diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8588 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 1400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.2919,
+      "step": 1
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4e-05,
+      "loss": 2.3664,
+      "step": 2
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6e-05,
+      "loss": 2.5039,
+      "step": 3
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 8e-05,
+      "loss": 2.3538,
+      "step": 4
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001,
+      "loss": 2.3211,
+      "step": 5
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.992831541218638e-05,
+      "loss": 2.1066,
+      "step": 6
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.985663082437276e-05,
+      "loss": 2.1203,
+      "step": 7
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.978494623655914e-05,
+      "loss": 2.1782,
+      "step": 8
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.971326164874553e-05,
+      "loss": 1.9107,
+      "step": 9
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.96415770609319e-05,
+      "loss": 2.0875,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.956989247311828e-05,
+      "loss": 1.8469,
+      "step": 11
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.949820788530466e-05,
+      "loss": 1.9198,
+      "step": 12
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.942652329749105e-05,
+      "loss": 1.8764,
+      "step": 13
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.935483870967742e-05,
+      "loss": 1.5738,
+      "step": 14
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.928315412186381e-05,
+      "loss": 1.7708,
+      "step": 15
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.921146953405019e-05,
+      "loss": 1.6138,
+      "step": 16
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.913978494623656e-05,
+      "loss": 1.4795,
+      "step": 17
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.906810035842294e-05,
+      "loss": 1.7306,
+      "step": 18
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.899641577060932e-05,
+      "loss": 1.7529,
+      "step": 19
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.892473118279571e-05,
+      "loss": 1.6125,
+      "step": 20
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.885304659498207e-05,
+      "loss": 1.4683,
+      "step": 21
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.878136200716847e-05,
+      "loss": 1.5695,
+      "step": 22
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.870967741935484e-05,
+      "loss": 1.6194,
+      "step": 23
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.863799283154122e-05,
+      "loss": 1.4536,
+      "step": 24
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.85663082437276e-05,
+      "loss": 1.6215,
+      "step": 25
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.849462365591399e-05,
+      "loss": 1.4687,
+      "step": 26
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.842293906810037e-05,
+      "loss": 1.5447,
+      "step": 27
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.835125448028674e-05,
+      "loss": 1.5991,
+      "step": 28
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.827956989247312e-05,
+      "loss": 1.3734,
+      "step": 29
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.820788530465951e-05,
+      "loss": 1.5202,
+      "step": 30
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.813620071684588e-05,
+      "loss": 1.3936,
+      "step": 31
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.806451612903226e-05,
+      "loss": 1.6183,
+      "step": 32
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.799283154121865e-05,
+      "loss": 1.7974,
+      "step": 33
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.792114695340502e-05,
+      "loss": 1.4831,
+      "step": 34
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.78494623655914e-05,
+      "loss": 1.3791,
+      "step": 35
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 1.5898,
+      "step": 36
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.770609318996417e-05,
+      "loss": 1.4183,
+      "step": 37
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.763440860215054e-05,
+      "loss": 1.4216,
+      "step": 38
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.756272401433693e-05,
+      "loss": 1.5615,
+      "step": 39
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.74910394265233e-05,
+      "loss": 1.4294,
+      "step": 40
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.741935483870968e-05,
+      "loss": 1.3912,
+      "step": 41
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.734767025089606e-05,
+      "loss": 1.7205,
+      "step": 42
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.727598566308245e-05,
+      "loss": 1.2721,
+      "step": 43
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.720430107526883e-05,
+      "loss": 1.1908,
+      "step": 44
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.713261648745519e-05,
+      "loss": 1.2543,
+      "step": 45
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.706093189964158e-05,
+      "loss": 1.2523,
+      "step": 46
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.698924731182796e-05,
+      "loss": 1.2512,
+      "step": 47
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.691756272401434e-05,
+      "loss": 1.0955,
+      "step": 48
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.684587813620072e-05,
+      "loss": 1.6172,
+      "step": 49
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.677419354838711e-05,
+      "loss": 1.345,
+      "step": 50
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.670250896057349e-05,
+      "loss": 1.4891,
+      "step": 51
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.663082437275986e-05,
+      "loss": 1.5186,
+      "step": 52
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.655913978494624e-05,
+      "loss": 1.4853,
+      "step": 53
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.648745519713263e-05,
+      "loss": 1.3203,
+      "step": 54
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.6415770609319e-05,
+      "loss": 1.4817,
+      "step": 55
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.634408602150539e-05,
+      "loss": 1.4074,
+      "step": 56
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.627240143369176e-05,
+      "loss": 1.0041,
+      "step": 57
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.620071684587814e-05,
+      "loss": 0.9736,
+      "step": 58
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.612903225806452e-05,
+      "loss": 1.1534,
+      "step": 59
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.60573476702509e-05,
+      "loss": 1.1337,
+      "step": 60
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.598566308243729e-05,
+      "loss": 1.4201,
+      "step": 61
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.591397849462365e-05,
+      "loss": 1.4095,
+      "step": 62
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.584229390681004e-05,
+      "loss": 1.0914,
+      "step": 63
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.577060931899642e-05,
+      "loss": 0.9977,
+      "step": 64
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 9.56989247311828e-05,
+      "loss": 1.1515,
+      "step": 65
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.562724014336918e-05,
+      "loss": 1.3172,
+      "step": 66
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.555555555555557e-05,
+      "loss": 0.8516,
+      "step": 67
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.548387096774195e-05,
+      "loss": 1.1091,
+      "step": 68
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.541218637992831e-05,
+      "loss": 0.9911,
+      "step": 69
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.53405017921147e-05,
+      "loss": 0.9427,
+      "step": 70
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.9063950777053833,
+      "eval_runtime": 29.2016,
+      "eval_samples_per_second": 4.109,
+      "eval_steps_per_second": 0.514,
+      "step": 70
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.526881720430108e-05,
+      "loss": 0.7439,
+      "step": 71
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.519713261648746e-05,
+      "loss": 0.6516,
+      "step": 72
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.512544802867383e-05,
+      "loss": 0.8568,
+      "step": 73
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.505376344086023e-05,
+      "loss": 0.7717,
+      "step": 74
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.49820788530466e-05,
+      "loss": 0.7564,
+      "step": 75
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.491039426523298e-05,
+      "loss": 0.593,
+      "step": 76
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.483870967741936e-05,
+      "loss": 0.8772,
+      "step": 77
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.476702508960575e-05,
+      "loss": 0.5562,
+      "step": 78
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.469534050179211e-05,
+      "loss": 0.7107,
+      "step": 79
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.46236559139785e-05,
+      "loss": 0.6734,
+      "step": 80
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.455197132616488e-05,
+      "loss": 0.8054,
+      "step": 81
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 9.448028673835126e-05,
+      "loss": 0.4643,
+      "step": 82
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 9.440860215053764e-05,
+      "loss": 0.5161,
+      "step": 83
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.433691756272402e-05,
+      "loss": 0.3869,
+      "step": 84
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.42652329749104e-05,
+      "loss": 0.5715,
+      "step": 85
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.419354838709677e-05,
+      "loss": 0.3854,
+      "step": 86
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.412186379928316e-05,
+      "loss": 0.4301,
+      "step": 87
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.405017921146954e-05,
+      "loss": 0.5839,
+      "step": 88
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.397849462365592e-05,
+      "loss": 0.3165,
+      "step": 89
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.39068100358423e-05,
+      "loss": 0.4854,
+      "step": 90
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.383512544802869e-05,
+      "loss": 0.7052,
+      "step": 91
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 9.376344086021506e-05,
+      "loss": 0.6839,
+      "step": 92
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.369175627240144e-05,
+      "loss": 0.6856,
+      "step": 93
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 9.362007168458782e-05,
+      "loss": 0.4064,
+      "step": 94
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 9.35483870967742e-05,
+      "loss": 0.3277,
+      "step": 95
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.347670250896057e-05,
+      "loss": 0.4991,
+      "step": 96
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 9.340501792114695e-05,
+      "loss": 0.5538,
+      "step": 97
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 0.3563,
+      "step": 98
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 9.326164874551971e-05,
+      "loss": 0.5073,
+      "step": 99
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.31899641577061e-05,
+      "loss": 0.4787,
+      "step": 100
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.311827956989248e-05,
+      "loss": 0.4723,
+      "step": 101
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.304659498207887e-05,
+      "loss": 0.4211,
+      "step": 102
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.297491039426523e-05,
+      "loss": 0.5097,
+      "step": 103
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.290322580645162e-05,
+      "loss": 0.4332,
+      "step": 104
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.2831541218638e-05,
+      "loss": 0.3212,
+      "step": 105
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.275985663082438e-05,
+      "loss": 0.5159,
+      "step": 106
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.268817204301076e-05,
+      "loss": 0.4215,
+      "step": 107
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.261648745519713e-05,
+      "loss": 0.2474,
+      "step": 108
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.254480286738351e-05,
+      "loss": 0.7533,
+      "step": 109
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.247311827956989e-05,
+      "loss": 0.409,
+      "step": 110
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.240143369175628e-05,
+      "loss": 0.2315,
+      "step": 111
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.232974910394266e-05,
+      "loss": 0.3415,
+      "step": 112
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.225806451612904e-05,
+      "loss": 0.4312,
+      "step": 113
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.218637992831541e-05,
+      "loss": 0.2231,
+      "step": 114
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.21146953405018e-05,
+      "loss": 0.2538,
+      "step": 115
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.204301075268817e-05,
+      "loss": 0.2535,
+      "step": 116
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.197132616487456e-05,
+      "loss": 0.2562,
+      "step": 117
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.189964157706094e-05,
+      "loss": 0.2448,
+      "step": 118
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.182795698924731e-05,
+      "loss": 0.2333,
+      "step": 119
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.175627240143369e-05,
+      "loss": 0.2554,
+      "step": 120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.168458781362007e-05,
+      "loss": 0.458,
+      "step": 121
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 9.161290322580646e-05,
+      "loss": 0.1911,
+      "step": 122
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 9.154121863799283e-05,
+      "loss": 0.3257,
+      "step": 123
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.146953405017922e-05,
+      "loss": 0.368,
+      "step": 124
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.13978494623656e-05,
+      "loss": 0.2023,
+      "step": 125
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 9.132616487455197e-05,
+      "loss": 0.2151,
+      "step": 126
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 9.125448028673835e-05,
+      "loss": 0.2101,
+      "step": 127
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.118279569892474e-05,
+      "loss": 0.2763,
+      "step": 128
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.111111111111112e-05,
+      "loss": 0.2098,
+      "step": 129
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.10394265232975e-05,
+      "loss": 0.303,
+      "step": 130
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.096774193548387e-05,
+      "loss": 0.5747,
+      "step": 131
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.089605734767026e-05,
+      "loss": 0.2052,
+      "step": 132
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.082437275985663e-05,
+      "loss": 0.4412,
+      "step": 133
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.0752688172043e-05,
+      "loss": 0.2414,
+      "step": 134
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.06810035842294e-05,
+      "loss": 0.2269,
+      "step": 135
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 9.060931899641578e-05,
+      "loss": 0.2451,
+      "step": 136
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 9.053763440860215e-05,
+      "loss": 0.1992,
+      "step": 137
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.046594982078853e-05,
+      "loss": 0.2655,
+      "step": 138
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.039426523297492e-05,
+      "loss": 0.306,
+      "step": 139
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.032258064516129e-05,
+      "loss": 0.2001,
+      "step": 140
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 2.630777597427368,
+      "eval_runtime": 29.2109,
+      "eval_samples_per_second": 4.108,
+      "eval_steps_per_second": 0.514,
+      "step": 140
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.025089605734768e-05,
+      "loss": 0.1669,
+      "step": 141
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.017921146953405e-05,
+      "loss": 0.1358,
+      "step": 142
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 9.010752688172043e-05,
+      "loss": 0.2005,
+      "step": 143
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.003584229390681e-05,
+      "loss": 0.1524,
+      "step": 144
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 8.99641577060932e-05,
+      "loss": 0.1667,
+      "step": 145
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 8.989247311827958e-05,
+      "loss": 0.1812,
+      "step": 146
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 8.982078853046594e-05,
+      "loss": 0.1708,
+      "step": 147
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 8.974910394265233e-05,
+      "loss": 0.1979,
+      "step": 148
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 8.967741935483871e-05,
+      "loss": 0.1459,
+      "step": 149
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 8.960573476702509e-05,
+      "loss": 0.1768,
+      "step": 150
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 8.953405017921147e-05,
+      "loss": 0.1606,
+      "step": 151
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 8.946236559139786e-05,
+      "loss": 0.1625,
+      "step": 152
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.939068100358424e-05,
+      "loss": 0.1414,
+      "step": 153
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 8.931899641577061e-05,
+      "loss": 0.1508,
+      "step": 154
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.924731182795699e-05,
+      "loss": 0.1811,
+      "step": 155
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.917562724014338e-05,
+      "loss": 0.1333,
+      "step": 156
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.910394265232975e-05,
+      "loss": 0.2154,
+      "step": 157
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.903225806451614e-05,
+      "loss": 0.1688,
+      "step": 158
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.896057347670252e-05,
+      "loss": 0.1293,
+      "step": 159
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.1286,
+      "step": 160
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.881720430107527e-05,
+      "loss": 0.1673,
+      "step": 161
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.874551971326165e-05,
+      "loss": 0.1949,
+      "step": 162
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.867383512544804e-05,
+      "loss": 0.1398,
+      "step": 163
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 8.86021505376344e-05,
+      "loss": 0.1165,
+      "step": 164
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.85304659498208e-05,
+      "loss": 0.1639,
+      "step": 165
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.845878136200717e-05,
+      "loss": 0.2318,
+      "step": 166
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.838709677419355e-05,
+      "loss": 0.1378,
+      "step": 167
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.831541218637993e-05,
+      "loss": 0.1299,
+      "step": 168
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.824372759856632e-05,
+      "loss": 0.1556,
+      "step": 169
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.81720430107527e-05,
+      "loss": 0.1871,
+      "step": 170
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.810035842293907e-05,
+      "loss": 0.1254,
+      "step": 171
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.802867383512545e-05,
+      "loss": 0.1314,
+      "step": 172
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.795698924731183e-05,
+      "loss": 0.1352,
+      "step": 173
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.788530465949821e-05,
+      "loss": 0.1462,
+      "step": 174
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.781362007168459e-05,
+      "loss": 0.1643,
+      "step": 175
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.774193548387098e-05,
+      "loss": 0.1427,
+      "step": 176
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.767025089605735e-05,
+      "loss": 0.1304,
+      "step": 177
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.759856630824373e-05,
+      "loss": 0.1738,
+      "step": 178
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 8.752688172043011e-05,
+      "loss": 0.1636,
+      "step": 179
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 8.74551971326165e-05,
+      "loss": 0.147,
+      "step": 180
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 8.738351254480286e-05,
+      "loss": 0.162,
+      "step": 181
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.731182795698926e-05,
+      "loss": 0.1702,
+      "step": 182
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.724014336917563e-05,
+      "loss": 0.1243,
+      "step": 183
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.716845878136201e-05,
+      "loss": 0.2052,
+      "step": 184
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.709677419354839e-05,
+      "loss": 0.1606,
+      "step": 185
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8.702508960573477e-05,
+      "loss": 0.1722,
+      "step": 186
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 8.695340501792116e-05,
+      "loss": 0.1113,
+      "step": 187
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 8.688172043010752e-05,
+      "loss": 0.1679,
+      "step": 188
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 8.681003584229391e-05,
+      "loss": 0.1209,
+      "step": 189
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 8.673835125448029e-05,
+      "loss": 0.1685,
+      "step": 190
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.2003,
+      "step": 191
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 8.659498207885305e-05,
+      "loss": 0.1669,
+      "step": 192
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.652329749103944e-05,
+      "loss": 0.1431,
+      "step": 193
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.645161290322581e-05,
+      "loss": 0.1572,
+      "step": 194
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.637992831541219e-05,
+      "loss": 0.1185,
+      "step": 195
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.630824372759857e-05,
+      "loss": 0.1722,
+      "step": 196
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.623655913978495e-05,
+      "loss": 0.1286,
+      "step": 197
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.616487455197133e-05,
+      "loss": 0.1147,
+      "step": 198
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 8.60931899641577e-05,
+      "loss": 0.1638,
+      "step": 199
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.60215053763441e-05,
+      "loss": 0.1337,
+      "step": 200
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.594982078853047e-05,
+      "loss": 0.1632,
+      "step": 201
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.587813620071685e-05,
+      "loss": 0.1508,
+      "step": 202
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.580645161290323e-05,
+      "loss": 0.1468,
+      "step": 203
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.573476702508962e-05,
+      "loss": 0.178,
+      "step": 204
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.566308243727598e-05,
+      "loss": 0.1672,
+      "step": 205
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 8.559139784946237e-05,
+      "loss": 0.153,
+      "step": 206
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.551971326164875e-05,
+      "loss": 0.1692,
+      "step": 207
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.544802867383513e-05,
+      "loss": 0.1739,
+      "step": 208
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.53763440860215e-05,
+      "loss": 0.1266,
+      "step": 209
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.530465949820788e-05,
+      "loss": 0.1129,
+      "step": 210
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 2.900357484817505,
+      "eval_runtime": 29.2063,
+      "eval_samples_per_second": 4.109,
+      "eval_steps_per_second": 0.514,
+      "step": 210
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 8.523297491039428e-05,
+      "loss": 0.097,
+      "step": 211
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 8.516129032258064e-05,
+      "loss": 0.0781,
+      "step": 212
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 8.508960573476703e-05,
+      "loss": 0.0905,
+      "step": 213
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 8.501792114695341e-05,
+      "loss": 0.0957,
+      "step": 214
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 8.494623655913979e-05,
+      "loss": 0.1099,
+      "step": 215
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 8.487455197132616e-05,
+      "loss": 0.1027,
+      "step": 216
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 8.480286738351255e-05,
+      "loss": 0.1097,
+      "step": 217
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 8.473118279569893e-05,
+      "loss": 0.1255,
+      "step": 218
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.465949820788531e-05,
+      "loss": 0.0926,
+      "step": 219
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.458781362007169e-05,
+      "loss": 0.101,
+      "step": 220
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.451612903225808e-05,
+      "loss": 0.0964,
+      "step": 221
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 0.0875,
+      "step": 222
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 8.437275985663082e-05,
+      "loss": 0.1222,
+      "step": 223
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 8.430107526881721e-05,
+      "loss": 0.0802,
+      "step": 224
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 8.422939068100359e-05,
+      "loss": 0.1002,
+      "step": 225
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 8.415770609318997e-05,
+      "loss": 0.1212,
+      "step": 226
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 8.408602150537634e-05,
+      "loss": 0.075,
+      "step": 227
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 8.401433691756274e-05,
+      "loss": 0.1154,
+      "step": 228
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 8.39426523297491e-05,
+      "loss": 0.0961,
+      "step": 229
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.387096774193549e-05,
+      "loss": 0.1074,
+      "step": 230
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.379928315412187e-05,
+      "loss": 0.1138,
+      "step": 231
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.372759856630825e-05,
+      "loss": 0.0955,
+      "step": 232
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.365591397849462e-05,
+      "loss": 0.1033,
+      "step": 233
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.358422939068102e-05,
+      "loss": 0.0716,
+      "step": 234
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 8.351254480286739e-05,
+      "loss": 0.0964,
+      "step": 235
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.344086021505376e-05,
+      "loss": 0.0918,
+      "step": 236
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 8.336917562724015e-05,
+      "loss": 0.1311,
+      "step": 237
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 8.329749103942653e-05,
+      "loss": 0.0905,
+      "step": 238
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 8.32258064516129e-05,
+      "loss": 0.0874,
+      "step": 239
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.315412186379928e-05,
+      "loss": 0.0739,
+      "step": 240
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.308243727598567e-05,
+      "loss": 0.1231,
+      "step": 241
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.301075268817205e-05,
+      "loss": 0.0936,
+      "step": 242
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.293906810035843e-05,
+      "loss": 0.0927,
+      "step": 243
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.28673835125448e-05,
+      "loss": 0.0757,
+      "step": 244
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.27956989247312e-05,
+      "loss": 0.1012,
+      "step": 245
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.272401433691756e-05,
+      "loss": 0.0965,
+      "step": 246
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 8.265232974910395e-05,
+      "loss": 0.1043,
+      "step": 247
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 8.258064516129033e-05,
+      "loss": 0.094,
+      "step": 248
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 8.250896057347671e-05,
+      "loss": 0.0806,
+      "step": 249
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.243727598566309e-05,
+      "loss": 0.0944,
+      "step": 250
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 8.236559139784946e-05,
+      "loss": 0.1048,
+      "step": 251
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 8.229390681003585e-05,
+      "loss": 0.1076,
+      "step": 252
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 0.1038,
+      "step": 253
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.215053763440861e-05,
+      "loss": 0.0992,
+      "step": 254
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 8.207885304659499e-05,
+      "loss": 0.097,
+      "step": 255
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.200716845878136e-05,
+      "loss": 0.1338,
+      "step": 256
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.193548387096774e-05,
+      "loss": 0.1194,
+      "step": 257
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.186379928315413e-05,
+      "loss": 0.1079,
+      "step": 258
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.179211469534051e-05,
+      "loss": 0.0796,
+      "step": 259
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.172043010752689e-05,
+      "loss": 0.1322,
+      "step": 260
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.164874551971327e-05,
+      "loss": 0.1085,
+      "step": 261
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.157706093189964e-05,
+      "loss": 0.0989,
+      "step": 262
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.150537634408602e-05,
+      "loss": 0.0964,
+      "step": 263
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.14336917562724e-05,
+      "loss": 0.1235,
+      "step": 264
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.136200716845879e-05,
+      "loss": 0.1125,
+      "step": 265
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 8.129032258064517e-05,
+      "loss": 0.0965,
+      "step": 266
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 8.121863799283155e-05,
+      "loss": 0.1099,
+      "step": 267
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.114695340501792e-05,
+      "loss": 0.1224,
+      "step": 268
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.107526881720431e-05,
+      "loss": 0.0729,
+      "step": 269
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 8.100358422939068e-05,
+      "loss": 0.1326,
+      "step": 270
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 8.093189964157707e-05,
+      "loss": 0.1136,
+      "step": 271
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 8.086021505376345e-05,
+      "loss": 0.0966,
+      "step": 272
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 8.078853046594983e-05,
+      "loss": 0.1128,
+      "step": 273
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 8.07168458781362e-05,
+      "loss": 0.0898,
+      "step": 274
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 8.064516129032258e-05,
+      "loss": 0.139,
+      "step": 275
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 8.057347670250897e-05,
+      "loss": 0.097,
+      "step": 276
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 8.050179211469534e-05,
+      "loss": 0.1424,
+      "step": 277
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 8.043010752688173e-05,
+      "loss": 0.1348,
+      "step": 278
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.03584229390681e-05,
+      "loss": 0.0942,
+      "step": 279
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 8.028673835125448e-05,
+      "loss": 0.0803,
+      "step": 280
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 3.033568859100342,
+      "eval_runtime": 29.2135,
+      "eval_samples_per_second": 4.108,
+      "eval_steps_per_second": 0.513,
+      "step": 280
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 8.021505376344086e-05,
+      "loss": 0.0755,
+      "step": 281
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 8.014336917562725e-05,
+      "loss": 0.056,
+      "step": 282
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.007168458781363e-05,
+      "loss": 0.081,
+      "step": 283
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8e-05,
+      "loss": 0.0863,
+      "step": 284
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 7.992831541218638e-05,
+      "loss": 0.0525,
+      "step": 285
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 7.985663082437278e-05,
+      "loss": 0.056,
+      "step": 286
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 7.978494623655914e-05,
+      "loss": 0.0897,
+      "step": 287
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 7.971326164874552e-05,
+      "loss": 0.0475,
+      "step": 288
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 7.964157706093191e-05,
+      "loss": 0.0704,
+      "step": 289
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 7.956989247311829e-05,
+      "loss": 0.0713,
+      "step": 290
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.949820788530466e-05,
+      "loss": 0.0592,
+      "step": 291
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.942652329749104e-05,
+      "loss": 0.0809,
+      "step": 292
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.935483870967743e-05,
+      "loss": 0.0689,
+      "step": 293
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 7.92831541218638e-05,
+      "loss": 0.0628,
+      "step": 294
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 7.921146953405019e-05,
+      "loss": 0.0584,
+      "step": 295
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 7.913978494623657e-05,
+      "loss": 0.0833,
+      "step": 296
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 7.906810035842294e-05,
+      "loss": 0.063,
+      "step": 297
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 7.899641577060932e-05,
+      "loss": 0.0608,
+      "step": 298
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.892473118279571e-05,
+      "loss": 0.0614,
+      "step": 299
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.885304659498209e-05,
+      "loss": 0.0634,
+      "step": 300
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 7.878136200716845e-05,
+      "loss": 0.0734,
+      "step": 301
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 7.870967741935484e-05,
+      "loss": 0.0673,
+      "step": 302
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 7.863799283154122e-05,
+      "loss": 0.0539,
+      "step": 303
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 7.85663082437276e-05,
+      "loss": 0.0743,
+      "step": 304
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 7.849462365591398e-05,
+      "loss": 0.0752,
+      "step": 305
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 7.842293906810037e-05,
+      "loss": 0.0601,
+      "step": 306
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.835125448028673e-05,
+      "loss": 0.0667,
+      "step": 307
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 7.827956989247312e-05,
+      "loss": 0.0802,
+      "step": 308
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.82078853046595e-05,
+      "loss": 0.0829,
+      "step": 309
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.81362007168459e-05,
+      "loss": 0.0643,
+      "step": 310
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 7.806451612903226e-05,
+      "loss": 0.0659,
+      "step": 311
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 7.799283154121865e-05,
+      "loss": 0.0587,
+      "step": 312
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 7.792114695340503e-05,
+      "loss": 0.0464,
+      "step": 313
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 7.784946236559139e-05,
+      "loss": 0.0794,
+      "step": 314
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.057,
+      "step": 315
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 7.770609318996416e-05,
+      "loss": 0.0612,
+      "step": 316
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 7.763440860215054e-05,
+      "loss": 0.0619,
+      "step": 317
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 7.756272401433691e-05,
+      "loss": 0.0646,
+      "step": 318
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 7.74910394265233e-05,
+      "loss": 0.0694,
+      "step": 319
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 7.741935483870968e-05,
+      "loss": 0.075,
+      "step": 320
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 7.734767025089606e-05,
+      "loss": 0.0767,
+      "step": 321
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 7.727598566308244e-05,
+      "loss": 0.062,
+      "step": 322
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 7.720430107526883e-05,
+      "loss": 0.0654,
+      "step": 323
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.71326164874552e-05,
+      "loss": 0.0712,
+      "step": 324
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.706093189964157e-05,
+      "loss": 0.0545,
+      "step": 325
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.698924731182796e-05,
+      "loss": 0.06,
+      "step": 326
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.691756272401434e-05,
+      "loss": 0.058,
+      "step": 327
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.684587813620072e-05,
+      "loss": 0.0797,
+      "step": 328
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.67741935483871e-05,
+      "loss": 0.0648,
+      "step": 329
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.670250896057349e-05,
+      "loss": 0.0687,
+      "step": 330
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.663082437275985e-05,
+      "loss": 0.0714,
+      "step": 331
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.655913978494624e-05,
+      "loss": 0.0738,
+      "step": 332
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.648745519713262e-05,
+      "loss": 0.0757,
+      "step": 333
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.6415770609319e-05,
+      "loss": 0.074,
+      "step": 334
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.634408602150538e-05,
+      "loss": 0.0899,
+      "step": 335
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.627240143369177e-05,
+      "loss": 0.0913,
+      "step": 336
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.620071684587814e-05,
+      "loss": 0.0754,
+      "step": 337
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.612903225806451e-05,
+      "loss": 0.0738,
+      "step": 338
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.60573476702509e-05,
+      "loss": 0.093,
+      "step": 339
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.598566308243728e-05,
+      "loss": 0.0849,
+      "step": 340
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.591397849462365e-05,
+      "loss": 0.0708,
+      "step": 341
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.584229390681003e-05,
+      "loss": 0.0894,
+      "step": 342
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.577060931899642e-05,
+      "loss": 0.0949,
+      "step": 343
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.56989247311828e-05,
+      "loss": 0.0684,
+      "step": 344
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.562724014336918e-05,
+      "loss": 0.0901,
+      "step": 345
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.555555555555556e-05,
+      "loss": 0.0637,
+      "step": 346
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.548387096774195e-05,
+      "loss": 0.0783,
+      "step": 347
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.541218637992831e-05,
+      "loss": 0.0674,
+      "step": 348
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.53405017921147e-05,
+      "loss": 0.0866,
+      "step": 349
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.526881720430108e-05,
+      "loss": 0.0665,
+      "step": 350
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 3.0398199558258057,
+      "eval_runtime": 29.2059,
+      "eval_samples_per_second": 4.109,
+      "eval_steps_per_second": 0.514,
+      "step": 350
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.519713261648746e-05,
+      "loss": 0.0533,
+      "step": 351
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.512544802867384e-05,
+      "loss": 0.0481,
+      "step": 352
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.505376344086021e-05,
+      "loss": 0.0618,
+      "step": 353
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.49820788530466e-05,
+      "loss": 0.0384,
+      "step": 354
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.491039426523297e-05,
+      "loss": 0.0684,
+      "step": 355
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.483870967741936e-05,
+      "loss": 0.0479,
+      "step": 356
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.476702508960574e-05,
+      "loss": 0.0481,
+      "step": 357
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.469534050179212e-05,
+      "loss": 0.0457,
+      "step": 358
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.462365591397849e-05,
+      "loss": 0.0407,
+      "step": 359
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.455197132616488e-05,
+      "loss": 0.0442,
+      "step": 360
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.448028673835126e-05,
+      "loss": 0.065,
+      "step": 361
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.440860215053764e-05,
+      "loss": 0.0533,
+      "step": 362
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.433691756272402e-05,
+      "loss": 0.0559,
+      "step": 363
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 7.42652329749104e-05,
+      "loss": 0.045,
+      "step": 364
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 7.419354838709677e-05,
+      "loss": 0.0498,
+      "step": 365
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 7.412186379928315e-05,
+      "loss": 0.0445,
+      "step": 366
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 7.405017921146954e-05,
+      "loss": 0.0536,
+      "step": 367
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 7.397849462365592e-05,
+      "loss": 0.0326,
+      "step": 368
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 7.39068100358423e-05,
+      "loss": 0.0532,
+      "step": 369
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 7.383512544802867e-05,
+      "loss": 0.0389,
+      "step": 370
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 7.376344086021507e-05,
+      "loss": 0.0578,
+      "step": 371
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 7.369175627240143e-05,
+      "loss": 0.0425,
+      "step": 372
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 7.362007168458782e-05,
+      "loss": 0.0514,
+      "step": 373
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 7.35483870967742e-05,
+      "loss": 0.0473,
+      "step": 374
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 7.347670250896058e-05,
+      "loss": 0.0527,
+      "step": 375
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 7.340501792114695e-05,
+      "loss": 0.0627,
+      "step": 376
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.0424,
+      "step": 377
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 7.326164874551972e-05,
+      "loss": 0.064,
+      "step": 378
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 7.318996415770609e-05,
+      "loss": 0.0523,
+      "step": 379
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 7.311827956989248e-05,
+      "loss": 0.056,
+      "step": 380
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 7.304659498207886e-05,
+      "loss": 0.0489,
+      "step": 381
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 7.297491039426523e-05,
+      "loss": 0.0569,
+      "step": 382
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 7.290322580645161e-05,
+      "loss": 0.07,
+      "step": 383
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 7.2831541218638e-05,
+      "loss": 0.0555,
+      "step": 384
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 7.275985663082438e-05,
+      "loss": 0.0479,
+      "step": 385
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 7.268817204301076e-05,
+      "loss": 0.0493,
+      "step": 386
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 7.261648745519714e-05,
+      "loss": 0.0703,
+      "step": 387
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 7.254480286738353e-05,
+      "loss": 0.0446,
+      "step": 388
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 7.247311827956989e-05,
+      "loss": 0.0477,
+      "step": 389
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 7.240143369175627e-05,
+      "loss": 0.0435,
+      "step": 390
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 7.232974910394266e-05,
+      "loss": 0.0527,
+      "step": 391
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 7.225806451612904e-05,
+      "loss": 0.0674,
+      "step": 392
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 7.218637992831541e-05,
+      "loss": 0.0377,
+      "step": 393
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 7.211469534050179e-05,
+      "loss": 0.0392,
+      "step": 394
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 7.204301075268818e-05,
+      "loss": 0.0566,
+      "step": 395
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 7.197132616487455e-05,
+      "loss": 0.0457,
+      "step": 396
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 7.189964157706094e-05,
+      "loss": 0.0562,
+      "step": 397
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 7.182795698924732e-05,
+      "loss": 0.0365,
+      "step": 398
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 7.17562724014337e-05,
+      "loss": 0.0502,
+      "step": 399
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 7.168458781362007e-05,
+      "loss": 0.0587,
+      "step": 400
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 7.161290322580646e-05,
+      "loss": 0.042,
+      "step": 401
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 7.154121863799284e-05,
+      "loss": 0.0493,
+      "step": 402
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 7.14695340501792e-05,
+      "loss": 0.0517,
+      "step": 403
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 7.13978494623656e-05,
+      "loss": 0.0534,
+      "step": 404
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 7.132616487455197e-05,
+      "loss": 0.0429,
+      "step": 405
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 7.125448028673835e-05,
+      "loss": 0.0468,
+      "step": 406
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 7.118279569892473e-05,
+      "loss": 0.0498,
+      "step": 407
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 7.111111111111112e-05,
+      "loss": 0.0539,
+      "step": 408
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 7.10394265232975e-05,
+      "loss": 0.0503,
+      "step": 409
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 7.096774193548388e-05,
+      "loss": 0.0576,
+      "step": 410
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 7.089605734767025e-05,
+      "loss": 0.0692,
+      "step": 411
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 7.082437275985664e-05,
+      "loss": 0.0446,
+      "step": 412
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 7.075268817204301e-05,
+      "loss": 0.0464,
+      "step": 413
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 7.06810035842294e-05,
+      "loss": 0.0496,
+      "step": 414
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 7.060931899641578e-05,
+      "loss": 0.0622,
+      "step": 415
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 7.053763440860215e-05,
+      "loss": 0.0487,
+      "step": 416
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 7.046594982078853e-05,
+      "loss": 0.0819,
+      "step": 417
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 7.039426523297491e-05,
+      "loss": 0.0515,
+      "step": 418
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 7.03225806451613e-05,
+      "loss": 0.0506,
+      "step": 419
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 7.025089605734767e-05,
+      "loss": 0.0396,
+      "step": 420
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 3.0768535137176514,
+      "eval_runtime": 29.2391,
+      "eval_samples_per_second": 4.104,
+      "eval_steps_per_second": 0.513,
+      "step": 420
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 7.017921146953406e-05,
+      "loss": 0.0488,
+      "step": 421
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 7.010752688172043e-05,
+      "loss": 0.0502,
+      "step": 422
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 7.003584229390681e-05,
+      "loss": 0.0448,
+      "step": 423
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 6.996415770609319e-05,
+      "loss": 0.0394,
+      "step": 424
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 6.989247311827958e-05,
+      "loss": 0.0339,
+      "step": 425
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 6.982078853046596e-05,
+      "loss": 0.0545,
+      "step": 426
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 6.974910394265234e-05,
+      "loss": 0.0426,
+      "step": 427
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 6.967741935483871e-05,
+      "loss": 0.0497,
+      "step": 428
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 6.960573476702509e-05,
+      "loss": 0.0474,
+      "step": 429
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 6.953405017921147e-05,
+      "loss": 0.0363,
+      "step": 430
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 6.946236559139785e-05,
+      "loss": 0.0407,
+      "step": 431
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 6.939068100358424e-05,
+      "loss": 0.0302,
+      "step": 432
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 6.931899641577062e-05,
+      "loss": 0.0422,
+      "step": 433
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 6.924731182795699e-05,
+      "loss": 0.0396,
+      "step": 434
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 6.917562724014337e-05,
+      "loss": 0.0476,
+      "step": 435
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 6.910394265232976e-05,
+      "loss": 0.0305,
+      "step": 436
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 6.903225806451613e-05,
+      "loss": 0.0492,
+      "step": 437
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 6.896057347670252e-05,
+      "loss": 0.0454,
+      "step": 438
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 6.88888888888889e-05,
+      "loss": 0.0423,
+      "step": 439
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 6.881720430107527e-05,
+      "loss": 0.0493,
+      "step": 440
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 6.874551971326165e-05,
+      "loss": 0.0362,
+      "step": 441
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 6.867383512544803e-05,
+      "loss": 0.0508,
+      "step": 442
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 6.860215053763442e-05,
+      "loss": 0.0374,
+      "step": 443
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 6.853046594982078e-05,
+      "loss": 0.0447,
+      "step": 444
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 6.845878136200717e-05,
+      "loss": 0.0343,
+      "step": 445
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 6.838709677419355e-05,
+      "loss": 0.049,
+      "step": 446
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 6.831541218637993e-05,
+      "loss": 0.0429,
+      "step": 447
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 6.824372759856631e-05,
+      "loss": 0.038,
+      "step": 448
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 6.81720430107527e-05,
+      "loss": 0.0396,
+      "step": 449
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 6.810035842293908e-05,
+      "loss": 0.0355,
+      "step": 450
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 6.802867383512545e-05,
+      "loss": 0.0349,
+      "step": 451
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 6.795698924731183e-05,
+      "loss": 0.0478,
+      "step": 452
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 6.788530465949821e-05,
+      "loss": 0.0359,
+      "step": 453
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 6.781362007168459e-05,
+      "loss": 0.0338,
+      "step": 454
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 6.774193548387096e-05,
+      "loss": 0.0393,
+      "step": 455
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 6.767025089605736e-05,
+      "loss": 0.0331,
+      "step": 456
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 6.759856630824373e-05,
+      "loss": 0.0368,
+      "step": 457
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 6.752688172043011e-05,
+      "loss": 0.0353,
+      "step": 458
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 6.745519713261649e-05,
+      "loss": 0.0534,
+      "step": 459
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 6.738351254480288e-05,
+      "loss": 0.0422,
+      "step": 460
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 6.731182795698924e-05,
+      "loss": 0.0528,
+      "step": 461
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 6.724014336917564e-05,
+      "loss": 0.0408,
+      "step": 462
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 6.716845878136201e-05,
+      "loss": 0.0351,
+      "step": 463
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 6.709677419354839e-05,
+      "loss": 0.0353,
+      "step": 464
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 6.702508960573477e-05,
+      "loss": 0.0283,
+      "step": 465
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 6.695340501792115e-05,
+      "loss": 0.0374,
+      "step": 466
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 6.688172043010754e-05,
+      "loss": 0.0542,
+      "step": 467
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 6.68100358422939e-05,
+      "loss": 0.0364,
+      "step": 468
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 6.673835125448029e-05,
+      "loss": 0.0391,
+      "step": 469
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.0888,
+      "step": 470
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 6.659498207885305e-05,
+      "loss": 0.0317,
+      "step": 471
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 6.652329749103943e-05,
+      "loss": 0.0484,
+      "step": 472
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 6.645161290322582e-05,
+      "loss": 0.0424,
+      "step": 473
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 6.63799283154122e-05,
+      "loss": 0.0398,
+      "step": 474
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 6.630824372759857e-05,
+      "loss": 0.0367,
+      "step": 475
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 6.623655913978495e-05,
+      "loss": 0.0311,
+      "step": 476
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 6.616487455197134e-05,
+      "loss": 0.0375,
+      "step": 477
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 6.60931899641577e-05,
+      "loss": 0.0507,
+      "step": 478
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 6.602150537634408e-05,
+      "loss": 0.0462,
+      "step": 479
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.594982078853047e-05,
+      "loss": 0.0378,
+      "step": 480
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 6.587813620071685e-05,
+      "loss": 0.0475,
+      "step": 481
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.580645161290323e-05,
+      "loss": 0.033,
+      "step": 482
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 6.57347670250896e-05,
+      "loss": 0.0366,
+      "step": 483
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 6.5663082437276e-05,
+      "loss": 0.04,
+      "step": 484
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 6.559139784946236e-05,
+      "loss": 0.0431,
+      "step": 485
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 6.551971326164875e-05,
+      "loss": 0.0531,
+      "step": 486
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 6.544802867383513e-05,
+      "loss": 0.0307,
+      "step": 487
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 6.537634408602151e-05,
+      "loss": 0.0458,
+      "step": 488
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 6.530465949820789e-05,
+      "loss": 0.0374,
+      "step": 489
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6.523297491039428e-05,
+      "loss": 0.0429,
+      "step": 490
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 3.150350332260132,
+      "eval_runtime": 29.1939,
+      "eval_samples_per_second": 4.11,
+      "eval_steps_per_second": 0.514,
+      "step": 490
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 6.516129032258065e-05,
+      "loss": 0.0268,
+      "step": 491
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 6.508960573476702e-05,
+      "loss": 0.0377,
+      "step": 492
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 6.501792114695341e-05,
+      "loss": 0.0309,
+      "step": 493
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 6.494623655913979e-05,
+      "loss": 0.0334,
+      "step": 494
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 6.487455197132617e-05,
+      "loss": 0.0266,
+      "step": 495
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 6.480286738351254e-05,
+      "loss": 0.0326,
+      "step": 496
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 6.473118279569893e-05,
+      "loss": 0.0319,
+      "step": 497
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 6.465949820788531e-05,
+      "loss": 0.0269,
+      "step": 498
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 6.458781362007169e-05,
+      "loss": 0.0335,
+      "step": 499
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 6.451612903225807e-05,
+      "loss": 0.0333,
+      "step": 500
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 6.444444444444446e-05,
+      "loss": 0.0297,
+      "step": 501
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 6.437275985663082e-05,
+      "loss": 0.0399,
+      "step": 502
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 6.430107526881721e-05,
+      "loss": 0.0341,
+      "step": 503
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 6.422939068100359e-05,
+      "loss": 0.0243,
+      "step": 504
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 6.415770609318996e-05,
+      "loss": 0.0423,
+      "step": 505
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 6.408602150537635e-05,
+      "loss": 0.0373,
+      "step": 506
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 6.401433691756272e-05,
+      "loss": 0.0416,
+      "step": 507
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 6.39426523297491e-05,
+      "loss": 0.0397,
+      "step": 508
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 6.387096774193548e-05,
+      "loss": 0.0279,
+      "step": 509
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 6.379928315412187e-05,
+      "loss": 0.0331,
+      "step": 510
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 6.372759856630825e-05,
+      "loss": 0.0296,
+      "step": 511
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 6.365591397849463e-05,
+      "loss": 0.0356,
+      "step": 512
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 6.3584229390681e-05,
+      "loss": 0.0301,
+      "step": 513
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 6.35125448028674e-05,
+      "loss": 0.0236,
+      "step": 514
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 6.344086021505376e-05,
+      "loss": 0.0299,
+      "step": 515
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 6.336917562724015e-05,
+      "loss": 0.0494,
+      "step": 516
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 6.329749103942653e-05,
+      "loss": 0.0372,
+      "step": 517
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 6.32258064516129e-05,
+      "loss": 0.0275,
+      "step": 518
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 6.315412186379928e-05,
+      "loss": 0.0298,
+      "step": 519
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 6.308243727598566e-05,
+      "loss": 0.0359,
+      "step": 520
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 6.301075268817205e-05,
+      "loss": 0.0318,
+      "step": 521
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 6.293906810035842e-05,
+      "loss": 0.0294,
+      "step": 522
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 6.286738351254481e-05,
+      "loss": 0.0369,
+      "step": 523
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 6.279569892473119e-05,
+      "loss": 0.0305,
+      "step": 524
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 6.272401433691756e-05,
+      "loss": 0.0315,
+      "step": 525
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 6.265232974910394e-05,
+      "loss": 0.033,
+      "step": 526
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 6.258064516129033e-05,
+      "loss": 0.0396,
+      "step": 527
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 6.250896057347671e-05,
+      "loss": 0.0416,
+      "step": 528
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 6.243727598566309e-05,
+      "loss": 0.0376,
+      "step": 529
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 6.236559139784946e-05,
+      "loss": 0.0428,
+      "step": 530
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 6.229390681003584e-05,
+      "loss": 0.0367,
+      "step": 531
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 0.0307,
+      "step": 532
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 6.21505376344086e-05,
+      "loss": 0.0283,
+      "step": 533
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 6.207885304659499e-05,
+      "loss": 0.0329,
+      "step": 534
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 6.200716845878137e-05,
+      "loss": 0.0372,
+      "step": 535
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 6.193548387096774e-05,
+      "loss": 0.0331,
+      "step": 536
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 6.186379928315412e-05,
+      "loss": 0.0283,
+      "step": 537
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 6.179211469534051e-05,
+      "loss": 0.0331,
+      "step": 538
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 6.172043010752688e-05,
+      "loss": 0.0415,
+      "step": 539
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 6.164874551971327e-05,
+      "loss": 0.0349,
+      "step": 540
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 6.157706093189965e-05,
+      "loss": 0.0315,
+      "step": 541
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 6.150537634408602e-05,
+      "loss": 0.0346,
+      "step": 542
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 6.14336917562724e-05,
+      "loss": 0.0349,
+      "step": 543
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 6.136200716845878e-05,
+      "loss": 0.0268,
+      "step": 544
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 6.129032258064517e-05,
+      "loss": 0.0258,
+      "step": 545
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 6.121863799283153e-05,
+      "loss": 0.0347,
+      "step": 546
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 6.114695340501793e-05,
+      "loss": 0.0314,
+      "step": 547
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 6.10752688172043e-05,
+      "loss": 0.0303,
+      "step": 548
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 6.100358422939069e-05,
+      "loss": 0.0392,
+      "step": 549
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 6.093189964157706e-05,
+      "loss": 0.0289,
+      "step": 550
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 6.086021505376345e-05,
+      "loss": 0.0389,
+      "step": 551
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 6.078853046594982e-05,
+      "loss": 0.0365,
+      "step": 552
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 6.071684587813621e-05,
+      "loss": 0.0386,
+      "step": 553
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 6.064516129032258e-05,
+      "loss": 0.0323,
+      "step": 554
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 6.057347670250897e-05,
+      "loss": 0.0402,
+      "step": 555
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 6.0501792114695344e-05,
+      "loss": 0.0351,
+      "step": 556
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 6.0430107526881715e-05,
+      "loss": 0.0323,
+      "step": 557
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 6.0358422939068106e-05,
+      "loss": 0.043,
+      "step": 558
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 6.028673835125448e-05,
+      "loss": 0.0244,
+      "step": 559
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 6.021505376344086e-05,
+      "loss": 0.0318,
+      "step": 560
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 3.2689783573150635,
+      "eval_runtime": 29.2298,
+      "eval_samples_per_second": 4.105,
+      "eval_steps_per_second": 0.513,
+      "step": 560
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 6.014336917562724e-05,
+      "loss": 0.0246,
+      "step": 561
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 6.0071684587813624e-05,
+      "loss": 0.0281,
+      "step": 562
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 6e-05,
+      "loss": 0.025,
+      "step": 563
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 5.9928315412186386e-05,
+      "loss": 0.0283,
+      "step": 564
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 5.9856630824372764e-05,
+      "loss": 0.026,
+      "step": 565
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 5.978494623655915e-05,
+      "loss": 0.037,
+      "step": 566
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 5.971326164874552e-05,
+      "loss": 0.0284,
+      "step": 567
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 5.964157706093191e-05,
+      "loss": 0.0316,
+      "step": 568
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 5.956989247311828e-05,
+      "loss": 0.027,
+      "step": 569
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 5.949820788530466e-05,
+      "loss": 0.0242,
+      "step": 570
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 5.942652329749104e-05,
+      "loss": 0.023,
+      "step": 571
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 5.935483870967742e-05,
+      "loss": 0.0243,
+      "step": 572
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 5.9283154121863805e-05,
+      "loss": 0.034,
+      "step": 573
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 5.9211469534050176e-05,
+      "loss": 0.027,
+      "step": 574
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 5.913978494623657e-05,
+      "loss": 0.0265,
+      "step": 575
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 5.906810035842294e-05,
+      "loss": 0.0269,
+      "step": 576
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 5.899641577060932e-05,
+      "loss": 0.0239,
+      "step": 577
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 5.89247311827957e-05,
+      "loss": 0.0331,
+      "step": 578
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 5.8853046594982085e-05,
+      "loss": 0.0244,
+      "step": 579
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 5.878136200716846e-05,
+      "loss": 0.0257,
+      "step": 580
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 5.870967741935483e-05,
+      "loss": 0.0279,
+      "step": 581
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 5.8637992831541224e-05,
+      "loss": 0.0235,
+      "step": 582
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 5.8566308243727595e-05,
+      "loss": 0.0265,
+      "step": 583
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 5.849462365591398e-05,
+      "loss": 0.0206,
+      "step": 584
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 5.842293906810036e-05,
+      "loss": 0.0278,
+      "step": 585
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 5.835125448028674e-05,
+      "loss": 0.0287,
+      "step": 586
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 5.827956989247312e-05,
+      "loss": 0.0283,
+      "step": 587
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 5.8207885304659504e-05,
+      "loss": 0.0315,
+      "step": 588
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 5.813620071684588e-05,
+      "loss": 0.0289,
+      "step": 589
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 5.8064516129032266e-05,
+      "loss": 0.0273,
+      "step": 590
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 5.799283154121864e-05,
+      "loss": 0.0328,
+      "step": 591
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 5.792114695340503e-05,
+      "loss": 0.0279,
+      "step": 592
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.78494623655914e-05,
+      "loss": 0.0273,
+      "step": 593
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.7777777777777776e-05,
+      "loss": 0.0355,
+      "step": 594
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 5.770609318996416e-05,
+      "loss": 0.0267,
+      "step": 595
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.763440860215054e-05,
+      "loss": 0.0256,
+      "step": 596
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5.756272401433692e-05,
+      "loss": 0.0289,
+      "step": 597
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5.7491039426523294e-05,
+      "loss": 0.0296,
+      "step": 598
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5.7419354838709685e-05,
+      "loss": 0.0245,
+      "step": 599
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.7347670250896056e-05,
+      "loss": 0.0348,
+      "step": 600
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 5.727598566308244e-05,
+      "loss": 0.025,
+      "step": 601
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5.720430107526882e-05,
+      "loss": 0.0278,
+      "step": 602
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.71326164874552e-05,
+      "loss": 0.034,
+      "step": 603
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.706093189964158e-05,
+      "loss": 0.0251,
+      "step": 604
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 5.6989247311827965e-05,
+      "loss": 0.0274,
+      "step": 605
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 5.691756272401434e-05,
+      "loss": 0.0292,
+      "step": 606
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5.684587813620071e-05,
+      "loss": 0.0299,
+      "step": 607
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 5.67741935483871e-05,
+      "loss": 0.0336,
+      "step": 608
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5.6702508960573475e-05,
+      "loss": 0.0332,
+      "step": 609
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5.663082437275986e-05,
+      "loss": 0.0334,
+      "step": 610
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 5.655913978494624e-05,
+      "loss": 0.0332,
+      "step": 611
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 5.648745519713262e-05,
+      "loss": 0.0252,
+      "step": 612
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 5.6415770609319e-05,
+      "loss": 0.027,
+      "step": 613
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 5.6344086021505384e-05,
+      "loss": 0.0356,
+      "step": 614
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 5.6272401433691755e-05,
+      "loss": 0.0308,
+      "step": 615
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 5.6200716845878146e-05,
+      "loss": 0.0293,
+      "step": 616
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 5.612903225806452e-05,
+      "loss": 0.0315,
+      "step": 617
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 5.60573476702509e-05,
+      "loss": 0.0319,
+      "step": 618
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 5.598566308243728e-05,
+      "loss": 0.0246,
+      "step": 619
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 5.5913978494623656e-05,
+      "loss": 0.0263,
+      "step": 620
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 5.584229390681004e-05,
+      "loss": 0.0241,
+      "step": 621
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 5.577060931899641e-05,
+      "loss": 0.0236,
+      "step": 622
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 5.56989247311828e-05,
+      "loss": 0.0257,
+      "step": 623
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5.5627240143369174e-05,
+      "loss": 0.0389,
+      "step": 624
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.0272,
+      "step": 625
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.5483870967741936e-05,
+      "loss": 0.0351,
+      "step": 626
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5.541218637992832e-05,
+      "loss": 0.0288,
+      "step": 627
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.53405017921147e-05,
+      "loss": 0.0295,
+      "step": 628
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.526881720430108e-05,
+      "loss": 0.0295,
+      "step": 629
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.519713261648746e-05,
+      "loss": 0.03,
+      "step": 630
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 3.481818675994873,
+      "eval_runtime": 29.2267,
+      "eval_samples_per_second": 4.106,
+      "eval_steps_per_second": 0.513,
+      "step": 630
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 5.5125448028673844e-05,
+      "loss": 0.0323,
+      "step": 631
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5.5053763440860215e-05,
+      "loss": 0.03,
+      "step": 632
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 5.498207885304659e-05,
+      "loss": 0.0231,
+      "step": 633
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 5.491039426523298e-05,
+      "loss": 0.0247,
+      "step": 634
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5.4838709677419355e-05,
+      "loss": 0.0241,
+      "step": 635
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 5.476702508960574e-05,
+      "loss": 0.0375,
+      "step": 636
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 5.469534050179212e-05,
+      "loss": 0.0218,
+      "step": 637
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 5.46236559139785e-05,
+      "loss": 0.0222,
+      "step": 638
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 5.455197132616487e-05,
+      "loss": 0.0236,
+      "step": 639
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 5.4480286738351264e-05,
+      "loss": 0.0324,
+      "step": 640
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 5.4408602150537635e-05,
+      "loss": 0.0234,
+      "step": 641
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 5.433691756272402e-05,
+      "loss": 0.0196,
+      "step": 642
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 5.4265232974910397e-05,
+      "loss": 0.0235,
+      "step": 643
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 5.419354838709678e-05,
+      "loss": 0.0196,
+      "step": 644
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 5.412186379928316e-05,
+      "loss": 0.0287,
+      "step": 645
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 5.405017921146953e-05,
+      "loss": 0.0237,
+      "step": 646
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 5.397849462365592e-05,
+      "loss": 0.0278,
+      "step": 647
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 5.390681003584229e-05,
+      "loss": 0.0234,
+      "step": 648
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 5.3835125448028676e-05,
+      "loss": 0.0285,
+      "step": 649
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 5.3763440860215054e-05,
+      "loss": 0.0292,
+      "step": 650
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 5.369175627240144e-05,
+      "loss": 0.0205,
+      "step": 651
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 5.3620071684587816e-05,
+      "loss": 0.0285,
+      "step": 652
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 5.35483870967742e-05,
+      "loss": 0.0353,
+      "step": 653
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 5.347670250896057e-05,
+      "loss": 0.0289,
+      "step": 654
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 5.340501792114696e-05,
+      "loss": 0.0231,
+      "step": 655
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.0329,
+      "step": 656
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 5.3261648745519724e-05,
+      "loss": 0.0257,
+      "step": 657
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 5.3189964157706095e-05,
+      "loss": 0.0275,
+      "step": 658
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 5.311827956989247e-05,
+      "loss": 0.0275,
+      "step": 659
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 5.304659498207886e-05,
+      "loss": 0.0304,
+      "step": 660
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 5.297491039426523e-05,
+      "loss": 0.0233,
+      "step": 661
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 5.290322580645162e-05,
+      "loss": 0.0253,
+      "step": 662
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 5.283154121863799e-05,
+      "loss": 0.0296,
+      "step": 663
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 5.2759856630824375e-05,
+      "loss": 0.0239,
+      "step": 664
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 5.268817204301075e-05,
+      "loss": 0.0315,
+      "step": 665
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 5.261648745519714e-05,
+      "loss": 0.0297,
+      "step": 666
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5.2544802867383514e-05,
+      "loss": 0.026,
+      "step": 667
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 5.24731182795699e-05,
+      "loss": 0.0235,
+      "step": 668
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 5.2401433691756277e-05,
+      "loss": 0.0279,
+      "step": 669
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 5.232974910394266e-05,
+      "loss": 0.0223,
+      "step": 670
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 5.225806451612903e-05,
+      "loss": 0.0229,
+      "step": 671
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 5.218637992831541e-05,
+      "loss": 0.0259,
+      "step": 672
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 5.2114695340501794e-05,
+      "loss": 0.0262,
+      "step": 673
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 5.204301075268817e-05,
+      "loss": 0.0266,
+      "step": 674
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 5.1971326164874556e-05,
+      "loss": 0.0217,
+      "step": 675
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 5.1899641577060934e-05,
+      "loss": 0.0243,
+      "step": 676
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 5.182795698924732e-05,
+      "loss": 0.0318,
+      "step": 677
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 5.175627240143369e-05,
+      "loss": 0.0239,
+      "step": 678
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 5.168458781362008e-05,
+      "loss": 0.0289,
+      "step": 679
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 5.161290322580645e-05,
+      "loss": 0.0282,
+      "step": 680
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 5.1541218637992835e-05,
+      "loss": 0.0288,
+      "step": 681
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 5.146953405017921e-05,
+      "loss": 0.0269,
+      "step": 682
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 5.13978494623656e-05,
+      "loss": 0.0259,
+      "step": 683
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 5.1326164874551975e-05,
+      "loss": 0.0306,
+      "step": 684
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 5.1254480286738346e-05,
+      "loss": 0.0246,
+      "step": 685
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 5.118279569892474e-05,
+      "loss": 0.0243,
+      "step": 686
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.0231,
+      "step": 687
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 5.103942652329749e-05,
+      "loss": 0.0317,
+      "step": 688
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 5.096774193548387e-05,
+      "loss": 0.0276,
+      "step": 689
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 5.0896057347670255e-05,
+      "loss": 0.0375,
+      "step": 690
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5.082437275985663e-05,
+      "loss": 0.033,
+      "step": 691
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5.075268817204302e-05,
+      "loss": 0.0272,
+      "step": 692
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5.0681003584229394e-05,
+      "loss": 0.0262,
+      "step": 693
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5.060931899641578e-05,
+      "loss": 0.0224,
+      "step": 694
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 5.053763440860215e-05,
+      "loss": 0.0313,
+      "step": 695
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 5.046594982078854e-05,
+      "loss": 0.0303,
+      "step": 696
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 5.039426523297491e-05,
+      "loss": 0.0337,
+      "step": 697
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 5.032258064516129e-05,
+      "loss": 0.0342,
+      "step": 698
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 5.0250896057347674e-05,
+      "loss": 0.0314,
+      "step": 699
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 5.017921146953405e-05,
+      "loss": 0.0258,
+      "step": 700
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 3.6010735034942627,
+      "eval_runtime": 29.22,
+      "eval_samples_per_second": 4.107,
+      "eval_steps_per_second": 0.513,
+      "step": 700
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 5.0107526881720436e-05,
+      "loss": 0.0205,
+      "step": 701
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 5.003584229390681e-05,
+      "loss": 0.0217,
+      "step": 702
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 4.996415770609319e-05,
+      "loss": 0.0279,
+      "step": 703
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 4.989247311827957e-05,
+      "loss": 0.0252,
+      "step": 704
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 4.982078853046595e-05,
+      "loss": 0.0241,
+      "step": 705
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 4.974910394265233e-05,
+      "loss": 0.0243,
+      "step": 706
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 4.967741935483871e-05,
+      "loss": 0.0208,
+      "step": 707
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 4.960573476702509e-05,
+      "loss": 0.0209,
+      "step": 708
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 4.953405017921147e-05,
+      "loss": 0.0275,
+      "step": 709
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 4.9462365591397855e-05,
+      "loss": 0.0237,
+      "step": 710
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 4.939068100358423e-05,
+      "loss": 0.0246,
+      "step": 711
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 4.931899641577061e-05,
+      "loss": 0.0242,
+      "step": 712
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 4.9247311827956995e-05,
+      "loss": 0.033,
+      "step": 713
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 4.917562724014337e-05,
+      "loss": 0.0309,
+      "step": 714
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 4.910394265232976e-05,
+      "loss": 0.0185,
+      "step": 715
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 4.903225806451613e-05,
+      "loss": 0.0205,
+      "step": 716
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 4.896057347670251e-05,
+      "loss": 0.0265,
+      "step": 717
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.0241,
+      "step": 718
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 4.881720430107527e-05,
+      "loss": 0.0248,
+      "step": 719
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 4.874551971326165e-05,
+      "loss": 0.0247,
+      "step": 720
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 4.867383512544803e-05,
+      "loss": 0.027,
+      "step": 721
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 4.8602150537634414e-05,
+      "loss": 0.0267,
+      "step": 722
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 4.853046594982079e-05,
+      "loss": 0.0281,
+      "step": 723
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 4.845878136200717e-05,
+      "loss": 0.0234,
+      "step": 724
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 4.8387096774193554e-05,
+      "loss": 0.0264,
+      "step": 725
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 4.831541218637993e-05,
+      "loss": 0.0274,
+      "step": 726
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 4.8243727598566316e-05,
+      "loss": 0.021,
+      "step": 727
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 4.8172043010752693e-05,
+      "loss": 0.0252,
+      "step": 728
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 4.810035842293907e-05,
+      "loss": 0.0262,
+      "step": 729
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 4.802867383512545e-05,
+      "loss": 0.0269,
+      "step": 730
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 4.7956989247311826e-05,
+      "loss": 0.0263,
+      "step": 731
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 4.788530465949821e-05,
+      "loss": 0.0317,
+      "step": 732
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 4.781362007168459e-05,
+      "loss": 0.0264,
+      "step": 733
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 4.774193548387097e-05,
+      "loss": 0.033,
+      "step": 734
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 4.767025089605735e-05,
+      "loss": 0.0234,
+      "step": 735
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 4.759856630824373e-05,
+      "loss": 0.026,
+      "step": 736
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 4.752688172043011e-05,
+      "loss": 0.0251,
+      "step": 737
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 4.745519713261649e-05,
+      "loss": 0.0251,
+      "step": 738
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 4.7383512544802875e-05,
+      "loss": 0.0246,
+      "step": 739
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 4.731182795698925e-05,
+      "loss": 0.0263,
+      "step": 740
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 4.724014336917563e-05,
+      "loss": 0.0304,
+      "step": 741
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 4.716845878136201e-05,
+      "loss": 0.029,
+      "step": 742
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 4.7096774193548385e-05,
+      "loss": 0.0192,
+      "step": 743
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 4.702508960573477e-05,
+      "loss": 0.0267,
+      "step": 744
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 4.695340501792115e-05,
+      "loss": 0.0286,
+      "step": 745
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 4.688172043010753e-05,
+      "loss": 0.0222,
+      "step": 746
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 4.681003584229391e-05,
+      "loss": 0.026,
+      "step": 747
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 4.673835125448029e-05,
+      "loss": 0.0248,
+      "step": 748
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.0209,
+      "step": 749
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 4.659498207885305e-05,
+      "loss": 0.0291,
+      "step": 750
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 4.6523297491039434e-05,
+      "loss": 0.0245,
+      "step": 751
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 4.645161290322581e-05,
+      "loss": 0.0255,
+      "step": 752
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 4.637992831541219e-05,
+      "loss": 0.0254,
+      "step": 753
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 4.630824372759857e-05,
+      "loss": 0.0247,
+      "step": 754
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 4.6236559139784944e-05,
+      "loss": 0.0236,
+      "step": 755
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 4.616487455197133e-05,
+      "loss": 0.023,
+      "step": 756
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 4.6093189964157706e-05,
+      "loss": 0.023,
+      "step": 757
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 4.6021505376344084e-05,
+      "loss": 0.0262,
+      "step": 758
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 4.594982078853047e-05,
+      "loss": 0.0207,
+      "step": 759
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 4.5878136200716846e-05,
+      "loss": 0.0248,
+      "step": 760
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 4.580645161290323e-05,
+      "loss": 0.0246,
+      "step": 761
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 4.573476702508961e-05,
+      "loss": 0.0307,
+      "step": 762
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 4.5663082437275986e-05,
+      "loss": 0.0215,
+      "step": 763
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 4.559139784946237e-05,
+      "loss": 0.0266,
+      "step": 764
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 4.551971326164875e-05,
+      "loss": 0.029,
+      "step": 765
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 4.544802867383513e-05,
+      "loss": 0.0233,
+      "step": 766
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 4.53763440860215e-05,
+      "loss": 0.0236,
+      "step": 767
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 4.530465949820789e-05,
+      "loss": 0.0293,
+      "step": 768
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 4.5232974910394265e-05,
+      "loss": 0.031,
+      "step": 769
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 4.516129032258064e-05,
+      "loss": 0.0247,
+      "step": 770
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 3.75777530670166,
+      "eval_runtime": 29.2261,
+      "eval_samples_per_second": 4.106,
+      "eval_steps_per_second": 0.513,
+      "step": 770
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.508960573476703e-05,
+      "loss": 0.0254,
+      "step": 771
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 4.5017921146953405e-05,
+      "loss": 0.0222,
+      "step": 772
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 4.494623655913979e-05,
+      "loss": 0.0321,
+      "step": 773
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 4.487455197132617e-05,
+      "loss": 0.0387,
+      "step": 774
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 4.4802867383512545e-05,
+      "loss": 0.0288,
+      "step": 775
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 4.473118279569893e-05,
+      "loss": 0.0198,
+      "step": 776
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 4.465949820788531e-05,
+      "loss": 0.0196,
+      "step": 777
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 4.458781362007169e-05,
+      "loss": 0.0276,
+      "step": 778
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 4.451612903225807e-05,
+      "loss": 0.0229,
+      "step": 779
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.022,
+      "step": 780
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 4.4372759856630824e-05,
+      "loss": 0.0204,
+      "step": 781
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 4.43010752688172e-05,
+      "loss": 0.0238,
+      "step": 782
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 4.4229390681003586e-05,
+      "loss": 0.0269,
+      "step": 783
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 4.4157706093189964e-05,
+      "loss": 0.032,
+      "step": 784
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 4.408602150537635e-05,
+      "loss": 0.0229,
+      "step": 785
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 4.4014336917562726e-05,
+      "loss": 0.0224,
+      "step": 786
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 4.3942652329749104e-05,
+      "loss": 0.0188,
+      "step": 787
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 4.387096774193549e-05,
+      "loss": 0.026,
+      "step": 788
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 4.3799283154121866e-05,
+      "loss": 0.0189,
+      "step": 789
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 4.372759856630825e-05,
+      "loss": 0.0224,
+      "step": 790
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 4.365591397849463e-05,
+      "loss": 0.0209,
+      "step": 791
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 4.3584229390681005e-05,
+      "loss": 0.0228,
+      "step": 792
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 4.351254480286738e-05,
+      "loss": 0.0236,
+      "step": 793
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 4.344086021505376e-05,
+      "loss": 0.0172,
+      "step": 794
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 4.3369175627240145e-05,
+      "loss": 0.0227,
+      "step": 795
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 4.329749103942652e-05,
+      "loss": 0.0258,
+      "step": 796
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 4.322580645161291e-05,
+      "loss": 0.0242,
+      "step": 797
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 4.3154121863799285e-05,
+      "loss": 0.0268,
+      "step": 798
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 4.308243727598566e-05,
+      "loss": 0.0251,
+      "step": 799
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 4.301075268817205e-05,
+      "loss": 0.027,
+      "step": 800
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 4.2939068100358425e-05,
+      "loss": 0.0239,
+      "step": 801
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 4.286738351254481e-05,
+      "loss": 0.0202,
+      "step": 802
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 4.279569892473119e-05,
+      "loss": 0.0277,
+      "step": 803
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 4.2724014336917564e-05,
+      "loss": 0.0277,
+      "step": 804
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 4.265232974910394e-05,
+      "loss": 0.0268,
+      "step": 805
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 4.258064516129032e-05,
+      "loss": 0.0223,
+      "step": 806
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 4.2508960573476704e-05,
+      "loss": 0.0273,
+      "step": 807
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 4.243727598566308e-05,
+      "loss": 0.0194,
+      "step": 808
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 4.2365591397849466e-05,
+      "loss": 0.0228,
+      "step": 809
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 4.2293906810035844e-05,
+      "loss": 0.031,
+      "step": 810
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.027,
+      "step": 811
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 4.2150537634408606e-05,
+      "loss": 0.0231,
+      "step": 812
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 4.2078853046594984e-05,
+      "loss": 0.0246,
+      "step": 813
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 4.200716845878137e-05,
+      "loss": 0.0249,
+      "step": 814
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 4.1935483870967746e-05,
+      "loss": 0.0217,
+      "step": 815
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 4.186379928315412e-05,
+      "loss": 0.0289,
+      "step": 816
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 4.179211469534051e-05,
+      "loss": 0.023,
+      "step": 817
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 4.172043010752688e-05,
+      "loss": 0.026,
+      "step": 818
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 4.164874551971326e-05,
+      "loss": 0.0175,
+      "step": 819
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 4.157706093189964e-05,
+      "loss": 0.0232,
+      "step": 820
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 4.1505376344086025e-05,
+      "loss": 0.0215,
+      "step": 821
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 4.14336917562724e-05,
+      "loss": 0.0237,
+      "step": 822
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 4.136200716845878e-05,
+      "loss": 0.0231,
+      "step": 823
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 4.1290322580645165e-05,
+      "loss": 0.0256,
+      "step": 824
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 4.121863799283154e-05,
+      "loss": 0.0221,
+      "step": 825
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 4.114695340501793e-05,
+      "loss": 0.0232,
+      "step": 826
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 4.1075268817204305e-05,
+      "loss": 0.0246,
+      "step": 827
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 4.100358422939068e-05,
+      "loss": 0.0292,
+      "step": 828
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 4.093189964157707e-05,
+      "loss": 0.0301,
+      "step": 829
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 4.0860215053763444e-05,
+      "loss": 0.0242,
+      "step": 830
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 4.078853046594982e-05,
+      "loss": 0.0223,
+      "step": 831
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 4.07168458781362e-05,
+      "loss": 0.0222,
+      "step": 832
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 4.0645161290322584e-05,
+      "loss": 0.0275,
+      "step": 833
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 4.057347670250896e-05,
+      "loss": 0.0253,
+      "step": 834
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 4.050179211469534e-05,
+      "loss": 0.0231,
+      "step": 835
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 4.0430107526881724e-05,
+      "loss": 0.0274,
+      "step": 836
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 4.03584229390681e-05,
+      "loss": 0.0239,
+      "step": 837
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 4.0286738351254486e-05,
+      "loss": 0.0269,
+      "step": 838
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 4.0215053763440864e-05,
+      "loss": 0.027,
+      "step": 839
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.014336917562724e-05,
+      "loss": 0.0287,
+      "step": 840
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 3.883384943008423,
+      "eval_runtime": 29.2074,
+      "eval_samples_per_second": 4.109,
+      "eval_steps_per_second": 0.514,
+      "step": 840
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.0071684587813626e-05,
+      "loss": 0.0214,
+      "step": 841
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 4e-05,
+      "loss": 0.03,
+      "step": 842
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 3.992831541218639e-05,
+      "loss": 0.0271,
+      "step": 843
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 3.985663082437276e-05,
+      "loss": 0.0249,
+      "step": 844
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 3.978494623655914e-05,
+      "loss": 0.022,
+      "step": 845
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 3.971326164874552e-05,
+      "loss": 0.0244,
+      "step": 846
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 3.96415770609319e-05,
+      "loss": 0.02,
+      "step": 847
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 3.956989247311828e-05,
+      "loss": 0.0193,
+      "step": 848
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 3.949820788530466e-05,
+      "loss": 0.0205,
+      "step": 849
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 3.9426523297491045e-05,
+      "loss": 0.0246,
+      "step": 850
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.0173,
+      "step": 851
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 3.92831541218638e-05,
+      "loss": 0.0189,
+      "step": 852
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 3.9211469534050185e-05,
+      "loss": 0.0186,
+      "step": 853
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 3.913978494623656e-05,
+      "loss": 0.0177,
+      "step": 854
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 3.906810035842295e-05,
+      "loss": 0.0209,
+      "step": 855
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 3.8996415770609324e-05,
+      "loss": 0.0245,
+      "step": 856
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 3.8924731182795695e-05,
+      "loss": 0.0221,
+      "step": 857
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 3.885304659498208e-05,
+      "loss": 0.0211,
+      "step": 858
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 3.878136200716846e-05,
+      "loss": 0.0284,
+      "step": 859
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.0191,
+      "step": 860
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 3.863799283154122e-05,
+      "loss": 0.0254,
+      "step": 861
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 3.85663082437276e-05,
+      "loss": 0.0228,
+      "step": 862
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 3.849462365591398e-05,
+      "loss": 0.0194,
+      "step": 863
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 3.842293906810036e-05,
+      "loss": 0.0218,
+      "step": 864
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 3.8351254480286743e-05,
+      "loss": 0.026,
+      "step": 865
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 3.827956989247312e-05,
+      "loss": 0.0201,
+      "step": 866
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 3.82078853046595e-05,
+      "loss": 0.0272,
+      "step": 867
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 3.813620071684588e-05,
+      "loss": 0.0207,
+      "step": 868
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 3.8064516129032254e-05,
+      "loss": 0.0268,
+      "step": 869
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 3.799283154121864e-05,
+      "loss": 0.0189,
+      "step": 870
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 3.7921146953405016e-05,
+      "loss": 0.0213,
+      "step": 871
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 3.78494623655914e-05,
+      "loss": 0.0227,
+      "step": 872
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.0241,
+      "step": 873
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 3.7706093189964156e-05,
+      "loss": 0.0228,
+      "step": 874
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 3.763440860215054e-05,
+      "loss": 0.0257,
+      "step": 875
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 3.756272401433692e-05,
+      "loss": 0.0192,
+      "step": 876
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 3.74910394265233e-05,
+      "loss": 0.0232,
+      "step": 877
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.026,
+      "step": 878
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 3.734767025089606e-05,
+      "loss": 0.0213,
+      "step": 879
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 3.727598566308244e-05,
+      "loss": 0.0345,
+      "step": 880
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 3.720430107526882e-05,
+      "loss": 0.0209,
+      "step": 881
+    },
+    {
+      "epoch": 12.6,
+      "learning_rate": 3.71326164874552e-05,
+      "loss": 0.0221,
+      "step": 882
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 3.7060931899641575e-05,
+      "loss": 0.0305,
+      "step": 883
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 3.698924731182796e-05,
+      "loss": 0.035,
+      "step": 884
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 3.691756272401434e-05,
+      "loss": 0.0365,
+      "step": 885
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 3.6845878136200715e-05,
+      "loss": 0.0228,
+      "step": 886
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.0245,
+      "step": 887
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 3.670250896057348e-05,
+      "loss": 0.0265,
+      "step": 888
+    },
+    {
+      "epoch": 12.7,
+      "learning_rate": 3.663082437275986e-05,
+      "loss": 0.0263,
+      "step": 889
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 3.655913978494624e-05,
+      "loss": 0.0278,
+      "step": 890
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 3.648745519713262e-05,
+      "loss": 0.0219,
+      "step": 891
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 3.6415770609319e-05,
+      "loss": 0.0247,
+      "step": 892
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 3.634408602150538e-05,
+      "loss": 0.0207,
+      "step": 893
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 3.627240143369176e-05,
+      "loss": 0.0322,
+      "step": 894
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 3.6200716845878134e-05,
+      "loss": 0.0242,
+      "step": 895
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.026,
+      "step": 896
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 3.6057347670250896e-05,
+      "loss": 0.0233,
+      "step": 897
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 3.5985663082437274e-05,
+      "loss": 0.0288,
+      "step": 898
+    },
+    {
+      "epoch": 12.84,
+      "learning_rate": 3.591397849462366e-05,
+      "loss": 0.0265,
+      "step": 899
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 3.5842293906810036e-05,
+      "loss": 0.0251,
+      "step": 900
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 3.577060931899642e-05,
+      "loss": 0.0247,
+      "step": 901
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 3.56989247311828e-05,
+      "loss": 0.022,
+      "step": 902
+    },
+    {
+      "epoch": 12.9,
+      "learning_rate": 3.5627240143369176e-05,
+      "loss": 0.024,
+      "step": 903
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 0.0294,
+      "step": 904
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.0293,
+      "step": 905
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 3.541218637992832e-05,
+      "loss": 0.0234,
+      "step": 906
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 3.53405017921147e-05,
+      "loss": 0.0248,
+      "step": 907
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 3.526881720430108e-05,
+      "loss": 0.0242,
+      "step": 908
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 3.5197132616487455e-05,
+      "loss": 0.0228,
+      "step": 909
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.512544802867383e-05,
+      "loss": 0.0257,
+      "step": 910
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 3.9491729736328125,
+      "eval_runtime": 29.1968,
+      "eval_samples_per_second": 4.11,
+      "eval_steps_per_second": 0.514,
+      "step": 910
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 3.505376344086022e-05,
+      "loss": 0.0185,
+      "step": 911
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 3.4982078853046595e-05,
+      "loss": 0.0244,
+      "step": 912
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 3.491039426523298e-05,
+      "loss": 0.0264,
+      "step": 913
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.0218,
+      "step": 914
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 3.4767025089605734e-05,
+      "loss": 0.0186,
+      "step": 915
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 3.469534050179212e-05,
+      "loss": 0.0277,
+      "step": 916
+    },
+    {
+      "epoch": 13.1,
+      "learning_rate": 3.4623655913978497e-05,
+      "loss": 0.0202,
+      "step": 917
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 3.455197132616488e-05,
+      "loss": 0.0188,
+      "step": 918
+    },
+    {
+      "epoch": 13.13,
+      "learning_rate": 3.448028673835126e-05,
+      "loss": 0.0208,
+      "step": 919
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 3.4408602150537636e-05,
+      "loss": 0.0218,
+      "step": 920
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 3.4336917562724014e-05,
+      "loss": 0.0199,
+      "step": 921
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 3.426523297491039e-05,
+      "loss": 0.0175,
+      "step": 922
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.0226,
+      "step": 923
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 3.4121863799283154e-05,
+      "loss": 0.0253,
+      "step": 924
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 3.405017921146954e-05,
+      "loss": 0.0178,
+      "step": 925
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 3.3978494623655916e-05,
+      "loss": 0.0253,
+      "step": 926
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 3.3906810035842293e-05,
+      "loss": 0.0248,
+      "step": 927
+    },
+    {
+      "epoch": 13.26,
+      "learning_rate": 3.383512544802868e-05,
+      "loss": 0.0276,
+      "step": 928
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 3.3763440860215055e-05,
+      "loss": 0.0308,
+      "step": 929
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 3.369175627240144e-05,
+      "loss": 0.0235,
+      "step": 930
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 3.362007168458782e-05,
+      "loss": 0.0256,
+      "step": 931
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.0201,
+      "step": 932
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 3.347670250896057e-05,
+      "loss": 0.0206,
+      "step": 933
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 3.340501792114695e-05,
+      "loss": 0.0213,
+      "step": 934
+    },
+    {
+      "epoch": 13.36,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.0246,
+      "step": 935
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 3.326164874551971e-05,
+      "loss": 0.0258,
+      "step": 936
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 3.31899641577061e-05,
+      "loss": 0.0239,
+      "step": 937
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 3.3118279569892475e-05,
+      "loss": 0.0242,
+      "step": 938
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 3.304659498207885e-05,
+      "loss": 0.0177,
+      "step": 939
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 3.297491039426524e-05,
+      "loss": 0.0227,
+      "step": 940
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.0209,
+      "step": 941
+    },
+    {
+      "epoch": 13.46,
+      "learning_rate": 3.2831541218638e-05,
+      "loss": 0.0226,
+      "step": 942
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 3.2759856630824376e-05,
+      "loss": 0.0266,
+      "step": 943
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 3.2688172043010754e-05,
+      "loss": 0.0289,
+      "step": 944
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 3.261648745519714e-05,
+      "loss": 0.023,
+      "step": 945
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 3.254480286738351e-05,
+      "loss": 0.0303,
+      "step": 946
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 3.2473118279569894e-05,
+      "loss": 0.0294,
+      "step": 947
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 3.240143369175627e-05,
+      "loss": 0.0265,
+      "step": 948
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 3.2329749103942656e-05,
+      "loss": 0.0215,
+      "step": 949
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.0228,
+      "step": 950
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 3.218637992831541e-05,
+      "loss": 0.0252,
+      "step": 951
+    },
+    {
+      "epoch": 13.6,
+      "learning_rate": 3.2114695340501796e-05,
+      "loss": 0.0209,
+      "step": 952
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 3.204301075268817e-05,
+      "loss": 0.0252,
+      "step": 953
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 3.197132616487455e-05,
+      "loss": 0.0262,
+      "step": 954
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 3.1899641577060935e-05,
+      "loss": 0.0246,
+      "step": 955
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 3.182795698924731e-05,
+      "loss": 0.0241,
+      "step": 956
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 3.17562724014337e-05,
+      "loss": 0.0251,
+      "step": 957
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 3.1684587813620075e-05,
+      "loss": 0.0265,
+      "step": 958
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.022,
+      "step": 959
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 3.154121863799283e-05,
+      "loss": 0.0217,
+      "step": 960
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 3.146953405017921e-05,
+      "loss": 0.0233,
+      "step": 961
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 3.139784946236559e-05,
+      "loss": 0.0267,
+      "step": 962
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 3.132616487455197e-05,
+      "loss": 0.0226,
+      "step": 963
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 3.1254480286738355e-05,
+      "loss": 0.0257,
+      "step": 964
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 3.118279569892473e-05,
+      "loss": 0.028,
+      "step": 965
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.0256,
+      "step": 966
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 3.1039426523297494e-05,
+      "loss": 0.0186,
+      "step": 967
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.0236,
+      "step": 968
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 3.0896057347670256e-05,
+      "loss": 0.021,
+      "step": 969
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 3.0824372759856634e-05,
+      "loss": 0.0242,
+      "step": 970
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 3.075268817204301e-05,
+      "loss": 0.0291,
+      "step": 971
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 3.068100358422939e-05,
+      "loss": 0.0224,
+      "step": 972
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 3.060931899641577e-05,
+      "loss": 0.0296,
+      "step": 973
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 3.053763440860215e-05,
+      "loss": 0.0279,
+      "step": 974
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 3.046594982078853e-05,
+      "loss": 0.03,
+      "step": 975
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 3.039426523297491e-05,
+      "loss": 0.0199,
+      "step": 976
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.02,
+      "step": 977
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 3.0250896057347672e-05,
+      "loss": 0.0217,
+      "step": 978
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 3.0179211469534053e-05,
+      "loss": 0.0229,
+      "step": 979
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.010752688172043e-05,
+      "loss": 0.0267,
+      "step": 980
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 3.9646317958831787,
+      "eval_runtime": 29.2015,
+      "eval_samples_per_second": 4.109,
+      "eval_steps_per_second": 0.514,
+      "step": 980
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.0035842293906812e-05,
+      "loss": 0.0197,
+      "step": 981
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 2.9964157706093193e-05,
+      "loss": 0.0187,
+      "step": 982
+    },
+    {
+      "epoch": 14.04,
+      "learning_rate": 2.9892473118279574e-05,
+      "loss": 0.0256,
+      "step": 983
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 2.9820788530465955e-05,
+      "loss": 0.0205,
+      "step": 984
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 2.974910394265233e-05,
+      "loss": 0.0239,
+      "step": 985
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 2.967741935483871e-05,
+      "loss": 0.0204,
+      "step": 986
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 2.9605734767025088e-05,
+      "loss": 0.0238,
+      "step": 987
+    },
+    {
+      "epoch": 14.11,
+      "learning_rate": 2.953405017921147e-05,
+      "loss": 0.0219,
+      "step": 988
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 2.946236559139785e-05,
+      "loss": 0.0346,
+      "step": 989
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 2.939068100358423e-05,
+      "loss": 0.0228,
+      "step": 990
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 2.9318996415770612e-05,
+      "loss": 0.029,
+      "step": 991
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 2.924731182795699e-05,
+      "loss": 0.0274,
+      "step": 992
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 2.917562724014337e-05,
+      "loss": 0.024,
+      "step": 993
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 2.9103942652329752e-05,
+      "loss": 0.0231,
+      "step": 994
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.0188,
+      "step": 995
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 2.8960573476702514e-05,
+      "loss": 0.022,
+      "step": 996
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.0241,
+      "step": 997
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 2.881720430107527e-05,
+      "loss": 0.0232,
+      "step": 998
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 2.8745519713261647e-05,
+      "loss": 0.0242,
+      "step": 999
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 2.8673835125448028e-05,
+      "loss": 0.0216,
+      "step": 1000
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 2.860215053763441e-05,
+      "loss": 0.0173,
+      "step": 1001
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 2.853046594982079e-05,
+      "loss": 0.0224,
+      "step": 1002
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 2.845878136200717e-05,
+      "loss": 0.0207,
+      "step": 1003
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 2.838709677419355e-05,
+      "loss": 0.0214,
+      "step": 1004
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 2.831541218637993e-05,
+      "loss": 0.0184,
+      "step": 1005
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 2.824372759856631e-05,
+      "loss": 0.0232,
+      "step": 1006
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 2.8172043010752692e-05,
+      "loss": 0.022,
+      "step": 1007
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 2.8100358422939073e-05,
+      "loss": 0.027,
+      "step": 1008
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 2.802867383512545e-05,
+      "loss": 0.029,
+      "step": 1009
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 2.7956989247311828e-05,
+      "loss": 0.0272,
+      "step": 1010
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 2.7885304659498206e-05,
+      "loss": 0.0286,
+      "step": 1011
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 2.7813620071684587e-05,
+      "loss": 0.0241,
+      "step": 1012
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 2.7741935483870968e-05,
+      "loss": 0.0269,
+      "step": 1013
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 2.767025089605735e-05,
+      "loss": 0.0236,
+      "step": 1014
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 2.759856630824373e-05,
+      "loss": 0.0291,
+      "step": 1015
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 2.7526881720430108e-05,
+      "loss": 0.0242,
+      "step": 1016
+    },
+    {
+      "epoch": 14.53,
+      "learning_rate": 2.745519713261649e-05,
+      "loss": 0.0209,
+      "step": 1017
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 2.738351254480287e-05,
+      "loss": 0.0228,
+      "step": 1018
+    },
+    {
+      "epoch": 14.56,
+      "learning_rate": 2.731182795698925e-05,
+      "loss": 0.0248,
+      "step": 1019
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 2.7240143369175632e-05,
+      "loss": 0.0249,
+      "step": 1020
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 2.716845878136201e-05,
+      "loss": 0.0212,
+      "step": 1021
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 2.709677419354839e-05,
+      "loss": 0.0209,
+      "step": 1022
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 2.7025089605734765e-05,
+      "loss": 0.0229,
+      "step": 1023
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 2.6953405017921146e-05,
+      "loss": 0.0245,
+      "step": 1024
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 2.6881720430107527e-05,
+      "loss": 0.021,
+      "step": 1025
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 2.6810035842293908e-05,
+      "loss": 0.0206,
+      "step": 1026
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 2.6738351254480286e-05,
+      "loss": 0.0258,
+      "step": 1027
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.0259,
+      "step": 1028
+    },
+    {
+      "epoch": 14.7,
+      "learning_rate": 2.6594982078853048e-05,
+      "loss": 0.0242,
+      "step": 1029
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 2.652329749103943e-05,
+      "loss": 0.0204,
+      "step": 1030
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 2.645161290322581e-05,
+      "loss": 0.0208,
+      "step": 1031
+    },
+    {
+      "epoch": 14.74,
+      "learning_rate": 2.6379928315412187e-05,
+      "loss": 0.0259,
+      "step": 1032
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 2.630824372759857e-05,
+      "loss": 0.0203,
+      "step": 1033
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 2.623655913978495e-05,
+      "loss": 0.0226,
+      "step": 1034
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 2.616487455197133e-05,
+      "loss": 0.0219,
+      "step": 1035
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 2.6093189964157705e-05,
+      "loss": 0.0272,
+      "step": 1036
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 2.6021505376344086e-05,
+      "loss": 0.0217,
+      "step": 1037
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 2.5949820788530467e-05,
+      "loss": 0.0215,
+      "step": 1038
+    },
+    {
+      "epoch": 14.84,
+      "learning_rate": 2.5878136200716844e-05,
+      "loss": 0.0203,
+      "step": 1039
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.0223,
+      "step": 1040
+    },
+    {
+      "epoch": 14.87,
+      "learning_rate": 2.5734767025089607e-05,
+      "loss": 0.0236,
+      "step": 1041
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 2.5663082437275988e-05,
+      "loss": 0.0243,
+      "step": 1042
+    },
+    {
+      "epoch": 14.9,
+      "learning_rate": 2.559139784946237e-05,
+      "loss": 0.0326,
+      "step": 1043
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 2.5519713261648746e-05,
+      "loss": 0.0222,
+      "step": 1044
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 2.5448028673835127e-05,
+      "loss": 0.0265,
+      "step": 1045
+    },
+    {
+      "epoch": 14.94,
+      "learning_rate": 2.537634408602151e-05,
+      "loss": 0.0247,
+      "step": 1046
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 2.530465949820789e-05,
+      "loss": 0.0229,
+      "step": 1047
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 2.523297491039427e-05,
+      "loss": 0.0224,
+      "step": 1048
+    },
+    {
+      "epoch": 14.99,
+      "learning_rate": 2.5161290322580645e-05,
+      "loss": 0.0192,
+      "step": 1049
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 2.5089605734767026e-05,
+      "loss": 0.0205,
+      "step": 1050
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 4.015718936920166,
+      "eval_runtime": 29.1779,
+      "eval_samples_per_second": 4.113,
+      "eval_steps_per_second": 0.514,
+      "step": 1050
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 2.5017921146953403e-05,
+      "loss": 0.0174,
+      "step": 1051
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 2.4946236559139784e-05,
+      "loss": 0.0207,
+      "step": 1052
+    },
+    {
+      "epoch": 15.04,
+      "learning_rate": 2.4874551971326165e-05,
+      "loss": 0.0196,
+      "step": 1053
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 2.4802867383512547e-05,
+      "loss": 0.0262,
+      "step": 1054
+    },
+    {
+      "epoch": 15.07,
+      "learning_rate": 2.4731182795698928e-05,
+      "loss": 0.0237,
+      "step": 1055
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 2.4659498207885305e-05,
+      "loss": 0.0225,
+      "step": 1056
+    },
+    {
+      "epoch": 15.1,
+      "learning_rate": 2.4587813620071686e-05,
+      "loss": 0.0202,
+      "step": 1057
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 2.4516129032258064e-05,
+      "loss": 0.0212,
+      "step": 1058
+    },
+    {
+      "epoch": 15.13,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.0243,
+      "step": 1059
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 2.4372759856630826e-05,
+      "loss": 0.0205,
+      "step": 1060
+    },
+    {
+      "epoch": 15.16,
+      "learning_rate": 2.4301075268817207e-05,
+      "loss": 0.0197,
+      "step": 1061
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 2.4229390681003585e-05,
+      "loss": 0.0202,
+      "step": 1062
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 2.4157706093189966e-05,
+      "loss": 0.0211,
+      "step": 1063
+    },
+    {
+      "epoch": 15.2,
+      "learning_rate": 2.4086021505376347e-05,
+      "loss": 0.0234,
+      "step": 1064
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 2.4014336917562724e-05,
+      "loss": 0.0237,
+      "step": 1065
+    },
+    {
+      "epoch": 15.23,
+      "learning_rate": 2.3942652329749105e-05,
+      "loss": 0.0213,
+      "step": 1066
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.0202,
+      "step": 1067
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 2.3799283154121864e-05,
+      "loss": 0.0219,
+      "step": 1068
+    },
+    {
+      "epoch": 15.27,
+      "learning_rate": 2.3727598566308245e-05,
+      "loss": 0.0197,
+      "step": 1069
+    },
+    {
+      "epoch": 15.29,
+      "learning_rate": 2.3655913978494626e-05,
+      "loss": 0.0247,
+      "step": 1070
+    },
+    {
+      "epoch": 15.3,
+      "learning_rate": 2.3584229390681004e-05,
+      "loss": 0.0218,
+      "step": 1071
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 2.3512544802867385e-05,
+      "loss": 0.0211,
+      "step": 1072
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 2.3440860215053766e-05,
+      "loss": 0.0244,
+      "step": 1073
+    },
+    {
+      "epoch": 15.34,
+      "learning_rate": 2.3369175627240144e-05,
+      "loss": 0.0237,
+      "step": 1074
+    },
+    {
+      "epoch": 15.36,
+      "learning_rate": 2.3297491039426525e-05,
+      "loss": 0.0208,
+      "step": 1075
+    },
+    {
+      "epoch": 15.37,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.021,
+      "step": 1076
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 2.3154121863799283e-05,
+      "loss": 0.0191,
+      "step": 1077
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 2.3082437275985664e-05,
+      "loss": 0.0242,
+      "step": 1078
+    },
+    {
+      "epoch": 15.41,
+      "learning_rate": 2.3010752688172042e-05,
+      "loss": 0.0199,
+      "step": 1079
+    },
+    {
+      "epoch": 15.43,
+      "learning_rate": 2.2939068100358423e-05,
+      "loss": 0.0269,
+      "step": 1080
+    },
+    {
+      "epoch": 15.44,
+      "learning_rate": 2.2867383512544804e-05,
+      "loss": 0.025,
+      "step": 1081
+    },
+    {
+      "epoch": 15.46,
+      "learning_rate": 2.2795698924731185e-05,
+      "loss": 0.0234,
+      "step": 1082
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 2.2724014336917566e-05,
+      "loss": 0.0202,
+      "step": 1083
+    },
+    {
+      "epoch": 15.49,
+      "learning_rate": 2.2652329749103944e-05,
+      "loss": 0.0227,
+      "step": 1084
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 2.258064516129032e-05,
+      "loss": 0.0223,
+      "step": 1085
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 2.2508960573476703e-05,
+      "loss": 0.0231,
+      "step": 1086
+    },
+    {
+      "epoch": 15.53,
+      "learning_rate": 2.2437275985663084e-05,
+      "loss": 0.0224,
+      "step": 1087
+    },
+    {
+      "epoch": 15.54,
+      "learning_rate": 2.2365591397849465e-05,
+      "loss": 0.0226,
+      "step": 1088
+    },
+    {
+      "epoch": 15.56,
+      "learning_rate": 2.2293906810035846e-05,
+      "loss": 0.0202,
+      "step": 1089
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.0242,
+      "step": 1090
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 2.21505376344086e-05,
+      "loss": 0.0263,
+      "step": 1091
+    },
+    {
+      "epoch": 15.6,
+      "learning_rate": 2.2078853046594982e-05,
+      "loss": 0.0265,
+      "step": 1092
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 2.2007168458781363e-05,
+      "loss": 0.0245,
+      "step": 1093
+    },
+    {
+      "epoch": 15.63,
+      "learning_rate": 2.1935483870967744e-05,
+      "loss": 0.0243,
+      "step": 1094
+    },
+    {
+      "epoch": 15.64,
+      "learning_rate": 2.1863799283154125e-05,
+      "loss": 0.0192,
+      "step": 1095
+    },
+    {
+      "epoch": 15.66,
+      "learning_rate": 2.1792114695340503e-05,
+      "loss": 0.0306,
+      "step": 1096
+    },
+    {
+      "epoch": 15.67,
+      "learning_rate": 2.172043010752688e-05,
+      "loss": 0.0348,
+      "step": 1097
+    },
+    {
+      "epoch": 15.69,
+      "learning_rate": 2.164874551971326e-05,
+      "loss": 0.0283,
+      "step": 1098
+    },
+    {
+      "epoch": 15.7,
+      "learning_rate": 2.1577060931899642e-05,
+      "loss": 0.0177,
+      "step": 1099
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 2.1505376344086024e-05,
+      "loss": 0.0235,
+      "step": 1100
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 2.1433691756272405e-05,
+      "loss": 0.0215,
+      "step": 1101
+    },
+    {
+      "epoch": 15.74,
+      "learning_rate": 2.1362007168458782e-05,
+      "loss": 0.0228,
+      "step": 1102
+    },
+    {
+      "epoch": 15.76,
+      "learning_rate": 2.129032258064516e-05,
+      "loss": 0.0295,
+      "step": 1103
+    },
+    {
+      "epoch": 15.77,
+      "learning_rate": 2.121863799283154e-05,
+      "loss": 0.0243,
+      "step": 1104
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 2.1146953405017922e-05,
+      "loss": 0.0182,
+      "step": 1105
+    },
+    {
+      "epoch": 15.8,
+      "learning_rate": 2.1075268817204303e-05,
+      "loss": 0.0284,
+      "step": 1106
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 2.1003584229390684e-05,
+      "loss": 0.021,
+      "step": 1107
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 2.093189964157706e-05,
+      "loss": 0.0302,
+      "step": 1108
+    },
+    {
+      "epoch": 15.84,
+      "learning_rate": 2.086021505376344e-05,
+      "loss": 0.0208,
+      "step": 1109
+    },
+    {
+      "epoch": 15.86,
+      "learning_rate": 2.078853046594982e-05,
+      "loss": 0.0218,
+      "step": 1110
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 2.07168458781362e-05,
+      "loss": 0.0219,
+      "step": 1111
+    },
+    {
+      "epoch": 15.89,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.0221,
+      "step": 1112
+    },
+    {
+      "epoch": 15.9,
+      "learning_rate": 2.0573476702508963e-05,
+      "loss": 0.0277,
+      "step": 1113
+    },
+    {
+      "epoch": 15.91,
+      "learning_rate": 2.050179211469534e-05,
+      "loss": 0.0267,
+      "step": 1114
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 2.0430107526881722e-05,
+      "loss": 0.0264,
+      "step": 1115
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 2.03584229390681e-05,
+      "loss": 0.0312,
+      "step": 1116
+    },
+    {
+      "epoch": 15.96,
+      "learning_rate": 2.028673835125448e-05,
+      "loss": 0.0271,
+      "step": 1117
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 2.0215053763440862e-05,
+      "loss": 0.0241,
+      "step": 1118
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 2.0143369175627243e-05,
+      "loss": 0.0193,
+      "step": 1119
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 2.007168458781362e-05,
+      "loss": 0.0202,
+      "step": 1120
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 4.051829814910889,
+      "eval_runtime": 29.1684,
+      "eval_samples_per_second": 4.114,
+      "eval_steps_per_second": 0.514,
+      "step": 1120
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 1121
+    },
+    {
+      "epoch": 16.03,
+      "learning_rate": 1.992831541218638e-05,
+      "loss": 0.0193,
+      "step": 1122
+    },
+    {
+      "epoch": 16.04,
+      "learning_rate": 1.985663082437276e-05,
+      "loss": 0.0151,
+      "step": 1123
+    },
+    {
+      "epoch": 16.06,
+      "learning_rate": 1.978494623655914e-05,
+      "loss": 0.0256,
+      "step": 1124
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 1.9713261648745522e-05,
+      "loss": 0.0297,
+      "step": 1125
+    },
+    {
+      "epoch": 16.09,
+      "learning_rate": 1.96415770609319e-05,
+      "loss": 0.0225,
+      "step": 1126
+    },
+    {
+      "epoch": 16.1,
+      "learning_rate": 1.956989247311828e-05,
+      "loss": 0.0208,
+      "step": 1127
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 1.9498207885304662e-05,
+      "loss": 0.0203,
+      "step": 1128
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 1.942652329749104e-05,
+      "loss": 0.0221,
+      "step": 1129
+    },
+    {
+      "epoch": 16.14,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.0217,
+      "step": 1130
+    },
+    {
+      "epoch": 16.16,
+      "learning_rate": 1.92831541218638e-05,
+      "loss": 0.0207,
+      "step": 1131
+    },
+    {
+      "epoch": 16.17,
+      "learning_rate": 1.921146953405018e-05,
+      "loss": 0.0193,
+      "step": 1132
+    },
+    {
+      "epoch": 16.19,
+      "learning_rate": 1.913978494623656e-05,
+      "loss": 0.0211,
+      "step": 1133
+    },
+    {
+      "epoch": 16.2,
+      "learning_rate": 1.906810035842294e-05,
+      "loss": 0.0286,
+      "step": 1134
+    },
+    {
+      "epoch": 16.21,
+      "learning_rate": 1.899641577060932e-05,
+      "loss": 0.0227,
+      "step": 1135
+    },
+    {
+      "epoch": 16.23,
+      "learning_rate": 1.89247311827957e-05,
+      "loss": 0.0237,
+      "step": 1136
+    },
+    {
+      "epoch": 16.24,
+      "learning_rate": 1.8853046594982078e-05,
+      "loss": 0.0241,
+      "step": 1137
+    },
+    {
+      "epoch": 16.26,
+      "learning_rate": 1.878136200716846e-05,
+      "loss": 0.0214,
+      "step": 1138
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.0212,
+      "step": 1139
+    },
+    {
+      "epoch": 16.29,
+      "learning_rate": 1.863799283154122e-05,
+      "loss": 0.0238,
+      "step": 1140
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 1.85663082437276e-05,
+      "loss": 0.0233,
+      "step": 1141
+    },
+    {
+      "epoch": 16.31,
+      "learning_rate": 1.849462365591398e-05,
+      "loss": 0.0238,
+      "step": 1142
+    },
+    {
+      "epoch": 16.33,
+      "learning_rate": 1.8422939068100357e-05,
+      "loss": 0.0246,
+      "step": 1143
+    },
+    {
+      "epoch": 16.34,
+      "learning_rate": 1.835125448028674e-05,
+      "loss": 0.0235,
+      "step": 1144
+    },
+    {
+      "epoch": 16.36,
+      "learning_rate": 1.827956989247312e-05,
+      "loss": 0.0243,
+      "step": 1145
+    },
+    {
+      "epoch": 16.37,
+      "learning_rate": 1.82078853046595e-05,
+      "loss": 0.019,
+      "step": 1146
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 1.813620071684588e-05,
+      "loss": 0.0197,
+      "step": 1147
+    },
+    {
+      "epoch": 16.4,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.0234,
+      "step": 1148
+    },
+    {
+      "epoch": 16.41,
+      "learning_rate": 1.7992831541218637e-05,
+      "loss": 0.0229,
+      "step": 1149
+    },
+    {
+      "epoch": 16.43,
+      "learning_rate": 1.7921146953405018e-05,
+      "loss": 0.0212,
+      "step": 1150
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 1.78494623655914e-05,
+      "loss": 0.0237,
+      "step": 1151
+    },
+    {
+      "epoch": 16.46,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.0236,
+      "step": 1152
+    },
+    {
+      "epoch": 16.47,
+      "learning_rate": 1.770609318996416e-05,
+      "loss": 0.0213,
+      "step": 1153
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 1.763440860215054e-05,
+      "loss": 0.0249,
+      "step": 1154
+    },
+    {
+      "epoch": 16.5,
+      "learning_rate": 1.7562724014336916e-05,
+      "loss": 0.027,
+      "step": 1155
+    },
+    {
+      "epoch": 16.51,
+      "learning_rate": 1.7491039426523297e-05,
+      "loss": 0.0217,
+      "step": 1156
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.0216,
+      "step": 1157
+    },
+    {
+      "epoch": 16.54,
+      "learning_rate": 1.734767025089606e-05,
+      "loss": 0.0221,
+      "step": 1158
+    },
+    {
+      "epoch": 16.56,
+      "learning_rate": 1.727598566308244e-05,
+      "loss": 0.0238,
+      "step": 1159
+    },
+    {
+      "epoch": 16.57,
+      "learning_rate": 1.7204301075268818e-05,
+      "loss": 0.0239,
+      "step": 1160
+    },
+    {
+      "epoch": 16.59,
+      "learning_rate": 1.7132616487455196e-05,
+      "loss": 0.0234,
+      "step": 1161
+    },
+    {
+      "epoch": 16.6,
+      "learning_rate": 1.7060931899641577e-05,
+      "loss": 0.0211,
+      "step": 1162
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 1.6989247311827958e-05,
+      "loss": 0.0228,
+      "step": 1163
+    },
+    {
+      "epoch": 16.63,
+      "learning_rate": 1.691756272401434e-05,
+      "loss": 0.0185,
+      "step": 1164
+    },
+    {
+      "epoch": 16.64,
+      "learning_rate": 1.684587813620072e-05,
+      "loss": 0.0236,
+      "step": 1165
+    },
+    {
+      "epoch": 16.66,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.0215,
+      "step": 1166
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 1.6702508960573475e-05,
+      "loss": 0.0209,
+      "step": 1167
+    },
+    {
+      "epoch": 16.69,
+      "learning_rate": 1.6630824372759856e-05,
+      "loss": 0.0209,
+      "step": 1168
+    },
+    {
+      "epoch": 16.7,
+      "learning_rate": 1.6559139784946237e-05,
+      "loss": 0.0205,
+      "step": 1169
+    },
+    {
+      "epoch": 16.71,
+      "learning_rate": 1.648745519713262e-05,
+      "loss": 0.0202,
+      "step": 1170
+    },
+    {
+      "epoch": 16.73,
+      "learning_rate": 1.6415770609319e-05,
+      "loss": 0.0262,
+      "step": 1171
+    },
+    {
+      "epoch": 16.74,
+      "learning_rate": 1.6344086021505377e-05,
+      "loss": 0.0203,
+      "step": 1172
+    },
+    {
+      "epoch": 16.76,
+      "learning_rate": 1.6272401433691755e-05,
+      "loss": 0.0243,
+      "step": 1173
+    },
+    {
+      "epoch": 16.77,
+      "learning_rate": 1.6200716845878136e-05,
+      "loss": 0.0182,
+      "step": 1174
+    },
+    {
+      "epoch": 16.79,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.0247,
+      "step": 1175
+    },
+    {
+      "epoch": 16.8,
+      "learning_rate": 1.6057347670250898e-05,
+      "loss": 0.0269,
+      "step": 1176
+    },
+    {
+      "epoch": 16.81,
+      "learning_rate": 1.5985663082437275e-05,
+      "loss": 0.0274,
+      "step": 1177
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 1.5913978494623657e-05,
+      "loss": 0.0243,
+      "step": 1178
+    },
+    {
+      "epoch": 16.84,
+      "learning_rate": 1.5842293906810038e-05,
+      "loss": 0.0298,
+      "step": 1179
+    },
+    {
+      "epoch": 16.86,
+      "learning_rate": 1.5770609318996415e-05,
+      "loss": 0.022,
+      "step": 1180
+    },
+    {
+      "epoch": 16.87,
+      "learning_rate": 1.5698924731182796e-05,
+      "loss": 0.0222,
+      "step": 1181
+    },
+    {
+      "epoch": 16.89,
+      "learning_rate": 1.5627240143369177e-05,
+      "loss": 0.0233,
+      "step": 1182
+    },
+    {
+      "epoch": 16.9,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.0226,
+      "step": 1183
+    },
+    {
+      "epoch": 16.91,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.0255,
+      "step": 1184
+    },
+    {
+      "epoch": 16.93,
+      "learning_rate": 1.5412186379928317e-05,
+      "loss": 0.0213,
+      "step": 1185
+    },
+    {
+      "epoch": 16.94,
+      "learning_rate": 1.5340501792114695e-05,
+      "loss": 0.0302,
+      "step": 1186
+    },
+    {
+      "epoch": 16.96,
+      "learning_rate": 1.5268817204301076e-05,
+      "loss": 0.0275,
+      "step": 1187
+    },
+    {
+      "epoch": 16.97,
+      "learning_rate": 1.5197132616487455e-05,
+      "loss": 0.0241,
+      "step": 1188
+    },
+    {
+      "epoch": 16.99,
+      "learning_rate": 1.5125448028673836e-05,
+      "loss": 0.0186,
+      "step": 1189
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 1.5053763440860215e-05,
+      "loss": 0.0222,
+      "step": 1190
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 4.085420608520508,
+      "eval_runtime": 29.2514,
+      "eval_samples_per_second": 4.102,
+      "eval_steps_per_second": 0.513,
+      "step": 1190
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 1.4982078853046596e-05,
+      "loss": 0.02,
+      "step": 1191
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 1.4910394265232978e-05,
+      "loss": 0.0217,
+      "step": 1192
+    },
+    {
+      "epoch": 17.04,
+      "learning_rate": 1.4838709677419355e-05,
+      "loss": 0.0275,
+      "step": 1193
+    },
+    {
+      "epoch": 17.06,
+      "learning_rate": 1.4767025089605735e-05,
+      "loss": 0.0238,
+      "step": 1194
+    },
+    {
+      "epoch": 17.07,
+      "learning_rate": 1.4695340501792116e-05,
+      "loss": 0.0218,
+      "step": 1195
+    },
+    {
+      "epoch": 17.09,
+      "learning_rate": 1.4623655913978495e-05,
+      "loss": 0.0191,
+      "step": 1196
+    },
+    {
+      "epoch": 17.1,
+      "learning_rate": 1.4551971326164876e-05,
+      "loss": 0.0213,
+      "step": 1197
+    },
+    {
+      "epoch": 17.11,
+      "learning_rate": 1.4480286738351257e-05,
+      "loss": 0.0236,
+      "step": 1198
+    },
+    {
+      "epoch": 17.13,
+      "learning_rate": 1.4408602150537635e-05,
+      "loss": 0.0191,
+      "step": 1199
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 1.4336917562724014e-05,
+      "loss": 0.019,
+      "step": 1200
+    },
+    {
+      "epoch": 17.16,
+      "learning_rate": 1.4265232974910395e-05,
+      "loss": 0.0238,
+      "step": 1201
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 1.4193548387096774e-05,
+      "loss": 0.0246,
+      "step": 1202
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 1.4121863799283155e-05,
+      "loss": 0.0197,
+      "step": 1203
+    },
+    {
+      "epoch": 17.2,
+      "learning_rate": 1.4050179211469536e-05,
+      "loss": 0.0196,
+      "step": 1204
+    },
+    {
+      "epoch": 17.21,
+      "learning_rate": 1.3978494623655914e-05,
+      "loss": 0.0233,
+      "step": 1205
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 1.3906810035842293e-05,
+      "loss": 0.0241,
+      "step": 1206
+    },
+    {
+      "epoch": 17.24,
+      "learning_rate": 1.3835125448028674e-05,
+      "loss": 0.0253,
+      "step": 1207
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 1.3763440860215054e-05,
+      "loss": 0.0215,
+      "step": 1208
+    },
+    {
+      "epoch": 17.27,
+      "learning_rate": 1.3691756272401435e-05,
+      "loss": 0.0183,
+      "step": 1209
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 1.3620071684587816e-05,
+      "loss": 0.0224,
+      "step": 1210
+    },
+    {
+      "epoch": 17.3,
+      "learning_rate": 1.3548387096774195e-05,
+      "loss": 0.0232,
+      "step": 1211
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 1.3476702508960573e-05,
+      "loss": 0.0216,
+      "step": 1212
+    },
+    {
+      "epoch": 17.33,
+      "learning_rate": 1.3405017921146954e-05,
+      "loss": 0.0225,
+      "step": 1213
+    },
+    {
+      "epoch": 17.34,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.0193,
+      "step": 1214
+    },
+    {
+      "epoch": 17.36,
+      "learning_rate": 1.3261648745519714e-05,
+      "loss": 0.0233,
+      "step": 1215
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 1.3189964157706094e-05,
+      "loss": 0.0214,
+      "step": 1216
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 1.3118279569892475e-05,
+      "loss": 0.0228,
+      "step": 1217
+    },
+    {
+      "epoch": 17.4,
+      "learning_rate": 1.3046594982078852e-05,
+      "loss": 0.0233,
+      "step": 1218
+    },
+    {
+      "epoch": 17.41,
+      "learning_rate": 1.2974910394265233e-05,
+      "loss": 0.025,
+      "step": 1219
+    },
+    {
+      "epoch": 17.43,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.0193,
+      "step": 1220
+    },
+    {
+      "epoch": 17.44,
+      "learning_rate": 1.2831541218637994e-05,
+      "loss": 0.0203,
+      "step": 1221
+    },
+    {
+      "epoch": 17.46,
+      "learning_rate": 1.2759856630824373e-05,
+      "loss": 0.026,
+      "step": 1222
+    },
+    {
+      "epoch": 17.47,
+      "learning_rate": 1.2688172043010754e-05,
+      "loss": 0.0223,
+      "step": 1223
+    },
+    {
+      "epoch": 17.49,
+      "learning_rate": 1.2616487455197135e-05,
+      "loss": 0.0232,
+      "step": 1224
+    },
+    {
+      "epoch": 17.5,
+      "learning_rate": 1.2544802867383513e-05,
+      "loss": 0.0229,
+      "step": 1225
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 1.2473118279569892e-05,
+      "loss": 0.0218,
+      "step": 1226
+    },
+    {
+      "epoch": 17.53,
+      "learning_rate": 1.2401433691756273e-05,
+      "loss": 0.0208,
+      "step": 1227
+    },
+    {
+      "epoch": 17.54,
+      "learning_rate": 1.2329749103942653e-05,
+      "loss": 0.0195,
+      "step": 1228
+    },
+    {
+      "epoch": 17.56,
+      "learning_rate": 1.2258064516129032e-05,
+      "loss": 0.02,
+      "step": 1229
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 1.2186379928315413e-05,
+      "loss": 0.0199,
+      "step": 1230
+    },
+    {
+      "epoch": 17.59,
+      "learning_rate": 1.2114695340501792e-05,
+      "loss": 0.0246,
+      "step": 1231
+    },
+    {
+      "epoch": 17.6,
+      "learning_rate": 1.2043010752688173e-05,
+      "loss": 0.0174,
+      "step": 1232
+    },
+    {
+      "epoch": 17.61,
+      "learning_rate": 1.1971326164874553e-05,
+      "loss": 0.0212,
+      "step": 1233
+    },
+    {
+      "epoch": 17.63,
+      "learning_rate": 1.1899641577060932e-05,
+      "loss": 0.0204,
+      "step": 1234
+    },
+    {
+      "epoch": 17.64,
+      "learning_rate": 1.1827956989247313e-05,
+      "loss": 0.0255,
+      "step": 1235
+    },
+    {
+      "epoch": 17.66,
+      "learning_rate": 1.1756272401433692e-05,
+      "loss": 0.0221,
+      "step": 1236
+    },
+    {
+      "epoch": 17.67,
+      "learning_rate": 1.1684587813620072e-05,
+      "loss": 0.0249,
+      "step": 1237
+    },
+    {
+      "epoch": 17.69,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.0242,
+      "step": 1238
+    },
+    {
+      "epoch": 17.7,
+      "learning_rate": 1.1541218637992832e-05,
+      "loss": 0.0228,
+      "step": 1239
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 1.1469534050179212e-05,
+      "loss": 0.0231,
+      "step": 1240
+    },
+    {
+      "epoch": 17.73,
+      "learning_rate": 1.1397849462365593e-05,
+      "loss": 0.0216,
+      "step": 1241
+    },
+    {
+      "epoch": 17.74,
+      "learning_rate": 1.1326164874551972e-05,
+      "loss": 0.0205,
+      "step": 1242
+    },
+    {
+      "epoch": 17.76,
+      "learning_rate": 1.1254480286738351e-05,
+      "loss": 0.0248,
+      "step": 1243
+    },
+    {
+      "epoch": 17.77,
+      "learning_rate": 1.1182795698924732e-05,
+      "loss": 0.0214,
+      "step": 1244
+    },
+    {
+      "epoch": 17.79,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.0198,
+      "step": 1245
+    },
+    {
+      "epoch": 17.8,
+      "learning_rate": 1.1039426523297491e-05,
+      "loss": 0.0271,
+      "step": 1246
+    },
+    {
+      "epoch": 17.81,
+      "learning_rate": 1.0967741935483872e-05,
+      "loss": 0.0212,
+      "step": 1247
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 1.0896057347670251e-05,
+      "loss": 0.0226,
+      "step": 1248
+    },
+    {
+      "epoch": 17.84,
+      "learning_rate": 1.082437275985663e-05,
+      "loss": 0.0271,
+      "step": 1249
+    },
+    {
+      "epoch": 17.86,
+      "learning_rate": 1.0752688172043012e-05,
+      "loss": 0.0249,
+      "step": 1250
+    },
+    {
+      "epoch": 17.87,
+      "learning_rate": 1.0681003584229391e-05,
+      "loss": 0.0229,
+      "step": 1251
+    },
+    {
+      "epoch": 17.89,
+      "learning_rate": 1.060931899641577e-05,
+      "loss": 0.0239,
+      "step": 1252
+    },
+    {
+      "epoch": 17.9,
+      "learning_rate": 1.0537634408602151e-05,
+      "loss": 0.027,
+      "step": 1253
+    },
+    {
+      "epoch": 17.91,
+      "learning_rate": 1.046594982078853e-05,
+      "loss": 0.0223,
+      "step": 1254
+    },
+    {
+      "epoch": 17.93,
+      "learning_rate": 1.039426523297491e-05,
+      "loss": 0.0228,
+      "step": 1255
+    },
+    {
+      "epoch": 17.94,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.0254,
+      "step": 1256
+    },
+    {
+      "epoch": 17.96,
+      "learning_rate": 1.025089605734767e-05,
+      "loss": 0.0223,
+      "step": 1257
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 1.017921146953405e-05,
+      "loss": 0.0225,
+      "step": 1258
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 1.0107526881720431e-05,
+      "loss": 0.0228,
+      "step": 1259
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 1.003584229390681e-05,
+      "loss": 0.0203,
+      "step": 1260
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 4.122922897338867,
+      "eval_runtime": 29.2501,
+      "eval_samples_per_second": 4.103,
+      "eval_steps_per_second": 0.513,
+      "step": 1260
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 9.96415770609319e-06,
+      "loss": 0.0217,
+      "step": 1261
+    },
+    {
+      "epoch": 18.03,
+      "learning_rate": 9.89247311827957e-06,
+      "loss": 0.0199,
+      "step": 1262
+    },
+    {
+      "epoch": 18.04,
+      "learning_rate": 9.82078853046595e-06,
+      "loss": 0.0254,
+      "step": 1263
+    },
+    {
+      "epoch": 18.06,
+      "learning_rate": 9.749103942652331e-06,
+      "loss": 0.0219,
+      "step": 1264
+    },
+    {
+      "epoch": 18.07,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.0207,
+      "step": 1265
+    },
+    {
+      "epoch": 18.09,
+      "learning_rate": 9.60573476702509e-06,
+      "loss": 0.0295,
+      "step": 1266
+    },
+    {
+      "epoch": 18.1,
+      "learning_rate": 9.53405017921147e-06,
+      "loss": 0.023,
+      "step": 1267
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 9.46236559139785e-06,
+      "loss": 0.0276,
+      "step": 1268
+    },
+    {
+      "epoch": 18.13,
+      "learning_rate": 9.39068100358423e-06,
+      "loss": 0.023,
+      "step": 1269
+    },
+    {
+      "epoch": 18.14,
+      "learning_rate": 9.31899641577061e-06,
+      "loss": 0.0226,
+      "step": 1270
+    },
+    {
+      "epoch": 18.16,
+      "learning_rate": 9.24731182795699e-06,
+      "loss": 0.0203,
+      "step": 1271
+    },
+    {
+      "epoch": 18.17,
+      "learning_rate": 9.17562724014337e-06,
+      "loss": 0.022,
+      "step": 1272
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 9.10394265232975e-06,
+      "loss": 0.0208,
+      "step": 1273
+    },
+    {
+      "epoch": 18.2,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.0152,
+      "step": 1274
+    },
+    {
+      "epoch": 18.21,
+      "learning_rate": 8.960573476702509e-06,
+      "loss": 0.023,
+      "step": 1275
+    },
+    {
+      "epoch": 18.23,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.0236,
+      "step": 1276
+    },
+    {
+      "epoch": 18.24,
+      "learning_rate": 8.81720430107527e-06,
+      "loss": 0.0197,
+      "step": 1277
+    },
+    {
+      "epoch": 18.26,
+      "learning_rate": 8.745519713261649e-06,
+      "loss": 0.0206,
+      "step": 1278
+    },
+    {
+      "epoch": 18.27,
+      "learning_rate": 8.67383512544803e-06,
+      "loss": 0.0212,
+      "step": 1279
+    },
+    {
+      "epoch": 18.29,
+      "learning_rate": 8.602150537634409e-06,
+      "loss": 0.0288,
+      "step": 1280
+    },
+    {
+      "epoch": 18.3,
+      "learning_rate": 8.530465949820788e-06,
+      "loss": 0.019,
+      "step": 1281
+    },
+    {
+      "epoch": 18.31,
+      "learning_rate": 8.45878136200717e-06,
+      "loss": 0.0224,
+      "step": 1282
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.0208,
+      "step": 1283
+    },
+    {
+      "epoch": 18.34,
+      "learning_rate": 8.315412186379928e-06,
+      "loss": 0.0157,
+      "step": 1284
+    },
+    {
+      "epoch": 18.36,
+      "learning_rate": 8.24372759856631e-06,
+      "loss": 0.0226,
+      "step": 1285
+    },
+    {
+      "epoch": 18.37,
+      "learning_rate": 8.172043010752689e-06,
+      "loss": 0.0204,
+      "step": 1286
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 8.100358422939068e-06,
+      "loss": 0.0228,
+      "step": 1287
+    },
+    {
+      "epoch": 18.4,
+      "learning_rate": 8.028673835125449e-06,
+      "loss": 0.0237,
+      "step": 1288
+    },
+    {
+      "epoch": 18.41,
+      "learning_rate": 7.956989247311828e-06,
+      "loss": 0.021,
+      "step": 1289
+    },
+    {
+      "epoch": 18.43,
+      "learning_rate": 7.885304659498208e-06,
+      "loss": 0.0264,
+      "step": 1290
+    },
+    {
+      "epoch": 18.44,
+      "learning_rate": 7.813620071684589e-06,
+      "loss": 0.0205,
+      "step": 1291
+    },
+    {
+      "epoch": 18.46,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.0275,
+      "step": 1292
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 7.670250896057347e-06,
+      "loss": 0.0228,
+      "step": 1293
+    },
+    {
+      "epoch": 18.49,
+      "learning_rate": 7.5985663082437275e-06,
+      "loss": 0.0186,
+      "step": 1294
+    },
+    {
+      "epoch": 18.5,
+      "learning_rate": 7.526881720430108e-06,
+      "loss": 0.0225,
+      "step": 1295
+    },
+    {
+      "epoch": 18.51,
+      "learning_rate": 7.455197132616489e-06,
+      "loss": 0.0272,
+      "step": 1296
+    },
+    {
+      "epoch": 18.53,
+      "learning_rate": 7.383512544802867e-06,
+      "loss": 0.0216,
+      "step": 1297
+    },
+    {
+      "epoch": 18.54,
+      "learning_rate": 7.3118279569892475e-06,
+      "loss": 0.0273,
+      "step": 1298
+    },
+    {
+      "epoch": 18.56,
+      "learning_rate": 7.2401433691756285e-06,
+      "loss": 0.0247,
+      "step": 1299
+    },
+    {
+      "epoch": 18.57,
+      "learning_rate": 7.168458781362007e-06,
+      "loss": 0.021,
+      "step": 1300
+    },
+    {
+      "epoch": 18.59,
+      "learning_rate": 7.096774193548387e-06,
+      "loss": 0.0183,
+      "step": 1301
+    },
+    {
+      "epoch": 18.6,
+      "learning_rate": 7.025089605734768e-06,
+      "loss": 0.0201,
+      "step": 1302
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 6.953405017921147e-06,
+      "loss": 0.0223,
+      "step": 1303
+    },
+    {
+      "epoch": 18.63,
+      "learning_rate": 6.881720430107527e-06,
+      "loss": 0.0193,
+      "step": 1304
+    },
+    {
+      "epoch": 18.64,
+      "learning_rate": 6.810035842293908e-06,
+      "loss": 0.0254,
+      "step": 1305
+    },
+    {
+      "epoch": 18.66,
+      "learning_rate": 6.7383512544802865e-06,
+      "loss": 0.0206,
+      "step": 1306
+    },
+    {
+      "epoch": 18.67,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.0187,
+      "step": 1307
+    },
+    {
+      "epoch": 18.69,
+      "learning_rate": 6.594982078853047e-06,
+      "loss": 0.0189,
+      "step": 1308
+    },
+    {
+      "epoch": 18.7,
+      "learning_rate": 6.523297491039426e-06,
+      "loss": 0.0242,
+      "step": 1309
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.0198,
+      "step": 1310
+    },
+    {
+      "epoch": 18.73,
+      "learning_rate": 6.3799283154121866e-06,
+      "loss": 0.0202,
+      "step": 1311
+    },
+    {
+      "epoch": 18.74,
+      "learning_rate": 6.308243727598568e-06,
+      "loss": 0.0237,
+      "step": 1312
+    },
+    {
+      "epoch": 18.76,
+      "learning_rate": 6.236559139784946e-06,
+      "loss": 0.021,
+      "step": 1313
+    },
+    {
+      "epoch": 18.77,
+      "learning_rate": 6.164874551971326e-06,
+      "loss": 0.0231,
+      "step": 1314
+    },
+    {
+      "epoch": 18.79,
+      "learning_rate": 6.0931899641577065e-06,
+      "loss": 0.0173,
+      "step": 1315
+    },
+    {
+      "epoch": 18.8,
+      "learning_rate": 6.021505376344087e-06,
+      "loss": 0.0229,
+      "step": 1316
+    },
+    {
+      "epoch": 18.81,
+      "learning_rate": 5.949820788530466e-06,
+      "loss": 0.0312,
+      "step": 1317
+    },
+    {
+      "epoch": 18.83,
+      "learning_rate": 5.878136200716846e-06,
+      "loss": 0.022,
+      "step": 1318
+    },
+    {
+      "epoch": 18.84,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.0205,
+      "step": 1319
+    },
+    {
+      "epoch": 18.86,
+      "learning_rate": 5.734767025089606e-06,
+      "loss": 0.0194,
+      "step": 1320
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 5.663082437275986e-06,
+      "loss": 0.0238,
+      "step": 1321
+    },
+    {
+      "epoch": 18.89,
+      "learning_rate": 5.591397849462366e-06,
+      "loss": 0.0314,
+      "step": 1322
+    },
+    {
+      "epoch": 18.9,
+      "learning_rate": 5.5197132616487455e-06,
+      "loss": 0.0232,
+      "step": 1323
+    },
+    {
+      "epoch": 18.91,
+      "learning_rate": 5.448028673835126e-06,
+      "loss": 0.0187,
+      "step": 1324
+    },
+    {
+      "epoch": 18.93,
+      "learning_rate": 5.376344086021506e-06,
+      "loss": 0.0252,
+      "step": 1325
+    },
+    {
+      "epoch": 18.94,
+      "learning_rate": 5.304659498207885e-06,
+      "loss": 0.0247,
+      "step": 1326
+    },
+    {
+      "epoch": 18.96,
+      "learning_rate": 5.232974910394265e-06,
+      "loss": 0.0204,
+      "step": 1327
+    },
+    {
+      "epoch": 18.97,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.0202,
+      "step": 1328
+    },
+    {
+      "epoch": 18.99,
+      "learning_rate": 5.089605734767025e-06,
+      "loss": 0.0261,
+      "step": 1329
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 5.017921146953405e-06,
+      "loss": 0.0231,
+      "step": 1330
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 4.1433234214782715,
+      "eval_runtime": 29.1935,
+      "eval_samples_per_second": 4.111,
+      "eval_steps_per_second": 0.514,
+      "step": 1330
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 4.946236559139785e-06,
+      "loss": 0.0252,
+      "step": 1331
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 4.8745519713261655e-06,
+      "loss": 0.0313,
+      "step": 1332
+    },
+    {
+      "epoch": 19.04,
+      "learning_rate": 4.802867383512545e-06,
+      "loss": 0.0241,
+      "step": 1333
+    },
+    {
+      "epoch": 19.06,
+      "learning_rate": 4.731182795698925e-06,
+      "loss": 0.0218,
+      "step": 1334
+    },
+    {
+      "epoch": 19.07,
+      "learning_rate": 4.659498207885305e-06,
+      "loss": 0.0273,
+      "step": 1335
+    },
+    {
+      "epoch": 19.09,
+      "learning_rate": 4.587813620071685e-06,
+      "loss": 0.0222,
+      "step": 1336
+    },
+    {
+      "epoch": 19.1,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.0215,
+      "step": 1337
+    },
+    {
+      "epoch": 19.11,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 0.0226,
+      "step": 1338
+    },
+    {
+      "epoch": 19.13,
+      "learning_rate": 4.372759856630824e-06,
+      "loss": 0.022,
+      "step": 1339
+    },
+    {
+      "epoch": 19.14,
+      "learning_rate": 4.3010752688172045e-06,
+      "loss": 0.0229,
+      "step": 1340
+    },
+    {
+      "epoch": 19.16,
+      "learning_rate": 4.229390681003585e-06,
+      "loss": 0.019,
+      "step": 1341
+    },
+    {
+      "epoch": 19.17,
+      "learning_rate": 4.157706093189964e-06,
+      "loss": 0.023,
+      "step": 1342
+    },
+    {
+      "epoch": 19.19,
+      "learning_rate": 4.086021505376344e-06,
+      "loss": 0.0214,
+      "step": 1343
+    },
+    {
+      "epoch": 19.2,
+      "learning_rate": 4.0143369175627245e-06,
+      "loss": 0.0263,
+      "step": 1344
+    },
+    {
+      "epoch": 19.21,
+      "learning_rate": 3.942652329749104e-06,
+      "loss": 0.0236,
+      "step": 1345
+    },
+    {
+      "epoch": 19.23,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.0184,
+      "step": 1346
+    },
+    {
+      "epoch": 19.24,
+      "learning_rate": 3.7992831541218638e-06,
+      "loss": 0.0179,
+      "step": 1347
+    },
+    {
+      "epoch": 19.26,
+      "learning_rate": 3.7275985663082444e-06,
+      "loss": 0.027,
+      "step": 1348
+    },
+    {
+      "epoch": 19.27,
+      "learning_rate": 3.6559139784946237e-06,
+      "loss": 0.0243,
+      "step": 1349
+    },
+    {
+      "epoch": 19.29,
+      "learning_rate": 3.5842293906810035e-06,
+      "loss": 0.0233,
+      "step": 1350
+    },
+    {
+      "epoch": 19.3,
+      "learning_rate": 3.512544802867384e-06,
+      "loss": 0.0179,
+      "step": 1351
+    },
+    {
+      "epoch": 19.31,
+      "learning_rate": 3.4408602150537635e-06,
+      "loss": 0.0214,
+      "step": 1352
+    },
+    {
+      "epoch": 19.33,
+      "learning_rate": 3.3691756272401432e-06,
+      "loss": 0.0195,
+      "step": 1353
+    },
+    {
+      "epoch": 19.34,
+      "learning_rate": 3.2974910394265234e-06,
+      "loss": 0.0214,
+      "step": 1354
+    },
+    {
+      "epoch": 19.36,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.0219,
+      "step": 1355
+    },
+    {
+      "epoch": 19.37,
+      "learning_rate": 3.154121863799284e-06,
+      "loss": 0.0197,
+      "step": 1356
+    },
+    {
+      "epoch": 19.39,
+      "learning_rate": 3.082437275985663e-06,
+      "loss": 0.0223,
+      "step": 1357
+    },
+    {
+      "epoch": 19.4,
+      "learning_rate": 3.0107526881720433e-06,
+      "loss": 0.0189,
+      "step": 1358
+    },
+    {
+      "epoch": 19.41,
+      "learning_rate": 2.939068100358423e-06,
+      "loss": 0.0179,
+      "step": 1359
+    },
+    {
+      "epoch": 19.43,
+      "learning_rate": 2.867383512544803e-06,
+      "loss": 0.0248,
+      "step": 1360
+    },
+    {
+      "epoch": 19.44,
+      "learning_rate": 2.795698924731183e-06,
+      "loss": 0.0196,
+      "step": 1361
+    },
+    {
+      "epoch": 19.46,
+      "learning_rate": 2.724014336917563e-06,
+      "loss": 0.0238,
+      "step": 1362
+    },
+    {
+      "epoch": 19.47,
+      "learning_rate": 2.6523297491039426e-06,
+      "loss": 0.0173,
+      "step": 1363
+    },
+    {
+      "epoch": 19.49,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.0264,
+      "step": 1364
+    },
+    {
+      "epoch": 19.5,
+      "learning_rate": 2.5089605734767026e-06,
+      "loss": 0.0185,
+      "step": 1365
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 2.4372759856630828e-06,
+      "loss": 0.0222,
+      "step": 1366
+    },
+    {
+      "epoch": 19.53,
+      "learning_rate": 2.3655913978494625e-06,
+      "loss": 0.0241,
+      "step": 1367
+    },
+    {
+      "epoch": 19.54,
+      "learning_rate": 2.2939068100358423e-06,
+      "loss": 0.0193,
+      "step": 1368
+    },
+    {
+      "epoch": 19.56,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 0.021,
+      "step": 1369
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 2.1505376344086023e-06,
+      "loss": 0.018,
+      "step": 1370
+    },
+    {
+      "epoch": 19.59,
+      "learning_rate": 2.078853046594982e-06,
+      "loss": 0.0182,
+      "step": 1371
+    },
+    {
+      "epoch": 19.6,
+      "learning_rate": 2.0071684587813622e-06,
+      "loss": 0.0199,
+      "step": 1372
+    },
+    {
+      "epoch": 19.61,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.0212,
+      "step": 1373
+    },
+    {
+      "epoch": 19.63,
+      "learning_rate": 1.8637992831541222e-06,
+      "loss": 0.0283,
+      "step": 1374
+    },
+    {
+      "epoch": 19.64,
+      "learning_rate": 1.7921146953405017e-06,
+      "loss": 0.0252,
+      "step": 1375
+    },
+    {
+      "epoch": 19.66,
+      "learning_rate": 1.7204301075268817e-06,
+      "loss": 0.024,
+      "step": 1376
+    },
+    {
+      "epoch": 19.67,
+      "learning_rate": 1.6487455197132617e-06,
+      "loss": 0.0192,
+      "step": 1377
+    },
+    {
+      "epoch": 19.69,
+      "learning_rate": 1.577060931899642e-06,
+      "loss": 0.0191,
+      "step": 1378
+    },
+    {
+      "epoch": 19.7,
+      "learning_rate": 1.5053763440860217e-06,
+      "loss": 0.0201,
+      "step": 1379
+    },
+    {
+      "epoch": 19.71,
+      "learning_rate": 1.4336917562724014e-06,
+      "loss": 0.0226,
+      "step": 1380
+    },
+    {
+      "epoch": 19.73,
+      "learning_rate": 1.3620071684587814e-06,
+      "loss": 0.0275,
+      "step": 1381
+    },
+    {
+      "epoch": 19.74,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.0246,
+      "step": 1382
+    },
+    {
+      "epoch": 19.76,
+      "learning_rate": 1.2186379928315414e-06,
+      "loss": 0.018,
+      "step": 1383
+    },
+    {
+      "epoch": 19.77,
+      "learning_rate": 1.1469534050179212e-06,
+      "loss": 0.0239,
+      "step": 1384
+    },
+    {
+      "epoch": 19.79,
+      "learning_rate": 1.0752688172043011e-06,
+      "loss": 0.0203,
+      "step": 1385
+    },
+    {
+      "epoch": 19.8,
+      "learning_rate": 1.0035842293906811e-06,
+      "loss": 0.0229,
+      "step": 1386
+    },
+    {
+      "epoch": 19.81,
+      "learning_rate": 9.318996415770611e-07,
+      "loss": 0.0251,
+      "step": 1387
+    },
+    {
+      "epoch": 19.83,
+      "learning_rate": 8.602150537634409e-07,
+      "loss": 0.0278,
+      "step": 1388
+    },
+    {
+      "epoch": 19.84,
+      "learning_rate": 7.88530465949821e-07,
+      "loss": 0.0207,
+      "step": 1389
+    },
+    {
+      "epoch": 19.86,
+      "learning_rate": 7.168458781362007e-07,
+      "loss": 0.0178,
+      "step": 1390
+    },
+    {
+      "epoch": 19.87,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.0185,
+      "step": 1391
+    },
+    {
+      "epoch": 19.89,
+      "learning_rate": 5.734767025089606e-07,
+      "loss": 0.0197,
+      "step": 1392
+    },
+    {
+      "epoch": 19.9,
+      "learning_rate": 5.017921146953406e-07,
+      "loss": 0.0221,
+      "step": 1393
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 4.3010752688172043e-07,
+      "loss": 0.0211,
+      "step": 1394
+    },
+    {
+      "epoch": 19.93,
+      "learning_rate": 3.5842293906810036e-07,
+      "loss": 0.0198,
+      "step": 1395
+    },
+    {
+      "epoch": 19.94,
+      "learning_rate": 2.867383512544803e-07,
+      "loss": 0.0197,
+      "step": 1396
+    },
+    {
+      "epoch": 19.96,
+      "learning_rate": 2.1505376344086022e-07,
+      "loss": 0.0206,
+      "step": 1397
+    },
+    {
+      "epoch": 19.97,
+      "learning_rate": 1.4336917562724014e-07,
+      "loss": 0.0219,
+      "step": 1398
+    },
+    {
+      "epoch": 19.99,
+      "learning_rate": 7.168458781362007e-08,
+      "loss": 0.0211,
+      "step": 1399
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 0.0,
+      "loss": 0.0199,
+      "step": 1400
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 4.1544904708862305,
+      "eval_runtime": 29.2342,
+      "eval_samples_per_second": 4.105,
+      "eval_steps_per_second": 0.513,
+      "step": 1400
+    },
+    {
+      "epoch": 20.0,
+      "step": 1400,
+      "total_flos": 1.342333327835136e+17,
+      "train_loss": 0.135591789316386,
+      "train_runtime": 4699.1287,
+      "train_samples_per_second": 1.192,
+      "train_steps_per_second": 0.298
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1400,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "total_flos": 1.342333327835136e+17,
+  "trial_name": null,
+  "trial_params": null
+}