
+      "loss": 0.7742,
+      "step": 1911
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.788168936758837e-05,
+      "loss": 0.7176,
+      "step": 1912
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7879487952304606e-05,
+      "loss": 0.6782,
+      "step": 1913
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7877285435632224e-05,
+      "loss": 0.8635,
+      "step": 1914
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.787508181765824e-05,
+      "loss": 1.0858,
+      "step": 1915
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.787287709846972e-05,
+      "loss": 1.0408,
+      "step": 1916
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.787067127815374e-05,
+      "loss": 0.7119,
+      "step": 1917
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.786846435679746e-05,
+      "loss": 0.7639,
+      "step": 1918
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.786625633448806e-05,
+      "loss": 0.9302,
+      "step": 1919
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.786404721131277e-05,
+      "loss": 0.5808,
+      "step": 1920
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.786183698735885e-05,
+      "loss": 1.3262,
+      "step": 1921
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.785962566271364e-05,
+      "loss": 0.7025,
+      "step": 1922
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.785741323746446e-05,
+      "loss": 1.2012,
+      "step": 1923
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.785519971169876e-05,
+      "loss": 1.0733,
+      "step": 1924
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.785298508550395e-05,
+      "loss": 0.5441,
+      "step": 1925
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.785076935896755e-05,
+      "loss": 0.4849,
+      "step": 1926
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.784855253217705e-05,
+      "loss": 0.7499,
+      "step": 1927
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.784633460522007e-05,
+      "loss": 0.3551,
+      "step": 1928
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.784411557818421e-05,
+      "loss": 0.8919,
+      "step": 1929
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.784189545115712e-05,
+      "loss": 0.6548,
+      "step": 1930
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.7839674224226545e-05,
+      "loss": 0.9593,
+      "step": 1931
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.78374518974802e-05,
+      "loss": 0.5882,
+      "step": 1932
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.783522847100589e-05,
+      "loss": 0.9708,
+      "step": 1933
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.783300394489146e-05,
+      "loss": 0.8031,
+      "step": 1934
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.783077831922478e-05,
+      "loss": 0.9799,
+      "step": 1935
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.7828551594093774e-05,
+      "loss": 1.4469,
+      "step": 1936
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.782632376958642e-05,
+      "loss": 0.7104,
+      "step": 1937
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.782409484579072e-05,
+      "loss": 0.2519,
+      "step": 1938
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.782186482279473e-05,
+      "loss": 0.5276,
+      "step": 1939
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.781963370068656e-05,
+      "loss": 0.3612,
+      "step": 1940
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.781740147955433e-05,
+      "loss": 0.5741,
+      "step": 1941
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.7815168159486235e-05,
+      "loss": 1.1865,
+      "step": 1942
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.781293374057051e-05,
+      "loss": 0.9762,
+      "step": 1943
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.781069822289541e-05,
+      "loss": 0.5092,
+      "step": 1944
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.780846160654927e-05,
+      "loss": 0.5546,
+      "step": 1945
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.780622389162043e-05,
+      "loss": 0.9617,
+      "step": 1946
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.7803985078197306e-05,
+      "loss": 0.5079,
+      "step": 1947
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.7801745166368325e-05,
+      "loss": 0.6805,
+      "step": 1948
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.779950415622199e-05,
+      "loss": 1.2057,
+      "step": 1949
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.779726204784683e-05,
+      "loss": 0.0116,
+      "step": 1950
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.779501884133143e-05,
+      "loss": 0.6094,
+      "step": 1951
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.779277453676438e-05,
+      "loss": 0.8318,
+      "step": 1952
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.779052913423438e-05,
+      "loss": 0.7887,
+      "step": 1953
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.778828263383011e-05,
+      "loss": 0.4857,
+      "step": 1954
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.778603503564032e-05,
+      "loss": 0.7917,
+      "step": 1955
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.778378633975381e-05,
+      "loss": 0.7449,
+      "step": 1956
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.778153654625941e-05,
+      "loss": 0.6301,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7779285655246e-05,
+      "loss": 0.6332,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.777703366680251e-05,
+      "loss": 0.283,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.77747805810179e-05,
+      "loss": 0.2687,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.777252639798117e-05,
+      "loss": 0.4423,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.777027111778139e-05,
+      "loss": 0.6347,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7768014740507636e-05,
+      "loss": 0.9838,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7765757266249056e-05,
+      "loss": 0.3895,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.776349869509484e-05,
+      "loss": 0.6065,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7761239027134204e-05,
+      "loss": 0.6749,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7758978262456425e-05,
+      "loss": 0.506,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.77567164011508e-05,
+      "loss": 0.5802,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7754453443306696e-05,
+      "loss": 0.8813,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.775218938901352e-05,
+      "loss": 1.0613,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.774992423836069e-05,
+      "loss": 0.7867,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.774765799143771e-05,
+      "loss": 0.9291,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.774539064833411e-05,
+      "loss": 0.8181,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.774312220913947e-05,
+      "loss": 0.5096,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.774085267394337e-05,
+      "loss": 1.0679,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.77385820428355e-05,
+      "loss": 0.6232,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.773631031590556e-05,
+      "loss": 1.0015,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.773403749324328e-05,
+      "loss": 0.5655,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7731763574938454e-05,
+      "loss": 0.7284,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.772948856108093e-05,
+      "loss": 0.8676,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.772721245176056e-05,
+      "loss": 1.106,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7724935247067275e-05,
+      "loss": 0.4336,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.772265694709104e-05,
+      "loss": 0.9631,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7720377551921845e-05,
+      "loss": 0.544,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.771809706164976e-05,
+      "loss": 0.8311,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.771581547636485e-05,
+      "loss": 1.2675,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.771353279615728e-05,
+      "loss": 0.5356,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.77112490211172e-05,
+      "loss": 0.776,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.770896415133485e-05,
+      "loss": 0.6384,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.770667818690049e-05,
+      "loss": 1.2002,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7704391127904426e-05,
+      "loss": 0.9704,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7702102974437006e-05,
+      "loss": 0.3663,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7699813726588625e-05,
+      "loss": 0.8273,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7697523384449726e-05,
+      "loss": 0.6121,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.769523194811079e-05,
+      "loss": 0.3925,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.769293941766233e-05,
+      "loss": 0.939,
+      "step": 1996
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.769064579319493e-05,
+      "loss": 0.0757,
+      "step": 1997
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.768835107479918e-05,
+      "loss": 0.409,
+      "step": 1998
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.768605526256575e-05,
+      "loss": 0.3258,
+      "step": 1999
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7683758356585336e-05,
+      "loss": 1.0481,
+      "step": 2000
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.7439136505126953,
+      "eval_runtime": 41.0344,
+      "eval_samples_per_second": 0.975,
+      "eval_steps_per_second": 0.975,
+      "step": 2000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.768146035694867e-05,
+      "loss": 0.5517,
+      "step": 2001
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.767916126374654e-05,
+      "loss": 0.3817,
+      "step": 2002
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.767686107706977e-05,
+      "loss": 0.8123,
+      "step": 2003
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7674559797009234e-05,
+      "loss": 1.0174,
+      "step": 2004
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.767225742365584e-05,
+      "loss": 0.557,
+      "step": 2005
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.766995395710055e-05,
+      "loss": 0.4383,
+      "step": 2006
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7667649397434355e-05,
+      "loss": 0.8115,
+      "step": 2007
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.76653437447483e-05,
+      "loss": 0.6093,
+      "step": 2008
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.766303699913347e-05,
+      "loss": 0.646,
+      "step": 2009
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7660729160681e-05,
+      "loss": 0.8168,
+      "step": 2010
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.765842022948206e-05,
+      "loss": 0.6851,
+      "step": 2011
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.765611020562786e-05,
+      "loss": 1.0448,
+      "step": 2012
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.765379908920966e-05,
+      "loss": 0.4805,
+      "step": 2013
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.765148688031877e-05,
+      "loss": 0.1915,
+      "step": 2014
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7649173579046525e-05,
+      "loss": 0.9332,
+      "step": 2015
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7646859185484305e-05,
+      "loss": 0.7994,
+      "step": 2016
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7644543699723555e-05,
+      "loss": 0.6647,
+      "step": 2017
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7642227121855755e-05,
+      "loss": 0.983,
+      "step": 2018
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.76399094519724e-05,
+      "loss": 0.8395,
+      "step": 2019
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.763759069016507e-05,
+      "loss": 0.7682,
+      "step": 2020
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.763527083652535e-05,
+      "loss": 0.2807,
+      "step": 2021
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7632949891144915e-05,
+      "loss": 0.4289,
+      "step": 2022
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.763062785411543e-05,
+      "loss": 1.027,
+      "step": 2023
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7628304725528635e-05,
+      "loss": 0.4979,
+      "step": 2024
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.762598050547631e-05,
+      "loss": 1.0727,
+      "step": 2025
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.762365519405026e-05,
+      "loss": 0.8148,
+      "step": 2026
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.762132879134236e-05,
+      "loss": 0.8263,
+      "step": 2027
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.761900129744452e-05,
+      "loss": 0.7777,
+      "step": 2028
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.761667271244868e-05,
+      "loss": 0.8664,
+      "step": 2029
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7614343036446824e-05,
+      "loss": 0.5069,
+      "step": 2030
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7612012269531e-05,
+      "loss": 0.9525,
+      "step": 2031
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7609680411793284e-05,
+      "loss": 0.714,
+      "step": 2032
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.76073474633258e-05,
+      "loss": 0.0,
+      "step": 2033
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7605013424220694e-05,
+      "loss": 1.3379,
+      "step": 2034
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.760267829457019e-05,
+      "loss": 0.4337,
+      "step": 2035
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7600342074466535e-05,
+      "loss": 0.6772,
+      "step": 2036
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.759800476400202e-05,
+      "loss": 0.9822,
+      "step": 2037
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7595666363268974e-05,
+      "loss": 1.0529,
+      "step": 2038
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.759332687235979e-05,
+      "loss": 0.3542,
+      "step": 2039
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7590986291366886e-05,
+      "loss": 1.1241,
+      "step": 2040
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.758864462038272e-05,
+      "loss": 1.1744,
+      "step": 2041
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.758630185949981e-05,
+      "loss": 0.6435,
+      "step": 2042
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.75839580088107e-05,
+      "loss": 0.4928,
+      "step": 2043
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7581613068407994e-05,
+      "loss": 0.4996,
+      "step": 2044
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.757926703838432e-05,
+      "loss": 0.2327,
+      "step": 2045
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7576919918832365e-05,
+      "loss": 0.6278,
+      "step": 2046
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7574571709844845e-05,
+      "loss": 0.8776,
+      "step": 2047
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.757222241151454e-05,
+      "loss": 1.0706,
+      "step": 2048
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.756987202393425e-05,
+      "loss": 0.6003,
+      "step": 2049
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7567520547196825e-05,
+      "loss": 0.7583,
+      "step": 2050
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.756516798139517e-05,
+      "loss": 0.4131,
+      "step": 2051
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7562814326622215e-05,
+      "loss": 0.8897,
+      "step": 2052
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.756045958297095e-05,
+      "loss": 0.8777,
+      "step": 2053
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7558103750534395e-05,
+      "loss": 0.6038,
+      "step": 2054
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7555746829405623e-05,
+      "loss": 0.6142,
+      "step": 2055
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.755338881967773e-05,
+      "loss": 0.3834,
+      "step": 2056
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.75510297214439e-05,
+      "loss": 1.5073,
+      "step": 2057
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.75486695347973e-05,
+      "loss": 0.9338,
+      "step": 2058
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.754630825983119e-05,
+      "loss": 0.4113,
+      "step": 2059
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.754394589663883e-05,
+      "loss": 0.5938,
+      "step": 2060
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.754158244531357e-05,
+      "loss": 0.7048,
+      "step": 2061
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7539217905948764e-05,
+      "loss": 0.613,
+      "step": 2062
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7536852278637834e-05,
+      "loss": 0.5176,
+      "step": 2063
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.753448556347423e-05,
+      "loss": 0.2665,
+      "step": 2064
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.753211776055145e-05,
+      "loss": 0.3395,
+      "step": 2065
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.752974886996303e-05,
+      "loss": 0.5451,
+      "step": 2066
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.752737889180256e-05,
+      "loss": 0.9085,
+      "step": 2067
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.752500782616367e-05,
+      "loss": 0.8152,
+      "step": 2068
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.752263567314003e-05,
+      "loss": 0.8035,
+      "step": 2069
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7520262432825333e-05,
+      "loss": 0.8421,
+      "step": 2070
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7517888105313355e-05,
+      "loss": 0.5483,
+      "step": 2071
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.75155126906979e-05,
+      "loss": 0.5585,
+      "step": 2072
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.751313618907278e-05,
+      "loss": 0.8828,
+      "step": 2073
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.751075860053191e-05,
+      "loss": 0.8177,
+      "step": 2074
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7508379925169205e-05,
+      "loss": 0.4159,
+      "step": 2075
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.750600016307864e-05,
+      "loss": 0.6047,
+      "step": 2076
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7503619314354216e-05,
+      "loss": 0.1451,
+      "step": 2077
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.750123737909e-05,
+      "loss": 0.5738,
+      "step": 2078
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7498854357380094e-05,
+      "loss": 0.9123,
+      "step": 2079
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.749647024931863e-05,
+      "loss": 0.4577,
+      "step": 2080
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7494085054999795e-05,
+      "loss": 0.573,
+      "step": 2081
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7491698774517825e-05,
+      "loss": 0.5301,
+      "step": 2082
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.748931140796698e-05,
+      "loss": 0.5394,
+      "step": 2083
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7486922955441584e-05,
+      "loss": 0.808,
+      "step": 2084
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7484533417035986e-05,
+      "loss": 0.7204,
+      "step": 2085
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.748214279284459e-05,
+      "loss": 0.2634,
+      "step": 2086
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.747975108296184e-05,
+      "loss": 0.6486,
+      "step": 2087
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7477358287482215e-05,
+      "loss": 0.7029,
+      "step": 2088
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.747496440650025e-05,
+      "loss": 0.2575,
+      "step": 2089
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7472569440110504e-05,
+      "loss": 0.5554,
+      "step": 2090
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7470173388407605e-05,
+      "loss": 1.0009,
+      "step": 2091
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.74677762514862e-05,
+      "loss": 0.8446,
+      "step": 2092
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7465378029440995e-05,
+      "loss": 0.6676,
+      "step": 2093
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.746297872236674e-05,
+      "loss": 1.0523,
+      "step": 2094
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.746057833035821e-05,
+      "loss": 0.8118,
+      "step": 2095
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7458176853510214e-05,
+      "loss": 0.5875,
+      "step": 2096
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7455774291917656e-05,
+      "loss": 0.847,
+      "step": 2097
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.745337064567544e-05,
+      "loss": 0.9207,
+      "step": 2098
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7450965914878525e-05,
+      "loss": 0.9504,
+      "step": 2099
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.74485600996219e-05,
+      "loss": 0.8046,
+      "step": 2100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.744615320000061e-05,
+      "loss": 0.5065,
+      "step": 2101
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.744374521610974e-05,
+      "loss": 0.7633,
+      "step": 2102
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.744133614804444e-05,
+      "loss": 0.0874,
+      "step": 2103
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.743892599589984e-05,
+      "loss": 0.7525,
+      "step": 2104
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7436514759771196e-05,
+      "loss": 0.845,
+      "step": 2105
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7434102439753734e-05,
+      "loss": 0.4839,
+      "step": 2106
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.743168903594277e-05,
+      "loss": 0.8494,
+      "step": 2107
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.742927454843364e-05,
+      "loss": 0.9966,
+      "step": 2108
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.742685897732173e-05,
+      "loss": 0.5649,
+      "step": 2109
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.742444232270247e-05,
+      "loss": 0.5512,
+      "step": 2110
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.742202458467133e-05,
+      "loss": 0.7609,
+      "step": 2111
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.741960576332383e-05,
+      "loss": 0.6711,
+      "step": 2112
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.741718585875551e-05,
+      "loss": 0.5871,
+      "step": 2113
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7414764871061984e-05,
+      "loss": 0.658,
+      "step": 2114
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.741234280033888e-05,
+      "loss": 0.7598,
+      "step": 2115
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.74099196466819e-05,
+      "loss": 0.4839,
+      "step": 2116
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.740749541018676e-05,
+      "loss": 0.7658,
+      "step": 2117
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.740507009094923e-05,
+      "loss": 0.5051,
+      "step": 2118
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.740264368906514e-05,
+      "loss": 0.251,
+      "step": 2119
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7400216204630316e-05,
+      "loss": 1.0443,
+      "step": 2120
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.739778763774067e-05,
+      "loss": 0.8632,
+      "step": 2121
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.739535798849216e-05,
+      "loss": 1.2181,
+      "step": 2122
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.739292725698075e-05,
+      "loss": 0.4233,
+      "step": 2123
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7390495443302464e-05,
+      "loss": 0.4808,
+      "step": 2124
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7388062547553386e-05,
+      "loss": 0.6412,
+      "step": 2125
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.738562856982963e-05,
+      "loss": 0.8029,
+      "step": 2126
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7383193510227335e-05,
+      "loss": 0.0429,
+      "step": 2127
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.738075736884272e-05,
+      "loss": 0.844,
+      "step": 2128
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7378320145772e-05,
+      "loss": 1.2704,
+      "step": 2129
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.737588184111147e-05,
+      "loss": 0.6759,
+      "step": 2130
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.737344245495747e-05,
+      "loss": 0.383,
+      "step": 2131
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.737100198740635e-05,
+      "loss": 1.0612,
+      "step": 2132
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.736856043855453e-05,
+      "loss": 0.9826,
+      "step": 2133
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.736611780849846e-05,
+      "loss": 0.5706,
+      "step": 2134
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.736367409733464e-05,
+      "loss": 0.2952,
+      "step": 2135
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.736122930515961e-05,
+      "loss": 0.78,
+      "step": 2136
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.735878343206995e-05,
+      "loss": 0.5994,
+      "step": 2137
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7356336478162286e-05,
+      "loss": 0.7896,
+      "step": 2138
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7353888443533286e-05,
+      "loss": 0.8117,
+      "step": 2139
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7351439328279655e-05,
+      "loss": 0.8533,
+      "step": 2140
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7348989132498165e-05,
+      "loss": 0.9205,
+      "step": 2141
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.734653785628559e-05,
+      "loss": 1.0827,
+      "step": 2142
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.734408549973877e-05,
+      "loss": 0.4475,
+      "step": 2143
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.73416320629546e-05,
+      "loss": 0.8022,
+      "step": 2144
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7339177546029996e-05,
+      "loss": 0.8749,
+      "step": 2145
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.733672194906193e-05,
+      "loss": 0.7165,
+      "step": 2146
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.73342652721474e-05,
+      "loss": 0.4837,
+      "step": 2147
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.733180751538346e-05,
+      "loss": 0.5648,
+      "step": 2148
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.732934867886722e-05,
+      "loss": 0.7924,
+      "step": 2149
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.73268887626958e-05,
+      "loss": 0.5426,
+      "step": 2150
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.732442776696639e-05,
+      "loss": 0.6529,
+      "step": 2151
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.732196569177621e-05,
+      "loss": 0.9436,
+      "step": 2152
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.731950253722251e-05,
+      "loss": 0.8755,
+      "step": 2153
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7317038303402625e-05,
+      "loss": 0.5558,
+      "step": 2154
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.731457299041389e-05,
+      "loss": 0.8949,
+      "step": 2155
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7312106598353696e-05,
+      "loss": 0.8774,
+      "step": 2156
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.730963912731948e-05,
+      "loss": 0.646,
+      "step": 2157
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.730717057740873e-05,
+      "loss": 0.4383,
+      "step": 2158
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7304700948718955e-05,
+      "loss": 0.2503,
+      "step": 2159
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7302230241347724e-05,
+      "loss": 1.0384,
+      "step": 2160
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7299758455392636e-05,
+      "loss": 0.9552,
+      "step": 2161
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.729728559095136e-05,
+      "loss": 0.5765,
+      "step": 2162
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.729481164812156e-05,
+      "loss": 0.5899,
+      "step": 2163
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7292336627000986e-05,
+      "loss": 1.0963,
+      "step": 2164
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.728986052768742e-05,
+      "loss": 0.3566,
+      "step": 2165
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.728738335027866e-05,
+      "loss": 0.7884,
+      "step": 2166
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.728490509487259e-05,
+      "loss": 0.7081,
+      "step": 2167
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.72824257615671e-05,
+      "loss": 0.6622,
+      "step": 2168
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.727994535046014e-05,
+      "loss": 1.0499,
+      "step": 2169
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.727746386164971e-05,
+      "loss": 0.7504,
+      "step": 2170
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7274981295233826e-05,
+      "loss": 0.744,
+      "step": 2171
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.727249765131057e-05,
+      "loss": 0.874,
+      "step": 2172
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.727001292997806e-05,
+      "loss": 0.8413,
+      "step": 2173
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.726752713133446e-05,
+      "loss": 0.9874,
+      "step": 2174
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.726504025547796e-05,
+      "loss": 0.1872,
+      "step": 2175
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.726255230250682e-05,
+      "loss": 0.9177,
+      "step": 2176
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7260063272519315e-05,
+      "loss": 0.7196,
+      "step": 2177
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.725757316561378e-05,
+      "loss": 0.892,
+      "step": 2178
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.725508198188858e-05,
+      "loss": 0.7574,
+      "step": 2179
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7252589721442144e-05,
+      "loss": 1.0191,
+      "step": 2180
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.725009638437292e-05,
+      "loss": 0.6196,
+      "step": 2181
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.724760197077941e-05,
+      "loss": 0.6625,
+      "step": 2182
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.724510648076016e-05,
+      "loss": 0.7974,
+      "step": 2183
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.724260991441376e-05,
+      "loss": 0.2781,
+      "step": 2184
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7240112271838826e-05,
+      "loss": 0.7951,
+      "step": 2185
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.723761355313403e-05,
+      "loss": 0.8387,
+      "step": 2186
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7235113758398084e-05,
+      "loss": 1.0115,
+      "step": 2187
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.723261288772975e-05,
+      "loss": 0.7597,
+      "step": 2188
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7230110941227816e-05,
+      "loss": 0.7704,
+      "step": 2189
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.722760791899114e-05,
+      "loss": 0.7434,
+      "step": 2190
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.722510382111858e-05,
+      "loss": 0.9915,
+      "step": 2191
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7222598647709084e-05,
+      "loss": 0.6625,
+      "step": 2192
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.722009239886161e-05,
+      "loss": 0.4962,
+      "step": 2193
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.721758507467517e-05,
+      "loss": 0.4149,
+      "step": 2194
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.721507667524881e-05,
+      "loss": 1.1105,
+      "step": 2195
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7212567200681626e-05,
+      "loss": 0.7678,
+      "step": 2196
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7210056651072765e-05,
+      "loss": 0.726,
+      "step": 2197
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.72075450265214e-05,
+      "loss": 1.0196,
+      "step": 2198
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7205032327126766e-05,
+      "loss": 0.8268,
+      "step": 2199
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.720251855298811e-05,
+      "loss": 1.049,
+      "step": 2200
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.747691810131073,
+      "eval_runtime": 38.297,
+      "eval_samples_per_second": 1.044,
+      "eval_steps_per_second": 1.044,
+      "step": 2200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.720000370420474e-05,
+      "loss": 1.0031,
+      "step": 2201
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.719748778087603e-05,
+      "loss": 0.9576,
+      "step": 2202
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.719497078310134e-05,
+      "loss": 0.4392,
+      "step": 2203
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.719245271098013e-05,
+      "loss": 0.3671,
+      "step": 2204
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7189933564611864e-05,
+      "loss": 1.0149,
+      "step": 2205
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7187413344096065e-05,
+      "loss": 0.6608,
+      "step": 2206
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.718489204953229e-05,
+      "loss": 0.7032,
+      "step": 2207
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.718236968102016e-05,
+      "loss": 0.5436,
+      "step": 2208
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.717984623865931e-05,
+      "loss": 0.6251,
+      "step": 2209
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.717732172254942e-05,
+      "loss": 0.6511,
+      "step": 2210
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.717479613279024e-05,
+      "loss": 1.0864,
+      "step": 2211
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.717226946948153e-05,
+      "loss": 1.1379,
+      "step": 2212
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.716974173272312e-05,
+      "loss": 0.8537,
+      "step": 2213
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.716721292261486e-05,
+      "loss": 0.2746,
+      "step": 2214
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7164683039256655e-05,
+      "loss": 0.5936,
+      "step": 2215
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.716215208274845e-05,
+      "loss": 0.596,
+      "step": 2216
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7159620053190236e-05,
+      "loss": 0.7659,
+      "step": 2217
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7157086950682015e-05,
+      "loss": 0.4044,
+      "step": 2218
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7154552775323895e-05,
+      "loss": 0.3012,
+      "step": 2219
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.715201752721597e-05,
+      "loss": 0.5737,
+      "step": 2220
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.714948120645839e-05,
+      "loss": 0.6769,
+      "step": 2221
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.714694381315137e-05,
+      "loss": 0.7979,
+      "step": 2222
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.714440534739514e-05,
+      "loss": 0.7717,
+      "step": 2223
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.714186580928998e-05,
+      "loss": 0.9479,
+      "step": 2224
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.713932519893622e-05,
+      "loss": 0.9896,
+      "step": 2225
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.713678351643424e-05,
+      "loss": 0.4586,
+      "step": 2226
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.713424076188443e-05,
+      "loss": 0.8681,
+      "step": 2227
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.713169693538725e-05,
+      "loss": 0.913,
+      "step": 2228
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.712915203704319e-05,
+      "loss": 0.7389,
+      "step": 2229
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.71266060669528e-05,
+      "loss": 0.4808,
+      "step": 2230
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7124059025216644e-05,
+      "loss": 0.901,
+      "step": 2231
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.712151091193536e-05,
+      "loss": 0.1735,
+      "step": 2232
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.71189617272096e-05,
+      "loss": 0.3673,
+      "step": 2233
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.711641147114007e-05,
+      "loss": 1.05,
+      "step": 2234
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.711386014382753e-05,
+      "loss": 0.89,
+      "step": 2235
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.711130774537276e-05,
+      "loss": 0.9195,
+      "step": 2236
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.71087542758766e-05,
+      "loss": 0.3625,
+      "step": 2237
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.710619973543992e-05,
+      "loss": 0.7313,
+      "step": 2238
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.710364412416363e-05,
+      "loss": 0.6822,
+      "step": 2239
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.710108744214872e-05,
+      "loss": 0.6843,
+      "step": 2240
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.709852968949616e-05,
+      "loss": 0.4911,
+      "step": 2241
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.709597086630701e-05,
+      "loss": 0.9581,
+      "step": 2242
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.709341097268236e-05,
+      "loss": 0.5664,
+      "step": 2243
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.709085000872333e-05,
+      "loss": 0.5561,
+      "step": 2244
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.70882879745311e-05,
+      "loss": 1.0127,
+      "step": 2245
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.708572487020689e-05,
+      "loss": 0.8096,
+      "step": 2246
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.708316069585193e-05,
+      "loss": 0.4087,
+      "step": 2247
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.708059545156755e-05,
+      "loss": 0.4984,
+      "step": 2248
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.707802913745507e-05,
+      "loss": 0.5912,
+      "step": 2249
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.707546175361587e-05,
+      "loss": 0.6874,
+      "step": 2250
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.70728933001514e-05,
+      "loss": 0.093,
+      "step": 2251
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7070323777163115e-05,
+      "loss": 0.4979,
+      "step": 2252
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.706775318475251e-05,
+      "loss": 0.2558,
+      "step": 2253
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.706518152302116e-05,
+      "loss": 0.5884,
+      "step": 2254
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.706260879207064e-05,
+      "loss": 0.5565,
+      "step": 2255
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.70600349920026e-05,
+      "loss": 1.0066,
+      "step": 2256
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7057460122918715e-05,
+      "loss": 0.2181,
+      "step": 2257
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.70548841849207e-05,
+      "loss": 0.7067,
+      "step": 2258
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.705230717811033e-05,
+      "loss": 0.3949,
+      "step": 2259
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.70497291025894e-05,
+      "loss": 0.1214,
+      "step": 2260
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.704714995845977e-05,
+      "loss": 0.5747,
+      "step": 2261
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.704456974582331e-05,
+      "loss": 0.6837,
+      "step": 2262
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7041988464781975e-05,
+      "loss": 0.558,
+      "step": 2263
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.703940611543773e-05,
+      "loss": 0.6545,
+      "step": 2264
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.703682269789258e-05,
+      "loss": 0.4056,
+      "step": 2265
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.70342382122486e-05,
+      "loss": 0.8755,
+      "step": 2266
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.703165265860788e-05,
+      "loss": 0.233,
+      "step": 2267
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.702906603707257e-05,
+      "loss": 0.9893,
+      "step": 2268
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7026478347744864e-05,
+      "loss": 0.3374,
+      "step": 2269
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.702388959072696e-05,
+      "loss": 0.8807,
+      "step": 2270
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.702129976612117e-05,
+      "loss": 0.7689,
+      "step": 2271
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.701870887402976e-05,
+      "loss": 0.392,
+      "step": 2272
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.701611691455512e-05,
+      "loss": 0.9926,
+      "step": 2273
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7013523887799635e-05,
+      "loss": 0.4582,
+      "step": 2274
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.701092979386573e-05,
+      "loss": 0.5868,
+      "step": 2275
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7008334632855904e-05,
+      "loss": 0.9924,
+      "step": 2276
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7005738404872675e-05,
+      "loss": 0.4243,
+      "step": 2277
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7003141110018604e-05,
+      "loss": 0.4405,
+      "step": 2278
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.700054274839629e-05,
+      "loss": 1.0751,
+      "step": 2279
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6997943320108404e-05,
+      "loss": 0.6668,
+      "step": 2280
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6995342825257615e-05,
+      "loss": 0.5205,
+      "step": 2281
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6992741263946674e-05,
+      "loss": 0.4986,
+      "step": 2282
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.699013863627835e-05,
+      "loss": 0.3716,
+      "step": 2283
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6987534942355456e-05,
+      "loss": 0.2552,
+      "step": 2284
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.698493018228085e-05,
+      "loss": 0.8408,
+      "step": 2285
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6982324356157456e-05,
+      "loss": 0.9304,
+      "step": 2286
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6979717464088185e-05,
+      "loss": 1.1357,
+      "step": 2287
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.697710950617604e-05,
+      "loss": 0.8219,
+      "step": 2288
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6974500482524056e-05,
+      "loss": 0.6497,
+      "step": 2289
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.69718903932353e-05,
+      "loss": 0.8149,
+      "step": 2290
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6969279238412866e-05,
+      "loss": 0.5708,
+      "step": 2291
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.696666701815993e-05,
+      "loss": 0.7996,
+      "step": 2292
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.69640537325797e-05,
+      "loss": 0.8896,
+      "step": 2293
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6961439381775374e-05,
+      "loss": 0.4361,
+      "step": 2294
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.695882396585026e-05,
+      "loss": 0.4794,
+      "step": 2295
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.695620748490768e-05,
+      "loss": 0.6153,
+      "step": 2296
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6953589939050994e-05,
+      "loss": 0.4225,
+      "step": 2297
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.695097132838361e-05,
+      "loss": 0.5902,
+      "step": 2298
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6948351653008976e-05,
+      "loss": 0.7349,
+      "step": 2299
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6945730913030585e-05,
+      "loss": 0.7233,
+      "step": 2300
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.694310910855197e-05,
+      "loss": 0.7246,
+      "step": 2301
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.694048623967671e-05,
+      "loss": 0.9065,
+      "step": 2302
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.693786230650841e-05,
+      "loss": 0.5343,
+      "step": 2303
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.693523730915074e-05,
+      "loss": 0.7733,
+      "step": 2304
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.693261124770741e-05,
+      "loss": 0.8233,
+      "step": 2305
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6929984122282143e-05,
+      "loss": 1.9124,
+      "step": 2306
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.692735593297873e-05,
+      "loss": 0.8655,
+      "step": 2307
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.692472667990101e-05,
+      "loss": 0.5369,
+      "step": 2308
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.692209636315285e-05,
+      "loss": 0.3681,
+      "step": 2309
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.691946498283815e-05,
+      "loss": 0.801,
+      "step": 2310
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.691683253906087e-05,
+      "loss": 0.933,
+      "step": 2311
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.6914199031925006e-05,
+      "loss": 1.0356,
+      "step": 2312
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.691156446153459e-05,
+      "loss": 0.6316,
+      "step": 2313
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6908928827993717e-05,
+      "loss": 1.0383,
+      "step": 2314
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6906292131406495e-05,
+      "loss": 0.717,
+      "step": 2315
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6903654371877086e-05,
+      "loss": 0.5581,
+      "step": 2316
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.690101554950971e-05,
+      "loss": 0.9974,
+      "step": 2317
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.68983756644086e-05,
+      "loss": 0.323,
+      "step": 2318
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.689573471667805e-05,
+      "loss": 0.3159,
+      "step": 2319
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.68930927064224e-05,
+      "loss": 0.6466,
+      "step": 2320
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6890449633746006e-05,
+      "loss": 0.6565,
+      "step": 2321
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.68878054987533e-05,
+      "loss": 0.7064,
+      "step": 2322
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6885160301548736e-05,
+      "loss": 0.4333,
+      "step": 2323
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6882514042236805e-05,
+      "loss": 0.6235,
+      "step": 2324
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.687986672092205e-05,
+      "loss": 0.7856,
+      "step": 2325
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6877218337709055e-05,
+      "loss": 0.5369,
+      "step": 2326
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.687456889270246e-05,
+      "loss": 0.3829,
+      "step": 2327
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6871918386006917e-05,
+      "loss": 0.942,
+      "step": 2328
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6869266817727133e-05,
+      "loss": 0.505,
+      "step": 2329
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6866614187967873e-05,
+      "loss": 1.3502,
+      "step": 2330
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.686396049683392e-05,
+      "loss": 0.5719,
+      "step": 2331
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.68613057444301e-05,
+      "loss": 0.5084,
+      "step": 2332
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.685864993086132e-05,
+      "loss": 0.1719,
+      "step": 2333
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6855993056232464e-05,
+      "loss": 0.8394,
+      "step": 2334
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6853335120648516e-05,
+      "loss": 1.3747,
+      "step": 2335
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.685067612421446e-05,
+      "loss": 0.5074,
+      "step": 2336
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.684801606703536e-05,
+      "loss": 0.8373,
+      "step": 2337
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6845354949216295e-05,
+      "loss": 0.7656,
+      "step": 2338
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6842692770862387e-05,
+      "loss": 0.8024,
+      "step": 2339
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.684002953207882e-05,
+      "loss": 0.1376,
+      "step": 2340
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.683736523297079e-05,
+      "loss": 0.8009,
+      "step": 2341
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.683469987364356e-05,
+      "loss": 0.889,
+      "step": 2342
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.683203345420243e-05,
+      "loss": 0.5783,
+      "step": 2343
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.682936597475273e-05,
+      "loss": 0.5267,
+      "step": 2344
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.682669743539984e-05,
+      "loss": 0.6188,
+      "step": 2345
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.682402783624918e-05,
+      "loss": 0.2344,
+      "step": 2346
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6821357177406226e-05,
+      "loss": 0.4927,
+      "step": 2347
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.681868545897646e-05,
+      "loss": 0.6907,
+      "step": 2348
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.681601268106546e-05,
+      "loss": 0.2962,
+      "step": 2349
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6813338843778787e-05,
+      "loss": 0.6271,
+      "step": 2350
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6810663947222085e-05,
+      "loss": 0.7627,
+      "step": 2351
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.680798799150103e-05,
+      "loss": 0.7257,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.680531097672133e-05,
+      "loss": 0.9039,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.680263290298874e-05,
+      "loss": 0.431,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.679995377040906e-05,
+      "loss": 0.4915,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.679727357908815e-05,
+      "loss": 0.3081,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.679459232913185e-05,
+      "loss": 0.14,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6791910020646104e-05,
+      "loss": 0.514,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6789226653736904e-05,
+      "loss": 1.0239,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6786542228510215e-05,
+      "loss": 0.819,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.678385674507211e-05,
+      "loss": 0.3792,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.678117020352868e-05,
+      "loss": 0.5434,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6778482603986034e-05,
+      "loss": 0.3595,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.677579394655039e-05,
+      "loss": 0.8625,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.677310423132792e-05,
+      "loss": 0.3897,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.677041345842491e-05,
+      "loss": 0.4804,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6767721627947644e-05,
+      "loss": 0.464,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.676502874000247e-05,
+      "loss": 0.233,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6762334794695794e-05,
+      "loss": 1.0191,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6759639792133995e-05,
+      "loss": 0.2482,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.675694373242357e-05,
+      "loss": 0.8987,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.675424661567103e-05,
+      "loss": 1.2915,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6751548441982915e-05,
+      "loss": 0.6828,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6748849211465816e-05,
+      "loss": 0.9688,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.674614892422638e-05,
+      "loss": 0.1226,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6743447580371266e-05,
+      "loss": 0.9641,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.674074518000719e-05,
+      "loss": 1.5561,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.673804172324094e-05,
+      "loss": 1.8179,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.673533721017929e-05,
+      "loss": 0.8036,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.673263164092909e-05,
+      "loss": 0.4941,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.672992501559722e-05,
+      "loss": 0.4453,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.672721733429062e-05,
+      "loss": 0.495,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.672450859711624e-05,
+      "loss": 0.5681,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.672179880418111e-05,
+      "loss": 0.9296,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.671908795559226e-05,
+      "loss": 0.7055,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.671637605145679e-05,
+      "loss": 0.4452,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.671366309188184e-05,
+      "loss": 0.6523,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.671094907697459e-05,
+      "loss": 0.7342,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6708234006842244e-05,
+      "loss": 0.5612,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6705517881592075e-05,
+      "loss": 0.8244,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6702800701331376e-05,
+      "loss": 0.9028,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.670008246616749e-05,
+      "loss": 1.0194,
+      "step": 2392
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6697363176207806e-05,
+      "loss": 0.9868,
+      "step": 2393
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.669464283155975e-05,
+      "loss": 0.3597,
+      "step": 2394
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6691921432330796e-05,
+      "loss": 0.8968,
+      "step": 2395
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.668919897862844e-05,
+      "loss": 0.7335,
+      "step": 2396
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.668647547056024e-05,
+      "loss": 0.5213,
+      "step": 2397
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.668375090823379e-05,
+      "loss": 1.1249,
+      "step": 2398
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.668102529175673e-05,
+      "loss": 0.6048,
+      "step": 2399
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6678298621236735e-05,
+      "loss": 0.9398,
+      "step": 2400
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.7398985028266907,
+      "eval_runtime": 38.33,
+      "eval_samples_per_second": 1.044,
+      "eval_steps_per_second": 1.044,
+      "step": 2400
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.667557089678151e-05,
+      "loss": 0.438,
+      "step": 2401
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6672842118498824e-05,
+      "loss": 0.0719,
+      "step": 2402
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6670112286496485e-05,
+      "loss": 1.0784,
+      "step": 2403
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.666738140088232e-05,
+      "loss": 0.7476,
+      "step": 2404
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.666464946176423e-05,
+      "loss": 0.7766,
+      "step": 2405
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.666191646925013e-05,
+      "loss": 0.4079,
+      "step": 2406
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.665918242344799e-05,
+      "loss": 0.4432,
+      "step": 2407
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.665644732446583e-05,
+      "loss": 0.7916,
+      "step": 2408
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.665371117241168e-05,
+      "loss": 0.9526,
+      "step": 2409
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6650973967393656e-05,
+      "loss": 0.5415,
+      "step": 2410
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6648235709519865e-05,
+      "loss": 1.1392,
+      "step": 2411
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.664549639889851e-05,
+      "loss": 0.339,
+      "step": 2412
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.66427560356378e-05,
+      "loss": 0.6707,
+      "step": 2413
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6640014619845976e-05,
+      "loss": 0.6715,
+      "step": 2414
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.663727215163137e-05,
+      "loss": 0.2399,
+      "step": 2415
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.66345286311023e-05,
+      "loss": 0.9361,
+      "step": 2416
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6631784058367155e-05,
+      "loss": 0.7349,
+      "step": 2417
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.662903843353436e-05,
+      "loss": 0.646,
+      "step": 2418
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.662629175671239e-05,
+      "loss": 0.8339,
+      "step": 2419
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.662354402800976e-05,
+      "loss": 0.4206,
+      "step": 2420
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.662079524753498e-05,
+      "loss": 1.1179,
+      "step": 2421
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6618045415396696e-05,
+      "loss": 0.999,
+      "step": 2422
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.66152945317035e-05,
+      "loss": 0.8294,
+      "step": 2423
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.661254259656408e-05,
+      "loss": 0.4251,
+      "step": 2424
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.660978961008716e-05,
+      "loss": 0.5322,
+      "step": 2425
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.660703557238149e-05,
+      "loss": 1.6909,
+      "step": 2426
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.660428048355587e-05,
+      "loss": 1.0169,
+      "step": 2427
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6601524343719135e-05,
+      "loss": 0.7277,
+      "step": 2428
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.659876715298018e-05,
+      "loss": 0.9047,
+      "step": 2429
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.659600891144792e-05,
+      "loss": 0.5045,
+      "step": 2430
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.659324961923132e-05,
+      "loss": 0.6952,
+      "step": 2431
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.659048927643938e-05,
+      "loss": 0.2713,
+      "step": 2432
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.658772788318117e-05,
+      "loss": 0.4438,
+      "step": 2433
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.658496543956576e-05,
+      "loss": 1.0218,
+      "step": 2434
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.65822019457023e-05,
+      "loss": 0.7881,
+      "step": 2435
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.657943740169994e-05,
+      "loss": 1.1147,
+      "step": 2436
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.65766718076679e-05,
+      "loss": 0.4019,
+      "step": 2437
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.657390516371545e-05,
+      "loss": 0.7867,
+      "step": 2438
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6571137469951876e-05,
+      "loss": 0.4034,
+      "step": 2439
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.656836872648651e-05,
+      "loss": 1.0349,
+      "step": 2440
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.656559893342875e-05,
+      "loss": 0.82,
+      "step": 2441
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.656282809088801e-05,
+      "loss": 0.3502,
+      "step": 2442
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6560056198973745e-05,
+      "loss": 1.0717,
+      "step": 2443
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6557283257795475e-05,
+      "loss": 0.492,
+      "step": 2444
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.655450926746273e-05,
+      "loss": 0.6875,
+      "step": 2445
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6551734228085104e-05,
+      "loss": 0.4935,
+      "step": 2446
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.654895813977222e-05,
+      "loss": 0.539,
+      "step": 2447
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.654618100263377e-05,
+      "loss": 1.0601,
+      "step": 2448
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.654340281677944e-05,
+      "loss": 0.7102,
+      "step": 2449
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.654062358231899e-05,
+      "loss": 0.4396,
+      "step": 2450
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.653784329936223e-05,
+      "loss": 0.7133,
+      "step": 2451
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.653506196801898e-05,
+      "loss": 0.0556,
+      "step": 2452
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.653227958839912e-05,
+      "loss": 0.7733,
+      "step": 2453
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.652949616061257e-05,
+      "loss": 0.6339,
+      "step": 2454
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.652671168476929e-05,
+      "loss": 0.8691,
+      "step": 2455
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.652392616097928e-05,
+      "loss": 0.745,
+      "step": 2456
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6521139589352585e-05,
+      "loss": 1.0928,
+      "step": 2457
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.65183519699993e-05,
+      "loss": 0.7104,
+      "step": 2458
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.651556330302954e-05,
+      "loss": 1.1016,
+      "step": 2459
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6512773588553455e-05,
+      "loss": 0.4732,
+      "step": 2460
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.650998282668128e-05,
+      "loss": 0.9604,
+      "step": 2461
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.650719101752326e-05,
+      "loss": 0.7245,
+      "step": 2462
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.650439816118969e-05,
+      "loss": 1.2018,
+      "step": 2463
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6501604257790885e-05,
+      "loss": 0.3063,
+      "step": 2464
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6498809307437235e-05,
+      "loss": 0.6478,
+      "step": 2465
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6496013310239144e-05,
+      "loss": 0.5003,
+      "step": 2466
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6493216266307084e-05,
+      "loss": 0.2542,
+      "step": 2467
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.649041817575154e-05,
+      "loss": 0.1678,
+      "step": 2468
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.648761903868305e-05,
+      "loss": 0.2975,
+      "step": 2469
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.64848188552122e-05,
+      "loss": 0.6275,
+      "step": 2470
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.648201762544962e-05,
+      "loss": 0.444,
+      "step": 2471
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.647921534950597e-05,
+      "loss": 0.3688,
+      "step": 2472
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6476412027491945e-05,
+      "loss": 0.6487,
+      "step": 2473
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.64736076595183e-05,
+      "loss": 0.4936,
+      "step": 2474
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.647080224569581e-05,
+      "loss": 0.9984,
+      "step": 2475
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.646799578613532e-05,
+      "loss": 1.1768,
+      "step": 2476
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.64651882809477e-05,
+      "loss": 0.7552,
+      "step": 2477
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6462379730243854e-05,
+      "loss": 0.7176,
+      "step": 2478
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.645957013413473e-05,
+      "loss": 1.083,
+      "step": 2479
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.645675949273132e-05,
+      "loss": 0.8687,
+      "step": 2480
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.645394780614469e-05,
+      "loss": 0.8902,
+      "step": 2481
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6451135074485877e-05,
+      "loss": 0.7378,
+      "step": 2482
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.644832129786602e-05,
+      "loss": 0.5959,
+      "step": 2483
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6445506476396276e-05,
+      "loss": 0.4796,
+      "step": 2484
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6442690610187844e-05,
+      "loss": 0.4247,
+      "step": 2485
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.643987369935196e-05,
+      "loss": 0.6027,
+      "step": 2486
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.643705574399991e-05,
+      "loss": 0.5143,
+      "step": 2487
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6434236744243034e-05,
+      "loss": 0.6931,
+      "step": 2488
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.643141670019267e-05,
+      "loss": 0.8012,
+      "step": 2489
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.642859561196025e-05,
+      "loss": 0.7486,
+      "step": 2490
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6425773479657205e-05,
+      "loss": 1.759,
+      "step": 2491
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.642295030339503e-05,
+      "loss": 1.0644,
+      "step": 2492
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.642012608328526e-05,
+      "loss": 0.2915,
+      "step": 2493
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.641730081943945e-05,
+      "loss": 0.0682,
+      "step": 2494
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.641447451196924e-05,
+      "loss": 0.8913,
+      "step": 2495
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.641164716098627e-05,
+      "loss": 0.6978,
+      "step": 2496
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6408818766602226e-05,
+      "loss": 0.92,
+      "step": 2497
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.640598932892886e-05,
+      "loss": 0.5048,
+      "step": 2498
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.640315884807794e-05,
+      "loss": 0.7655,
+      "step": 2499
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6400327324161294e-05,
+      "loss": 0.9582,
+      "step": 2500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.639749475729077e-05,
+      "loss": 0.7815,
+      "step": 2501
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.639466114757828e-05,
+      "loss": 0.5898,
+      "step": 2502
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.639182649513577e-05,
+      "loss": 0.2491,
+      "step": 2503
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6388990800075214e-05,
+      "loss": 0.7748,
+      "step": 2504
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.638615406250864e-05,
+      "loss": 0.482,
+      "step": 2505
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6383316282548104e-05,
+      "loss": 0.4752,
+      "step": 2506
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.638047746030574e-05,
+      "loss": 0.9914,
+      "step": 2507
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.637763759589368e-05,
+      "loss": 0.0826,
+      "step": 2508
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.63747966894241e-05,
+      "loss": 0.2833,
+      "step": 2509
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.637195474100926e-05,
+      "loss": 0.4729,
+      "step": 2510
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.636911175076141e-05,
+      "loss": 0.8242,
+      "step": 2511
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.636626771879287e-05,
+      "loss": 0.5326,
+      "step": 2512
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6363422645216e-05,
+      "loss": 0.4099,
+      "step": 2513
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.636057653014319e-05,
+      "loss": 0.6549,
+      "step": 2514
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6357729373686884e-05,
+      "loss": 0.3685,
+      "step": 2515
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.635488117595954e-05,
+      "loss": 0.2407,
+      "step": 2516
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6352031937073705e-05,
+      "loss": 0.6308,
+      "step": 2517
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.634918165714191e-05,
+      "loss": 0.9507,
+      "step": 2518
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6346330336276784e-05,
+      "loss": 0.993,
+      "step": 2519
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6343477974590956e-05,
+      "loss": 0.5062,
+      "step": 2520
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.63406245721971e-05,
+      "loss": 0.1279,
+      "step": 2521
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6337770129207963e-05,
+      "loss": 0.659,
+      "step": 2522
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6334914645736286e-05,
+      "loss": 0.8075,
+      "step": 2523
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.63320581218949e-05,
+      "loss": 0.5421,
+      "step": 2524
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6329200557796637e-05,
+      "loss": 0.6782,
+      "step": 2525
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.632634195355438e-05,
+      "loss": 0.8697,
+      "step": 2526
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6323482309281085e-05,
+      "loss": 0.6651,
+      "step": 2527
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.63206216250897e-05,
+      "loss": 0.7261,
+      "step": 2528
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.631775990109325e-05,
+      "loss": 1.2606,
+      "step": 2529
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6314897137404774e-05,
+      "loss": 0.6161,
+      "step": 2530
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.631203333413738e-05,
+      "loss": 0.8564,
+      "step": 2531
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6309168491404205e-05,
+      "loss": 1.2581,
+      "step": 2532
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6306302609318414e-05,
+      "loss": 0.3392,
+      "step": 2533
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.630343568799323e-05,
+      "loss": 0.4906,
+      "step": 2534
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6300567727541905e-05,
+      "loss": 0.7462,
+      "step": 2535
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.629769872807776e-05,
+      "loss": 0.4723,
+      "step": 2536
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.629482868971412e-05,
+      "loss": 0.3268,
+      "step": 2537
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.629195761256436e-05,
+      "loss": 1.0663,
+      "step": 2538
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.628908549674192e-05,
+      "loss": 0.9861,
+      "step": 2539
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.628621234236025e-05,
+      "loss": 0.4528,
+      "step": 2540
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.628333814953286e-05,
+      "loss": 0.6076,
+      "step": 2541
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.62804629183733e-05,
+      "loss": 0.7486,
+      "step": 2542
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6277586648995146e-05,
+      "loss": 1.1104,
+      "step": 2543
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6274709341512046e-05,
+      "loss": 0.8749,
+      "step": 2544
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6271830996037645e-05,
+      "loss": 0.8388,
+      "step": 2545
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.626895161268567e-05,
+      "loss": 0.9039,
+      "step": 2546
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6266071191569864e-05,
+      "loss": 0.4994,
+      "step": 2547
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.626318973280402e-05,
+      "loss": 0.6607,
+      "step": 2548
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.626030723650197e-05,
+      "loss": 0.3826,
+      "step": 2549
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6257423702777596e-05,
+      "loss": 0.6967,
+      "step": 2550
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.625453913174481e-05,
+      "loss": 0.5235,
+      "step": 2551
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6251653523517556e-05,
+      "loss": 0.7532,
+      "step": 2552
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.624876687820984e-05,
+      "loss": 0.5449,
+      "step": 2553
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.624587919593571e-05,
+      "loss": 0.6076,
+      "step": 2554
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.624299047680923e-05,
+      "loss": 0.8811,
+      "step": 2555
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.624010072094452e-05,
+      "loss": 0.7986,
+      "step": 2556
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6237209928455747e-05,
+      "loss": 0.9333,
+      "step": 2557
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.623431809945712e-05,
+      "loss": 0.672,
+      "step": 2558
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6231425234062864e-05,
+      "loss": 0.3952,
+      "step": 2559
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.622853133238728e-05,
+      "loss": 0.432,
+      "step": 2560
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.622563639454467e-05,
+      "loss": 0.4932,
+      "step": 2561
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6222740420649425e-05,
+      "loss": 0.426,
+      "step": 2562
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.621984341081593e-05,
+      "loss": 1.0328,
+      "step": 2563
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.621694536515865e-05,
+      "loss": 0.9681,
+      "step": 2564
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6214046283792073e-05,
+      "loss": 0.8951,
+      "step": 2565
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.62111461668307e-05,
+      "loss": 0.6665,
+      "step": 2566
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.620824501438914e-05,
+      "loss": 0.8054,
+      "step": 2567
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6205342826581984e-05,
+      "loss": 1.1863,
+      "step": 2568
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6202439603523886e-05,
+      "loss": 0.9683,
+      "step": 2569
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.619953534532953e-05,
+      "loss": 1.1291,
+      "step": 2570
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.619663005211366e-05,
+      "loss": 0.3408,
+      "step": 2571
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.619372372399106e-05,
+      "loss": 0.8327,
+      "step": 2572
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.619081636107652e-05,
+      "loss": 0.6481,
+      "step": 2573
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.618790796348492e-05,
+      "loss": 0.6558,
+      "step": 2574
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.618499853133115e-05,
+      "loss": 0.9452,
+      "step": 2575
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6182088064730135e-05,
+      "loss": 0.8246,
+      "step": 2576
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.617917656379687e-05,
+      "loss": 0.5309,
+      "step": 2577
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.617626402864638e-05,
+      "loss": 0.7243,
+      "step": 2578
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.61733504593937e-05,
+      "loss": 0.6082,
+      "step": 2579
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6170435856153956e-05,
+      "loss": 0.9301,
+      "step": 2580
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.616752021904228e-05,
+      "loss": 0.494,
+      "step": 2581
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6164603548173854e-05,
+      "loss": 1.1012,
+      "step": 2582
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.616168584366391e-05,
+      "loss": 0.6566,
+      "step": 2583
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.615876710562771e-05,
+      "loss": 0.8778,
+      "step": 2584
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.615584733418054e-05,
+      "loss": 0.1998,
+      "step": 2585
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.615292652943779e-05,
+      "loss": 0.4657,
+      "step": 2586
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6150004691514805e-05,
+      "loss": 0.6115,
+      "step": 2587
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6147081820527034e-05,
+      "loss": 0.8981,
+      "step": 2588
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.614415791658994e-05,
+      "loss": 0.4744,
+      "step": 2589
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.614123297981903e-05,
+      "loss": 0.6616,
+      "step": 2590
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.613830701032987e-05,
+      "loss": 0.9437,
+      "step": 2591
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.613538000823803e-05,
+      "loss": 0.9831,
+      "step": 2592
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6132451973659166e-05,
+      "loss": 0.7578,
+      "step": 2593
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.612952290670893e-05,
+      "loss": 1.0596,
+      "step": 2594
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6126592807503035e-05,
+      "loss": 0.832,
+      "step": 2595
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6123661676157256e-05,
+      "loss": 1.4584,
+      "step": 2596
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.612072951278737e-05,
+      "loss": 0.4823,
+      "step": 2597
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6117796317509225e-05,
+      "loss": 0.9099,
+      "step": 2598
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.611486209043869e-05,
+      "loss": 0.7404,
+      "step": 2599
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.611192683169168e-05,
+      "loss": 1.0692,
+      "step": 2600
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.740403950214386,
+      "eval_runtime": 40.3357,
+      "eval_samples_per_second": 0.992,
+      "eval_steps_per_second": 0.992,
+      "step": 2600
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.610899054138416e-05,
+      "loss": 0.3694,
+      "step": 2601
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.610605321963213e-05,
+      "loss": 1.4945,
+      "step": 2602
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.610311486655163e-05,
+      "loss": 0.9259,
+      "step": 2603
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.610017548225874e-05,
+      "loss": 0.7662,
+      "step": 2604
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.609723506686957e-05,
+      "loss": 0.4011,
+      "step": 2605
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.609429362050029e-05,
+      "loss": 0.5497,
+      "step": 2606
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6091351143267114e-05,
+      "loss": 0.4994,
+      "step": 2607
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.608840763528628e-05,
+      "loss": 0.8098,
+      "step": 2608
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6085463096674055e-05,
+      "loss": 0.8168,
+      "step": 2609
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.608251752754679e-05,
+      "loss": 0.7585,
+      "step": 2610
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.607957092802082e-05,
+      "loss": 0.8301,
+      "step": 2611
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6076623298212586e-05,
+      "loss": 1.0009,
+      "step": 2612
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6073674638238506e-05,
+      "loss": 0.5298,
+      "step": 2613
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6070724948215096e-05,
+      "loss": 0.6872,
+      "step": 2614
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.606777422825885e-05,
+      "loss": 0.4239,
+      "step": 2615
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.606482247848636e-05,
+      "loss": 1.0141,
+      "step": 2616
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6061869699014246e-05,
+      "loss": 0.4917,
+      "step": 2617
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6058915889959134e-05,
+      "loss": 0.8281,
+      "step": 2618
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.605596105143773e-05,
+      "loss": 0.7479,
+      "step": 2619
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.605300518356675e-05,
+      "loss": 1.3429,
+      "step": 2620
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.605004828646299e-05,
+      "loss": 0.9133,
+      "step": 2621
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6047090360243255e-05,
+      "loss": 0.6668,
+      "step": 2622
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.604413140502438e-05,
+      "loss": 0.8339,
+      "step": 2623
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.604117142092329e-05,
+      "loss": 0.7596,
+      "step": 2624
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.60382104080569e-05,
+      "loss": 0.835,
+      "step": 2625
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.603524836654218e-05,
+      "loss": 0.8846,
+      "step": 2626
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6032285296496166e-05,
+      "loss": 0.7849,
+      "step": 2627
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.602932119803591e-05,
+      "loss": 1.4144,
+      "step": 2628
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.60263560712785e-05,
+      "loss": 0.7265,
+      "step": 2629
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.602338991634108e-05,
+      "loss": 0.4988,
+      "step": 2630
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.6020422733340834e-05,
+      "loss": 0.9521,
+      "step": 2631
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.601745452239497e-05,
+      "loss": 0.6448,
+      "step": 2632
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.601448528362076e-05,
+      "loss": 0.8554,
+      "step": 2633
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.60115150171355e-05,
+      "loss": 0.4319,
+      "step": 2634
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.6008543723056525e-05,
+      "loss": 0.7985,
+      "step": 2635
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.600557140150123e-05,
+      "loss": 0.2205,
+      "step": 2636
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.600259805258703e-05,
+      "loss": 0.3314,
+      "step": 2637
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5999623676431385e-05,
+      "loss": 0.8448,
+      "step": 2638
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5996648273151805e-05,
+      "loss": 0.6188,
+      "step": 2639
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.599367184286583e-05,
+      "loss": 0.801,
+      "step": 2640
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.599069438569104e-05,
+      "loss": 0.9416,
+      "step": 2641
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.598771590174508e-05,
+      "loss": 0.8165,
+      "step": 2642
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5984736391145596e-05,
+      "loss": 0.9796,
+      "step": 2643
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5981755854010314e-05,
+      "loss": 0.9144,
+      "step": 2644
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.597877429045696e-05,
+      "loss": 0.7698,
+      "step": 2645
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.597579170060333e-05,
+      "loss": 0.5247,
+      "step": 2646
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.597280808456725e-05,
+      "loss": 0.8622,
+      "step": 2647
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.596982344246659e-05,
+      "loss": 0.5959,
+      "step": 2648
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.596683777441927e-05,
+      "loss": 0.9834,
+      "step": 2649
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5963851080543225e-05,
+      "loss": 0.6929,
+      "step": 2650
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.596086336095646e-05,
+      "loss": 1.2457,
+      "step": 2651
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.595787461577698e-05,
+      "loss": 1.4643,
+      "step": 2652
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5954884845122886e-05,
+      "loss": 0.4714,
+      "step": 2653
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.595189404911227e-05,
+      "loss": 0.7982,
+      "step": 2654
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5948902227863295e-05,
+      "loss": 0.896,
+      "step": 2655
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5945909381494146e-05,
+      "loss": 1.1211,
+      "step": 2656
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.594291551012307e-05,
+      "loss": 0.9164,
+      "step": 2657
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.593992061386833e-05,
+      "loss": 0.792,
+      "step": 2658
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.593692469284823e-05,
+      "loss": 0.8595,
+      "step": 2659
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5933927747181157e-05,
+      "loss": 0.7249,
+      "step": 2660
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5930929776985473e-05,
+      "loss": 1.0078,
+      "step": 2661
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5927930782379626e-05,
+      "loss": 0.8958,
+      "step": 2662
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5924930763482104e-05,
+      "loss": 1.1392,
+      "step": 2663
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.592192972041141e-05,
+      "loss": 0.6773,
+      "step": 2664
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5918927653286105e-05,
+      "loss": 0.2168,
+      "step": 2665
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5915924562224776e-05,
+      "loss": 0.7126,
+      "step": 2666
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.591292044734608e-05,
+      "loss": 0.201,
+      "step": 2667
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.590991530876868e-05,
+      "loss": 0.5272,
+      "step": 2668
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5906909146611316e-05,
+      "loss": 0.3646,
+      "step": 2669
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.590390196099273e-05,
+      "loss": 0.6297,
+      "step": 2670
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.590089375203172e-05,
+      "loss": 0.8491,
+      "step": 2671
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.589788451984713e-05,
+      "loss": 0.9103,
+      "step": 2672
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.589487426455784e-05,
+      "loss": 0.9414,
+      "step": 2673
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5891862986282774e-05,
+      "loss": 1.302,
+      "step": 2674
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5888850685140905e-05,
+      "loss": 1.0952,
+      "step": 2675
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.588583736125121e-05,
+      "loss": 0.5646,
+      "step": 2676
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.588282301473275e-05,
+      "loss": 0.5328,
+      "step": 2677
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5879807645704606e-05,
+      "loss": 0.9143,
+      "step": 2678
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.587679125428589e-05,
+      "loss": 0.5563,
+      "step": 2679
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.587377384059578e-05,
+      "loss": 0.1785,
+      "step": 2680
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.587075540475347e-05,
+      "loss": 0.4303,
+      "step": 2681
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5867735946878206e-05,
+      "loss": 0.0728,
+      "step": 2682
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.586471546708927e-05,
+      "loss": 0.4494,
+      "step": 2683
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5861693965506004e-05,
+      "loss": 0.9872,
+      "step": 2684
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5858671442247755e-05,
+      "loss": 0.2863,
+      "step": 2685
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.585564789743393e-05,
+      "loss": 1.138,
+      "step": 2686
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.585262333118398e-05,
+      "loss": 0.9854,
+      "step": 2687
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.584959774361739e-05,
+      "loss": 0.7945,
+      "step": 2688
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.584657113485369e-05,
+      "loss": 0.6924,
+      "step": 2689
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.584354350501245e-05,
+      "loss": 0.3746,
+      "step": 2690
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5840514854213266e-05,
+      "loss": 0.4885,
+      "step": 2691
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.583748518257579e-05,
+      "loss": 0.8365,
+      "step": 2692
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.583445449021972e-05,
+      "loss": 0.4134,
+      "step": 2693
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.583142277726477e-05,
+      "loss": 0.8944,
+      "step": 2694
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5828390043830724e-05,
+      "loss": 0.9145,
+      "step": 2695
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.582535629003737e-05,
+      "loss": 0.7204,
+      "step": 2696
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.582232151600459e-05,
+      "loss": 0.828,
+      "step": 2697
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.581928572185224e-05,
+      "loss": 0.6473,
+      "step": 2698
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5816248907700267e-05,
+      "loss": 0.7084,
+      "step": 2699
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.581321107366864e-05,
+      "loss": 0.8451,
+      "step": 2700
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5810172219877366e-05,
+      "loss": 0.8896,
+      "step": 2701
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.580713234644651e-05,
+      "loss": 0.1794,
+      "step": 2702
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.580409145349614e-05,
+      "loss": 0.8617,
+      "step": 2703
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.58010495411464e-05,
+      "loss": 0.7904,
+      "step": 2704
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.579800660951746e-05,
+      "loss": 0.5144,
+      "step": 2705
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.579496265872954e-05,
+      "loss": 0.7133,
+      "step": 2706
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5791917688902884e-05,
+      "loss": 0.4214,
+      "step": 2707
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.578887170015779e-05,
+      "loss": 0.8009,
+      "step": 2708
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5785824692614575e-05,
+      "loss": 0.5799,
+      "step": 2709
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.578277666639363e-05,
+      "loss": 0.5378,
+      "step": 2710
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.577972762161537e-05,
+      "loss": 0.6473,
+      "step": 2711
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5776677558400224e-05,
+      "loss": 0.7238,
+      "step": 2712
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.577362647686871e-05,
+      "loss": 0.5207,
+      "step": 2713
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.577057437714136e-05,
+      "loss": 0.1684,
+      "step": 2714
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5767521259338736e-05,
+      "loss": 0.5498,
+      "step": 2715
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.576446712358146e-05,
+      "loss": 0.7326,
+      "step": 2716
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.576141196999018e-05,
+      "loss": 0.8034,
+      "step": 2717
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.575835579868561e-05,
+      "loss": 1.0225,
+      "step": 2718
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5755298609788464e-05,
+      "loss": 0.6405,
+      "step": 2719
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.575224040341952e-05,
+      "loss": 0.9315,
+      "step": 2720
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.574918117969961e-05,
+      "loss": 1.2511,
+      "step": 2721
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.574612093874957e-05,
+      "loss": 0.9338,
+      "step": 2722
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5743059680690306e-05,
+      "loss": 0.7168,
+      "step": 2723
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5739997405642754e-05,
+      "loss": 0.2641,
+      "step": 2724
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.573693411372789e-05,
+      "loss": 0.7783,
+      "step": 2725
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.573386980506673e-05,
+      "loss": 1.2056,
+      "step": 2726
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.573080447978032e-05,
+      "loss": 0.6101,
+      "step": 2727
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.572773813798977e-05,
+      "loss": 0.8378,
+      "step": 2728
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.572467077981622e-05,
+      "loss": 0.9987,
+      "step": 2729
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.572160240538083e-05,
+      "loss": 0.3702,
+      "step": 2730
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.571853301480483e-05,
+      "loss": 0.9587,
+      "step": 2731
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.571546260820948e-05,
+      "loss": 0.9841,
+      "step": 2732
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.571239118571607e-05,
+      "loss": 1.2965,
+      "step": 2733
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.570931874744593e-05,
+      "loss": 0.5222,
+      "step": 2734
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.570624529352046e-05,
+      "loss": 0.2083,
+      "step": 2735
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.570317082406106e-05,
+      "loss": 1.022,
+      "step": 2736
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5700095339189186e-05,
+      "loss": 0.417,
+      "step": 2737
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5697018839026346e-05,
+      "loss": 0.8354,
+      "step": 2738
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.569394132369407e-05,
+      "loss": 0.9717,
+      "step": 2739
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.569086279331396e-05,
+      "loss": 0.8938,
+      "step": 2740
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.56877832480076e-05,
+      "loss": 1.1888,
+      "step": 2741
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.568470268789667e-05,
+      "loss": 1.1929,
+      "step": 2742
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.568162111310286e-05,
+      "loss": 0.9341,
+      "step": 2743
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.567853852374791e-05,
+      "loss": 1.2538,
+      "step": 2744
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5675454919953614e-05,
+      "loss": 0.6754,
+      "step": 2745
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.567237030184176e-05,
+      "loss": 0.8345,
+      "step": 2746
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.566928466953423e-05,
+      "loss": 0.6313,
+      "step": 2747
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.566619802315292e-05,
+      "loss": 0.4971,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.566311036281976e-05,
+      "loss": 0.6619,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5660021688656736e-05,
+      "loss": 0.7883,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.565693200078588e-05,
+      "loss": 0.6715,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5653841299329224e-05,
+      "loss": 0.8304,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.565074958440888e-05,
+      "loss": 0.4299,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.564765685614699e-05,
+      "loss": 0.747,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.564456311466573e-05,
+      "loss": 0.4458,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5641468360087324e-05,
+      "loss": 0.6672,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.563837259253401e-05,
+      "loss": 0.7994,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.563527581212813e-05,
+      "loss": 0.8134,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.563217801899198e-05,
+      "loss": 0.529,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.562907921324797e-05,
+      "loss": 0.7182,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.562597939501851e-05,
+      "loss": 0.6786,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.562287856442604e-05,
+      "loss": 0.5928,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.561977672159309e-05,
+      "loss": 0.4252,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.561667386664218e-05,
+      "loss": 0.7723,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.561356999969589e-05,
+      "loss": 0.3987,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.561046512087684e-05,
+      "loss": 0.8957,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.560735923030771e-05,
+      "loss": 0.4647,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.560425232811117e-05,
+      "loss": 0.711,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5601144414409974e-05,
+      "loss": 0.7016,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.559803548932691e-05,
+      "loss": 0.775,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.559492555298477e-05,
+      "loss": 0.513,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.559181460550644e-05,
+      "loss": 0.0907,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.55887026470148e-05,
+      "loss": 1.0756,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.55855896776328e-05,
+      "loss": 0.1336,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5582475697483416e-05,
+      "loss": 0.6336,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.557936070668967e-05,
+      "loss": 0.923,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.557624470537461e-05,
+      "loss": 0.5313,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.557312769366134e-05,
+      "loss": 0.8948,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5570009671673007e-05,
+      "loss": 0.7485,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.556689063953279e-05,
+      "loss": 0.9683,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5563770597363883e-05,
+      "loss": 0.7147,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.556064954528958e-05,
+      "loss": 0.4458,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.555752748343315e-05,
+      "loss": 0.7152,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.555440441191795e-05,
+      "loss": 0.8594,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.555128033086734e-05,
+      "loss": 1.1122,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.554815524040476e-05,
+      "loss": 0.5213,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.554502914065365e-05,
+      "loss": 0.63,
+      "step": 2787
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.554190203173751e-05,
+      "loss": 0.2924,
+      "step": 2788
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5538773913779886e-05,
+      "loss": 0.3763,
+      "step": 2789
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.553564478690435e-05,
+      "loss": 1.1953,
+      "step": 2790
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.553251465123452e-05,
+      "loss": 0.7393,
+      "step": 2791
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5529383506894064e-05,
+      "loss": 0.4953,
+      "step": 2792
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.552625135400665e-05,
+      "loss": 0.8053,
+      "step": 2793
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.552311819269604e-05,
+      "loss": 0.7984,
+      "step": 2794
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5519984023086e-05,
+      "loss": 0.6323,
+      "step": 2795
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.551684884530035e-05,
+      "loss": 0.5494,
+      "step": 2796
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5513712659462945e-05,
+      "loss": 0.8841,
+      "step": 2797
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5510575465697686e-05,
+      "loss": 0.2233,
+      "step": 2798
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.55074372641285e-05,
+      "loss": 0.5584,
+      "step": 2799
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.550429805487938e-05,
+      "loss": 0.8817,
+      "step": 2800
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.7403446435928345,
+      "eval_runtime": 38.6491,
+      "eval_samples_per_second": 1.035,
+      "eval_steps_per_second": 1.035,
+      "step": 2800
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.550115783807431e-05,
+      "loss": 0.4584,
+      "step": 2801
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.549801661383738e-05,
+      "loss": 0.4966,
+      "step": 2802
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.549487438229265e-05,
+      "loss": 0.6471,
+      "step": 2803
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.549173114356429e-05,
+      "loss": 0.7103,
+      "step": 2804
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5488586897776466e-05,
+      "loss": 0.8751,
+      "step": 2805
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.548544164505337e-05,
+      "loss": 0.7686,
+      "step": 2806
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5482295385519276e-05,
+      "loss": 0.5753,
+      "step": 2807
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.547914811929848e-05,
+      "loss": 0.8469,
+      "step": 2808
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.54759998465153e-05,
+      "loss": 0.6103,
+      "step": 2809
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.547285056729412e-05,
+      "loss": 0.8992,
+      "step": 2810
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5469700281759366e-05,
+      "loss": 0.6271,
+      "step": 2811
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5466548990035466e-05,
+      "loss": 0.8411,
+      "step": 2812
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.546339669224692e-05,
+      "loss": 0.8199,
+      "step": 2813
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5460243388518284e-05,
+      "loss": 0.3338,
+      "step": 2814
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.545708907897409e-05,
+      "loss": 0.6047,
+      "step": 2815
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5453933763738985e-05,
+      "loss": 0.7361,
+      "step": 2816
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.54507774429376e-05,
+      "loss": 0.3097,
+      "step": 2817
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.544762011669465e-05,
+      "loss": 0.7856,
+      "step": 2818
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.544446178513483e-05,
+      "loss": 0.6044,
+      "step": 2819
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.544130244838294e-05,
+      "loss": 0.4443,
+      "step": 2820
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.543814210656378e-05,
+      "loss": 0.8526,
+      "step": 2821
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.543498075980219e-05,
+      "loss": 0.5445,
+      "step": 2822
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.543181840822309e-05,
+      "loss": 0.8913,
+      "step": 2823
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5428655051951394e-05,
+      "loss": 0.9649,
+      "step": 2824
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.542549069111206e-05,
+      "loss": 0.6135,
+      "step": 2825
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.542232532583011e-05,
+      "loss": 0.886,
+      "step": 2826
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.541915895623058e-05,
+      "loss": 0.4568,
+      "step": 2827
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.541599158243858e-05,
+      "loss": 0.9995,
+      "step": 2828
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.541282320457922e-05,
+      "loss": 0.4056,
+      "step": 2829
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.540965382277768e-05,
+      "loss": 0.8204,
+      "step": 2830
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.540648343715916e-05,
+      "loss": 0.5762,
+      "step": 2831
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.54033120478489e-05,
+      "loss": 1.3645,
+      "step": 2832
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.54001396549722e-05,
+      "loss": 1.133,
+      "step": 2833
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5396966258654384e-05,
+      "loss": 0.8179,
+      "step": 2834
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5393791859020815e-05,
+      "loss": 0.5222,
+      "step": 2835
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5390616456196904e-05,
+      "loss": 1.0603,
+      "step": 2836
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5387440050308086e-05,
+      "loss": 1.1159,
+      "step": 2837
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.538426264147985e-05,
+      "loss": 0.608,
+      "step": 2838
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.538108422983773e-05,
+      "loss": 0.5927,
+      "step": 2839
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.537790481550728e-05,
+      "loss": 0.5505,
+      "step": 2840
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.53747243986141e-05,
+      "loss": 0.6466,
+      "step": 2841
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.537154297928385e-05,
+      "loss": 0.9664,
+      "step": 2842
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5368360557642203e-05,
+      "loss": 0.3492,
+      "step": 2843
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5365177133814875e-05,
+      "loss": 0.3178,
+      "step": 2844
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5361992707927634e-05,
+      "loss": 0.4977,
+      "step": 2845
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.535880728010629e-05,
+      "loss": 0.6525,
+      "step": 2846
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.535562085047667e-05,
+      "loss": 0.6099,
+      "step": 2847
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5352433419164675e-05,
+      "loss": 0.7919,
+      "step": 2848
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.53492449862962e-05,
+      "loss": 0.8781,
+      "step": 2849
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.534605555199722e-05,
+      "loss": 0.6365,
+      "step": 2850
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.534286511639373e-05,
+      "loss": 0.8928,
+      "step": 2851
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.533967367961178e-05,
+      "loss": 0.9464,
+      "step": 2852
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.533648124177744e-05,
+      "loss": 0.7186,
+      "step": 2853
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.533328780301682e-05,
+      "loss": 0.2892,
+      "step": 2854
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.533009336345609e-05,
+      "loss": 0.8736,
+      "step": 2855
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.532689792322145e-05,
+      "loss": 0.4926,
+      "step": 2856
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.532370148243913e-05,
+      "loss": 1.0912,
+      "step": 2857
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.532050404123541e-05,
+      "loss": 0.8584,
+      "step": 2858
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.53173055997366e-05,
+      "loss": 0.623,
+      "step": 2859
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5314106158069056e-05,
+      "loss": 0.5798,
+      "step": 2860
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5310905716359175e-05,
+      "loss": 1.0203,
+      "step": 2861
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5307704274733405e-05,
+      "loss": 0.5275,
+      "step": 2862
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.53045018333182e-05,
+      "loss": 0.4693,
+      "step": 2863
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.530129839224008e-05,
+      "loss": 0.1431,
+      "step": 2864
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.52980939516256e-05,
+      "loss": 0.8735,
+      "step": 2865
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.529488851160135e-05,
+      "loss": 0.219,
+      "step": 2866
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.529168207229397e-05,
+      "loss": 0.7407,
+      "step": 2867
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.528847463383012e-05,
+      "loss": 0.9038,
+      "step": 2868
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.528526619633652e-05,
+      "loss": 0.5467,
+      "step": 2869
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.528205675993992e-05,
+      "loss": 0.1654,
+      "step": 2870
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.52788463247671e-05,
+      "loss": 0.4749,
+      "step": 2871
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.527563489094489e-05,
+      "loss": 0.6795,
+      "step": 2872
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5272422458600175e-05,
+      "loss": 0.4224,
+      "step": 2873
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5269209027859856e-05,
+      "loss": 0.7675,
+      "step": 2874
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.526599459885087e-05,
+      "loss": 0.3677,
+      "step": 2875
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.526277917170021e-05,
+      "loss": 0.6892,
+      "step": 2876
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5259562746534905e-05,
+      "loss": 0.7742,
+      "step": 2877
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.525634532348203e-05,
+      "loss": 0.3498,
+      "step": 2878
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5253126902668666e-05,
+      "loss": 0.4194,
+      "step": 2879
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.524990748422198e-05,
+      "loss": 0.8409,
+      "step": 2880
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.524668706826915e-05,
+      "loss": 0.1791,
+      "step": 2881
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.524346565493739e-05,
+      "loss": 0.6634,
+      "step": 2882
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.524024324435399e-05,
+      "loss": 0.2235,
+      "step": 2883
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.523701983664622e-05,
+      "loss": 0.5054,
+      "step": 2884
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5233795431941435e-05,
+      "loss": 0.7186,
+      "step": 2885
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5230570030367016e-05,
+      "loss": 0.7078,
+      "step": 2886
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.522734363205039e-05,
+      "loss": 0.4584,
+      "step": 2887
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5224116237119e-05,
+      "loss": 0.488,
+      "step": 2888
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5220887845700364e-05,
+      "loss": 0.806,
+      "step": 2889
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5217658457922013e-05,
+      "loss": 0.8494,
+      "step": 2890
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5214428073911534e-05,
+      "loss": 0.1994,
+      "step": 2891
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.521119669379653e-05,
+      "loss": 0.4935,
+      "step": 2892
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5207964317704656e-05,
+      "loss": 0.5873,
+      "step": 2893
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.520473094576363e-05,
+      "loss": 0.7344,
+      "step": 2894
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.520149657810116e-05,
+      "loss": 0.717,
+      "step": 2895
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.519826121484503e-05,
+      "loss": 1.1473,
+      "step": 2896
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.519502485612307e-05,
+      "loss": 0.863,
+      "step": 2897
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.519178750206312e-05,
+      "loss": 1.2885,
+      "step": 2898
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.518854915279307e-05,
+      "loss": 1.1247,
+      "step": 2899
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.518530980844086e-05,
+      "loss": 0.8393,
+      "step": 2900
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.518206946913446e-05,
+      "loss": 0.8985,
+      "step": 2901
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5178828135001864e-05,
+      "loss": 0.9599,
+      "step": 2902
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.517558580617115e-05,
+      "loss": 0.9212,
+      "step": 2903
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.517234248277039e-05,
+      "loss": 0.4637,
+      "step": 2904
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.516909816492771e-05,
+      "loss": 0.7557,
+      "step": 2905
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.516585285277129e-05,
+      "loss": 0.844,
+      "step": 2906
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.516260654642934e-05,
+      "loss": 0.8947,
+      "step": 2907
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5159359246030085e-05,
+      "loss": 0.0911,
+      "step": 2908
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.515611095170183e-05,
+      "loss": 0.9027,
+      "step": 2909
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.515286166357289e-05,
+      "loss": 0.8989,
+      "step": 2910
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.514961138177163e-05,
+      "loss": 1.4128,
+      "step": 2911
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.514636010642646e-05,
+      "loss": 0.5964,
+      "step": 2912
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.514310783766582e-05,
+      "loss": 0.5549,
+      "step": 2913
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5139854575618185e-05,
+      "loss": 2.347,
+      "step": 2914
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.513660032041209e-05,
+      "loss": 0.7151,
+      "step": 2915
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.513334507217608e-05,
+      "loss": 0.2195,
+      "step": 2916
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.513008883103877e-05,
+      "loss": 1.1179,
+      "step": 2917
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.512683159712879e-05,
+      "loss": 0.5824,
+      "step": 2918
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.512357337057481e-05,
+      "loss": 0.8829,
+      "step": 2919
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5120314151505564e-05,
+      "loss": 0.3276,
+      "step": 2920
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.51170539400498e-05,
+      "loss": 1.051,
+      "step": 2921
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.511379273633632e-05,
+      "loss": 1.0756,
+      "step": 2922
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.511053054049394e-05,
+      "loss": 0.8713,
+      "step": 2923
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.510726735265156e-05,
+      "loss": 1.2761,
+      "step": 2924
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.510400317293808e-05,
+      "loss": 0.9131,
+      "step": 2925
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.510073800148245e-05,
+      "loss": 0.5954,
+      "step": 2926
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.509747183841366e-05,
+      "loss": 0.8177,
+      "step": 2927
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.509420468386076e-05,
+      "loss": 1.2571,
+      "step": 2928
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5090936537952796e-05,
+      "loss": 0.4041,
+      "step": 2929
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.508766740081889e-05,
+      "loss": 0.7195,
+      "step": 2930
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5084397272588195e-05,
+      "loss": 1.1907,
+      "step": 2931
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.508112615338988e-05,
+      "loss": 0.2312,
+      "step": 2932
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.50778540433532e-05,
+      "loss": 0.3552,
+      "step": 2933
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.507458094260739e-05,
+      "loss": 0.8225,
+      "step": 2934
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.507130685128177e-05,
+      "loss": 1.0851,
+      "step": 2935
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.506803176950568e-05,
+      "loss": 0.6033,
+      "step": 2936
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.506475569740851e-05,
+      "loss": 0.3744,
+      "step": 2937
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.506147863511969e-05,
+      "loss": 0.5638,
+      "step": 2938
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5058200582768664e-05,
+      "loss": 0.7863,
+      "step": 2939
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5054921540484925e-05,
+      "loss": 0.8681,
+      "step": 2940
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.505164150839804e-05,
+      "loss": 0.5817,
+      "step": 2941
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.504836048663757e-05,
+      "loss": 0.4994,
+      "step": 2942
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.504507847533314e-05,
+      "loss": 1.1053,
+      "step": 2943
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.50417954746144e-05,
+      "loss": 0.2764,
+      "step": 2944
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.503851148461105e-05,
+      "loss": 0.3483,
+      "step": 2945
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.503522650545283e-05,
+      "loss": 0.5396,
+      "step": 2946
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.50319405372695e-05,
+      "loss": 0.5045,
+      "step": 2947
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.502865358019089e-05,
+      "loss": 0.4548,
+      "step": 2948
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.502536563434685e-05,
+      "loss": 0.713,
+      "step": 2949
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.502207669986726e-05,
+      "loss": 0.5257,
+      "step": 2950
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5018786776882056e-05,
+      "loss": 0.0691,
+      "step": 2951
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5015495865521205e-05,
+      "loss": 0.6204,
+      "step": 2952
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.501220396591472e-05,
+      "loss": 1.0191,
+      "step": 2953
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.500891107819265e-05,
+      "loss": 0.5814,
+      "step": 2954
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.500561720248508e-05,
+      "loss": 0.9147,
+      "step": 2955
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.500232233892215e-05,
+      "loss": 0.2704,
+      "step": 2956
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.499902648763399e-05,
+      "loss": 0.5607,
+      "step": 2957
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.499572964875083e-05,
+      "loss": 0.1052,
+      "step": 2958
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.499243182240289e-05,
+      "loss": 0.5387,
+      "step": 2959
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.498913300872049e-05,
+      "loss": 1.0517,
+      "step": 2960
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.498583320783392e-05,
+      "loss": 0.9147,
+      "step": 2961
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.498253241987355e-05,
+      "loss": 0.5889,
+      "step": 2962
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.497923064496978e-05,
+      "loss": 0.4055,
+      "step": 2963
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4975927883253035e-05,
+      "loss": 1.0599,
+      "step": 2964
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.497262413485381e-05,
+      "loss": 0.8497,
+      "step": 2965
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4969319399902606e-05,
+      "loss": 1.0238,
+      "step": 2966
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4966013678529984e-05,
+      "loss": 0.6329,
+      "step": 2967
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4962706970866546e-05,
+      "loss": 0.4699,
+      "step": 2968
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.495939927704291e-05,
+      "loss": 0.9966,
+      "step": 2969
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.495609059718976e-05,
+      "loss": 0.7012,
+      "step": 2970
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4952780931437794e-05,
+      "loss": 0.561,
+      "step": 2971
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.494947027991776e-05,
+      "loss": 1.1056,
+      "step": 2972
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.494615864276047e-05,
+      "loss": 1.3964,
+      "step": 2973
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.494284602009672e-05,
+      "loss": 0.5928,
+      "step": 2974
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4939532412057404e-05,
+      "loss": 0.2823,
+      "step": 2975
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.493621781877341e-05,
+      "loss": 1.0105,
+      "step": 2976
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.493290224037569e-05,
+      "loss": 0.4167,
+      "step": 2977
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.492958567699522e-05,
+      "loss": 0.2757,
+      "step": 2978
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.492626812876303e-05,
+      "loss": 0.9337,
+      "step": 2979
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4922949595810175e-05,
+      "loss": 0.4713,
+      "step": 2980
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.491963007826775e-05,
+      "loss": 1.0502,
+      "step": 2981
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.491630957626692e-05,
+      "loss": 0.5147,
+      "step": 2982
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.491298808993883e-05,
+      "loss": 0.5897,
+      "step": 2983
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.49096656194147e-05,
+      "loss": 0.5244,
+      "step": 2984
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.490634216482581e-05,
+      "loss": 1.1204,
+      "step": 2985
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.490301772630343e-05,
+      "loss": 0.9476,
+      "step": 2986
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.48996923039789e-05,
+      "loss": 0.4735,
+      "step": 2987
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.48963658979836e-05,
+      "loss": 0.5072,
+      "step": 2988
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4893038508448937e-05,
+      "loss": 0.757,
+      "step": 2989
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.488971013550636e-05,
+      "loss": 0.5013,
+      "step": 2990
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.488638077928734e-05,
+      "loss": 0.3742,
+      "step": 2991
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4883050439923434e-05,
+      "loss": 0.3561,
+      "step": 2992
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.487971911754619e-05,
+      "loss": 0.505,
+      "step": 2993
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.487638681228722e-05,
+      "loss": 0.264,
+      "step": 2994
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4873053524278164e-05,
+      "loss": 0.8168,
+      "step": 2995
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4869719253650704e-05,
+      "loss": 1.0818,
+      "step": 2996
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.486638400053656e-05,
+      "loss": 0.5114,
+      "step": 2997
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.48630477650675e-05,
+      "loss": 0.8417,
+      "step": 2998
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.485971054737532e-05,
+      "loss": 0.4379,
+      "step": 2999
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.485637234759186e-05,
+      "loss": 0.5023,
+      "step": 3000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.7280268669128418,
+      "eval_runtime": 38.6284,
+      "eval_samples_per_second": 1.036,
+      "eval_steps_per_second": 1.036,
+      "step": 3000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.485303316584899e-05,
+      "loss": 0.9588,
+      "step": 3001
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.484969300227863e-05,
+      "loss": 0.5124,
+      "step": 3002
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4846351857012727e-05,
+      "loss": 1.2751,
+      "step": 3003
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.484300973018329e-05,
+      "loss": 0.9571,
+      "step": 3004
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.483966662192233e-05,
+      "loss": 0.2591,
+      "step": 3005
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.483632253236194e-05,
+      "loss": 0.4209,
+      "step": 3006
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.483297746163421e-05,
+      "loss": 0.5461,
+      "step": 3007
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.48296314098713e-05,
+      "loss": 0.8433,
+      "step": 3008
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.48262843772054e-05,
+      "loss": 0.6279,
+      "step": 3009
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.482293636376871e-05,
+      "loss": 0.9299,
+      "step": 3010
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.481958736969353e-05,
+      "loss": 0.8793,
+      "step": 3011
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.481623739511214e-05,
+      "loss": 0.3524,
+      "step": 3012
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.481288644015689e-05,
+      "loss": 0.8381,
+      "step": 3013
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.480953450496016e-05,
+      "loss": 1.0984,
+      "step": 3014
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.480618158965437e-05,
+      "loss": 0.4202,
+      "step": 3015
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4802827694371966e-05,
+      "loss": 0.4368,
+      "step": 3016
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.479947281924546e-05,
+      "loss": 0.5632,
+      "step": 3017
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4796116964407386e-05,
+      "loss": 0.8584,
+      "step": 3018
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.479276012999032e-05,
+      "loss": 0.5602,
+      "step": 3019
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4789402316126865e-05,
+      "loss": 0.3246,
+      "step": 3020
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4786043522949676e-05,
+      "loss": 0.444,
+      "step": 3021
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.478268375059144e-05,
+      "loss": 0.6322,
+      "step": 3022
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4779322999184885e-05,
+      "loss": 0.4996,
+      "step": 3023
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.47759612688628e-05,
+      "loss": 0.9241,
+      "step": 3024
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.477259855975797e-05,
+      "loss": 0.3668,
+      "step": 3025
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4769234872003245e-05,
+      "loss": 0.9836,
+      "step": 3026
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.476587020573151e-05,
+      "loss": 0.4652,
+      "step": 3027
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.476250456107568e-05,
+      "loss": 0.7807,
+      "step": 3028
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.475913793816873e-05,
+      "loss": 0.3978,
+      "step": 3029
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.475577033714365e-05,
+      "loss": 0.4851,
+      "step": 3030
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4752401758133476e-05,
+      "loss": 0.7568,
+      "step": 3031
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.474903220127129e-05,
+      "loss": 0.8829,
+      "step": 3032
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.474566166669021e-05,
+      "loss": 0.8943,
+      "step": 3033
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.474229015452338e-05,
+      "loss": 1.3107,
+      "step": 3034
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4738917664904e-05,
+      "loss": 0.3767,
+      "step": 3035
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4735544197965306e-05,
+      "loss": 0.7276,
+      "step": 3036
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.473216975384055e-05,
+      "loss": 1.0727,
+      "step": 3037
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.472879433266307e-05,
+      "loss": 1.0051,
+      "step": 3038
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.472541793456618e-05,
+      "loss": 0.2108,
+      "step": 3039
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.472204055968329e-05,
+      "loss": 1.3184,
+      "step": 3040
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.471866220814782e-05,
+      "loss": 0.704,
+      "step": 3041
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4715282880093224e-05,
+      "loss": 0.1048,
+      "step": 3042
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.471190257565301e-05,
+      "loss": 0.7021,
+      "step": 3043
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4708521294960715e-05,
+      "loss": 0.7052,
+      "step": 3044
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.470513903814992e-05,
+      "loss": 0.5723,
+      "step": 3045
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4701755805354246e-05,
+      "loss": 0.1213,
+      "step": 3046
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4698371596707334e-05,
+      "loss": 0.8055,
+      "step": 3047
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.46949864123429e-05,
+      "loss": 0.1421,
+      "step": 3048
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.469160025239466e-05,
+      "loss": 0.5678,
+      "step": 3049
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.46882131169964e-05,
+      "loss": 0.3878,
+      "step": 3050
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4684825006281905e-05,
+      "loss": 0.1688,
+      "step": 3051
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.468143592038505e-05,
+      "loss": 0.9611,
+      "step": 3052
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.467804585943971e-05,
+      "loss": 0.5776,
+      "step": 3053
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4674654823579805e-05,
+      "loss": 0.9621,
+      "step": 3054
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.467126281293931e-05,
+      "loss": 0.976,
+      "step": 3055
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.466786982765223e-05,
+      "loss": 0.9865,
+      "step": 3056
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.466447586785259e-05,
+      "loss": 0.4811,
+      "step": 3057
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.466108093367448e-05,
+      "loss": 0.98,
+      "step": 3058
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.465768502525202e-05,
+      "loss": 1.0707,
+      "step": 3059
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.465428814271937e-05,
+      "loss": 0.5524,
+      "step": 3060
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4650890286210706e-05,
+      "loss": 1.6206,
+      "step": 3061
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.464749145586028e-05,
+      "loss": 0.4653,
+      "step": 3062
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.464409165180235e-05,
+      "loss": 0.8268,
+      "step": 3063
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4640690874171245e-05,
+      "loss": 0.8332,
+      "step": 3064
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.46372891231013e-05,
+      "loss": 0.5488,
+      "step": 3065
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4633886398726896e-05,
+      "loss": 1.2634,
+      "step": 3066
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.463048270118248e-05,
+      "loss": 0.925,
+      "step": 3067
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.46270780306025e-05,
+      "loss": 0.6362,
+      "step": 3068
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.462367238712146e-05,
+      "loss": 0.5082,
+      "step": 3069
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.462026577087392e-05,
+      "loss": 0.5567,
+      "step": 3070
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.461685818199442e-05,
+      "loss": 1.028,
+      "step": 3071
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.461344962061762e-05,
+      "loss": 0.892,
+      "step": 3072
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.461004008687815e-05,
+      "loss": 1.2561,
+      "step": 3073
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4606629580910705e-05,
+      "loss": 0.8722,
+      "step": 3074
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4603218102850036e-05,
+      "loss": 0.4266,
+      "step": 3075
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4599805652830904e-05,
+      "loss": 0.8097,
+      "step": 3076
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4596392230988124e-05,
+      "loss": 1.0272,
+      "step": 3077
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4592977837456536e-05,
+      "loss": 0.6401,
+      "step": 3078
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.458956247237103e-05,
+      "loss": 0.6202,
+      "step": 3079
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.458614613586653e-05,
+      "loss": 0.889,
+      "step": 3080
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.458272882807801e-05,
+      "loss": 0.6812,
+      "step": 3081
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.457931054914045e-05,
+      "loss": 1.0657,
+      "step": 3082
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.457589129918891e-05,
+      "loss": 0.4634,
+      "step": 3083
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.457247107835847e-05,
+      "loss": 0.8467,
+      "step": 3084
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.456904988678422e-05,
+      "loss": 0.5685,
+      "step": 3085
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4565627724601354e-05,
+      "loss": 0.814,
+      "step": 3086
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.456220459194504e-05,
+      "loss": 0.688,
+      "step": 3087
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.455878048895051e-05,
+      "loss": 0.6196,
+      "step": 3088
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.455535541575305e-05,
+      "loss": 1.0224,
+      "step": 3089
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.455192937248795e-05,
+      "loss": 0.6827,
+      "step": 3090
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4548502359290566e-05,
+      "loss": 1.0978,
+      "step": 3091
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4545074376296294e-05,
+      "loss": 0.3153,
+      "step": 3092
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.454164542364054e-05,
+      "loss": 0.4346,
+      "step": 3093
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.453821550145877e-05,
+      "loss": 0.4797,
+      "step": 3094
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.453478460988649e-05,
+      "loss": 0.58,
+      "step": 3095
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.453135274905925e-05,
+      "loss": 0.3789,
+      "step": 3096
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.452791991911259e-05,
+      "loss": 1.201,
+      "step": 3097
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.452448612018217e-05,
+      "loss": 0.7043,
+      "step": 3098
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.45210513524036e-05,
+      "loss": 0.8728,
+      "step": 3099
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4517615615912603e-05,
+      "loss": 0.4016,
+      "step": 3100
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.45141789108449e-05,
+      "loss": 0.288,
+      "step": 3101
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.451074123733626e-05,
+      "loss": 0.0865,
+      "step": 3102
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4507302595522476e-05,
+      "loss": 1.0665,
+      "step": 3103
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.450386298553942e-05,
+      "loss": 0.811,
+      "step": 3104
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4500422407522944e-05,
+      "loss": 0.5559,
+      "step": 3105
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4496980861609e-05,
+      "loss": 0.8713,
+      "step": 3106
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4493538347933515e-05,
+      "loss": 0.7921,
+      "step": 3107
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4490094866632516e-05,
+      "loss": 0.7815,
+      "step": 3108
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.448665041784202e-05,
+      "loss": 0.8483,
+      "step": 3109
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.448320500169811e-05,
+      "loss": 0.7779,
+      "step": 3110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4479758618336896e-05,
+      "loss": 0.361,
+      "step": 3111
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.447631126789453e-05,
+      "loss": 0.9717,
+      "step": 3112
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.44728629505072e-05,
+      "loss": 0.4266,
+      "step": 3113
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.446941366631112e-05,
+      "loss": 0.593,
+      "step": 3114
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.446596341544258e-05,
+      "loss": 0.5754,
+      "step": 3115
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4462512198037864e-05,
+      "loss": 0.9578,
+      "step": 3116
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4459060014233324e-05,
+      "loss": 1.0044,
+      "step": 3117
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.445560686416534e-05,
+      "loss": 0.7558,
+      "step": 3118
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.445215274797031e-05,
+      "loss": 0.9069,
+      "step": 3119
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.444869766578472e-05,
+      "loss": 0.4702,
+      "step": 3120
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.444524161774504e-05,
+      "loss": 0.556,
+      "step": 3121
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.444178460398783e-05,
+      "loss": 1.0173,
+      "step": 3122
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.443832662464962e-05,
+      "loss": 0.9782,
+      "step": 3123
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.443486767986705e-05,
+      "loss": 0.7552,
+      "step": 3124
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.443140776977677e-05,
+      "loss": 0.7535,
+      "step": 3125
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.442794689451543e-05,
+      "loss": 0.1883,
+      "step": 3126
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.442448505421979e-05,
+      "loss": 0.3214,
+      "step": 3127
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.44210222490266e-05,
+      "loss": 0.5157,
+      "step": 3128
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.441755847907265e-05,
+      "loss": 0.4972,
+      "step": 3129
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.441409374449478e-05,
+      "loss": 0.8394,
+      "step": 3130
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.441062804542988e-05,
+      "loss": 0.6961,
+      "step": 3131
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.440716138201485e-05,
+      "loss": 0.9062,
+      "step": 3132
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.440369375438664e-05,
+      "loss": 0.6021,
+      "step": 3133
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.440022516268225e-05,
+      "loss": 1.0425,
+      "step": 3134
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.43967556070387e-05,
+      "loss": 0.4365,
+      "step": 3135
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.439328508759306e-05,
+      "loss": 0.5547,
+      "step": 3136
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4389813604482435e-05,
+      "loss": 0.695,
+      "step": 3137
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4386341157843956e-05,
+      "loss": 1.227,
+      "step": 3138
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4382867747814815e-05,
+      "loss": 0.8937,
+      "step": 3139
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.437939337453223e-05,
+      "loss": 0.5649,
+      "step": 3140
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4375918038133455e-05,
+      "loss": 0.4458,
+      "step": 3141
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4372441738755784e-05,
+      "loss": 0.6665,
+      "step": 3142
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.436896447653654e-05,
+      "loss": 0.4964,
+      "step": 3143
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.436548625161311e-05,
+      "loss": 1.4488,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.43620070641229e-05,
+      "loss": 0.9119,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.435852691420334e-05,
+      "loss": 0.9003,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.435504580199193e-05,
+      "loss": 0.8223,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4351563727626186e-05,
+      "loss": 0.5377,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.434808069124368e-05,
+      "loss": 0.2301,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4344596692982e-05,
+      "loss": 0.3322,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4341111732978766e-05,
+      "loss": 0.4061,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.433762581137169e-05,
+      "loss": 0.556,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4334138928298456e-05,
+      "loss": 0.5519,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.433065108389683e-05,
+      "loss": 0.3453,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.432716227830459e-05,
+      "loss": 0.7426,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.432367251165957e-05,
+      "loss": 0.674,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.432018178409962e-05,
+      "loss": 0.9096,
+      "step": 3157
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.431669009576266e-05,
+      "loss": 0.8107,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.431319744678663e-05,
+      "loss": 0.6386,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.43097038373095e-05,
+      "loss": 0.5591,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.430620926746928e-05,
+      "loss": 0.6914,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4302713737404034e-05,
+      "loss": 1.0227,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4299217247251855e-05,
+      "loss": 0.8136,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4295719797150867e-05,
+      "loss": 0.581,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.429222138723925e-05,
+      "loss": 1.9821,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.42887220176552e-05,
+      "loss": 0.6194,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.428522168853695e-05,
+      "loss": 0.8181,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.428172040002281e-05,
+      "loss": 0.6897,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.427821815225107e-05,
+      "loss": 0.8814,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.42747149453601e-05,
+      "loss": 0.6303,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.42712107794883e-05,
+      "loss": 0.883,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.426770565477411e-05,
+      "loss": 0.7148,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4264199571355976e-05,
+      "loss": 1.0046,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.426069252937243e-05,
+      "loss": 1.161,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4257184528962e-05,
+      "loss": 0.851,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4253675570263296e-05,
+      "loss": 0.85,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.425016565341491e-05,
+      "loss": 1.4817,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.424665477855553e-05,
+      "loss": 0.3775,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4243142945823834e-05,
+      "loss": 0.392,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.423963015535858e-05,
+      "loss": 1.0479,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.423611640729851e-05,
+      "loss": 0.3936,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.423260170178246e-05,
+      "loss": 0.2337,
+      "step": 3182
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.422908603894928e-05,
+      "loss": 0.6632,
+      "step": 3183
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4225569418937854e-05,
+      "loss": 0.9378,
+      "step": 3184
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.42220518418871e-05,
+      "loss": 0.819,
+      "step": 3185
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.421853330793598e-05,
+      "loss": 0.4915,
+      "step": 3186
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.421501381722351e-05,
+      "loss": 0.7667,
+      "step": 3187
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.421149336988872e-05,
+      "loss": 0.1432,
+      "step": 3188
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4207971966070676e-05,
+      "loss": 0.5431,
+      "step": 3189
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.420444960590851e-05,
+      "loss": 0.5955,
+      "step": 3190
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.420092628954136e-05,
+      "loss": 0.616,
+      "step": 3191
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.419740201710842e-05,
+      "loss": 0.4043,
+      "step": 3192
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.419387678874893e-05,
+      "loss": 0.4643,
+      "step": 3193
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.419035060460215e-05,
+      "loss": 0.2962,
+      "step": 3194
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4186823464807355e-05,
+      "loss": 0.4782,
+      "step": 3195
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4183295369503933e-05,
+      "loss": 0.7512,
+      "step": 3196
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.417976631883123e-05,
+      "loss": 0.8741,
+      "step": 3197
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.417623631292866e-05,
+      "loss": 0.7446,
+      "step": 3198
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4172705351935695e-05,
+      "loss": 0.4414,
+      "step": 3199
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.416917343599182e-05,
+      "loss": 0.6488,
+      "step": 3200
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.7208492159843445,
+      "eval_runtime": 38.2967,
+      "eval_samples_per_second": 1.044,
+      "eval_steps_per_second": 1.044,
+      "step": 3200
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.416564056523657e-05,
+      "loss": 1.018,
+      "step": 3201
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.41621067398095e-05,
+      "loss": 0.7746,
+      "step": 3202
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.415857195985022e-05,
+      "loss": 0.7934,
+      "step": 3203
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4155036225498376e-05,
+      "loss": 0.3417,
+      "step": 3204
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.415149953689365e-05,
+      "loss": 0.706,
+      "step": 3205
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.414796189417575e-05,
+      "loss": 0.9483,
+      "step": 3206
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.414442329748444e-05,
+      "loss": 0.3318,
+      "step": 3207
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.414088374695951e-05,
+      "loss": 0.6205,
+      "step": 3208
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4137343242740796e-05,
+      "loss": 0.9273,
+      "step": 3209
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.413380178496817e-05,
+      "loss": 0.3018,
+      "step": 3210
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.413025937378152e-05,
+      "loss": 0.4125,
+      "step": 3211
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4126716009320807e-05,
+      "loss": 0.4286,
+      "step": 3212
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.412317169172601e-05,
+      "loss": 0.3053,
+      "step": 3213
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.411962642113715e-05,
+      "loss": 0.4669,
+      "step": 3214
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.411608019769428e-05,
+      "loss": 0.3423,
+      "step": 3215
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.411253302153749e-05,
+      "loss": 0.9115,
+      "step": 3216
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.410898489280692e-05,
+      "loss": 1.0167,
+      "step": 3217
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.410543581164274e-05,
+      "loss": 0.4514,
+      "step": 3218
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.410188577818515e-05,
+      "loss": 0.3755,
+      "step": 3219
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.409833479257441e-05,
+      "loss": 0.5022,
+      "step": 3220
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.409478285495078e-05,
+      "loss": 1.3336,
+      "step": 3221
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4091229965454605e-05,
+      "loss": 0.3204,
+      "step": 3222
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.408767612422623e-05,
+      "loss": 0.2836,
+      "step": 3223
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4084121331406056e-05,
+      "loss": 0.8006,
+      "step": 3224
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.408056558713451e-05,
+      "loss": 0.913,
+      "step": 3225
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.407700889155206e-05,
+      "loss": 1.0137,
+      "step": 3226
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4073451244799226e-05,
+      "loss": 0.3066,
+      "step": 3227
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.406989264701655e-05,
+      "loss": 0.8879,
+      "step": 3228
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4066333098344616e-05,
+      "loss": 0.2451,
+      "step": 3229
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.406277259892404e-05,
+      "loss": 0.7969,
+      "step": 3230
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.405921114889549e-05,
+      "loss": 0.4008,
+      "step": 3231
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.405564874839965e-05,
+      "loss": 0.7398,
+      "step": 3232
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4052085397577265e-05,
+      "loss": 0.7657,
+      "step": 3233
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.40485210965691e-05,
+      "loss": 0.6623,
+      "step": 3234
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.404495584551597e-05,
+      "loss": 0.7376,
+      "step": 3235
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.404138964455871e-05,
+      "loss": 0.407,
+      "step": 3236
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.403782249383821e-05,
+      "loss": 0.7089,
+      "step": 3237
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4034254393495396e-05,
+      "loss": 0.7224,
+      "step": 3238
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.403068534367122e-05,
+      "loss": 1.1915,
+      "step": 3239
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4027115344506695e-05,
+      "loss": 1.455,
+      "step": 3240
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4023544396142824e-05,
+      "loss": 0.3578,
+      "step": 3241
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.401997249872071e-05,
+      "loss": 0.6214,
+      "step": 3242
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4016399652381446e-05,
+      "loss": 0.2223,
+      "step": 3243
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4012825857266177e-05,
+      "loss": 0.41,
+      "step": 3244
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.400925111351609e-05,
+      "loss": 0.2066,
+      "step": 3245
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4005675421272414e-05,
+      "loss": 0.6506,
+      "step": 3246
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.40020987806764e-05,
+      "loss": 0.6177,
+      "step": 3247
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3998521191869346e-05,
+      "loss": 0.3476,
+      "step": 3248
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.399494265499258e-05,
+      "loss": 0.5229,
+      "step": 3249
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3991363170187484e-05,
+      "loss": 1.0443,
+      "step": 3250
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3987782737595466e-05,
+      "loss": 0.9631,
+      "step": 3251
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3984201357357956e-05,
+      "loss": 0.7261,
+      "step": 3252
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.398061902961646e-05,
+      "loss": 1.606,
+      "step": 3253
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.397703575451249e-05,
+      "loss": 1.0777,
+      "step": 3254
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3973451532187596e-05,
+      "loss": 0.5906,
+      "step": 3255
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3969866362783385e-05,
+      "loss": 0.8331,
+      "step": 3256
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3966280246441496e-05,
+      "loss": 1.0392,
+      "step": 3257
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.396269318330357e-05,
+      "loss": 0.4898,
+      "step": 3258
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.395910517351135e-05,
+      "loss": 0.4852,
+      "step": 3259
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.395551621720656e-05,
+      "loss": 0.3743,
+      "step": 3260
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.395192631453099e-05,
+      "loss": 0.5949,
+      "step": 3261
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.394833546562647e-05,
+      "loss": 0.5596,
+      "step": 3262
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3944743670634836e-05,
+      "loss": 0.6763,
+      "step": 3263
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3941150929698004e-05,
+      "loss": 0.4464,
+      "step": 3264
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3937557242957895e-05,
+      "loss": 0.807,
+      "step": 3265
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.393396261055648e-05,
+      "loss": 0.4801,
+      "step": 3266
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.393036703263577e-05,
+      "loss": 0.3351,
+      "step": 3267
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3926770509337796e-05,
+      "loss": 1.5932,
+      "step": 3268
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3923173040804665e-05,
+      "loss": 0.3802,
+      "step": 3269
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.391957462717848e-05,
+      "loss": 0.5456,
+      "step": 3270
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3915975268601386e-05,
+      "loss": 0.4325,
+      "step": 3271
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.39123749652156e-05,
+      "loss": 0.7807,
+      "step": 3272
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.390877371716334e-05,
+      "loss": 0.6045,
+      "step": 3273
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.390517152458689e-05,
+      "loss": 0.7485,
+      "step": 3274
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.390156838762854e-05,
+      "loss": 0.4863,
+      "step": 3275
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.389796430643063e-05,
+      "loss": 1.2019,
+      "step": 3276
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.389435928113555e-05,
+      "loss": 0.3548,
+      "step": 3277
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.389075331188572e-05,
+      "loss": 0.7256,
+      "step": 3278
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3887146398823586e-05,
+      "loss": 0.6203,
+      "step": 3279
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3883538542091644e-05,
+      "loss": 1.031,
+      "step": 3280
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.387992974183242e-05,
+      "loss": 0.5983,
+      "step": 3281
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3876319998188494e-05,
+      "loss": 0.8434,
+      "step": 3282
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3872709311302465e-05,
+      "loss": 1.0936,
+      "step": 3283
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3869097681316956e-05,
+      "loss": 0.3602,
+      "step": 3284
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3865485108374674e-05,
+      "loss": 1.4682,
+      "step": 3285
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.386187159261831e-05,
+      "loss": 0.6852,
+      "step": 3286
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.385825713419065e-05,
+      "loss": 0.857,
+      "step": 3287
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.385464173323444e-05,
+      "loss": 0.5733,
+      "step": 3288
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.385102538989254e-05,
+      "loss": 0.835,
+      "step": 3289
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3847408104307795e-05,
+      "loss": 0.7676,
+      "step": 3290
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3843789876623126e-05,
+      "loss": 0.9583,
+      "step": 3291
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.384017070698147e-05,
+      "loss": 0.3593,
+      "step": 3292
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.383655059552579e-05,
+      "loss": 0.7108,
+      "step": 3293
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.383292954239911e-05,
+      "loss": 0.6116,
+      "step": 3294
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3829307547744476e-05,
+      "loss": 0.6764,
+      "step": 3295
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.382568461170497e-05,
+      "loss": 1.0224,
+      "step": 3296
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.382206073442374e-05,
+      "loss": 0.9734,
+      "step": 3297
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.381843591604393e-05,
+      "loss": 1.0006,
+      "step": 3298
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.381481015670874e-05,
+      "loss": 0.4584,
+      "step": 3299
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3811183456561415e-05,
+      "loss": 0.5114,
+      "step": 3300
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.380755581574522e-05,
+      "loss": 0.5592,
+      "step": 3301
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.380392723440347e-05,
+      "loss": 1.743,
+      "step": 3302
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3800297712679514e-05,
+      "loss": 0.6821,
+      "step": 3303
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.379666725071674e-05,
+      "loss": 0.2264,
+      "step": 3304
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3793035848658573e-05,
+      "loss": 0.3753,
+      "step": 3305
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.378940350664846e-05,
+      "loss": 0.6766,
+      "step": 3306
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.378577022482991e-05,
+      "loss": 0.4161,
+      "step": 3307
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.378213600334645e-05,
+      "loss": 0.0524,
+      "step": 3308
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.377850084234165e-05,
+      "loss": 0.0804,
+      "step": 3309
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3774864741959134e-05,
+      "loss": 0.9512,
+      "step": 3310
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.377122770234253e-05,
+      "loss": 0.5302,
+      "step": 3311
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.376758972363553e-05,
+      "loss": 0.6701,
+      "step": 3312
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.376395080598184e-05,
+      "loss": 0.6839,
+      "step": 3313
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.376031094952523e-05,
+      "loss": 0.7423,
+      "step": 3314
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3756670154409496e-05,
+      "loss": 0.3259,
+      "step": 3315
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.375302842077845e-05,
+      "loss": 0.519,
+      "step": 3316
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.374938574877598e-05,
+      "loss": 0.9365,
+      "step": 3317
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.374574213854599e-05,
+      "loss": 0.4666,
+      "step": 3318
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.374209759023241e-05,
+      "loss": 0.3409,
+      "step": 3319
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3738452103979226e-05,
+      "loss": 0.4993,
+      "step": 3320
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.373480567993045e-05,
+      "loss": 0.7271,
+      "step": 3321
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3731158318230146e-05,
+      "loss": 0.801,
+      "step": 3322
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.372751001902239e-05,
+      "loss": 1.0016,
+      "step": 3323
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3723860782451326e-05,
+      "loss": 0.6488,
+      "step": 3324
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.37202106086611e-05,
+      "loss": 1.1687,
+      "step": 3325
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.371655949779592e-05,
+      "loss": 0.6837,
+      "step": 3326
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.371290745000003e-05,
+      "loss": 0.5532,
+      "step": 3327
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3709254465417694e-05,
+      "loss": 0.9052,
+      "step": 3328
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3705600544193233e-05,
+      "loss": 0.2561,
+      "step": 3329
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3701945686471006e-05,
+      "loss": 0.5168,
+      "step": 3330
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.369828989239539e-05,
+      "loss": 0.4738,
+      "step": 3331
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.36946331621108e-05,
+      "loss": 1.0325,
+      "step": 3332
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.36909754957617e-05,
+      "loss": 0.3049,
+      "step": 3333
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.368731689349259e-05,
+      "loss": 1.1607,
+      "step": 3334
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3683657355448e-05,
+      "loss": 0.5444,
+      "step": 3335
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.367999688177253e-05,
+      "loss": 0.5181,
+      "step": 3336
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.367633547261074e-05,
+      "loss": 1.2521,
+      "step": 3337
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3672673128107314e-05,
+      "loss": 0.4529,
+      "step": 3338
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.366900984840692e-05,
+      "loss": 0.8869,
+      "step": 3339
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.366534563365426e-05,
+      "loss": 0.8981,
+      "step": 3340
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3661680483994125e-05,
+      "loss": 0.7846,
+      "step": 3341
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3658014399571284e-05,
+      "loss": 0.8712,
+      "step": 3342
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3654347380530576e-05,
+      "loss": 0.5242,
+      "step": 3343
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3650679427016855e-05,
+      "loss": 0.9338,
+      "step": 3344
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.364701053917504e-05,
+      "loss": 0.8475,
+      "step": 3345
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.364334071715007e-05,
+      "loss": 1.0934,
+      "step": 3346
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.363966996108692e-05,
+      "loss": 0.7465,
+      "step": 3347
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.36359982711306e-05,
+      "loss": 1.4941,
+      "step": 3348
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.363232564742617e-05,
+      "loss": 0.5525,
+      "step": 3349
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.36286520901187e-05,
+      "loss": 0.2847,
+      "step": 3350
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.362497759935333e-05,
+      "loss": 0.4996,
+      "step": 3351
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.362130217527523e-05,
+      "loss": 0.2092,
+      "step": 3352
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3617625818029585e-05,
+      "loss": 0.5751,
+      "step": 3353
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3613948527761635e-05,
+      "loss": 0.0121,
+      "step": 3354
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.361027030461665e-05,
+      "loss": 0.435,
+      "step": 3355
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3606591148739946e-05,
+      "loss": 0.4447,
+      "step": 3356
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.360291106027686e-05,
+      "loss": 0.9547,
+      "step": 3357
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.359923003937278e-05,
+      "loss": 0.52,
+      "step": 3358
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3595548086173124e-05,
+      "loss": 0.6472,
+      "step": 3359
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3591865200823355e-05,
+      "loss": 0.891,
+      "step": 3360
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.358818138346896e-05,
+      "loss": 1.1117,
+      "step": 3361
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3584496634255476e-05,
+      "loss": 0.7682,
+      "step": 3362
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.358081095332847e-05,
+      "loss": 1.1576,
+      "step": 3363
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3577124340833533e-05,
+      "loss": 0.5085,
+      "step": 3364
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3573436796916315e-05,
+      "loss": 0.5947,
+      "step": 3365
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3569748321722495e-05,
+      "loss": 0.4844,
+      "step": 3366
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.356605891539779e-05,
+      "loss": 0.6594,
+      "step": 3367
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.356236857808794e-05,
+      "loss": 0.7195,
+      "step": 3368
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3558677309938744e-05,
+      "loss": 0.4734,
+      "step": 3369
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.355498511109603e-05,
+      "loss": 0.7897,
+      "step": 3370
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.355129198170565e-05,
+      "loss": 0.8076,
+      "step": 3371
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.35475979219135e-05,
+      "loss": 0.1413,
+      "step": 3372
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3543902931865526e-05,
+      "loss": 1.1439,
+      "step": 3373
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.354020701170769e-05,
+      "loss": 0.1683,
+      "step": 3374
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.353651016158601e-05,
+      "loss": 0.4158,
+      "step": 3375
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.353281238164652e-05,
+      "loss": 0.2603,
+      "step": 3376
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.352911367203531e-05,
+      "loss": 0.8478,
+      "step": 3377
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.35254140328985e-05,
+      "loss": 0.876,
+      "step": 3378
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.352171346438224e-05,
+      "loss": 0.4498,
+      "step": 3379
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.351801196663272e-05,
+      "loss": 0.4925,
+      "step": 3380
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3514309539796176e-05,
+      "loss": 0.6683,
+      "step": 3381
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.351060618401887e-05,
+      "loss": 0.7832,
+      "step": 3382
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.350690189944711e-05,
+      "loss": 0.9386,
+      "step": 3383
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.350319668622723e-05,
+      "loss": 0.3819,
+      "step": 3384
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.349949054450561e-05,
+      "loss": 1.1619,
+      "step": 3385
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.349578347442865e-05,
+      "loss": 0.2793,
+      "step": 3386
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3492075476142814e-05,
+      "loss": 0.379,
+      "step": 3387
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.348836654979458e-05,
+      "loss": 0.7434,
+      "step": 3388
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3484656695530465e-05,
+      "loss": 0.3639,
+      "step": 3389
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.348094591349704e-05,
+      "loss": 0.6584,
+      "step": 3390
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.34772342038409e-05,
+      "loss": 0.866,
+      "step": 3391
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.347352156670867e-05,
+      "loss": 0.4678,
+      "step": 3392
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.346980800224702e-05,
+      "loss": 1.2861,
+      "step": 3393
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.346609351060265e-05,
+      "loss": 0.7545,
+      "step": 3394
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.346237809192232e-05,
+      "loss": 0.4552,
+      "step": 3395
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.34586617463528e-05,
+      "loss": 0.5515,
+      "step": 3396
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.34549444740409e-05,
+      "loss": 0.8673,
+      "step": 3397
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3451226275133464e-05,
+      "loss": 0.5398,
+      "step": 3398
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.344750714977741e-05,
+      "loss": 0.3132,
+      "step": 3399
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.344378709811964e-05,
+      "loss": 0.8749,
+      "step": 3400
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.725633978843689,
+      "eval_runtime": 39.3767,
+      "eval_samples_per_second": 1.016,
+      "eval_steps_per_second": 1.016,
+      "step": 3400
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.344006612030713e-05,
+      "loss": 0.8014,
+      "step": 3401
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.343634421648686e-05,
+      "loss": 1.3071,
+      "step": 3402
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.343262138680588e-05,
+      "loss": 0.8307,
+      "step": 3403
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3428897631411255e-05,
+      "loss": 0.2553,
+      "step": 3404
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3425172950450095e-05,
+      "loss": 0.7748,
+      "step": 3405
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3421447344069545e-05,
+      "loss": 0.5818,
+      "step": 3406
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3417720812416794e-05,
+      "loss": 1.0699,
+      "step": 3407
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.341399335563904e-05,
+      "loss": 0.9068,
+      "step": 3408
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3410264973883565e-05,
+      "loss": 1.2054,
+      "step": 3409
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.340653566729763e-05,
+      "loss": 0.5712,
+      "step": 3410
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.340280543602859e-05,
+      "loss": 1.1512,
+      "step": 3411
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3399074280223785e-05,
+      "loss": 1.1641,
+      "step": 3412
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3395342200030634e-05,
+      "loss": 0.8695,
+      "step": 3413
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3391609195596574e-05,
+      "loss": 0.7074,
+      "step": 3414
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.338787526706906e-05,
+      "loss": 0.4792,
+      "step": 3415
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3384140414595616e-05,
+      "loss": 0.8192,
+      "step": 3416
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.338040463832379e-05,
+      "loss": 0.6519,
+      "step": 3417
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.337666793840116e-05,
+      "loss": 0.9922,
+      "step": 3418
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.337293031497535e-05,
+      "loss": 0.3876,
+      "step": 3419
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.336919176819401e-05,
+      "loss": 0.7201,
+      "step": 3420
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.336545229820484e-05,
+      "loss": 0.1321,
+      "step": 3421
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.336171190515555e-05,
+      "loss": 0.7805,
+      "step": 3422
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3357970589193934e-05,
+      "loss": 0.7097,
+      "step": 3423
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.335422835046778e-05,
+      "loss": 0.5142,
+      "step": 3424
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3350485189124916e-05,
+      "loss": 0.5235,
+      "step": 3425
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.334674110531324e-05,
+      "loss": 0.2199,
+      "step": 3426
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.334299609918065e-05,
+      "loss": 0.4326,
+      "step": 3427
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.333925017087509e-05,
+      "loss": 0.4566,
+      "step": 3428
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.333550332054455e-05,
+      "loss": 0.3398,
+      "step": 3429
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.333175554833705e-05,
+      "loss": 0.9853,
+      "step": 3430
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.332800685440065e-05,
+      "loss": 0.2915,
+      "step": 3431
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.332425723888343e-05,
+      "loss": 0.6872,
+      "step": 3432
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3320506701933535e-05,
+      "loss": 1.1327,
+      "step": 3433
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.331675524369913e-05,
+      "loss": 1.0633,
+      "step": 3434
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.331300286432841e-05,
+      "loss": 0.7035,
+      "step": 3435
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.330924956396961e-05,
+      "loss": 0.688,
+      "step": 3436
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3305495342771025e-05,
+      "loss": 0.4871,
+      "step": 3437
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3301740200880946e-05,
+      "loss": 0.5239,
+      "step": 3438
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.329798413844774e-05,
+      "loss": 0.7047,
+      "step": 3439
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3294227155619776e-05,
+      "loss": 0.5788,
+      "step": 3440
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.329046925254548e-05,
+      "loss": 0.6298,
+      "step": 3441
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.32867104293733e-05,
+      "loss": 0.5022,
+      "step": 3442
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.328295068625175e-05,
+      "loss": 0.015,
+      "step": 3443
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.327919002332935e-05,
+      "loss": 0.9126,
+      "step": 3444
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3275428440754656e-05,
+      "loss": 0.9235,
+      "step": 3445
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3271665938676286e-05,
+      "loss": 0.914,
+      "step": 3446
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3267902517242876e-05,
+      "loss": 0.4995,
+      "step": 3447
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.326413817660309e-05,
+      "loss": 0.5298,
+      "step": 3448
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.326037291690565e-05,
+      "loss": 0.4252,
+      "step": 3449
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3256606738299305e-05,
+      "loss": 0.7496,
+      "step": 3450
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3252839640932836e-05,
+      "loss": 0.7157,
+      "step": 3451
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3249071624955056e-05,
+      "loss": 0.9304,
+      "step": 3452
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3245302690514834e-05,
+      "loss": 0.8718,
+      "step": 3453
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.324153283776106e-05,
+      "loss": 0.8921,
+      "step": 3454
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.323776206684266e-05,
+      "loss": 0.6936,
+      "step": 3455
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.32339903779086e-05,
+      "loss": 1.3287,
+      "step": 3456
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.323021777110788e-05,
+      "loss": 1.479,
+      "step": 3457
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.322644424658954e-05,
+      "loss": 0.9109,
+      "step": 3458
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.322266980450266e-05,
+      "loss": 0.4809,
+      "step": 3459
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.321889444499634e-05,
+      "loss": 0.9358,
+      "step": 3460
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.321511816821974e-05,
+      "loss": 0.7182,
+      "step": 3461
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.321134097432203e-05,
+      "loss": 0.4781,
+      "step": 3462
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.320756286345243e-05,
+      "loss": 0.7361,
+      "step": 3463
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.32037838357602e-05,
+      "loss": 0.5129,
+      "step": 3464
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3200003891394645e-05,
+      "loss": 0.8359,
+      "step": 3465
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.319622303050508e-05,
+      "loss": 1.1793,
+      "step": 3466
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.319244125324086e-05,
+      "loss": 1.0207,
+      "step": 3467
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.318865855975139e-05,
+      "loss": 1.0147,
+      "step": 3468
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.318487495018612e-05,
+      "loss": 0.9419,
+      "step": 3469
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.318109042469452e-05,
+      "loss": 0.5881,
+      "step": 3470
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.317730498342608e-05,
+      "loss": 0.2698,
+      "step": 3471
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.317351862653037e-05,
+      "loss": 0.7937,
+      "step": 3472
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3169731354156955e-05,
+      "loss": 0.4977,
+      "step": 3473
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.316594316645545e-05,
+      "loss": 0.6191,
+      "step": 3474
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.316215406357552e-05,
+      "loss": 0.9784,
+      "step": 3475
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.315836404566686e-05,
+      "loss": 0.596,
+      "step": 3476
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3154573112879175e-05,
+      "loss": 0.4489,
+      "step": 3477
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.315078126536224e-05,
+      "loss": 1.0268,
+      "step": 3478
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.314698850326585e-05,
+      "loss": 0.3221,
+      "step": 3479
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.314319482673985e-05,
+      "loss": 0.89,
+      "step": 3480
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.313940023593409e-05,
+      "loss": 0.5153,
+      "step": 3481
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.31356047309985e-05,
+      "loss": 0.2799,
+      "step": 3482
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.313180831208301e-05,
+      "loss": 1.1572,
+      "step": 3483
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.312801097933759e-05,
+      "loss": 0.8357,
+      "step": 3484
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3124212732912274e-05,
+      "loss": 0.8193,
+      "step": 3485
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3120413572957097e-05,
+      "loss": 0.8435,
+      "step": 3486
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.311661349962214e-05,
+      "loss": 0.3095,
+      "step": 3487
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3112812513057563e-05,
+      "loss": 0.5246,
+      "step": 3488
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.310901061341349e-05,
+      "loss": 0.8992,
+      "step": 3489
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.310520780084013e-05,
+      "loss": 0.6956,
+      "step": 3490
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.310140407548771e-05,
+      "loss": 0.8402,
+      "step": 3491
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.30975994375065e-05,
+      "loss": 0.4784,
+      "step": 3492
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.30937938870468e-05,
+      "loss": 0.7958,
+      "step": 3493
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3089987424258955e-05,
+      "loss": 0.7143,
+      "step": 3494
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3086180049293345e-05,
+      "loss": 0.5034,
+      "step": 3495
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.308237176230036e-05,
+      "loss": 0.7458,
+      "step": 3496
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.307856256343047e-05,
+      "loss": 1.0625,
+      "step": 3497
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.307475245283415e-05,
+      "loss": 0.4905,
+      "step": 3498
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.307094143066193e-05,
+      "loss": 0.5112,
+      "step": 3499
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.306712949706434e-05,
+      "loss": 0.417,
+      "step": 3500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3063316652192e-05,
+      "loss": 1.1063,
+      "step": 3501
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.305950289619553e-05,
+      "loss": 0.6648,
+      "step": 3502
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3055688229225576e-05,
+      "loss": 0.8473,
+      "step": 3503
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3051872651432856e-05,
+      "loss": 0.7181,
+      "step": 3504
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.30480561629681e-05,
+      "loss": 0.6478,
+      "step": 3505
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3044238763982085e-05,
+      "loss": 1.0976,
+      "step": 3506
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.304042045462561e-05,
+      "loss": 1.0611,
+      "step": 3507
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3036601235049524e-05,
+      "loss": 0.7681,
+      "step": 3508
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3032781105404705e-05,
+      "loss": 0.7169,
+      "step": 3509
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3028960065842074e-05,
+      "loss": 0.6438,
+      "step": 3510
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.302513811651258e-05,
+      "loss": 1.0733,
+      "step": 3511
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.302131525756719e-05,
+      "loss": 1.1169,
+      "step": 3512
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.301749148915697e-05,
+      "loss": 0.2038,
+      "step": 3513
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.301366681143293e-05,
+      "loss": 0.9086,
+      "step": 3514
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.30098412245462e-05,
+      "loss": 0.4583,
+      "step": 3515
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.30060147286479e-05,
+      "loss": 0.8715,
+      "step": 3516
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.30021873238892e-05,
+      "loss": 1.0059,
+      "step": 3517
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2998359010421296e-05,
+      "loss": 0.9278,
+      "step": 3518
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.299452978839544e-05,
+      "loss": 0.9969,
+      "step": 3519
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.29906996579629e-05,
+      "loss": 0.4336,
+      "step": 3520
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2986868619274975e-05,
+      "loss": 1.0015,
+      "step": 3521
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.298303667248302e-05,
+      "loss": 0.2659,
+      "step": 3522
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2979203817738425e-05,
+      "loss": 0.5956,
+      "step": 3523
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2975370055192605e-05,
+      "loss": 0.7326,
+      "step": 3524
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2971535384997e-05,
+      "loss": 0.8478,
+      "step": 3525
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.296769980730312e-05,
+      "loss": 0.4719,
+      "step": 3526
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2963863322262475e-05,
+      "loss": 1.1381,
+      "step": 3527
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2960025930026634e-05,
+      "loss": 0.8951,
+      "step": 3528
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2956187630747195e-05,
+      "loss": 0.5949,
+      "step": 3529
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2952348424575796e-05,
+      "loss": 0.7986,
+      "step": 3530
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.29485083116641e-05,
+      "loss": 0.2395,
+      "step": 3531
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2944667292163806e-05,
+      "loss": 0.3314,
+      "step": 3532
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.294082536622666e-05,
+      "loss": 0.3799,
+      "step": 3533
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.293698253400444e-05,
+      "loss": 0.5639,
+      "step": 3534
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.293313879564897e-05,
+      "loss": 0.4453,
+      "step": 3535
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2929294151312074e-05,
+      "loss": 0.9894,
+      "step": 3536
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.292544860114565e-05,
+      "loss": 0.8338,
+      "step": 3537
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.292160214530162e-05,
+      "loss": 1.1136,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.291775478393193e-05,
+      "loss": 0.4256,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2913906517188585e-05,
+      "loss": 1.0887,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2910057345223595e-05,
+      "loss": 0.6334,
+      "step": 3541
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.290620726818903e-05,
+      "loss": 0.3948,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2902356286237e-05,
+      "loss": 0.309,
+      "step": 3543
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.289850439951962e-05,
+      "loss": 0.2926,
+      "step": 3544
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2894651608189076e-05,
+      "loss": 0.4748,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.289079791239757e-05,
+      "loss": 0.4593,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2886943312297335e-05,
+      "loss": 0.2054,
+      "step": 3547
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.288308780804066e-05,
+      "loss": 0.4874,
+      "step": 3548
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2879231399779844e-05,
+      "loss": 0.4157,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.287537408766725e-05,
+      "loss": 0.3576,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.287151587185525e-05,
+      "loss": 0.8498,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.286765675249628e-05,
+      "loss": 1.2763,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.286379672974278e-05,
+      "loss": 0.7994,
+      "step": 3553
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.285993580374725e-05,
+      "loss": 1.1675,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2856073974662215e-05,
+      "loss": 0.439,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2852211242640244e-05,
+      "loss": 0.9228,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.284834760783392e-05,
+      "loss": 0.8843,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.284448307039589e-05,
+      "loss": 0.9161,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.284061763047882e-05,
+      "loss": 0.593,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.283675128823541e-05,
+      "loss": 0.5997,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.283288404381842e-05,
+      "loss": 0.87,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.282901589738061e-05,
+      "loss": 0.4255,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.28251468490748e-05,
+      "loss": 0.5387,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.282127689905383e-05,
+      "loss": 0.7921,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.281740604747059e-05,
+      "loss": 0.4139,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2813534294477995e-05,
+      "loss": 0.4672,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2809661640229e-05,
+      "loss": 0.3007,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.28057880848766e-05,
+      "loss": 0.5976,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.280191362857382e-05,
+      "loss": 1.0077,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2798038271473716e-05,
+      "loss": 1.2887,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.27941620137294e-05,
+      "loss": 0.0365,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.279028485549399e-05,
+      "loss": 0.5519,
+      "step": 3572
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2786406796920665e-05,
+      "loss": 0.3655,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2782527838162614e-05,
+      "loss": 0.5126,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.277864797937309e-05,
+      "loss": 0.5872,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.277476722070536e-05,
+      "loss": 0.6607,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.277088556231275e-05,
+      "loss": 0.9301,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2767003004348585e-05,
+      "loss": 1.5268,
+      "step": 3578
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.276311954696626e-05,
+      "loss": 1.0978,
+      "step": 3579
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.275923519031919e-05,
+      "loss": 0.7171,
+      "step": 3580
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2755349934560826e-05,
+      "loss": 0.8239,
+      "step": 3581
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.275146377984466e-05,
+      "loss": 0.7501,
+      "step": 3582
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.274757672632421e-05,
+      "loss": 0.9405,
+      "step": 3583
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.274368877415304e-05,
+      "loss": 0.789,
+      "step": 3584
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.273979992348474e-05,
+      "loss": 0.3646,
+      "step": 3585
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.273591017447294e-05,
+      "loss": 0.8797,
+      "step": 3586
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.273201952727132e-05,
+      "loss": 0.3029,
+      "step": 3587
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2728127982033575e-05,
+      "loss": 1.0024,
+      "step": 3588
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.272423553891343e-05,
+      "loss": 0.4731,
+      "step": 3589
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.272034219806466e-05,
+      "loss": 0.8609,
+      "step": 3590
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2716447959641076e-05,
+      "loss": 0.5639,
+      "step": 3591
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.271255282379653e-05,
+      "loss": 0.8175,
+      "step": 3592
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.27086567906849e-05,
+      "loss": 0.8533,
+      "step": 3593
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2704759860460085e-05,
+      "loss": 1.224,
+      "step": 3594
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.270086203327605e-05,
+      "loss": 1.0387,
+      "step": 3595
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.269696330928677e-05,
+      "loss": 0.8409,
+      "step": 3596
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.269306368864626e-05,
+      "loss": 0.4962,
+      "step": 3597
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2689163171508596e-05,
+      "loss": 0.415,
+      "step": 3598
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.268526175802785e-05,
+      "loss": 0.9795,
+      "step": 3599
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.268135944835816e-05,
+      "loss": 0.6285,
+      "step": 3600
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.7129235863685608,
+      "eval_runtime": 39.3994,
+      "eval_samples_per_second": 1.015,
+      "eval_steps_per_second": 1.015,
+      "step": 3600
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.267745624265369e-05,
+      "loss": 0.4407,
+      "step": 3601
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2673552141068627e-05,
+      "loss": 1.8464,
+      "step": 3602
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.26696471437572e-05,
+      "loss": 0.464,
+      "step": 3603
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2665741250873695e-05,
+      "loss": 0.9757,
+      "step": 3604
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2661834462572415e-05,
+      "loss": 1.1594,
+      "step": 3605
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2657926779007684e-05,
+      "loss": 0.7128,
+      "step": 3606
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.265401820033388e-05,
+      "loss": 0.4104,
+      "step": 3607
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.265010872670541e-05,
+      "loss": 0.1611,
+      "step": 3608
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2646198358276734e-05,
+      "loss": 0.2237,
+      "step": 3609
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.264228709520232e-05,
+      "loss": 0.8584,
+      "step": 3610
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.263837493763669e-05,
+      "loss": 0.336,
+      "step": 3611
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.263446188573439e-05,
+      "loss": 0.8804,
+      "step": 3612
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2630547939650015e-05,
+      "loss": 1.1606,
+      "step": 3613
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.262663309953817e-05,
+      "loss": 0.5024,
+      "step": 3614
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.262271736555353e-05,
+      "loss": 0.8945,
+      "step": 3615
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.261880073785077e-05,
+      "loss": 0.4442,
+      "step": 3616
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2614883216584634e-05,
+      "loss": 1.0253,
+      "step": 3617
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2610964801909886e-05,
+      "loss": 0.8943,
+      "step": 3618
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.26070454939813e-05,
+      "loss": 0.606,
+      "step": 3619
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2603125292953744e-05,
+      "loss": 1.048,
+      "step": 3620
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.259920419898206e-05,
+      "loss": 0.5965,
+      "step": 3621
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.259528221222116e-05,
+      "loss": 0.6008,
+      "step": 3622
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.259135933282599e-05,
+      "loss": 0.6631,
+      "step": 3623
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.258743556095152e-05,
+      "loss": 0.7027,
+      "step": 3624
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2583510896752754e-05,
+      "loss": 1.2302,
+      "step": 3625
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.257958534038475e-05,
+      "loss": 0.9102,
+      "step": 3626
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2575658892002574e-05,
+      "loss": 0.2544,
+      "step": 3627
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2571731551761356e-05,
+      "loss": 0.2817,
+      "step": 3628
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2567803319816225e-05,
+      "loss": 0.4916,
+      "step": 3629
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.25638741963224e-05,
+      "loss": 0.6823,
+      "step": 3630
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.255994418143508e-05,
+      "loss": 0.6308,
+      "step": 3631
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.255601327530951e-05,
+      "loss": 0.2739,
+      "step": 3632
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.255208147810102e-05,
+      "loss": 1.1199,
+      "step": 3633
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2548148789964906e-05,
+      "loss": 0.9593,
+      "step": 3634
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.254421521105653e-05,
+      "loss": 0.7379,
+      "step": 3635
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2540280741531315e-05,
+      "loss": 0.7658,
+      "step": 3636
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.253634538154465e-05,
+      "loss": 0.6847,
+      "step": 3637
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2532409131252055e-05,
+      "loss": 0.7729,
+      "step": 3638
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2528471990808995e-05,
+      "loss": 1.1256,
+      "step": 3639
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.252453396037102e-05,
+      "loss": 0.2506,
+      "step": 3640
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.252059504009371e-05,
+      "loss": 0.2964,
+      "step": 3641
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2516655230132674e-05,
+      "loss": 0.9781,
+      "step": 3642
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.251271453064353e-05,
+      "loss": 0.9225,
+      "step": 3643
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.250877294178198e-05,
+      "loss": 0.8189,
+      "step": 3644
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2504830463703736e-05,
+      "loss": 0.5569,
+      "step": 3645
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2500887096564556e-05,
+      "loss": 0.4043,
+      "step": 3646
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.249694284052019e-05,
+      "loss": 0.8345,
+      "step": 3647
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.249299769572649e-05,
+      "loss": 0.8475,
+      "step": 3648
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.24890516623393e-05,
+      "loss": 0.632,
+      "step": 3649
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.248510474051451e-05,
+      "loss": 0.5624,
+      "step": 3650
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.248115693040804e-05,
+      "loss": 0.6354,
+      "step": 3651
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.247720823217585e-05,
+      "loss": 0.743,
+      "step": 3652
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2473258645973946e-05,
+      "loss": 0.5231,
+      "step": 3653
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2469308171958345e-05,
+      "loss": 0.4204,
+      "step": 3654
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2465356810285114e-05,
+      "loss": 0.6811,
+      "step": 3655
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.246140456111036e-05,
+      "loss": 0.685,
+      "step": 3656
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.24574514245902e-05,
+      "loss": 0.3865,
+      "step": 3657
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.245349740088083e-05,
+      "loss": 0.3178,
+      "step": 3658
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2449542490138435e-05,
+      "loss": 0.6737,
+      "step": 3659
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2445586692519264e-05,
+      "loss": 0.7399,
+      "step": 3660
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.244163000817959e-05,
+      "loss": 0.4019,
+      "step": 3661
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.243767243727573e-05,
+      "loss": 0.5983,
+      "step": 3662
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2433713979964016e-05,
+      "loss": 1.0174,
+      "step": 3663
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.242975463640083e-05,
+      "loss": 1.0816,
+      "step": 3664
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.24257944067426e-05,
+      "loss": 0.6179,
+      "step": 3665
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.242183329114576e-05,
+      "loss": 0.82,
+      "step": 3666
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2417871289766806e-05,
+      "loss": 0.4649,
+      "step": 3667
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2413908402762254e-05,
+      "loss": 0.9004,
+      "step": 3668
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2409944630288676e-05,
+      "loss": 0.008,
+      "step": 3669
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.240597997250263e-05,
+      "loss": 0.3861,
+      "step": 3670
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.240201442956076e-05,
+      "loss": 0.6436,
+      "step": 3671
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.239804800161973e-05,
+      "loss": 0.4255,
+      "step": 3672
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.239408068883623e-05,
+      "loss": 0.891,
+      "step": 3673
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.239011249136698e-05,
+      "loss": 0.5793,
+      "step": 3674
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.238614340936876e-05,
+      "loss": 0.7292,
+      "step": 3675
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.238217344299837e-05,
+      "loss": 0.3244,
+      "step": 3676
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2378202592412626e-05,
+      "loss": 0.7129,
+      "step": 3677
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2374230857768425e-05,
+      "loss": 0.2367,
+      "step": 3678
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2370258239222644e-05,
+      "loss": 0.3388,
+      "step": 3679
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.236628473693224e-05,
+      "loss": 0.4763,
+      "step": 3680
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.236231035105419e-05,
+      "loss": 0.5631,
+      "step": 3681
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.235833508174549e-05,
+      "loss": 0.8622,
+      "step": 3682
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.235435892916319e-05,
+      "loss": 0.6206,
+      "step": 3683
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2350381893464374e-05,
+      "loss": 0.818,
+      "step": 3684
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.234640397480616e-05,
+      "loss": 0.3709,
+      "step": 3685
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.234242517334568e-05,
+      "loss": 0.2377,
+      "step": 3686
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2338445489240135e-05,
+      "loss": 0.7954,
+      "step": 3687
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.233446492264673e-05,
+      "loss": 0.6962,
+      "step": 3688
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.233048347372273e-05,
+      "loss": 0.5347,
+      "step": 3689
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.232650114262542e-05,
+      "loss": 0.8303,
+      "step": 3690
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.232251792951212e-05,
+      "loss": 0.9549,
+      "step": 3691
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2318533834540195e-05,
+      "loss": 0.726,
+      "step": 3692
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.231454885786702e-05,
+      "loss": 0.4004,
+      "step": 3693
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2310562999650046e-05,
+      "loss": 0.9915,
+      "step": 3694
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2306576260046724e-05,
+      "loss": 0.6239,
+      "step": 3695
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2302588639214556e-05,
+      "loss": 0.8646,
+      "step": 3696
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.229860013731107e-05,
+      "loss": 0.4984,
+      "step": 3697
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.229461075449383e-05,
+      "loss": 0.519,
+      "step": 3698
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.229062049092044e-05,
+      "loss": 0.823,
+      "step": 3699
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.228662934674855e-05,
+      "loss": 0.7315,
+      "step": 3700
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.228263732213581e-05,
+      "loss": 0.3998,
+      "step": 3701
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.227864441723995e-05,
+      "loss": 0.2311,
+      "step": 3702
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.227465063221869e-05,
+      "loss": 1.0758,
+      "step": 3703
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2270655967229814e-05,
+      "loss": 0.5592,
+      "step": 3704
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.226666042243114e-05,
+      "loss": 0.6969,
+      "step": 3705
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.22626639979805e-05,
+      "loss": 0.8782,
+      "step": 3706
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.225866669403578e-05,
+      "loss": 0.7152,
+      "step": 3707
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2254668510754896e-05,
+      "loss": 0.1641,
+      "step": 3708
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.225066944829579e-05,
+      "loss": 0.8981,
+      "step": 3709
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.224666950681647e-05,
+      "loss": 0.504,
+      "step": 3710
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.224266868647492e-05,
+      "loss": 0.0284,
+      "step": 3711
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.223866698742922e-05,
+      "loss": 0.4944,
+      "step": 3712
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2234664409837445e-05,
+      "loss": 0.5959,
+      "step": 3713
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.223066095385773e-05,
+      "loss": 0.7704,
+      "step": 3714
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.222665661964822e-05,
+      "loss": 0.6158,
+      "step": 3715
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2222651407367116e-05,
+      "loss": 0.9501,
+      "step": 3716
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2218645317172644e-05,
+      "loss": 0.9379,
+      "step": 3717
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.221463834922307e-05,
+      "loss": 0.7047,
+      "step": 3718
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2210630503676677e-05,
+      "loss": 1.419,
+      "step": 3719
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.220662178069181e-05,
+      "loss": 0.4588,
+      "step": 3720
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.220261218042683e-05,
+      "loss": 0.3439,
+      "step": 3721
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2198601703040126e-05,
+      "loss": 0.9524,
+      "step": 3722
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2194590348690155e-05,
+      "loss": 0.2461,
+      "step": 3723
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.219057811753537e-05,
+      "loss": 0.9081,
+      "step": 3724
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2186565009734296e-05,
+      "loss": 0.3287,
+      "step": 3725
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.218255102544545e-05,
+      "loss": 0.4077,
+      "step": 3726
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2178536164827415e-05,
+      "loss": 0.4659,
+      "step": 3727
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.21745204280388e-05,
+      "loss": 0.6746,
+      "step": 3728
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2170503815238244e-05,
+      "loss": 0.4907,
+      "step": 3729
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2166486326584426e-05,
+      "loss": 0.5413,
+      "step": 3730
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.216246796223606e-05,
+      "loss": 0.5885,
+      "step": 3731
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.21584487223519e-05,
+      "loss": 0.4434,
+      "step": 3732
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.215442860709071e-05,
+      "loss": 0.8082,
+      "step": 3733
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2150407616611316e-05,
+      "loss": 1.0219,
+      "step": 3734
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2146385751072576e-05,
+      "loss": 0.225,
+      "step": 3735
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.214236301063337e-05,
+      "loss": 0.8234,
+      "step": 3736
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.21383393954526e-05,
+      "loss": 0.6432,
+      "step": 3737
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2134314905689245e-05,
+      "loss": 0.888,
+      "step": 3738
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.213028954150228e-05,
+      "loss": 1.0844,
+      "step": 3739
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.212626330305074e-05,
+      "loss": 0.3958,
+      "step": 3740
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.212223619049367e-05,
+      "loss": 0.5732,
+      "step": 3741
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.211820820399017e-05,
+      "loss": 0.1322,
+      "step": 3742
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.211417934369937e-05,
+      "loss": 0.193,
+      "step": 3743
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.211014960978042e-05,
+      "loss": 0.8671,
+      "step": 3744
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2106119002392534e-05,
+      "loss": 0.1784,
+      "step": 3745
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.210208752169492e-05,
+      "loss": 0.7235,
+      "step": 3746
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2098055167846855e-05,
+      "loss": 0.8032,
+      "step": 3747
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.209402194100764e-05,
+      "loss": 1.1338,
+      "step": 3748
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.20899878413366e-05,
+      "loss": 0.359,
+      "step": 3749
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.208595286899312e-05,
+      "loss": 0.752,
+      "step": 3750
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.208191702413659e-05,
+      "loss": 0.514,
+      "step": 3751
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.207788030692645e-05,
+      "loss": 0.6878,
+      "step": 3752
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.207384271752217e-05,
+      "loss": 0.4686,
+      "step": 3753
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.206980425608326e-05,
+      "loss": 0.9417,
+      "step": 3754
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.206576492276926e-05,
+      "loss": 0.4901,
+      "step": 3755
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.206172471773974e-05,
+      "loss": 0.3302,
+      "step": 3756
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.205768364115432e-05,
+      "loss": 1.3846,
+      "step": 3757
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.205364169317264e-05,
+      "loss": 0.4962,
+      "step": 3758
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.204959887395438e-05,
+      "loss": 0.4783,
+      "step": 3759
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.204555518365924e-05,
+      "loss": 0.8041,
+      "step": 3760
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2041510622446986e-05,
+      "loss": 0.786,
+      "step": 3761
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.203746519047738e-05,
+      "loss": 1.5589,
+      "step": 3762
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.203341888791026e-05,
+      "loss": 0.4996,
+      "step": 3763
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2029371714905465e-05,
+      "loss": 0.9363,
+      "step": 3764
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2025323671622874e-05,
+      "loss": 0.5598,
+      "step": 3765
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2021274758222415e-05,
+      "loss": 0.4665,
+      "step": 3766
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.201722497486404e-05,
+      "loss": 0.5558,
+      "step": 3767
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.201317432170774e-05,
+      "loss": 0.4595,
+      "step": 3768
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.200912279891353e-05,
+      "loss": 0.9506,
+      "step": 3769
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2005070406641476e-05,
+      "loss": 1.0597,
+      "step": 3770
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.200101714505166e-05,
+      "loss": 0.7195,
+      "step": 3771
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.1996963014304214e-05,
+      "loss": 0.5605,
+      "step": 3772
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.1992908014559294e-05,
+      "loss": 0.5638,
+      "step": 3773
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.198885214597709e-05,
+      "loss": 0.5146,
+      "step": 3774
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.198479540871785e-05,
+      "loss": 0.0186,
+      "step": 3775
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.1980737802941816e-05,
+      "loss": 0.6012,
+      "step": 3776
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.197667932880929e-05,
+      "loss": 0.6563,
+      "step": 3777
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.19726199864806e-05,
+      "loss": 1.0235,
+      "step": 3778
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.196855977611613e-05,
+      "loss": 1.2089,
+      "step": 3779
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1964498697876254e-05,
+      "loss": 0.8276,
+      "step": 3780
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.196043675192142e-05,
+      "loss": 0.7866,
+      "step": 3781
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.195637393841211e-05,
+      "loss": 0.3248,
+      "step": 3782
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.19523102575088e-05,
+      "loss": 0.1243,
+      "step": 3783
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.194824570937204e-05,
+      "loss": 0.726,
+      "step": 3784
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.194418029416241e-05,
+      "loss": 1.0145,
+      "step": 3785
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.19401140120405e-05,
+      "loss": 0.9096,
+      "step": 3786
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.193604686316696e-05,
+      "loss": 1.2137,
+      "step": 3787
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.193197884770246e-05,
+      "loss": 0.7508,
+      "step": 3788
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.19279099658077e-05,
+      "loss": 0.7674,
+      "step": 3789
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.192384021764344e-05,
+      "loss": 0.6921,
+      "step": 3790
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.191976960337045e-05,
+      "loss": 0.4439,
+      "step": 3791
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1915698123149534e-05,
+      "loss": 2.1799,
+      "step": 3792
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.191162577714154e-05,
+      "loss": 0.8316,
+      "step": 3793
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.190755256550736e-05,
+      "loss": 0.4809,
+      "step": 3794
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1903478488407886e-05,
+      "loss": 0.4678,
+      "step": 3795
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.189940354600409e-05,
+      "loss": 0.2402,
+      "step": 3796
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.189532773845693e-05,
+      "loss": 1.1694,
+      "step": 3797
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.189125106592744e-05,
+      "loss": 1.0106,
+      "step": 3798
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.188717352857667e-05,
+      "loss": 0.9424,
+      "step": 3799
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.188309512656568e-05,
+      "loss": 0.9172,
+      "step": 3800
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.7068687677383423,
+      "eval_runtime": 38.6058,
+      "eval_samples_per_second": 1.036,
+      "eval_steps_per_second": 1.036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1879015860055615e-05,
+      "loss": 0.202,
+      "step": 3801
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1874935729207625e-05,
+      "loss": 0.7236,
+      "step": 3802
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.187085473418288e-05,
+      "loss": 0.8824,
+      "step": 3803
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.186677287514262e-05,
+      "loss": 0.9819,
+      "step": 3804
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.18626901522481e-05,
+      "loss": 0.7829,
+      "step": 3805
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.18586065656606e-05,
+      "loss": 0.9443,
+      "step": 3806
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1854522115541437e-05,
+      "loss": 0.1814,
+      "step": 3807
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.185043680205198e-05,
+      "loss": 0.485,
+      "step": 3808
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.184635062535362e-05,
+      "loss": 0.971,
+      "step": 3809
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1842263585607784e-05,
+      "loss": 0.5221,
+      "step": 3810
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.183817568297592e-05,
+      "loss": 0.7243,
+      "step": 3811
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.183408691761954e-05,
+      "loss": 0.2954,
+      "step": 3812
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.182999728970016e-05,
+      "loss": 0.6198,
+      "step": 3813
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.182590679937934e-05,
+      "loss": 0.9995,
+      "step": 3814
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.182181544681868e-05,
+      "loss": 0.6493,
+      "step": 3815
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.181772323217982e-05,
+      "loss": 1.0047,
+      "step": 3816
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.181363015562442e-05,
+      "loss": 0.8992,
+      "step": 3817
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1809536217314154e-05,
+      "loss": 0.4238,
+      "step": 3818
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1805441417410785e-05,
+      "loss": 0.5992,
+      "step": 3819
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.180134575607608e-05,
+      "loss": 0.7634,
+      "step": 3820
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.179724923347181e-05,
+      "loss": 0.9533,
+      "step": 3821
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.179315184975985e-05,
+      "loss": 1.0454,
+      "step": 3822
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.178905360510203e-05,
+      "loss": 0.9379,
+      "step": 3823
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.178495449966027e-05,
+      "loss": 0.3796,
+      "step": 3824
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.178085453359651e-05,
+      "loss": 0.6975,
+      "step": 3825
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.177675370707271e-05,
+      "loss": 0.7348,
+      "step": 3826
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1772652020250886e-05,
+      "loss": 0.526,
+      "step": 3827
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.176854947329307e-05,
+      "loss": 0.9978,
+      "step": 3828
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.176444606636133e-05,
+      "loss": 0.5621,
+      "step": 3829
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.176034179961778e-05,
+      "loss": 0.8661,
+      "step": 3830
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.175623667322457e-05,
+      "loss": 0.1445,
+      "step": 3831
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1752130687343856e-05,
+      "loss": 1.0188,
+      "step": 3832
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.174802384213785e-05,
+      "loss": 0.6349,
+      "step": 3833
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1743916137768806e-05,
+      "loss": 0.7953,
+      "step": 3834
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.173980757439898e-05,
+      "loss": 0.5184,
+      "step": 3835
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.173569815219071e-05,
+      "loss": 1.0153,
+      "step": 3836
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1731587871306305e-05,
+      "loss": 0.5447,
+      "step": 3837
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1727476731908176e-05,
+      "loss": 0.6395,
+      "step": 3838
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1723364734158714e-05,
+      "loss": 1.1625,
+      "step": 3839
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.171925187822037e-05,
+      "loss": 0.5657,
+      "step": 3840
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.171513816425564e-05,
+      "loss": 0.8479,
+      "step": 3841
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1711023592427017e-05,
+      "loss": 0.6761,
+      "step": 3842
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.170690816289705e-05,
+      "loss": 1.2483,
+      "step": 3843
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.170279187582833e-05,
+      "loss": 0.6193,
+      "step": 3844
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.169867473138347e-05,
+      "loss": 0.6833,
+      "step": 3845
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1694556729725116e-05,
+      "loss": 0.8732,
+      "step": 3846
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.169043787101594e-05,
+      "loss": 1.1187,
+      "step": 3847
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.168631815541869e-05,
+      "loss": 1.0256,
+      "step": 3848
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.168219758309608e-05,
+      "loss": 0.2231,
+      "step": 3849
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1678076154210924e-05,
+      "loss": 0.146,
+      "step": 3850
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.167395386892602e-05,
+      "loss": 0.4955,
+      "step": 3851
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.166983072740424e-05,
+      "loss": 0.7829,
+      "step": 3852
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.166570672980846e-05,
+      "loss": 1.1394,
+      "step": 3853
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.166158187630159e-05,
+      "loss": 0.6922,
+      "step": 3854
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.165745616704659e-05,
+      "loss": 1.1855,
+      "step": 3855
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.165332960220646e-05,
+      "loss": 0.9306,
+      "step": 3856
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.164920218194421e-05,
+      "loss": 1.3252,
+      "step": 3857
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.164507390642289e-05,
+      "loss": 0.9182,
+      "step": 3858
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.16409447758056e-05,
+      "loss": 0.3857,
+      "step": 3859
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.163681479025545e-05,
+      "loss": 0.6279,
+      "step": 3860
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1632683949935624e-05,
+      "loss": 0.3312,
+      "step": 3861
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.162855225500928e-05,
+      "loss": 1.3036,
+      "step": 3862
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.162441970563966e-05,
+      "loss": 0.7376,
+      "step": 3863
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.162028630199001e-05,
+      "loss": 0.2734,
+      "step": 3864
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.161615204422364e-05,
+      "loss": 0.6383,
+      "step": 3865
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.161201693250386e-05,
+      "loss": 0.5823,
+      "step": 3866
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1607880966994025e-05,
+      "loss": 1.2311,
+      "step": 3867
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1603744147857545e-05,
+      "loss": 0.3499,
+      "step": 3868
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.159960647525783e-05,
+      "loss": 0.3979,
+      "step": 3869
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.159546794935835e-05,
+      "loss": 0.8119,
+      "step": 3870
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1591328570322596e-05,
+      "loss": 0.4546,
+      "step": 3871
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1587188338314094e-05,
+      "loss": 0.4876,
+      "step": 3872
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1583047253496415e-05,
+      "loss": 0.4697,
+      "step": 3873
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.157890531603314e-05,
+      "loss": 0.7081,
+      "step": 3874
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.15747625260879e-05,
+      "loss": 0.4921,
+      "step": 3875
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.157061888382436e-05,
+      "loss": 0.8587,
+      "step": 3876
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.156647438940622e-05,
+      "loss": 0.8182,
+      "step": 3877
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.15623290429972e-05,
+      "loss": 0.5531,
+      "step": 3878
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.155818284476108e-05,
+      "loss": 0.7432,
+      "step": 3879
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.155403579486163e-05,
+      "loss": 0.7248,
+      "step": 3880
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1549887893462716e-05,
+      "loss": 1.0512,
+      "step": 3881
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.154573914072818e-05,
+      "loss": 1.0135,
+      "step": 3882
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1541589536821914e-05,
+      "loss": 0.122,
+      "step": 3883
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.153743908190787e-05,
+      "loss": 0.766,
+      "step": 3884
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.153328777614999e-05,
+      "loss": 0.7285,
+      "step": 3885
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.152913561971229e-05,
+      "loss": 0.1854,
+      "step": 3886
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.15249826127588e-05,
+      "loss": 1.0266,
+      "step": 3887
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.152082875545358e-05,
+      "loss": 0.9554,
+      "step": 3888
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1516674047960737e-05,
+      "loss": 0.7442,
+      "step": 3889
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.15125184904444e-05,
+      "loss": 0.9784,
+      "step": 3890
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1508362083068723e-05,
+      "loss": 1.0187,
+      "step": 3891
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1504204825997926e-05,
+      "loss": 0.6609,
+      "step": 3892
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.150004671939624e-05,
+      "loss": 0.8382,
+      "step": 3893
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.149588776342793e-05,
+      "loss": 0.46,
+      "step": 3894
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1491727958257285e-05,
+      "loss": 0.7691,
+      "step": 3895
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.148756730404866e-05,
+      "loss": 0.6192,
+      "step": 3896
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1483405800966406e-05,
+      "loss": 1.0954,
+      "step": 3897
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.147924344917493e-05,
+      "loss": 0.4424,
+      "step": 3898
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1475080248838666e-05,
+      "loss": 0.7669,
+      "step": 3899
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1470916200122084e-05,
+      "loss": 1.6543,
+      "step": 3900
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1466751303189696e-05,
+      "loss": 0.9579,
+      "step": 3901
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1462585558206025e-05,
+      "loss": 0.4744,
+      "step": 3902
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1458418965335645e-05,
+      "loss": 0.3825,
+      "step": 3903
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.145425152474315e-05,
+      "loss": 0.9006,
+      "step": 3904
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1450083236593186e-05,
+      "loss": 0.8053,
+      "step": 3905
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1445914101050434e-05,
+      "loss": 0.8426,
+      "step": 3906
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1441744118279565e-05,
+      "loss": 1.4554,
+      "step": 3907
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1437573288445344e-05,
+      "loss": 0.3537,
+      "step": 3908
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.143340161171253e-05,
+      "loss": 1.071,
+      "step": 3909
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.142922908824592e-05,
+      "loss": 0.9288,
+      "step": 3910
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.142505571821036e-05,
+      "loss": 1.1033,
+      "step": 3911
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1420881501770726e-05,
+      "loss": 0.578,
+      "step": 3912
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1416706439091914e-05,
+      "loss": 1.041,
+      "step": 3913
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.141253053033886e-05,
+      "loss": 1.2123,
+      "step": 3914
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.140835377567653e-05,
+      "loss": 0.5769,
+      "step": 3915
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.140417617526995e-05,
+      "loss": 0.0268,
+      "step": 3916
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.139999772928413e-05,
+      "loss": 0.3836,
+      "step": 3917
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1395818437884154e-05,
+      "loss": 1.5328,
+      "step": 3918
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.139163830123513e-05,
+      "loss": 0.5704,
+      "step": 3919
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.138745731950219e-05,
+      "loss": 1.0944,
+      "step": 3920
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.13832754928505e-05,
+      "loss": 0.5706,
+      "step": 3921
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.137909282144528e-05,
+      "loss": 1.0364,
+      "step": 3922
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1374909305451756e-05,
+      "loss": 0.7496,
+      "step": 3923
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.137072494503519e-05,
+      "loss": 0.7298,
+      "step": 3924
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.136653974036092e-05,
+      "loss": 0.6882,
+      "step": 3925
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.136235369159425e-05,
+      "loss": 0.6503,
+      "step": 3926
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.135816679890056e-05,
+      "loss": 0.6383,
+      "step": 3927
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.135397906244527e-05,
+      "loss": 0.5155,
+      "step": 3928
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.134979048239379e-05,
+      "loss": 0.7279,
+      "step": 3929
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.134560105891162e-05,
+      "loss": 0.4909,
+      "step": 3930
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.134141079216425e-05,
+      "loss": 0.4976,
+      "step": 3931
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.133721968231721e-05,
+      "loss": 0.5839,
+      "step": 3932
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.13330277295361e-05,
+      "loss": 0.8637,
+      "step": 3933
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1328834933986485e-05,
+      "loss": 0.4177,
+      "step": 3934
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1324641295834035e-05,
+      "loss": 0.6061,
+      "step": 3935
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.132044681524441e-05,
+      "loss": 0.4126,
+      "step": 3936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.131625149238331e-05,
+      "loss": 0.1857,
+      "step": 3937
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1312055327416476e-05,
+      "loss": 0.4918,
+      "step": 3938
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.130785832050969e-05,
+      "loss": 0.5545,
+      "step": 3939
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.130366047182873e-05,
+      "loss": 0.8627,
+      "step": 3940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.129946178153945e-05,
+      "loss": 0.5403,
+      "step": 3941
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.129526224980773e-05,
+      "loss": 1.1377,
+      "step": 3942
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.129106187679947e-05,
+      "loss": 1.037,
+      "step": 3943
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.128686066268059e-05,
+      "loss": 0.8454,
+      "step": 3944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1282658607617073e-05,
+      "loss": 0.914,
+      "step": 3945
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.127845571177492e-05,
+      "loss": 0.7956,
+      "step": 3946
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1274251975320164e-05,
+      "loss": 0.4805,
+      "step": 3947
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1270047398418885e-05,
+      "loss": 0.9907,
+      "step": 3948
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1265841981237185e-05,
+      "loss": 0.5953,
+      "step": 3949
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.126163572394119e-05,
+      "loss": 0.9378,
+      "step": 3950
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1257428626697084e-05,
+      "loss": 0.0748,
+      "step": 3951
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.125322068967105e-05,
+      "loss": 0.7179,
+      "step": 3952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.124901191302934e-05,
+      "loss": 0.2094,
+      "step": 3953
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.124480229693823e-05,
+      "loss": 0.8815,
+      "step": 3954
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 15816,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "total_flos": 5.928791139896525e+17,
+  "trial_name": null,
+  "trial_params": null
+}