diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
deleted file mode 100644--- "a/last-checkpoint/trainer_state.json"
+++ /dev/null
@@ -1,7416 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 15.27020630048712,
-  "global_step": 1000000,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 0.8784,
-      "step": 1000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 5.999999999999999e-06,
-      "loss": 0.77,
-      "step": 2000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 8.999999999999999e-06,
-      "loss": 0.7664,
-      "step": 3000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.1999999999999999e-05,
-      "loss": 0.7655,
-      "step": 4000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.4999999999999999e-05,
-      "loss": 0.765,
-      "step": 5000
-    },
-    {
-      "epoch": 0.08,
-      "eval_runtime": 1.3797,
-      "eval_samples_per_second": 724.791,
-      "eval_steps_per_second": 11.597,
-      "step": 5000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.7999999999999997e-05,
-      "loss": 0.7647,
-      "step": 6000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 2.1e-05,
-      "loss": 0.7644,
-      "step": 7000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 2.3999999999999997e-05,
-      "loss": 0.7638,
-      "step": 8000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 2.6999999999999996e-05,
-      "loss": 0.7633,
-      "step": 9000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 2.9999999999999997e-05,
-      "loss": 0.76,
-      "step": 10000
-    },
-    {
-      "epoch": 0.15,
-      "eval_runtime": 1.1376,
-      "eval_samples_per_second": 879.066,
-      "eval_steps_per_second": 14.065,
-      "step": 10000
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 3.2999999999999996e-05,
-      "loss": 0.7148,
-      "step": 11000
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 3.5999999999999994e-05,
-      "loss": 0.6963,
-      "step": 12000
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 3.9e-05,
-      "loss": 0.6755,
-      "step": 13000
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 4.2e-05,
-      "loss": 0.6516,
-      "step": 14000
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 4.4999999999999996e-05,
-      "loss": 0.6412,
-      "step": 15000
-    },
-    {
-      "epoch": 0.23,
-      "eval_runtime": 1.1689,
-      "eval_samples_per_second": 855.472,
-      "eval_steps_per_second": 13.688,
-      "step": 15000
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 4.7999999999999994e-05,
-      "loss": 0.6348,
-      "step": 16000
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 5.1e-05,
-      "loss": 0.6295,
-      "step": 17000
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 5.399999999999999e-05,
-      "loss": 0.6224,
-      "step": 18000
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 5.6999999999999996e-05,
-      "loss": 0.6169,
-      "step": 19000
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 5.9999999999999995e-05,
-      "loss": 0.6113,
-      "step": 20000
-    },
-    {
-      "epoch": 0.31,
-      "eval_runtime": 1.0179,
-      "eval_samples_per_second": 982.441,
-      "eval_steps_per_second": 15.719,
-      "step": 20000
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.299999999999999e-05,
-      "loss": 0.6074,
-      "step": 21000
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 6.599999999999999e-05,
-      "loss": 0.6039,
-      "step": 22000
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.9e-05,
-      "loss": 0.6005,
-      "step": 23000
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 7.199999999999999e-05,
-      "loss": 0.5968,
-      "step": 24000
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 7.5e-05,
-      "loss": 0.5932,
-      "step": 25000
-    },
-    {
-      "epoch": 0.38,
-      "eval_runtime": 1.1249,
-      "eval_samples_per_second": 888.989,
-      "eval_steps_per_second": 14.224,
-      "step": 25000
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 7.8e-05,
-      "loss": 0.5912,
-      "step": 26000
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 8.1e-05,
-      "loss": 0.58,
-      "step": 27000
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 8.4e-05,
-      "loss": 0.5698,
-      "step": 28000
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 8.699999999999999e-05,
-      "loss": 0.5639,
-      "step": 29000
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 8.999999999999999e-05,
-      "loss": 0.5601,
-      "step": 30000
-    },
-    {
-      "epoch": 0.46,
-      "eval_runtime": 1.0096,
-      "eval_samples_per_second": 990.512,
-      "eval_steps_per_second": 15.848,
-      "step": 30000
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 9.3e-05,
-      "loss": 0.5536,
-      "step": 31000
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 9.599999999999999e-05,
-      "loss": 0.5496,
-      "step": 32000
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 9.9e-05,
-      "loss": 0.5458,
-      "step": 33000
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 0.000102,
-      "loss": 0.5426,
-      "step": 34000
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 0.00010499999999999999,
-      "loss": 0.5394,
-      "step": 35000
-    },
-    {
-      "epoch": 0.53,
-      "eval_runtime": 1.3102,
-      "eval_samples_per_second": 763.27,
-      "eval_steps_per_second": 12.212,
-      "step": 35000
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 0.00010799999999999998,
-      "loss": 0.5345,
-      "step": 36000
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 0.00011099999999999999,
-      "loss": 0.5302,
-      "step": 37000
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 0.00011399999999999999,
-      "loss": 0.527,
-      "step": 38000
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 0.000117,
-      "loss": 0.5232,
-      "step": 39000
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 0.00011999999999999999,
-      "loss": 0.5202,
-      "step": 40000
-    },
-    {
-      "epoch": 0.61,
-      "eval_runtime": 1.0146,
-      "eval_samples_per_second": 985.598,
-      "eval_steps_per_second": 15.77,
-      "step": 40000
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 0.00012299999999999998,
-      "loss": 0.5163,
-      "step": 41000
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 0.00012599999999999997,
-      "loss": 0.5126,
-      "step": 42000
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 0.000129,
-      "loss": 0.5094,
-      "step": 43000
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 0.00013199999999999998,
-      "loss": 0.5061,
-      "step": 44000
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 0.000135,
-      "loss": 0.5036,
-      "step": 45000
-    },
-    {
-      "epoch": 0.69,
-      "eval_runtime": 1.0362,
-      "eval_samples_per_second": 965.067,
-      "eval_steps_per_second": 15.441,
-      "step": 45000
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 0.000138,
-      "loss": 0.4995,
-      "step": 46000
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 0.00014099999999999998,
-      "loss": 0.4967,
-      "step": 47000
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 0.00014399999999999998,
-      "loss": 0.4934,
-      "step": 48000
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 0.000147,
-      "loss": 0.4898,
-      "step": 49000
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 0.00015,
-      "loss": 0.4863,
-      "step": 50000
-    },
-    {
-      "epoch": 0.76,
-      "eval_runtime": 1.0374,
-      "eval_samples_per_second": 963.96,
-      "eval_steps_per_second": 15.423,
-      "step": 50000
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 0.0001499996172456075,
-      "loss": 0.4824,
-      "step": 51000
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 0.00014999846898661572,
-      "loss": 0.4778,
-      "step": 52000
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 0.00014999655523558183,
-      "loss": 0.474,
-      "step": 53000
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 0.00014999387601343436,
-      "loss": 0.4694,
-      "step": 54000
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 0.00014999043134947282,
-      "loss": 0.4651,
-      "step": 55000
-    },
-    {
-      "epoch": 0.84,
-      "eval_runtime": 1.0465,
-      "eval_samples_per_second": 955.566,
-      "eval_steps_per_second": 15.289,
-      "step": 55000
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 0.00014998622128136748,
-      "loss": 0.4608,
-      "step": 56000
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 0.000149981245855159,
-      "loss": 0.4566,
-      "step": 57000
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 0.00014997550512525784,
-      "loss": 0.4523,
-      "step": 58000
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 0.0001499689991544437,
-      "loss": 0.4483,
-      "step": 59000
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 0.00014996172801386482,
-      "loss": 0.4447,
-      "step": 60000
-    },
-    {
-      "epoch": 0.92,
-      "eval_runtime": 1.2288,
-      "eval_samples_per_second": 813.826,
-      "eval_steps_per_second": 13.021,
-      "step": 60000
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 0.00014995369178303722,
-      "loss": 0.4408,
-      "step": 61000
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 0.0001499448905498439,
-      "loss": 0.4381,
-      "step": 62000
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 0.00014993532441053364,
-      "loss": 0.434,
-      "step": 63000
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 0.0001499249934697203,
-      "loss": 0.4316,
-      "step": 64000
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 0.0001499138978403813,
-      "loss": 0.4275,
-      "step": 65000
-    },
-    {
-      "epoch": 0.99,
-      "eval_runtime": 1.0345,
-      "eval_samples_per_second": 966.655,
-      "eval_steps_per_second": 15.466,
-      "step": 65000
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 0.00014990203764385677,
-      "loss": 0.425,
-      "step": 66000
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 0.00014988941300984784,
-      "loss": 0.422,
-      "step": 67000
-    },
-    {
-      "epoch": 1.04,
-      "learning_rate": 0.0001498760240764155,
-      "loss": 0.4191,
-      "step": 68000
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 0.000149861870989979,
-      "loss": 0.4164,
-      "step": 69000
-    },
-    {
-      "epoch": 1.07,
-      "learning_rate": 0.0001498469539053142,
-      "loss": 0.4138,
-      "step": 70000
-    },
-    {
-      "epoch": 1.07,
-      "eval_runtime": 1.1341,
-      "eval_samples_per_second": 881.784,
-      "eval_steps_per_second": 14.109,
-      "step": 70000
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 0.00014983127298555198,
-      "loss": 0.4114,
-      "step": 71000
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 0.00014981482840217632,
-      "loss": 0.4086,
-      "step": 72000
-    },
-    {
-      "epoch": 1.11,
-      "learning_rate": 0.00014979762033502262,
-      "loss": 0.4066,
-      "step": 73000
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 0.00014977964897227547,
-      "loss": 0.4042,
-      "step": 74000
-    },
-    {
-      "epoch": 1.15,
-      "learning_rate": 0.00014976091451046687,
-      "loss": 0.402,
-      "step": 75000
-    },
-    {
-      "epoch": 1.15,
-      "eval_runtime": 1.0331,
-      "eval_samples_per_second": 967.957,
-      "eval_steps_per_second": 15.487,
-      "step": 75000
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 0.00014974141715447386,
-      "loss": 0.3999,
-      "step": 76000
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 0.00014972115711751644,
-      "loss": 0.398,
-      "step": 77000
-    },
-    {
-      "epoch": 1.19,
-      "learning_rate": 0.00014970013462115505,
-      "loss": 0.3971,
-      "step": 78000
-    },
-    {
-      "epoch": 1.21,
-      "learning_rate": 0.00014967834989528843,
-      "loss": 0.3942,
-      "step": 79000
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 0.00014965580317815078,
-      "loss": 0.3926,
-      "step": 80000
-    },
-    {
-      "epoch": 1.22,
-      "eval_runtime": 1.084,
-      "eval_samples_per_second": 922.521,
-      "eval_steps_per_second": 14.76,
-      "step": 80000
-    },
-    {
-      "epoch": 1.24,
-      "learning_rate": 0.00014963249471630944,
-      "loss": 0.3906,
-      "step": 81000
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 0.000149608424764662,
-      "loss": 0.391,
-      "step": 82000
-    },
-    {
-      "epoch": 1.27,
-      "learning_rate": 0.0001495835935864336,
-      "loss": 0.3875,
-      "step": 83000
-    },
-    {
-      "epoch": 1.28,
-      "learning_rate": 0.00014955800145317397,
-      "loss": 0.3861,
-      "step": 84000
-    },
-    {
-      "epoch": 1.3,
-      "learning_rate": 0.00014953164864475466,
-      "loss": 0.3844,
-      "step": 85000
-    },
-    {
-      "epoch": 1.3,
-      "eval_runtime": 1.0992,
-      "eval_samples_per_second": 909.734,
-      "eval_steps_per_second": 14.556,
-      "step": 85000
-    },
-    {
-      "epoch": 1.31,
-      "learning_rate": 0.0001495045354493657,
-      "loss": 0.3829,
-      "step": 86000
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 0.00014947666216351272,
-      "loss": 0.3815,
-      "step": 87000
-    },
-    {
-      "epoch": 1.34,
-      "learning_rate": 0.00014944802909201344,
-      "loss": 0.38,
-      "step": 88000
-    },
-    {
-      "epoch": 1.36,
-      "learning_rate": 0.00014941863654799456,
-      "loss": 0.3789,
-      "step": 89000
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 0.00014938848485288825,
-      "loss": 0.3785,
-      "step": 90000
-    },
-    {
-      "epoch": 1.37,
-      "eval_runtime": 0.9266,
-      "eval_samples_per_second": 1079.167,
-      "eval_steps_per_second": 17.267,
-      "step": 90000
-    },
-    {
-      "epoch": 1.39,
-      "learning_rate": 0.0001493575743364286,
-      "loss": 0.3766,
-      "step": 91000
-    },
-    {
-      "epoch": 1.4,
-      "learning_rate": 0.00014932590533664808,
-      "loss": 0.3745,
-      "step": 92000
-    },
-    {
-      "epoch": 1.42,
-      "learning_rate": 0.0001492934781998738,
-      "loss": 0.3741,
-      "step": 93000
-    },
-    {
-      "epoch": 1.44,
-      "learning_rate": 0.0001492602932807237,
-      "loss": 0.3729,
-      "step": 94000
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 0.00014922635094210277,
-      "loss": 0.3709,
-      "step": 95000
-    },
-    {
-      "epoch": 1.45,
-      "eval_runtime": 0.9895,
-      "eval_samples_per_second": 1010.579,
-      "eval_steps_per_second": 16.169,
-      "step": 95000
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 0.000149191651555199,
-      "loss": 0.3699,
-      "step": 96000
-    },
-    {
-      "epoch": 1.48,
-      "learning_rate": 0.0001491561954994793,
-      "loss": 0.3688,
-      "step": 97000
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 0.00014911998316268537,
-      "loss": 0.3678,
-      "step": 98000
-    },
-    {
-      "epoch": 1.51,
-      "learning_rate": 0.00014908301494082963,
-      "loss": 0.3666,
-      "step": 99000
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 0.00014904529123819054,
-      "loss": 0.3654,
-      "step": 100000
-    },
-    {
-      "epoch": 1.53,
-      "eval_runtime": 1.0046,
-      "eval_samples_per_second": 995.424,
-      "eval_steps_per_second": 15.927,
-      "step": 100000
-    },
-    {
-      "epoch": 1.54,
-      "learning_rate": 0.00014900681246730852,
-      "loss": 0.3643,
-      "step": 101000
-    },
-    {
-      "epoch": 1.56,
-      "learning_rate": 0.00014896757904898125,
-      "loss": 0.3646,
-      "step": 102000
-    },
-    {
-      "epoch": 1.57,
-      "learning_rate": 0.00014892759141225904,
-      "loss": 0.3628,
-      "step": 103000
-    },
-    {
-      "epoch": 1.59,
-      "learning_rate": 0.00014888684999444035,
-      "loss": 0.3616,
-      "step": 104000
-    },
-    {
-      "epoch": 1.6,
-      "learning_rate": 0.00014884535524106675,
-      "loss": 0.3604,
-      "step": 105000
-    },
-    {
-      "epoch": 1.6,
-      "eval_runtime": 1.0499,
-      "eval_samples_per_second": 952.499,
-      "eval_steps_per_second": 15.24,
-      "step": 105000
-    },
-    {
-      "epoch": 1.62,
-      "learning_rate": 0.00014880310760591824,
-      "loss": 0.3594,
-      "step": 106000
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 0.0001487601075510082,
-      "loss": 0.3597,
-      "step": 107000
-    },
-    {
-      "epoch": 1.65,
-      "learning_rate": 0.0001487163555465783,
-      "loss": 0.3583,
-      "step": 108000
-    },
-    {
-      "epoch": 1.66,
-      "learning_rate": 0.0001486718520710935,
-      "loss": 0.3583,
-      "step": 109000
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 0.00014862659761123663,
-      "loss": 0.3558,
-      "step": 110000
-    },
-    {
-      "epoch": 1.68,
-      "eval_runtime": 1.0153,
-      "eval_samples_per_second": 984.91,
-      "eval_steps_per_second": 15.759,
-      "step": 110000
-    },
-    {
-      "epoch": 1.69,
-      "learning_rate": 0.00014858059266190327,
-      "loss": 0.3552,
-      "step": 111000
-    },
-    {
-      "epoch": 1.71,
-      "learning_rate": 0.00014853383772619612,
-      "loss": 0.3544,
-      "step": 112000
-    },
-    {
-      "epoch": 1.73,
-      "learning_rate": 0.00014848633331541967,
-      "loss": 0.3537,
-      "step": 113000
-    },
-    {
-      "epoch": 1.74,
-      "learning_rate": 0.0001484380799490746,
-      "loss": 0.3524,
-      "step": 114000
-    },
-    {
-      "epoch": 1.76,
-      "learning_rate": 0.00014838907815485194,
-      "loss": 0.3519,
-      "step": 115000
-    },
-    {
-      "epoch": 1.76,
-      "eval_runtime": 1.003,
-      "eval_samples_per_second": 997.001,
-      "eval_steps_per_second": 15.952,
-      "step": 115000
-    },
-    {
-      "epoch": 1.77,
-      "learning_rate": 0.00014833932846862748,
-      "loss": 0.3511,
-      "step": 116000
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 0.00014828883143445582,
-      "loss": 0.3502,
-      "step": 117000
-    },
-    {
-      "epoch": 1.8,
-      "learning_rate": 0.0001482375876045644,
-      "loss": 0.3493,
-      "step": 118000
-    },
-    {
-      "epoch": 1.82,
-      "learning_rate": 0.0001481855975393476,
-      "loss": 0.3489,
-      "step": 119000
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 0.0001481328618073604,
-      "loss": 0.3482,
-      "step": 120000
-    },
-    {
-      "epoch": 1.83,
-      "eval_runtime": 1.0596,
-      "eval_samples_per_second": 943.744,
-      "eval_steps_per_second": 15.1,
-      "step": 120000
-    },
-    {
-      "epoch": 1.85,
-      "learning_rate": 0.0001480793809853123,
-      "loss": 0.3478,
-      "step": 121000
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 0.00014802515565806107,
-      "loss": 0.3468,
-      "step": 122000
-    },
-    {
-      "epoch": 1.88,
-      "learning_rate": 0.00014797018641860612,
-      "loss": 0.346,
-      "step": 123000
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 0.0001479144738680823,
-      "loss": 0.3474,
-      "step": 124000
-    },
-    {
-      "epoch": 1.91,
-      "learning_rate": 0.00014785801861575312,
-      "loss": 0.3447,
-      "step": 125000
-    },
-    {
-      "epoch": 1.91,
-      "eval_runtime": 0.9375,
-      "eval_samples_per_second": 1066.699,
-      "eval_steps_per_second": 17.067,
-      "step": 125000
-    },
-    {
-      "epoch": 1.92,
-      "learning_rate": 0.00014780082127900416,
-      "loss": 0.3439,
-      "step": 126000
-    },
-    {
-      "epoch": 1.94,
-      "learning_rate": 0.00014774288248333635,
-      "loss": 0.3436,
-      "step": 127000
-    },
-    {
-      "epoch": 1.95,
-      "learning_rate": 0.00014768420286235908,
-      "loss": 0.3429,
-      "step": 128000
-    },
-    {
-      "epoch": 1.97,
-      "learning_rate": 0.00014762478305778328,
-      "loss": 0.3422,
-      "step": 129000
-    },
-    {
-      "epoch": 1.99,
-      "learning_rate": 0.0001475646237194144,
-      "loss": 0.3414,
-      "step": 130000
-    },
-    {
-      "epoch": 1.99,
-      "eval_runtime": 1.0085,
-      "eval_samples_per_second": 991.553,
-      "eval_steps_per_second": 15.865,
-      "step": 130000
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 0.00014750372550514533,
-      "loss": 0.3409,
-      "step": 131000
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 0.0001474420890809492,
-      "loss": 0.3401,
-      "step": 132000
-    },
-    {
-      "epoch": 2.03,
-      "learning_rate": 0.00014737971512087202,
-      "loss": 0.3396,
-      "step": 133000
-    },
-    {
-      "epoch": 2.05,
-      "learning_rate": 0.00014731660430702552,
-      "loss": 0.339,
-      "step": 134000
-    },
-    {
-      "epoch": 2.06,
-      "learning_rate": 0.00014725275732957937,
-      "loss": 0.3402,
-      "step": 135000
-    },
-    {
-      "epoch": 2.06,
-      "eval_runtime": 1.1005,
-      "eval_samples_per_second": 908.652,
-      "eval_steps_per_second": 14.538,
-      "step": 135000
-    },
-    {
-      "epoch": 2.08,
-      "learning_rate": 0.00014718817488675387,
-      "loss": 0.3379,
-      "step": 136000
-    },
-    {
-      "epoch": 2.09,
-      "learning_rate": 0.00014712285768481235,
-      "loss": 0.3371,
-      "step": 137000
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 0.00014705680643805323,
-      "loss": 0.3368,
-      "step": 138000
-    },
-    {
-      "epoch": 2.12,
-      "learning_rate": 0.00014699002186880232,
-      "loss": 0.3363,
-      "step": 139000
-    },
-    {
-      "epoch": 2.14,
-      "learning_rate": 0.00014692250470740503,
-      "loss": 0.3361,
-      "step": 140000
-    },
-    {
-      "epoch": 2.14,
-      "eval_runtime": 1.0104,
-      "eval_samples_per_second": 989.716,
-      "eval_steps_per_second": 15.835,
-      "step": 140000
-    },
-    {
-      "epoch": 2.15,
-      "learning_rate": 0.00014685425569221819,
-      "loss": 0.3353,
-      "step": 141000
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 0.00014678527556960207,
-      "loss": 0.3346,
-      "step": 142000
-    },
-    {
-      "epoch": 2.18,
-      "learning_rate": 0.0001467155650939123,
-      "loss": 0.3342,
-      "step": 143000
-    },
-    {
-      "epoch": 2.2,
-      "learning_rate": 0.00014664512502749141,
-      "loss": 0.3338,
-      "step": 144000
-    },
-    {
-      "epoch": 2.21,
-      "learning_rate": 0.00014657395614066075,
-      "loss": 0.3334,
-      "step": 145000
-    },
-    {
-      "epoch": 2.21,
-      "eval_runtime": 1.0369,
-      "eval_samples_per_second": 964.439,
-      "eval_steps_per_second": 15.431,
-      "step": 145000
-    },
-    {
-      "epoch": 2.23,
-      "learning_rate": 0.0001465020592117118,
-      "loss": 0.3327,
-      "step": 146000
-    },
-    {
-      "epoch": 2.24,
-      "learning_rate": 0.0001464294350268979,
-      "loss": 0.3324,
-      "step": 147000
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 0.00014635608438042546,
-      "loss": 0.3319,
-      "step": 148000
-    },
-    {
-      "epoch": 2.28,
-      "learning_rate": 0.00014628200807444543,
-      "loss": 0.3313,
-      "step": 149000
-    },
-    {
-      "epoch": 2.29,
-      "learning_rate": 0.0001462072069190444,
-      "loss": 0.3307,
-      "step": 150000
-    },
-    {
-      "epoch": 2.29,
-      "eval_runtime": 1.0431,
-      "eval_samples_per_second": 958.687,
-      "eval_steps_per_second": 15.339,
-      "step": 150000
-    },
-    {
-      "epoch": 2.31,
-      "learning_rate": 0.00014613168173223585,
-      "loss": 0.3308,
-      "step": 151000
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 0.00014605543333995113,
-      "loss": 0.3302,
-      "step": 152000
-    },
-    {
-      "epoch": 2.34,
-      "learning_rate": 0.00014597846257603038,
-      "loss": 0.3294,
-      "step": 153000
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 0.0001459007702822136,
-      "loss": 0.329,
-      "step": 154000
-    },
-    {
-      "epoch": 2.37,
-      "learning_rate": 0.00014582235730813128,
-      "loss": 0.3283,
-      "step": 155000
-    },
-    {
-      "epoch": 2.37,
-      "eval_runtime": 1.0629,
-      "eval_samples_per_second": 940.817,
-      "eval_steps_per_second": 15.053,
-      "step": 155000
-    },
-    {
-      "epoch": 2.38,
-      "learning_rate": 0.00014574322451129507,
-      "loss": 0.3281,
-      "step": 156000
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 0.00014566337275708863,
-      "loss": 0.328,
-      "step": 157000
-    },
-    {
-      "epoch": 2.41,
-      "learning_rate": 0.0001455828029187579,
-      "loss": 0.3272,
-      "step": 158000
-    },
-    {
-      "epoch": 2.43,
-      "learning_rate": 0.00014550151587740178,
-      "loss": 0.3269,
-      "step": 159000
-    },
-    {
-      "epoch": 2.44,
-      "learning_rate": 0.00014541951252196225,
-      "loss": 0.3265,
-      "step": 160000
-    },
-    {
-      "epoch": 2.44,
-      "eval_runtime": 1.0199,
-      "eval_samples_per_second": 980.452,
-      "eval_steps_per_second": 15.687,
-      "step": 160000
-    },
-    {
-      "epoch": 2.46,
-      "learning_rate": 0.00014533679374921493,
-      "loss": 0.3259,
-      "step": 161000
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 0.00014525336046375905,
-      "loss": 0.3254,
-      "step": 162000
-    },
-    {
-      "epoch": 2.49,
-      "learning_rate": 0.00014516921357800766,
-      "loss": 0.3251,
-      "step": 163000
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 0.00014508435401217759,
-      "loss": 0.3244,
-      "step": 164000
-    },
-    {
-      "epoch": 2.52,
-      "learning_rate": 0.00014499878269427948,
-      "loss": 0.3243,
-      "step": 165000
-    },
-    {
-      "epoch": 2.52,
-      "eval_runtime": 1.0655,
-      "eval_samples_per_second": 938.486,
-      "eval_steps_per_second": 15.016,
-      "step": 165000
-    },
-    {
-      "epoch": 2.53,
-      "learning_rate": 0.00014491250056010758,
-      "loss": 0.3236,
-      "step": 166000
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 0.00014482550855322943,
-      "loss": 0.3233,
-      "step": 167000
-    },
-    {
-      "epoch": 2.57,
-      "learning_rate": 0.0001447378076249757,
-      "loss": 0.3231,
-      "step": 168000
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 0.00014464939873442973,
-      "loss": 0.3228,
-      "step": 169000
-    },
-    {
-      "epoch": 2.6,
-      "learning_rate": 0.00014456028284841693,
-      "loss": 0.3221,
-      "step": 170000
-    },
-    {
-      "epoch": 2.6,
-      "eval_runtime": 1.1756,
-      "eval_samples_per_second": 850.656,
-      "eval_steps_per_second": 13.611,
-      "step": 170000
-    },
-    {
-      "epoch": 2.61,
-      "learning_rate": 0.00014447046094149437,
-      "loss": 0.3221,
-      "step": 171000
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 0.00014437993399594003,
-      "loss": 0.3216,
-      "step": 172000
-    },
-    {
-      "epoch": 2.64,
-      "learning_rate": 0.0001442887030017421,
-      "loss": 0.3217,
-      "step": 173000
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 0.00014419676895658807,
-      "loss": 0.3208,
-      "step": 174000
-    },
-    {
-      "epoch": 2.67,
-      "learning_rate": 0.000144104132865854,
-      "loss": 0.3207,
-      "step": 175000
-    },
-    {
-      "epoch": 2.67,
-      "eval_runtime": 1.0679,
-      "eval_samples_per_second": 936.423,
-      "eval_steps_per_second": 14.983,
-      "step": 175000
-    },
-    {
-      "epoch": 2.69,
-      "learning_rate": 0.0001440107957425933,
-      "loss": 0.3203,
-      "step": 176000
-    },
-    {
-      "epoch": 2.7,
-      "learning_rate": 0.0001439167586075258,
-      "loss": 0.3201,
-      "step": 177000
-    },
-    {
-      "epoch": 2.72,
-      "learning_rate": 0.0001438220224890265,
-      "loss": 0.3191,
-      "step": 178000
-    },
-    {
-      "epoch": 2.73,
-      "learning_rate": 0.00014372658842311449,
-      "loss": 0.3195,
-      "step": 179000
-    },
-    {
-      "epoch": 2.75,
-      "learning_rate": 0.00014363045745344137,
-      "loss": 0.3191,
-      "step": 180000
-    },
-    {
-      "epoch": 2.75,
-      "eval_runtime": 1.0169,
-      "eval_samples_per_second": 983.42,
-      "eval_steps_per_second": 15.735,
-      "step": 180000
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 0.00014353363063128005,
-      "loss": 0.3183,
-      "step": 181000
-    },
-    {
-      "epoch": 2.78,
-      "learning_rate": 0.0001434361090155131,
-      "loss": 0.3177,
-      "step": 182000
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 0.00014333789367262136,
-      "loss": 0.3178,
-      "step": 183000
-    },
-    {
-      "epoch": 2.81,
-      "learning_rate": 0.00014323898567667202,
-      "loss": 0.3177,
-      "step": 184000
-    },
-    {
-      "epoch": 2.82,
-      "learning_rate": 0.00014313938610930712,
-      "loss": 0.3171,
-      "step": 185000
-    },
-    {
-      "epoch": 2.82,
-      "eval_runtime": 1.0441,
-      "eval_samples_per_second": 957.721,
-      "eval_steps_per_second": 15.324,
-      "step": 185000
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 0.00014303909605973154,
-      "loss": 0.3167,
-      "step": 186000
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 0.0001429381166247012,
-      "loss": 0.3168,
-      "step": 187000
-    },
-    {
-      "epoch": 2.87,
-      "learning_rate": 0.00014283644890851103,
-      "loss": 0.3164,
-      "step": 188000
-    },
-    {
-      "epoch": 2.89,
-      "learning_rate": 0.00014273409402298291,
-      "loss": 0.3161,
-      "step": 189000
-    },
-    {
-      "epoch": 2.9,
-      "learning_rate": 0.00014263105308745343,
-      "loss": 0.3155,
-      "step": 190000
-    },
-    {
-      "epoch": 2.9,
-      "eval_runtime": 1.0119,
-      "eval_samples_per_second": 988.212,
-      "eval_steps_per_second": 15.811,
-      "step": 190000
-    },
-    {
-      "epoch": 2.92,
-      "learning_rate": 0.00014252732722876176,
-      "loss": 0.3149,
-      "step": 191000
-    },
-    {
-      "epoch": 2.93,
-      "learning_rate": 0.0001424229175812373,
-      "loss": 0.3149,
-      "step": 192000
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 0.00014231782528668717,
-      "loss": 0.3146,
-      "step": 193000
-    },
-    {
-      "epoch": 2.96,
-      "learning_rate": 0.00014221205149438394,
-      "loss": 0.3145,
-      "step": 194000
-    },
-    {
-      "epoch": 2.98,
-      "learning_rate": 0.0001421055973610528,
-      "loss": 0.3138,
-      "step": 195000
-    },
-    {
-      "epoch": 2.98,
-      "eval_runtime": 1.0908,
-      "eval_samples_per_second": 916.734,
-      "eval_steps_per_second": 14.668,
-      "step": 195000
-    },
-    {
-      "epoch": 2.99,
-      "learning_rate": 0.00014199846405085913,
-      "loss": 0.3137,
-      "step": 196000
-    },
-    {
-      "epoch": 3.01,
-      "learning_rate": 0.00014189065273539564,
-      "loss": 0.3135,
-      "step": 197000
-    },
-    {
-      "epoch": 3.02,
-      "learning_rate": 0.00014178216459366958,
-      "loss": 0.3137,
-      "step": 198000
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 0.00014167300081208988,
-      "loss": 0.3131,
-      "step": 199000
-    },
-    {
-      "epoch": 3.05,
-      "learning_rate": 0.00014156316258445421,
-      "loss": 0.3125,
-      "step": 200000
-    },
-    {
-      "epoch": 3.05,
-      "eval_runtime": 1.1346,
-      "eval_samples_per_second": 881.333,
-      "eval_steps_per_second": 14.101,
-      "step": 200000
-    },
-    {
-      "epoch": 3.07,
-      "learning_rate": 0.00014145265111193583,
-      "loss": 0.3121,
-      "step": 201000
-    },
-    {
-      "epoch": 3.08,
-      "learning_rate": 0.00014134146760307043,
-      "loss": 0.3122,
-      "step": 202000
-    },
-    {
-      "epoch": 3.1,
-      "learning_rate": 0.00014122961327374313,
-      "loss": 0.3131,
-      "step": 203000
-    },
-    {
-      "epoch": 3.12,
-      "learning_rate": 0.0001411170893471749,
-      "loss": 0.3116,
-      "step": 204000
-    },
-    {
-      "epoch": 3.13,
-      "learning_rate": 0.00014100389705390938,
-      "loss": 0.311,
-      "step": 205000
-    },
-    {
-      "epoch": 3.13,
-      "eval_runtime": 1.1239,
-      "eval_samples_per_second": 889.731,
-      "eval_steps_per_second": 14.236,
-      "step": 205000
-    },
-    {
-      "epoch": 3.15,
-      "learning_rate": 0.0001408900376317994,
-      "loss": 0.311,
-      "step": 206000
-    },
-    {
-      "epoch": 3.16,
-      "learning_rate": 0.0001407755123259933,
-      "loss": 0.3108,
-      "step": 207000
-    },
-    {
-      "epoch": 3.18,
-      "learning_rate": 0.00014066032238892152,
-      "loss": 0.3104,
-      "step": 208000
-    },
-    {
-      "epoch": 3.19,
-      "learning_rate": 0.00014054446908028272,
-      "loss": 0.3102,
-      "step": 209000
-    },
-    {
-      "epoch": 3.21,
-      "learning_rate": 0.00014042795366703018,
-      "loss": 0.3097,
-      "step": 210000
-    },
-    {
-      "epoch": 3.21,
-      "eval_runtime": 1.0233,
-      "eval_samples_per_second": 977.233,
-      "eval_steps_per_second": 15.636,
-      "step": 210000
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 0.0001403107774233577,
-      "loss": 0.3098,
-      "step": 211000
-    },
-    {
-      "epoch": 3.24,
-      "learning_rate": 0.00014019294163068597,
-      "loss": 0.3093,
-      "step": 212000
-    },
-    {
-      "epoch": 3.25,
-      "learning_rate": 0.00014007444757764835,
-      "loss": 0.3093,
-      "step": 213000
-    },
-    {
-      "epoch": 3.27,
-      "learning_rate": 0.0001399552965600768,
-      "loss": 0.3088,
-      "step": 214000
-    },
-    {
-      "epoch": 3.28,
-      "learning_rate": 0.0001398354898809877,
-      "loss": 0.3089,
-      "step": 215000
-    },
-    {
-      "epoch": 3.28,
-      "eval_runtime": 1.0098,
-      "eval_samples_per_second": 990.287,
-      "eval_steps_per_second": 15.845,
-      "step": 215000
-    },
-    {
-      "epoch": 3.3,
-      "learning_rate": 0.0001397150288505678,
-      "loss": 0.3315,
-      "step": 216000
-    },
-    {
-      "epoch": 3.31,
-      "learning_rate": 0.00013959391478615959,
-      "loss": 0.628,
-      "step": 217000
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 0.00013947214901224706,
-      "loss": 0.3112,
-      "step": 218000
-    },
-    {
-      "epoch": 3.34,
-      "learning_rate": 0.0001393497328604412,
-      "loss": 0.3094,
-      "step": 219000
-    },
-    {
-      "epoch": 3.36,
-      "learning_rate": 0.00013922666766946545,
-      "loss": 0.3082,
-      "step": 220000
-    },
-    {
-      "epoch": 3.36,
-      "eval_runtime": 1.0751,
-      "eval_samples_per_second": 930.139,
-      "eval_steps_per_second": 14.882,
-      "step": 220000
-    },
-    {
-      "epoch": 3.37,
-      "learning_rate": 0.00013910295478514106,
-      "loss": 0.3079,
-      "step": 221000
-    },
-    {
-      "epoch": 3.39,
-      "learning_rate": 0.0001389785955603722,
-      "loss": 0.3077,
-      "step": 222000
-    },
-    {
-      "epoch": 3.41,
-      "learning_rate": 0.00013885359135513154,
-      "loss": 0.3073,
-      "step": 223000
-    },
-    {
-      "epoch": 3.42,
-      "learning_rate": 0.000138727943536445,
-      "loss": 0.3064,
-      "step": 224000
-    },
-    {
-      "epoch": 3.44,
-      "learning_rate": 0.00013860165347837698,
-      "loss": 0.3066,
-      "step": 225000
-    },
-    {
-      "epoch": 3.44,
-      "eval_runtime": 1.0901,
-      "eval_samples_per_second": 917.309,
-      "eval_steps_per_second": 14.677,
-      "step": 225000
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 0.00013847472256201535,
-      "loss": 0.306,
-      "step": 226000
-    },
-    {
-      "epoch": 3.47,
-      "learning_rate": 0.00013834715217545625,
-      "loss": 0.3058,
-      "step": 227000
-    },
-    {
-      "epoch": 3.48,
-      "learning_rate": 0.000138218943713789,
-      "loss": 0.3056,
-      "step": 228000
-    },
-    {
-      "epoch": 3.5,
-      "learning_rate": 0.0001380900985790808,
-      "loss": 0.3054,
-      "step": 229000
-    },
-    {
-      "epoch": 3.51,
-      "learning_rate": 0.00013796061818036138,
-      "loss": 0.3051,
-      "step": 230000
-    },
-    {
-      "epoch": 3.51,
-      "eval_runtime": 1.0217,
-      "eval_samples_per_second": 978.715,
-      "eval_steps_per_second": 15.659,
-      "step": 230000
-    },
-    {
-      "epoch": 3.53,
-      "learning_rate": 0.00013783050393360768,
-      "loss": 0.3048,
-      "step": 231000
-    },
-    {
-      "epoch": 3.54,
-      "learning_rate": 0.0001376997572617282,
-      "loss": 0.305,
-      "step": 232000
-    },
-    {
-      "epoch": 3.56,
-      "learning_rate": 0.00013756837959454766,
-      "loss": 0.3042,
-      "step": 233000
-    },
-    {
-      "epoch": 3.57,
-      "learning_rate": 0.0001374363723687911,
-      "loss": 0.3042,
-      "step": 234000
-    },
-    {
-      "epoch": 3.59,
-      "learning_rate": 0.00013730373702806846,
-      "loss": 0.304,
-      "step": 235000
-    },
-    {
-      "epoch": 3.59,
-      "eval_runtime": 1.0392,
-      "eval_samples_per_second": 962.319,
-      "eval_steps_per_second": 15.397,
-      "step": 235000
-    },
-    {
-      "epoch": 3.6,
-      "learning_rate": 0.00013717047502285855,
-      "loss": 0.3036,
-      "step": 236000
-    },
-    {
-      "epoch": 3.62,
-      "learning_rate": 0.0001370365878104933,
-      "loss": 0.3036,
-      "step": 237000
-    },
-    {
-      "epoch": 3.63,
-      "learning_rate": 0.00013690207685514185,
-      "loss": 0.3031,
-      "step": 238000
-    },
-    {
-      "epoch": 3.65,
-      "learning_rate": 0.0001367669436277944,
-      "loss": 0.3032,
-      "step": 239000
-    },
-    {
-      "epoch": 3.66,
-      "learning_rate": 0.0001366311896062463,
-      "loss": 0.3036,
-      "step": 240000
-    },
-    {
-      "epoch": 3.66,
-      "eval_runtime": 1.0097,
-      "eval_samples_per_second": 990.396,
-      "eval_steps_per_second": 15.846,
-      "step": 240000
-    },
-    {
-      "epoch": 3.68,
-      "learning_rate": 0.00013649481627508181,
-      "loss": 0.3031,
-      "step": 241000
-    },
-    {
-      "epoch": 3.7,
-      "learning_rate": 0.0001363578251256578,
-      "loss": 0.3023,
-      "step": 242000
-    },
-    {
-      "epoch": 3.71,
-      "learning_rate": 0.00013622021765608754,
-      "loss": 0.3022,
-      "step": 243000
-    },
-    {
-      "epoch": 3.73,
-      "learning_rate": 0.00013608199537122425,
-      "loss": 0.3017,
-      "step": 244000
-    },
-    {
-      "epoch": 3.74,
-      "learning_rate": 0.0001359431597826447,
-      "loss": 0.3019,
-      "step": 245000
-    },
-    {
-      "epoch": 3.74,
-      "eval_runtime": 1.0744,
-      "eval_samples_per_second": 930.717,
-      "eval_steps_per_second": 14.891,
-      "step": 245000
-    },
-    {
-      "epoch": 3.76,
-      "learning_rate": 0.0001358037124086327,
-      "loss": 0.3015,
-      "step": 246000
-    },
-    {
-      "epoch": 3.77,
-      "learning_rate": 0.00013566365477416233,
-      "loss": 0.3018,
-      "step": 247000
-    },
-    {
-      "epoch": 3.79,
-      "learning_rate": 0.00013552298841088144,
-      "loss": 0.3013,
-      "step": 248000
-    },
-    {
-      "epoch": 3.8,
-      "learning_rate": 0.00013538171485709486,
-      "loss": 0.3006,
-      "step": 249000
-    },
-    {
-      "epoch": 3.82,
-      "learning_rate": 0.00013523983565774753,
-      "loss": 0.3008,
-      "step": 250000
-    },
-    {
-      "epoch": 3.82,
-      "eval_runtime": 1.0168,
-      "eval_samples_per_second": 983.434,
-      "eval_steps_per_second": 15.735,
-      "step": 250000
-    },
-    {
-      "epoch": 3.83,
-      "learning_rate": 0.00013509735236440766,
-      "loss": 0.3003,
-      "step": 251000
-    },
-    {
-      "epoch": 3.85,
-      "learning_rate": 0.00013495426653524972,
-      "loss": 0.3,
-      "step": 252000
-    },
-    {
-      "epoch": 3.86,
-      "learning_rate": 0.00013481057973503742,
-      "loss": 0.3,
-      "step": 253000
-    },
-    {
-      "epoch": 3.88,
-      "learning_rate": 0.00013466629353510651,
-      "loss": 0.2997,
-      "step": 254000
-    },
-    {
-      "epoch": 3.89,
-      "learning_rate": 0.00013452140951334787,
-      "loss": 0.2995,
-      "step": 255000
-    },
-    {
-      "epoch": 3.89,
-      "eval_runtime": 0.8192,
-      "eval_samples_per_second": 1220.744,
-      "eval_steps_per_second": 19.532,
-      "step": 255000
-    },
-    {
-      "epoch": 3.91,
-      "learning_rate": 0.00013437592925418985,
-      "loss": 0.2996,
-      "step": 256000
-    },
-    {
-      "epoch": 3.92,
-      "learning_rate": 0.00013422985434858133,
-      "loss": 0.299,
-      "step": 257000
-    },
-    {
-      "epoch": 3.94,
-      "learning_rate": 0.00013408318639397405,
-      "loss": 0.2987,
-      "step": 258000
-    },
-    {
-      "epoch": 3.95,
-      "learning_rate": 0.00013393592699430525,
-      "loss": 0.2986,
-      "step": 259000
-    },
-    {
-      "epoch": 3.97,
-      "learning_rate": 0.00013378807775998012,
-      "loss": 0.2984,
-      "step": 260000
-    },
-    {
-      "epoch": 3.97,
-      "eval_runtime": 1.0461,
-      "eval_samples_per_second": 955.963,
-      "eval_steps_per_second": 15.295,
-      "step": 260000
-    },
-    {
-      "epoch": 3.99,
-      "learning_rate": 0.00013363964030785422,
-      "loss": 0.2983,
-      "step": 261000
-    },
-    {
-      "epoch": 4.0,
-      "learning_rate": 0.00013349061626121578,
-      "loss": 0.2982,
-      "step": 262000
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 0.00013334100724976783,
-      "loss": 0.2977,
-      "step": 263000
-    },
-    {
-      "epoch": 4.03,
-      "learning_rate": 0.0001331908149096106,
-      "loss": 0.2976,
-      "step": 264000
-    },
-    {
-      "epoch": 4.05,
-      "learning_rate": 0.00013304004088322342,
-      "loss": 0.2978,
-      "step": 265000
-    },
-    {
-      "epoch": 4.05,
-      "eval_runtime": 1.0225,
-      "eval_samples_per_second": 978.001,
-      "eval_steps_per_second": 15.648,
-      "step": 265000
-    },
-    {
-      "epoch": 4.06,
-      "learning_rate": 0.00013288868681944692,
-      "loss": 0.2971,
-      "step": 266000
-    },
-    {
-      "epoch": 4.08,
-      "learning_rate": 0.00013273675437346487,
-      "loss": 0.2972,
-      "step": 267000
-    },
-    {
-      "epoch": 4.09,
-      "learning_rate": 0.00013258424520678618,
-      "loss": 0.2969,
-      "step": 268000
-    },
-    {
-      "epoch": 4.11,
-      "learning_rate": 0.00013243116098722663,
-      "loss": 0.2968,
-      "step": 269000
-    },
-    {
-      "epoch": 4.12,
-      "learning_rate": 0.00013227750338889077,
-      "loss": 0.2966,
-      "step": 270000
-    },
-    {
-      "epoch": 4.12,
-      "eval_runtime": 1.1084,
-      "eval_samples_per_second": 902.192,
-      "eval_steps_per_second": 14.435,
-      "step": 270000
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 0.00013212327409215343,
-      "loss": 0.296,
-      "step": 271000
-    },
-    {
-      "epoch": 4.15,
-      "learning_rate": 0.0001319684747836415,
-      "loss": 0.2958,
-      "step": 272000
-    },
-    {
-      "epoch": 4.17,
-      "learning_rate": 0.0001318131071562154,
-      "loss": 0.2961,
-      "step": 273000
-    },
-    {
-      "epoch": 4.18,
-      "learning_rate": 0.00013165717290895067,
-      "loss": 0.2957,
-      "step": 274000
-    },
-    {
-      "epoch": 4.2,
-      "learning_rate": 0.0001315006737471192,
-      "loss": 0.2955,
-      "step": 275000
-    },
-    {
-      "epoch": 4.2,
-      "eval_runtime": 1.0552,
-      "eval_samples_per_second": 947.654,
-      "eval_steps_per_second": 15.162,
-      "step": 275000
-    },
-    {
-      "epoch": 4.21,
-      "learning_rate": 0.0001313436113821708,
-      "loss": 0.2952,
-      "step": 276000
-    },
-    {
-      "epoch": 4.23,
-      "learning_rate": 0.00013118598753171425,
-      "loss": 0.2951,
-      "step": 277000
-    },
-    {
-      "epoch": 4.25,
-      "learning_rate": 0.0001310278039194988,
-      "loss": 0.2951,
-      "step": 278000
-    },
-    {
-      "epoch": 4.26,
-      "learning_rate": 0.00013086906227539506,
-      "loss": 0.2952,
-      "step": 279000
-    },
-    {
-      "epoch": 4.28,
-      "learning_rate": 0.00013070976433537623,
-      "loss": 0.2946,
-      "step": 280000
-    },
-    {
-      "epoch": 4.28,
-      "eval_runtime": 1.0293,
-      "eval_samples_per_second": 971.532,
-      "eval_steps_per_second": 15.545,
-      "step": 280000
-    },
-    {
-      "epoch": 4.29,
-      "learning_rate": 0.00013054991184149905,
-      "loss": 0.2946,
-      "step": 281000
-    },
-    {
-      "epoch": 4.31,
-      "learning_rate": 0.00013038950654188476,
-      "loss": 0.2942,
-      "step": 282000
-    },
-    {
-      "epoch": 4.32,
-      "learning_rate": 0.00013022855019070005,
-      "loss": 0.2941,
-      "step": 283000
-    },
-    {
-      "epoch": 4.34,
-      "learning_rate": 0.0001300670445481378,
-      "loss": 0.2937,
-      "step": 284000
-    },
-    {
-      "epoch": 4.35,
-      "learning_rate": 0.0001299049913803978,
-      "loss": 0.2937,
-      "step": 285000
-    },
-    {
-      "epoch": 4.35,
-      "eval_runtime": 1.0469,
-      "eval_samples_per_second": 955.197,
-      "eval_steps_per_second": 15.283,
-      "step": 285000
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 0.00012974239245966754,
-      "loss": 0.2934,
-      "step": 286000
-    },
-    {
-      "epoch": 4.38,
-      "learning_rate": 0.0001295792495641028,
-      "loss": 0.2962,
-      "step": 287000
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 0.00012941556447780813,
-      "loss": 0.2931,
-      "step": 288000
-    },
-    {
-      "epoch": 4.41,
-      "learning_rate": 0.0001292513389908174,
-      "loss": 0.2931,
-      "step": 289000
-    },
-    {
-      "epoch": 4.43,
-      "learning_rate": 0.0001290865748990742,
-      "loss": 0.2932,
-      "step": 290000
-    },
-    {
-      "epoch": 4.43,
-      "eval_runtime": 1.0143,
-      "eval_samples_per_second": 985.898,
-      "eval_steps_per_second": 15.774,
-      "step": 290000
-    },
-    {
-      "epoch": 4.44,
-      "learning_rate": 0.00012892127400441228,
-      "loss": 0.2923,
-      "step": 291000
-    },
-    {
-      "epoch": 4.46,
-      "learning_rate": 0.00012875543811453576,
-      "loss": 0.2919,
-      "step": 292000
-    },
-    {
-      "epoch": 4.47,
-      "learning_rate": 0.0001285890690429993,
-      "loss": 0.2931,
-      "step": 293000
-    },
-    {
-      "epoch": 4.49,
-      "learning_rate": 0.00012842216860918846,
-      "loss": 0.292,
-      "step": 294000
-    },
-    {
-      "epoch": 4.5,
-      "learning_rate": 0.0001282547386382996,
-      "loss": 0.2914,
-      "step": 295000
-    },
-    {
-      "epoch": 4.5,
-      "eval_runtime": 1.0329,
-      "eval_samples_per_second": 968.135,
-      "eval_steps_per_second": 15.49,
-      "step": 295000
-    },
-    {
-      "epoch": 4.52,
-      "learning_rate": 0.0001280867809613201,
-      "loss": 0.2919,
-      "step": 296000
-    },
-    {
-      "epoch": 4.54,
-      "learning_rate": 0.0001279182974150082,
-      "loss": 0.2915,
-      "step": 297000
-    },
-    {
-      "epoch": 4.55,
-      "learning_rate": 0.00012774928984187297,
-      "loss": 0.2914,
-      "step": 298000
-    },
-    {
-      "epoch": 4.57,
-      "learning_rate": 0.00012757976009015413,
-      "loss": 0.2908,
-      "step": 299000
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 0.0001274097100138019,
-      "loss": 0.2909,
-      "step": 300000
-    },
-    {
-      "epoch": 4.58,
-      "eval_runtime": 1.0054,
-      "eval_samples_per_second": 994.612,
-      "eval_steps_per_second": 15.914,
-      "step": 300000
-    },
-    {
-      "epoch": 4.6,
-      "learning_rate": 0.00012723914147245663,
-      "loss": 0.2906,
-      "step": 301000
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 0.00012706805633142863,
-      "loss": 0.2906,
-      "step": 302000
-    },
-    {
-      "epoch": 4.63,
-      "learning_rate": 0.00012689645646167755,
-      "loss": 0.2902,
-      "step": 303000
-    },
-    {
-      "epoch": 4.64,
-      "learning_rate": 0.00012672434373979207,
-      "loss": 0.291,
-      "step": 304000
-    },
-    {
-      "epoch": 4.66,
-      "learning_rate": 0.00012655172004796936,
-      "loss": 0.2899,
-      "step": 305000
-    },
-    {
-      "epoch": 4.66,
-      "eval_runtime": 1.0975,
-      "eval_samples_per_second": 911.158,
-      "eval_steps_per_second": 14.579,
-      "step": 305000
-    },
-    {
-      "epoch": 4.67,
-      "learning_rate": 0.00012637858727399448,
-      "loss": 0.2898,
-      "step": 306000
-    },
-    {
-      "epoch": 4.69,
-      "learning_rate": 0.00012620494731121966,
-      "loss": 0.2896,
-      "step": 307000
-    },
-    {
-      "epoch": 4.7,
-      "learning_rate": 0.00012603080205854372,
-      "loss": 0.2894,
-      "step": 308000
-    },
-    {
-      "epoch": 4.72,
-      "learning_rate": 0.00012585615342039126,
-      "loss": 0.2894,
-      "step": 309000
-    },
-    {
-      "epoch": 4.73,
-      "learning_rate": 0.0001256810033066918,
-      "loss": 0.2894,
-      "step": 310000
-    },
-    {
-      "epoch": 4.73,
-      "eval_runtime": 1.0481,
-      "eval_samples_per_second": 954.11,
-      "eval_steps_per_second": 15.266,
-      "step": 310000
-    },
-    {
-      "epoch": 4.75,
-      "learning_rate": 0.0001255053536328589,
-      "loss": 0.2887,
-      "step": 311000
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 0.0001253292063197693,
-      "loss": 0.2887,
-      "step": 312000
-    },
-    {
-      "epoch": 4.78,
-      "learning_rate": 0.0001251525632937418,
-      "loss": 0.2886,
-      "step": 313000
-    },
-    {
-      "epoch": 4.79,
-      "learning_rate": 0.00012497542648651615,
-      "loss": 0.2887,
-      "step": 314000
-    },
-    {
-      "epoch": 4.81,
-      "learning_rate": 0.00012479779783523216,
-      "loss": 0.2883,
-      "step": 315000
-    },
-    {
-      "epoch": 4.81,
-      "eval_runtime": 1.0333,
-      "eval_samples_per_second": 967.804,
-      "eval_steps_per_second": 15.485,
-      "step": 315000
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 0.00012461967928240828,
-      "loss": 0.2883,
-      "step": 316000
-    },
-    {
-      "epoch": 4.84,
-      "learning_rate": 0.00012444107277592047,
-      "loss": 0.2877,
-      "step": 317000
-    },
-    {
-      "epoch": 4.86,
-      "learning_rate": 0.0001242619802689809,
-      "loss": 0.2879,
-      "step": 318000
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 0.00012408240372011647,
-      "loss": 0.2876,
-      "step": 319000
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 0.0001239023450931476,
-      "loss": 0.2874,
-      "step": 320000
-    },
-    {
-      "epoch": 4.89,
-      "eval_runtime": 1.04,
-      "eval_samples_per_second": 961.537,
-      "eval_steps_per_second": 15.385,
-      "step": 320000
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 0.00012372180635716656,
-      "loss": 0.2874,
-      "step": 321000
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 0.00012354078948651604,
-      "loss": 0.2873,
-      "step": 322000
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 0.00012335929646076758,
-      "loss": 0.2868,
-      "step": 323000
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 0.00012317732926469976,
-      "loss": 0.2871,
-      "step": 324000
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 0.00012299488988827675,
-      "loss": 0.2869,
-      "step": 325000
-    },
-    {
-      "epoch": 4.96,
-      "eval_runtime": 1.3977,
-      "eval_samples_per_second": 715.452,
-      "eval_steps_per_second": 11.447,
-      "step": 325000
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 0.0001228119803266263,
-      "loss": 0.2867,
-      "step": 326000
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 0.0001226286025800181,
-      "loss": 0.2866,
-      "step": 327000
-    },
-    {
-      "epoch": 5.01,
-      "learning_rate": 0.00012244475865384177,
-      "loss": 0.2862,
-      "step": 328000
-    },
-    {
-      "epoch": 5.02,
-      "learning_rate": 0.00012226045055858505,
-      "loss": 0.2858,
-      "step": 329000
-    },
-    {
-      "epoch": 5.04,
-      "learning_rate": 0.00012207568030981174,
-      "loss": 0.2859,
-      "step": 330000
-    },
-    {
-      "epoch": 5.04,
-      "eval_runtime": 1.1314,
-      "eval_samples_per_second": 883.862,
-      "eval_steps_per_second": 14.142,
-      "step": 330000
-    },
-    {
-      "epoch": 5.05,
-      "learning_rate": 0.00012189044992813972,
-      "loss": 0.2858,
-      "step": 331000
-    },
-    {
-      "epoch": 5.07,
-      "learning_rate": 0.0001217047614392187,
-      "loss": 0.2857,
-      "step": 332000
-    },
-    {
-      "epoch": 5.08,
-      "learning_rate": 0.00012151861687370828,
-      "loss": 0.2857,
-      "step": 333000
-    },
-    {
-      "epoch": 5.1,
-      "learning_rate": 0.00012133201826725558,
-      "loss": 0.2852,
-      "step": 334000
-    },
-    {
-      "epoch": 5.12,
-      "learning_rate": 0.0001211449676604731,
-      "loss": 0.2853,
-      "step": 335000
-    },
-    {
-      "epoch": 5.12,
-      "eval_runtime": 1.3419,
-      "eval_samples_per_second": 745.216,
-      "eval_steps_per_second": 11.923,
-      "step": 335000
-    },
-    {
-      "epoch": 5.13,
-      "learning_rate": 0.00012095746709891632,
-      "loss": 0.2852,
-      "step": 336000
-    },
-    {
-      "epoch": 5.15,
-      "learning_rate": 0.00012076951863306127,
-      "loss": 0.285,
-      "step": 337000
-    },
-    {
-      "epoch": 5.16,
-      "learning_rate": 0.0001205811243182823,
-      "loss": 0.2848,
-      "step": 338000
-    },
-    {
-      "epoch": 5.18,
-      "learning_rate": 0.00012039228621482949,
-      "loss": 0.2858,
-      "step": 339000
-    },
-    {
-      "epoch": 5.19,
-      "learning_rate": 0.00012020300638780604,
-      "loss": 0.2845,
-      "step": 340000
-    },
-    {
-      "epoch": 5.19,
-      "eval_runtime": 1.2559,
-      "eval_samples_per_second": 796.26,
-      "eval_steps_per_second": 12.74,
-      "step": 340000
-    },
-    {
-      "epoch": 5.21,
-      "learning_rate": 0.00012001328690714582,
-      "loss": 0.284,
-      "step": 341000
-    },
-    {
-      "epoch": 5.22,
-      "learning_rate": 0.00011982312984759068,
-      "loss": 0.2845,
-      "step": 342000
-    },
-    {
-      "epoch": 5.24,
-      "learning_rate": 0.00011963253728866778,
-      "loss": 0.2841,
-      "step": 343000
-    },
-    {
-      "epoch": 5.25,
-      "learning_rate": 0.00011944151131466675,
-      "loss": 0.284,
-      "step": 344000
-    },
-    {
-      "epoch": 5.27,
-      "learning_rate": 0.00011925005401461709,
-      "loss": 0.2836,
-      "step": 345000
-    },
-    {
-      "epoch": 5.27,
-      "eval_runtime": 1.1037,
-      "eval_samples_per_second": 906.031,
-      "eval_steps_per_second": 14.496,
-      "step": 345000
-    },
-    {
-      "epoch": 5.28,
-      "learning_rate": 0.00011905816748226513,
-      "loss": 0.2834,
-      "step": 346000
-    },
-    {
-      "epoch": 5.3,
-      "learning_rate": 0.00011886585381605125,
-      "loss": 0.2835,
-      "step": 347000
-    },
-    {
-      "epoch": 5.31,
-      "learning_rate": 0.00011867311511908693,
-      "loss": 0.2832,
-      "step": 348000
-    },
-    {
-      "epoch": 5.33,
-      "learning_rate": 0.00011847995349913162,
-      "loss": 0.2828,
-      "step": 349000
-    },
-    {
-      "epoch": 5.34,
-      "learning_rate": 0.00011828637106856989,
-      "loss": 0.2828,
-      "step": 350000
-    },
-    {
-      "epoch": 5.34,
-      "eval_runtime": 1.0295,
-      "eval_samples_per_second": 971.32,
-      "eval_steps_per_second": 15.541,
-      "step": 350000
-    },
-    {
-      "epoch": 5.36,
-      "learning_rate": 0.00011809236994438816,
-      "loss": 0.2831,
-      "step": 351000
-    },
-    {
-      "epoch": 5.38,
-      "learning_rate": 0.00011789795224815164,
-      "loss": 0.2827,
-      "step": 352000
-    },
-    {
-      "epoch": 5.39,
-      "learning_rate": 0.00011770312010598116,
-      "loss": 0.282,
-      "step": 353000
-    },
-    {
-      "epoch": 5.41,
-      "learning_rate": 0.00011750787564852973,
-      "loss": 0.2822,
-      "step": 354000
-    },
-    {
-      "epoch": 5.42,
-      "learning_rate": 0.00011731222101095955,
-      "loss": 0.2825,
-      "step": 355000
-    },
-    {
-      "epoch": 5.42,
-      "eval_runtime": 1.0697,
-      "eval_samples_per_second": 934.885,
-      "eval_steps_per_second": 14.958,
-      "step": 355000
-    },
-    {
-      "epoch": 5.44,
-      "learning_rate": 0.00011711615833291833,
-      "loss": 0.2822,
-      "step": 356000
-    },
-    {
-      "epoch": 5.45,
-      "learning_rate": 0.0001169196897585161,
-      "loss": 0.2824,
-      "step": 357000
-    },
-    {
-      "epoch": 5.47,
-      "learning_rate": 0.00011672281743630175,
-      "loss": 0.2818,
-      "step": 358000
-    },
-    {
-      "epoch": 5.48,
-      "learning_rate": 0.0001165255435192394,
-      "loss": 0.2815,
-      "step": 359000
-    },
-    {
-      "epoch": 5.5,
-      "learning_rate": 0.00011632787016468506,
-      "loss": 0.2819,
-      "step": 360000
-    },
-    {
-      "epoch": 5.5,
-      "eval_runtime": 1.1008,
-      "eval_samples_per_second": 908.433,
-      "eval_steps_per_second": 14.535,
-      "step": 360000
-    },
-    {
-      "epoch": 5.51,
-      "learning_rate": 0.0001161297995343628,
-      "loss": 0.2815,
-      "step": 361000
-    },
-    {
-      "epoch": 5.53,
-      "learning_rate": 0.00011593133379434138,
-      "loss": 0.2815,
-      "step": 362000
-    },
-    {
-      "epoch": 5.54,
-      "learning_rate": 0.00011573247511501028,
-      "loss": 0.2811,
-      "step": 363000
-    },
-    {
-      "epoch": 5.56,
-      "learning_rate": 0.00011553322567105619,
-      "loss": 0.2807,
-      "step": 364000
-    },
-    {
-      "epoch": 5.57,
-      "learning_rate": 0.00011533358764143905,
-      "loss": 0.2808,
-      "step": 365000
-    },
-    {
-      "epoch": 5.57,
-      "eval_runtime": 1.1301,
-      "eval_samples_per_second": 884.842,
-      "eval_steps_per_second": 14.157,
-      "step": 365000
-    },
-    {
-      "epoch": 5.59,
-      "learning_rate": 0.00011513356320936841,
-      "loss": 0.2808,
-      "step": 366000
-    },
-    {
-      "epoch": 5.6,
-      "learning_rate": 0.00011493315456227943,
-      "loss": 0.2817,
-      "step": 367000
-    },
-    {
-      "epoch": 5.62,
-      "learning_rate": 0.00011473236389180894,
-      "loss": 0.2803,
-      "step": 368000
-    },
-    {
-      "epoch": 5.63,
-      "learning_rate": 0.00011453119339377154,
-      "loss": 0.2803,
-      "step": 369000
-    },
-    {
-      "epoch": 5.65,
-      "learning_rate": 0.00011432964526813558,
-      "loss": 0.2817,
-      "step": 370000
-    },
-    {
-      "epoch": 5.65,
-      "eval_runtime": 1.2187,
-      "eval_samples_per_second": 820.56,
-      "eval_steps_per_second": 13.129,
-      "step": 370000
-    },
-    {
-      "epoch": 5.67,
-      "learning_rate": 0.00011412772171899904,
-      "loss": 0.2819,
-      "step": 371000
-    },
-    {
-      "epoch": 5.68,
-      "learning_rate": 0.00011392542495456556,
-      "loss": 0.28,
-      "step": 372000
-    },
-    {
-      "epoch": 5.7,
-      "learning_rate": 0.00011372275718712006,
-      "loss": 0.2797,
-      "step": 373000
-    },
-    {
-      "epoch": 5.71,
-      "learning_rate": 0.00011351972063300484,
-      "loss": 0.2797,
-      "step": 374000
-    },
-    {
-      "epoch": 5.73,
-      "learning_rate": 0.00011331631751259515,
-      "loss": 0.2801,
-      "step": 375000
-    },
-    {
-      "epoch": 5.73,
-      "eval_runtime": 1.0146,
-      "eval_samples_per_second": 985.631,
-      "eval_steps_per_second": 15.77,
-      "step": 375000
-    },
-    {
-      "epoch": 5.74,
-      "learning_rate": 0.00011311255005027487,
-      "loss": 0.2789,
-      "step": 376000
-    },
-    {
-      "epoch": 5.76,
-      "learning_rate": 0.00011290842047441232,
-      "loss": 0.2791,
-      "step": 377000
-    },
-    {
-      "epoch": 5.77,
-      "learning_rate": 0.00011270393101733585,
-      "loss": 0.279,
-      "step": 378000
-    },
-    {
-      "epoch": 5.79,
-      "learning_rate": 0.00011249908391530946,
-      "loss": 0.279,
-      "step": 379000
-    },
-    {
-      "epoch": 5.8,
-      "learning_rate": 0.00011229388140850814,
-      "loss": 0.279,
-      "step": 380000
-    },
-    {
-      "epoch": 5.8,
-      "eval_runtime": 1.2375,
-      "eval_samples_per_second": 808.112,
-      "eval_steps_per_second": 12.93,
-      "step": 380000
-    },
-    {
-      "epoch": 5.82,
-      "learning_rate": 0.00011208832574099368,
-      "loss": 0.2788,
-      "step": 381000
-    },
-    {
-      "epoch": 5.83,
-      "learning_rate": 0.00011188241916068993,
-      "loss": 0.2785,
-      "step": 382000
-    },
-    {
-      "epoch": 5.85,
-      "learning_rate": 0.00011167616391935826,
-      "loss": 0.2783,
-      "step": 383000
-    },
-    {
-      "epoch": 5.86,
-      "learning_rate": 0.00011146956227257293,
-      "loss": 0.2785,
-      "step": 384000
-    },
-    {
-      "epoch": 5.88,
-      "learning_rate": 0.00011126261647969645,
-      "loss": 0.2781,
-      "step": 385000
-    },
-    {
-      "epoch": 5.88,
-      "eval_runtime": 1.0191,
-      "eval_samples_per_second": 981.273,
-      "eval_steps_per_second": 15.7,
-      "step": 385000
-    },
-    {
-      "epoch": 5.89,
-      "learning_rate": 0.00011105532880385487,
-      "loss": 0.2782,
-      "step": 386000
-    },
-    {
-      "epoch": 5.91,
-      "learning_rate": 0.00011084770151191299,
-      "loss": 0.2782,
-      "step": 387000
-    },
-    {
-      "epoch": 5.92,
-      "learning_rate": 0.00011063973687444962,
-      "loss": 0.2779,
-      "step": 388000
-    },
-    {
-      "epoch": 5.94,
-      "learning_rate": 0.00011043143716573272,
-      "loss": 0.2774,
-      "step": 389000
-    },
-    {
-      "epoch": 5.96,
-      "learning_rate": 0.00011022280466369448,
-      "loss": 0.2776,
-      "step": 390000
-    },
-    {
-      "epoch": 5.96,
-      "eval_runtime": 1.0236,
-      "eval_samples_per_second": 976.954,
-      "eval_steps_per_second": 15.631,
-      "step": 390000
-    },
-    {
-      "epoch": 5.97,
-      "learning_rate": 0.00011001384164990662,
-      "loss": 0.2775,
-      "step": 391000
-    },
-    {
-      "epoch": 5.99,
-      "learning_rate": 0.00010980455040955506,
-      "loss": 0.2769,
-      "step": 392000
-    },
-    {
-      "epoch": 6.0,
-      "learning_rate": 0.00010959493323141538,
-      "loss": 0.2773,
-      "step": 393000
-    },
-    {
-      "epoch": 6.02,
-      "learning_rate": 0.00010938499240782739,
-      "loss": 0.277,
-      "step": 394000
-    },
-    {
-      "epoch": 6.03,
-      "learning_rate": 0.00010917473023467032,
-      "loss": 0.277,
-      "step": 395000
-    },
-    {
-      "epoch": 6.03,
-      "eval_runtime": 1.0769,
-      "eval_samples_per_second": 928.59,
-      "eval_steps_per_second": 14.857,
-      "step": 395000
-    },
-    {
-      "epoch": 6.05,
-      "learning_rate": 0.00010896414901133761,
-      "loss": 0.2766,
-      "step": 396000
-    },
-    {
-      "epoch": 6.06,
-      "learning_rate": 0.00010875325104071177,
-      "loss": 0.2768,
-      "step": 397000
-    },
-    {
-      "epoch": 6.08,
-      "learning_rate": 0.00010854203862913927,
-      "loss": 0.2765,
-      "step": 398000
-    },
-    {
-      "epoch": 6.09,
-      "learning_rate": 0.00010833051408640509,
-      "loss": 0.2763,
-      "step": 399000
-    },
-    {
-      "epoch": 6.11,
-      "learning_rate": 0.00010811867972570786,
-      "loss": 0.2767,
-      "step": 400000
-    },
-    {
-      "epoch": 6.11,
-      "eval_runtime": 1.1081,
-      "eval_samples_per_second": 902.417,
-      "eval_steps_per_second": 14.439,
-      "step": 400000
-    },
-    {
-      "epoch": 6.12,
-      "learning_rate": 0.00010790653786363416,
-      "loss": 0.2759,
-      "step": 401000
-    },
-    {
-      "epoch": 6.14,
-      "learning_rate": 0.00010769409082013337,
-      "loss": 0.2759,
-      "step": 402000
-    },
-    {
-      "epoch": 6.15,
-      "learning_rate": 0.00010748134091849238,
-      "loss": 0.2757,
-      "step": 403000
-    },
-    {
-      "epoch": 6.17,
-      "learning_rate": 0.00010726829048531,
-      "loss": 0.2762,
-      "step": 404000
-    },
-    {
-      "epoch": 6.18,
-      "learning_rate": 0.00010705494185047165,
-      "loss": 0.276,
-      "step": 405000
-    },
-    {
-      "epoch": 6.18,
-      "eval_runtime": 1.1676,
-      "eval_samples_per_second": 856.476,
-      "eval_steps_per_second": 13.704,
-      "step": 405000
-    },
-    {
-      "epoch": 6.2,
-      "learning_rate": 0.0001068412973471238,
-      "loss": 0.2754,
-      "step": 406000
-    },
-    {
-      "epoch": 6.21,
-      "learning_rate": 0.00010662735931164853,
-      "loss": 0.2755,
-      "step": 407000
-    },
-    {
-      "epoch": 6.23,
-      "learning_rate": 0.0001064131300836379,
-      "loss": 0.2752,
-      "step": 408000
-    },
-    {
-      "epoch": 6.25,
-      "learning_rate": 0.0001061986120058684,
-      "loss": 0.2748,
-      "step": 409000
-    },
-    {
-      "epoch": 6.26,
-      "learning_rate": 0.00010598380742427543,
-      "loss": 0.2749,
-      "step": 410000
-    },
-    {
-      "epoch": 6.26,
-      "eval_runtime": 1.0797,
-      "eval_samples_per_second": 926.22,
-      "eval_steps_per_second": 14.82,
-      "step": 410000
-    },
-    {
-      "epoch": 6.28,
-      "learning_rate": 0.00010576871868792746,
-      "loss": 0.275,
-      "step": 411000
-    },
-    {
-      "epoch": 6.29,
-      "learning_rate": 0.0001055533481490004,
-      "loss": 0.2746,
-      "step": 412000
-    },
-    {
-      "epoch": 6.31,
-      "learning_rate": 0.000105337698162752,
-      "loss": 0.2741,
-      "step": 413000
-    },
-    {
-      "epoch": 6.32,
-      "learning_rate": 0.00010512177108749594,
-      "loss": 0.2746,
-      "step": 414000
-    },
-    {
-      "epoch": 6.34,
-      "learning_rate": 0.00010490556928457616,
-      "loss": 0.2743,
-      "step": 415000
-    },
-    {
-      "epoch": 6.34,
-      "eval_runtime": 1.0107,
-      "eval_samples_per_second": 989.389,
-      "eval_steps_per_second": 15.83,
-      "step": 415000
-    },
-    {
-      "epoch": 6.35,
-      "learning_rate": 0.00010468909511834088,
-      "loss": 0.2741,
-      "step": 416000
-    },
-    {
-      "epoch": 6.37,
-      "learning_rate": 0.00010447235095611692,
-      "loss": 0.2738,
-      "step": 417000
-    },
-    {
-      "epoch": 6.38,
-      "learning_rate": 0.00010425533916818376,
-      "loss": 0.2738,
-      "step": 418000
-    },
-    {
-      "epoch": 6.4,
-      "learning_rate": 0.00010403806212774747,
-      "loss": 0.2742,
-      "step": 419000
-    },
-    {
-      "epoch": 6.41,
-      "learning_rate": 0.000103820522210915,
-      "loss": 0.2737,
-      "step": 420000
-    },
-    {
-      "epoch": 6.41,
-      "eval_runtime": 1.055,
-      "eval_samples_per_second": 947.861,
-      "eval_steps_per_second": 15.166,
-      "step": 420000
-    },
-    {
-      "epoch": 6.43,
-      "learning_rate": 0.00010360272179666802,
-      "loss": 0.2742,
-      "step": 421000
-    },
-    {
-      "epoch": 6.44,
-      "learning_rate": 0.00010338466326683697,
-      "loss": 0.2733,
-      "step": 422000
-    },
-    {
-      "epoch": 6.46,
-      "learning_rate": 0.00010316634900607497,
-      "loss": 0.2737,
-      "step": 423000
-    },
-    {
-      "epoch": 6.47,
-      "learning_rate": 0.00010294778140183182,
-      "loss": 0.2732,
-      "step": 424000
-    },
-    {
-      "epoch": 6.49,
-      "learning_rate": 0.00010272896284432785,
-      "loss": 0.2733,
-      "step": 425000
-    },
-    {
-      "epoch": 6.49,
-      "eval_runtime": 1.0035,
-      "eval_samples_per_second": 996.544,
-      "eval_steps_per_second": 15.945,
-      "step": 425000
-    },
-    {
-      "epoch": 6.51,
-      "learning_rate": 0.00010250989572652766,
-      "loss": 0.2728,
-      "step": 426000
-    },
-    {
-      "epoch": 6.52,
-      "learning_rate": 0.00010229058244411427,
-      "loss": 0.2729,
-      "step": 427000
-    },
-    {
-      "epoch": 6.54,
-      "learning_rate": 0.00010207102539546251,
-      "loss": 0.2728,
-      "step": 428000
-    },
-    {
-      "epoch": 6.55,
-      "learning_rate": 0.00010185122698161311,
-      "loss": 0.2726,
-      "step": 429000
-    },
-    {
-      "epoch": 6.57,
-      "learning_rate": 0.00010163118960624632,
-      "loss": 0.2725,
-      "step": 430000
-    },
-    {
-      "epoch": 6.57,
-      "eval_runtime": 1.0983,
-      "eval_samples_per_second": 910.508,
-      "eval_steps_per_second": 14.568,
-      "step": 430000
-    },
-    {
-      "epoch": 6.58,
-      "learning_rate": 0.00010141091567565561,
-      "loss": 0.2727,
-      "step": 431000
-    },
-    {
-      "epoch": 6.6,
-      "learning_rate": 0.00010119040759872142,
-      "loss": 0.2725,
-      "step": 432000
-    },
-    {
-      "epoch": 6.61,
-      "learning_rate": 0.00010096966778688472,
-      "loss": 0.2721,
-      "step": 433000
-    },
-    {
-      "epoch": 6.63,
-      "learning_rate": 0.00010074869865412074,
-      "loss": 0.272,
-      "step": 434000
-    },
-    {
-      "epoch": 6.64,
-      "learning_rate": 0.00010052750261691254,
-      "loss": 0.2721,
-      "step": 435000
-    },
-    {
-      "epoch": 6.64,
-      "eval_runtime": 0.9895,
-      "eval_samples_per_second": 1010.612,
-      "eval_steps_per_second": 16.17,
-      "step": 435000
-    },
-    {
-      "epoch": 6.66,
-      "learning_rate": 0.0001003060820942245,
-      "loss": 0.2716,
-      "step": 436000
-    },
-    {
-      "epoch": 6.67,
-      "learning_rate": 0.00010008443950747599,
-      "loss": 0.2716,
-      "step": 437000
-    },
-    {
-      "epoch": 6.69,
-      "learning_rate": 9.986257728051483e-05,
-      "loss": 0.2717,
-      "step": 438000
-    },
-    {
-      "epoch": 6.7,
-      "learning_rate": 9.964049783959082e-05,
-      "loss": 0.2716,
-      "step": 439000
-    },
-    {
-      "epoch": 6.72,
-      "learning_rate": 9.94182036133291e-05,
-      "loss": 0.2715,
-      "step": 440000
-    },
-    {
-      "epoch": 6.72,
-      "eval_runtime": 1.0245,
-      "eval_samples_per_second": 976.131,
-      "eval_steps_per_second": 15.618,
-      "step": 440000
-    },
-    {
-      "epoch": 6.73,
-      "learning_rate": 9.919569703270376e-05,
-      "loss": 0.2716,
-      "step": 441000
-    },
-    {
-      "epoch": 6.75,
-      "learning_rate": 9.89729805310111e-05,
-      "loss": 0.2711,
-      "step": 442000
-    },
-    {
-      "epoch": 6.76,
-      "learning_rate": 9.875005654384307e-05,
-      "loss": 0.2712,
-      "step": 443000
-    },
-    {
-      "epoch": 6.78,
-      "learning_rate": 9.852692750906071e-05,
-      "loss": 0.2717,
-      "step": 444000
-    },
-    {
-      "epoch": 6.8,
-      "learning_rate": 9.830359586676737e-05,
-      "loss": 0.2722,
-      "step": 445000
-    },
-    {
-      "epoch": 6.8,
-      "eval_runtime": 1.1145,
-      "eval_samples_per_second": 897.295,
-      "eval_steps_per_second": 14.357,
-      "step": 445000
-    },
-    {
-      "epoch": 6.81,
-      "learning_rate": 9.808006405928215e-05,
-      "loss": 0.2703,
-      "step": 446000
-    },
-    {
-      "epoch": 6.83,
-      "learning_rate": 9.785633453111306e-05,
-      "loss": 0.2705,
-      "step": 447000
-    },
-    {
-      "epoch": 6.84,
-      "learning_rate": 9.763240972893037e-05,
-      "loss": 0.27,
-      "step": 448000
-    },
-    {
-      "epoch": 6.86,
-      "learning_rate": 9.740829210153984e-05,
-      "loss": 0.2703,
-      "step": 449000
-    },
-    {
-      "epoch": 6.87,
-      "learning_rate": 9.718398409985593e-05,
-      "loss": 0.27,
-      "step": 450000
-    },
-    {
-      "epoch": 6.87,
-      "eval_runtime": 0.9938,
-      "eval_samples_per_second": 1006.215,
-      "eval_steps_per_second": 16.099,
-      "step": 450000
-    },
-    {
-      "epoch": 6.89,
-      "learning_rate": 9.695948817687504e-05,
-      "loss": 0.2699,
-      "step": 451000
-    },
-    {
-      "epoch": 6.9,
-      "learning_rate": 9.673480678764858e-05,
-      "loss": 0.2698,
-      "step": 452000
-    },
-    {
-      "epoch": 6.92,
-      "learning_rate": 9.650994238925626e-05,
-      "loss": 0.2699,
-      "step": 453000
-    },
-    {
-      "epoch": 6.93,
-      "learning_rate": 9.628489744077911e-05,
-      "loss": 0.2696,
-      "step": 454000
-    },
-    {
-      "epoch": 6.95,
-      "learning_rate": 9.60596744032726e-05,
-      "loss": 0.2699,
-      "step": 455000
-    },
-    {
-      "epoch": 6.95,
-      "eval_runtime": 1.0008,
-      "eval_samples_per_second": 999.165,
-      "eval_steps_per_second": 15.987,
-      "step": 455000
-    },
-    {
-      "epoch": 6.96,
-      "learning_rate": 9.583427573973982e-05,
-      "loss": 0.2696,
-      "step": 456000
-    },
-    {
-      "epoch": 6.98,
-      "learning_rate": 9.560870391510441e-05,
-      "loss": 0.2695,
-      "step": 457000
-    },
-    {
-      "epoch": 6.99,
-      "learning_rate": 9.538296139618371e-05,
-      "loss": 0.2691,
-      "step": 458000
-    },
-    {
-      "epoch": 7.01,
-      "learning_rate": 9.515705065166178e-05,
-      "loss": 0.2693,
-      "step": 459000
-    },
-    {
-      "epoch": 7.02,
-      "learning_rate": 9.493097415206228e-05,
-      "loss": 0.2688,
-      "step": 460000
-    },
-    {
-      "epoch": 7.02,
-      "eval_runtime": 1.0225,
-      "eval_samples_per_second": 978.034,
-      "eval_steps_per_second": 15.649,
-      "step": 460000
-    },
-    {
-      "epoch": 7.04,
-      "learning_rate": 9.47047343697216e-05,
-      "loss": 0.269,
-      "step": 461000
-    },
-    {
-      "epoch": 7.05,
-      "learning_rate": 9.447833377876176e-05,
-      "loss": 0.269,
-      "step": 462000
-    },
-    {
-      "epoch": 7.07,
-      "learning_rate": 9.425177485506336e-05,
-      "loss": 0.2688,
-      "step": 463000
-    },
-    {
-      "epoch": 7.09,
-      "learning_rate": 9.402506007623848e-05,
-      "loss": 0.269,
-      "step": 464000
-    },
-    {
-      "epoch": 7.1,
-      "learning_rate": 9.379819192160362e-05,
-      "loss": 0.2692,
-      "step": 465000
-    },
-    {
-      "epoch": 7.1,
-      "eval_runtime": 1.1401,
-      "eval_samples_per_second": 877.142,
-      "eval_steps_per_second": 14.034,
-      "step": 465000
-    },
-    {
-      "epoch": 7.12,
-      "learning_rate": 9.357117287215258e-05,
-      "loss": 0.2682,
-      "step": 466000
-    },
-    {
-      "epoch": 7.13,
-      "learning_rate": 9.334400541052928e-05,
-      "loss": 0.2683,
-      "step": 467000
-    },
-    {
-      "epoch": 7.15,
-      "learning_rate": 9.311669202100073e-05,
-      "loss": 0.2693,
-      "step": 468000
-    },
-    {
-      "epoch": 7.16,
-      "learning_rate": 9.288923518942968e-05,
-      "loss": 0.2683,
-      "step": 469000
-    },
-    {
-      "epoch": 7.18,
-      "learning_rate": 9.26616374032477e-05,
-      "loss": 0.2677,
-      "step": 470000
-    },
-    {
-      "epoch": 7.18,
-      "eval_runtime": 0.8954,
-      "eval_samples_per_second": 1116.774,
-      "eval_steps_per_second": 17.868,
-      "step": 470000
-    },
-    {
-      "epoch": 7.19,
-      "learning_rate": 9.243390115142761e-05,
-      "loss": 0.2678,
-      "step": 471000
-    },
-    {
-      "epoch": 7.21,
-      "learning_rate": 9.220602892445661e-05,
-      "loss": 0.2678,
-      "step": 472000
-    },
-    {
-      "epoch": 7.22,
-      "learning_rate": 9.197802321430889e-05,
-      "loss": 0.2679,
-      "step": 473000
-    },
-    {
-      "epoch": 7.24,
-      "learning_rate": 9.174988651441833e-05,
-      "loss": 0.2673,
-      "step": 474000
-    },
-    {
-      "epoch": 7.25,
-      "learning_rate": 9.152162131965137e-05,
-      "loss": 0.2675,
-      "step": 475000
-    },
-    {
-      "epoch": 7.25,
-      "eval_runtime": 1.0353,
-      "eval_samples_per_second": 965.922,
-      "eval_steps_per_second": 15.455,
-      "step": 475000
-    },
-    {
-      "epoch": 7.27,
-      "learning_rate": 9.129323012627956e-05,
-      "loss": 0.2693,
-      "step": 476000
-    },
-    {
-      "epoch": 7.28,
-      "learning_rate": 9.106471543195244e-05,
-      "loss": 0.2675,
-      "step": 477000
-    },
-    {
-      "epoch": 7.3,
-      "learning_rate": 9.08360797356701e-05,
-      "loss": 0.2679,
-      "step": 478000
-    },
-    {
-      "epoch": 7.31,
-      "learning_rate": 9.060732553775582e-05,
-      "loss": 0.2672,
-      "step": 479000
-    },
-    {
-      "epoch": 7.33,
-      "learning_rate": 9.037845533982892e-05,
-      "loss": 0.267,
-      "step": 480000
-    },
-    {
-      "epoch": 7.33,
-      "eval_runtime": 1.0347,
-      "eval_samples_per_second": 966.468,
-      "eval_steps_per_second": 15.463,
-      "step": 480000
-    },
-    {
-      "epoch": 7.34,
-      "learning_rate": 9.014947164477721e-05,
-      "loss": 0.2663,
-      "step": 481000
-    },
-    {
-      "epoch": 7.36,
-      "learning_rate": 8.992037695672967e-05,
-      "loss": 0.267,
-      "step": 482000
-    },
-    {
-      "epoch": 7.38,
-      "learning_rate": 8.969117378102912e-05,
-      "loss": 0.2665,
-      "step": 483000
-    },
-    {
-      "epoch": 7.39,
-      "learning_rate": 8.946186462420478e-05,
-      "loss": 0.2662,
-      "step": 484000
-    },
-    {
-      "epoch": 7.41,
-      "learning_rate": 8.923245199394482e-05,
-      "loss": 0.2662,
-      "step": 485000
-    },
-    {
-      "epoch": 7.41,
-      "eval_runtime": 1.0079,
-      "eval_samples_per_second": 992.191,
-      "eval_steps_per_second": 15.875,
-      "step": 485000
-    },
-    {
-      "epoch": 7.42,
-      "learning_rate": 8.900293839906903e-05,
-      "loss": 0.2664,
-      "step": 486000
-    },
-    {
-      "epoch": 7.44,
-      "learning_rate": 8.87733263495013e-05,
-      "loss": 0.2658,
-      "step": 487000
-    },
-    {
-      "epoch": 7.45,
-      "learning_rate": 8.85436183562422e-05,
-      "loss": 0.2659,
-      "step": 488000
-    },
-    {
-      "epoch": 7.47,
-      "learning_rate": 8.83138169313416e-05,
-      "loss": 0.2663,
-      "step": 489000
-    },
-    {
-      "epoch": 7.48,
-      "learning_rate": 8.808392458787103e-05,
-      "loss": 0.2656,
-      "step": 490000
-    },
-    {
-      "epoch": 7.48,
-      "eval_runtime": 1.075,
-      "eval_samples_per_second": 930.213,
-      "eval_steps_per_second": 14.883,
-      "step": 490000
-    },
-    {
-      "epoch": 7.5,
-      "learning_rate": 8.78539438398963e-05,
-      "loss": 0.2655,
-      "step": 491000
-    },
-    {
-      "epoch": 7.51,
-      "learning_rate": 8.762387720245008e-05,
-      "loss": 0.2656,
-      "step": 492000
-    },
-    {
-      "epoch": 7.53,
-      "learning_rate": 8.73937271915042e-05,
-      "loss": 0.2655,
-      "step": 493000
-    },
-    {
-      "epoch": 7.54,
-      "learning_rate": 8.716349632394235e-05,
-      "loss": 0.2652,
-      "step": 494000
-    },
-    {
-      "epoch": 7.56,
-      "learning_rate": 8.69331871175324e-05,
-      "loss": 0.2651,
-      "step": 495000
-    },
-    {
-      "epoch": 7.56,
-      "eval_runtime": 1.1978,
-      "eval_samples_per_second": 834.871,
-      "eval_steps_per_second": 13.358,
-      "step": 495000
-    },
-    {
-      "epoch": 7.57,
-      "learning_rate": 8.67028020908989e-05,
-      "loss": 0.2647,
-      "step": 496000
-    },
-    {
-      "epoch": 7.59,
-      "learning_rate": 8.647234376349565e-05,
-      "loss": 0.2653,
-      "step": 497000
-    },
-    {
-      "epoch": 7.6,
-      "learning_rate": 8.624181465557794e-05,
-      "loss": 0.2649,
-      "step": 498000
-    },
-    {
-      "epoch": 7.62,
-      "learning_rate": 8.601121728817519e-05,
-      "loss": 0.2647,
-      "step": 499000
-    },
-    {
-      "epoch": 7.64,
-      "learning_rate": 8.578055418306327e-05,
-      "loss": 0.2654,
-      "step": 500000
-    },
-    {
-      "epoch": 7.64,
-      "eval_runtime": 1.1022,
-      "eval_samples_per_second": 907.298,
-      "eval_steps_per_second": 14.517,
-      "step": 500000
-    },
-    {
-      "epoch": 7.65,
-      "learning_rate": 8.55498278627369e-05,
-      "loss": 0.2646,
-      "step": 501000
-    },
-    {
-      "epoch": 7.67,
-      "learning_rate": 8.531904085038221e-05,
-      "loss": 0.2646,
-      "step": 502000
-    },
-    {
-      "epoch": 7.68,
-      "learning_rate": 8.508819566984897e-05,
-      "loss": 0.2641,
-      "step": 503000
-    },
-    {
-      "epoch": 7.7,
-      "learning_rate": 8.485729484562307e-05,
-      "loss": 0.2641,
-      "step": 504000
-    },
-    {
-      "epoch": 7.71,
-      "learning_rate": 8.462634090279895e-05,
-      "loss": 0.264,
-      "step": 505000
-    },
-    {
-      "epoch": 7.71,
-      "eval_runtime": 1.0129,
-      "eval_samples_per_second": 987.309,
-      "eval_steps_per_second": 15.797,
-      "step": 505000
-    },
-    {
-      "epoch": 7.73,
-      "learning_rate": 8.439533636705194e-05,
-      "loss": 0.2635,
-      "step": 506000
-    },
-    {
-      "epoch": 7.74,
-      "learning_rate": 8.416428376461061e-05,
-      "loss": 0.2644,
-      "step": 507000
-    },
-    {
-      "epoch": 7.76,
-      "learning_rate": 8.393318562222916e-05,
-      "loss": 0.2642,
-      "step": 508000
-    },
-    {
-      "epoch": 7.77,
-      "learning_rate": 8.370204446715997e-05,
-      "loss": 0.2638,
-      "step": 509000
-    },
-    {
-      "epoch": 7.79,
-      "learning_rate": 8.347086282712556e-05,
-      "loss": 0.2637,
-      "step": 510000
-    },
-    {
-      "epoch": 7.79,
-      "eval_runtime": 1.1071,
-      "eval_samples_per_second": 903.278,
-      "eval_steps_per_second": 14.452,
-      "step": 510000
-    },
-    {
-      "epoch": 7.8,
-      "learning_rate": 8.323964323029136e-05,
-      "loss": 0.2633,
-      "step": 511000
-    },
-    {
-      "epoch": 7.82,
-      "learning_rate": 8.300838820523784e-05,
-      "loss": 0.2634,
-      "step": 512000
-    },
-    {
-      "epoch": 7.83,
-      "learning_rate": 8.277710028093289e-05,
-      "loss": 0.263,
-      "step": 513000
-    },
-    {
-      "epoch": 7.85,
-      "learning_rate": 8.254578198670421e-05,
-      "loss": 0.2632,
-      "step": 514000
-    },
-    {
-      "epoch": 7.86,
-      "learning_rate": 8.231443585221157e-05,
-      "loss": 0.2629,
-      "step": 515000
-    },
-    {
-      "epoch": 7.86,
-      "eval_runtime": 1.0457,
-      "eval_samples_per_second": 956.256,
-      "eval_steps_per_second": 15.3,
-      "step": 515000
-    },
-    {
-      "epoch": 7.88,
-      "learning_rate": 8.208306440741926e-05,
-      "loss": 0.2626,
-      "step": 516000
-    },
-    {
-      "epoch": 7.89,
-      "learning_rate": 8.185167018256834e-05,
-      "loss": 0.2629,
-      "step": 517000
-    },
-    {
-      "epoch": 7.91,
-      "learning_rate": 8.162025570814896e-05,
-      "loss": 0.2625,
-      "step": 518000
-    },
-    {
-      "epoch": 7.93,
-      "learning_rate": 8.138882351487275e-05,
-      "loss": 0.2623,
-      "step": 519000
-    },
-    {
-      "epoch": 7.94,
-      "learning_rate": 8.115737613364511e-05,
-      "loss": 0.2626,
-      "step": 520000
-    },
-    {
-      "epoch": 7.94,
-      "eval_runtime": 1.0504,
-      "eval_samples_per_second": 952.036,
-      "eval_steps_per_second": 15.233,
-      "step": 520000
-    },
-    {
-      "epoch": 7.96,
-      "learning_rate": 8.092591609553747e-05,
-      "loss": 0.2623,
-      "step": 521000
-    },
-    {
-      "epoch": 7.97,
-      "learning_rate": 8.069444593175975e-05,
-      "loss": 0.2622,
-      "step": 522000
-    },
-    {
-      "epoch": 7.99,
-      "learning_rate": 8.046296817363259e-05,
-      "loss": 0.262,
-      "step": 523000
-    },
-    {
-      "epoch": 8.0,
-      "learning_rate": 8.023148535255965e-05,
-      "loss": 0.2619,
-      "step": 524000
-    },
-    {
-      "epoch": 8.02,
-      "learning_rate": 7.999999999999999e-05,
-      "loss": 0.262,
-      "step": 525000
-    },
-    {
-      "epoch": 8.02,
-      "eval_runtime": 1.1375,
-      "eval_samples_per_second": 879.131,
-      "eval_steps_per_second": 14.066,
-      "step": 525000
-    },
-    {
-      "epoch": 8.03,
-      "learning_rate": 7.976851464744033e-05,
-      "loss": 0.2616,
-      "step": 526000
-    },
-    {
-      "epoch": 8.05,
-      "learning_rate": 7.953703182636741e-05,
-      "loss": 0.2616,
-      "step": 527000
-    },
-    {
-      "epoch": 8.06,
-      "learning_rate": 7.930555406824026e-05,
-      "loss": 0.2617,
-      "step": 528000
-    },
-    {
-      "epoch": 8.08,
-      "learning_rate": 7.907408390446254e-05,
-      "loss": 0.2614,
-      "step": 529000
-    },
-    {
-      "epoch": 8.09,
-      "learning_rate": 7.884262386635489e-05,
-      "loss": 0.2607,
-      "step": 530000
-    },
-    {
-      "epoch": 8.09,
-      "eval_runtime": 1.0134,
-      "eval_samples_per_second": 986.75,
-      "eval_steps_per_second": 15.788,
-      "step": 530000
-    },
-    {
-      "epoch": 8.11,
-      "learning_rate": 7.861117648512725e-05,
-      "loss": 0.2613,
-      "step": 531000
-    },
-    {
-      "epoch": 8.12,
-      "learning_rate": 7.837974429185103e-05,
-      "loss": 0.2614,
-      "step": 532000
-    },
-    {
-      "epoch": 8.14,
-      "learning_rate": 7.814832981743164e-05,
-      "loss": 0.2614,
-      "step": 533000
-    },
-    {
-      "epoch": 8.15,
-      "learning_rate": 7.791693559258072e-05,
-      "loss": 0.2608,
-      "step": 534000
-    },
-    {
-      "epoch": 8.17,
-      "learning_rate": 7.768556414778842e-05,
-      "loss": 0.2606,
-      "step": 535000
-    },
-    {
-      "epoch": 8.17,
-      "eval_runtime": 1.097,
-      "eval_samples_per_second": 911.552,
-      "eval_steps_per_second": 14.585,
-      "step": 535000
-    },
-    {
-      "epoch": 8.18,
-      "learning_rate": 7.74542180132958e-05,
-      "loss": 0.2606,
-      "step": 536000
-    },
-    {
-      "epoch": 8.2,
-      "learning_rate": 7.72228997190671e-05,
-      "loss": 0.2608,
-      "step": 537000
-    },
-    {
-      "epoch": 8.22,
-      "learning_rate": 7.699161179476217e-05,
-      "loss": 0.2604,
-      "step": 538000
-    },
-    {
-      "epoch": 8.23,
-      "learning_rate": 7.676035676970863e-05,
-      "loss": 0.2606,
-      "step": 539000
-    },
-    {
-      "epoch": 8.25,
-      "learning_rate": 7.652913717287443e-05,
-      "loss": 0.2604,
-      "step": 540000
-    },
-    {
-      "epoch": 8.25,
-      "eval_runtime": 1.1778,
-      "eval_samples_per_second": 849.063,
-      "eval_steps_per_second": 13.585,
-      "step": 540000
-    },
-    {
-      "epoch": 8.26,
-      "learning_rate": 7.629795553284005e-05,
-      "loss": 0.2602,
-      "step": 541000
-    },
-    {
-      "epoch": 8.28,
-      "learning_rate": 7.606681437777081e-05,
-      "loss": 0.2605,
-      "step": 542000
-    },
-    {
-      "epoch": 8.29,
-      "learning_rate": 7.583571623538939e-05,
-      "loss": 0.26,
-      "step": 543000
-    },
-    {
-      "epoch": 8.31,
-      "learning_rate": 7.560466363294806e-05,
-      "loss": 0.2596,
-      "step": 544000
-    },
-    {
-      "epoch": 8.32,
-      "learning_rate": 7.537365909720104e-05,
-      "loss": 0.2595,
-      "step": 545000
-    },
-    {
-      "epoch": 8.32,
-      "eval_runtime": 1.1629,
-      "eval_samples_per_second": 859.911,
-      "eval_steps_per_second": 13.759,
-      "step": 545000
-    },
-    {
-      "epoch": 8.34,
-      "learning_rate": 7.514270515437691e-05,
-      "loss": 0.2595,
-      "step": 546000
-    },
-    {
-      "epoch": 8.35,
-      "learning_rate": 7.491180433015101e-05,
-      "loss": 0.2594,
-      "step": 547000
-    },
-    {
-      "epoch": 8.37,
-      "learning_rate": 7.468095914961777e-05,
-      "loss": 0.2596,
-      "step": 548000
-    },
-    {
-      "epoch": 8.38,
-      "learning_rate": 7.445017213726307e-05,
-      "loss": 0.2596,
-      "step": 549000
-    },
-    {
-      "epoch": 8.4,
-      "learning_rate": 7.421944581693674e-05,
-      "loss": 0.2594,
-      "step": 550000
-    },
-    {
-      "epoch": 8.4,
-      "eval_runtime": 0.9899,
-      "eval_samples_per_second": 1010.184,
-      "eval_steps_per_second": 16.163,
-      "step": 550000
-    },
-    {
-      "epoch": 8.41,
-      "learning_rate": 7.39887827118248e-05,
-      "loss": 0.259,
-      "step": 551000
-    },
-    {
-      "epoch": 8.43,
-      "learning_rate": 7.375818534442207e-05,
-      "loss": 0.2588,
-      "step": 552000
-    },
-    {
-      "epoch": 8.44,
-      "learning_rate": 7.352765623650435e-05,
-      "loss": 0.259,
-      "step": 553000
-    },
-    {
-      "epoch": 8.46,
-      "learning_rate": 7.329719790910108e-05,
-      "loss": 0.2587,
-      "step": 554000
-    },
-    {
-      "epoch": 8.47,
-      "learning_rate": 7.30668128824676e-05,
-      "loss": 0.2587,
-      "step": 555000
-    },
-    {
-      "epoch": 8.47,
-      "eval_runtime": 1.1635,
-      "eval_samples_per_second": 859.466,
-      "eval_steps_per_second": 13.751,
-      "step": 555000
-    },
-    {
-      "epoch": 8.49,
-      "learning_rate": 7.283650367605764e-05,
-      "loss": 0.2584,
-      "step": 556000
-    },
-    {
-      "epoch": 8.51,
-      "learning_rate": 7.260627280849581e-05,
-      "loss": 0.2585,
-      "step": 557000
-    },
-    {
-      "epoch": 8.52,
-      "learning_rate": 7.23761227975499e-05,
-      "loss": 0.2584,
-      "step": 558000
-    },
-    {
-      "epoch": 8.54,
-      "learning_rate": 7.21460561601037e-05,
-      "loss": 0.2584,
-      "step": 559000
-    },
-    {
-      "epoch": 8.55,
-      "learning_rate": 7.191607541212897e-05,
-      "loss": 0.2585,
-      "step": 560000
-    },
-    {
-      "epoch": 8.55,
-      "eval_runtime": 1.1711,
-      "eval_samples_per_second": 853.863,
-      "eval_steps_per_second": 13.662,
-      "step": 560000
-    },
-    {
-      "epoch": 8.57,
-      "learning_rate": 7.168618306865838e-05,
-      "loss": 0.2583,
-      "step": 561000
-    },
-    {
-      "epoch": 8.58,
-      "learning_rate": 7.145638164375779e-05,
-      "loss": 0.2588,
-      "step": 562000
-    },
-    {
-      "epoch": 8.6,
-      "learning_rate": 7.122667365049869e-05,
-      "loss": 0.2578,
-      "step": 563000
-    },
-    {
-      "epoch": 8.61,
-      "learning_rate": 7.099706160093098e-05,
-      "loss": 0.2578,
-      "step": 564000
-    },
-    {
-      "epoch": 8.63,
-      "learning_rate": 7.076754800605516e-05,
-      "loss": 0.2579,
-      "step": 565000
-    },
-    {
-      "epoch": 8.63,
-      "eval_runtime": 1.0129,
-      "eval_samples_per_second": 987.305,
-      "eval_steps_per_second": 15.797,
-      "step": 565000
-    },
-    {
-      "epoch": 8.64,
-      "learning_rate": 7.053813537579523e-05,
-      "loss": 0.2581,
-      "step": 566000
-    },
-    {
-      "epoch": 8.66,
-      "learning_rate": 7.030882621897088e-05,
-      "loss": 0.2575,
-      "step": 567000
-    },
-    {
-      "epoch": 8.67,
-      "learning_rate": 7.00796230432703e-05,
-      "loss": 0.2574,
-      "step": 568000
-    },
-    {
-      "epoch": 8.69,
-      "learning_rate": 6.985052835522279e-05,
-      "loss": 0.2572,
-      "step": 569000
-    },
-    {
-      "epoch": 8.7,
-      "learning_rate": 6.962154466017105e-05,
-      "loss": 0.2572,
-      "step": 570000
-    },
-    {
-      "epoch": 8.7,
-      "eval_runtime": 1.048,
-      "eval_samples_per_second": 954.187,
-      "eval_steps_per_second": 15.267,
-      "step": 570000
-    },
-    {
-      "epoch": 8.72,
-      "learning_rate": 6.939267446224418e-05,
-      "loss": 0.2569,
-      "step": 571000
-    },
-    {
-      "epoch": 8.73,
-      "learning_rate": 6.91639202643299e-05,
-      "loss": 0.2569,
-      "step": 572000
-    },
-    {
-      "epoch": 8.75,
-      "learning_rate": 6.893528456804756e-05,
-      "loss": 0.2569,
-      "step": 573000
-    },
-    {
-      "epoch": 8.77,
-      "learning_rate": 6.870676987372044e-05,
-      "loss": 0.2568,
-      "step": 574000
-    },
-    {
-      "epoch": 8.78,
-      "learning_rate": 6.847837868034861e-05,
-      "loss": 0.257,
-      "step": 575000
-    },
-    {
-      "epoch": 8.78,
-      "eval_runtime": 1.0002,
-      "eval_samples_per_second": 999.79,
-      "eval_steps_per_second": 15.997,
-      "step": 575000
-    },
-    {
-      "epoch": 8.8,
-      "learning_rate": 6.825011348558167e-05,
-      "loss": 0.2573,
-      "step": 576000
-    },
-    {
-      "epoch": 8.81,
-      "learning_rate": 6.802197678569109e-05,
-      "loss": 0.2566,
-      "step": 577000
-    },
-    {
-      "epoch": 8.83,
-      "learning_rate": 6.779397107554339e-05,
-      "loss": 0.2562,
-      "step": 578000
-    },
-    {
-      "epoch": 8.84,
-      "learning_rate": 6.756609884857239e-05,
-      "loss": 0.2566,
-      "step": 579000
-    },
-    {
-      "epoch": 8.86,
-      "learning_rate": 6.733836259675233e-05,
-      "loss": 0.2564,
-      "step": 580000
-    },
-    {
-      "epoch": 8.86,
-      "eval_runtime": 1.0727,
-      "eval_samples_per_second": 932.263,
-      "eval_steps_per_second": 14.916,
-      "step": 580000
-    },
-    {
-      "epoch": 8.87,
-      "learning_rate": 6.71107648105703e-05,
-      "loss": 0.2564,
-      "step": 581000
-    },
-    {
-      "epoch": 8.89,
-      "learning_rate": 6.688330797899925e-05,
-      "loss": 0.2562,
-      "step": 582000
-    },
-    {
-      "epoch": 8.9,
-      "learning_rate": 6.665599458947072e-05,
-      "loss": 0.2562,
-      "step": 583000
-    },
-    {
-      "epoch": 8.92,
-      "learning_rate": 6.642882712784742e-05,
-      "loss": 0.2561,
-      "step": 584000
-    },
-    {
-      "epoch": 8.93,
-      "learning_rate": 6.620180807839639e-05,
-      "loss": 0.2561,
-      "step": 585000
-    },
-    {
-      "epoch": 8.93,
-      "eval_runtime": 0.9936,
-      "eval_samples_per_second": 1006.405,
-      "eval_steps_per_second": 16.102,
-      "step": 585000
-    },
-    {
-      "epoch": 8.95,
-      "learning_rate": 6.597493992376152e-05,
-      "loss": 0.2557,
-      "step": 586000
-    },
-    {
-      "epoch": 8.96,
-      "learning_rate": 6.574822514493664e-05,
-      "loss": 0.2554,
-      "step": 587000
-    },
-    {
-      "epoch": 8.98,
-      "learning_rate": 6.552166622123824e-05,
-      "loss": 0.2554,
-      "step": 588000
-    },
-    {
-      "epoch": 8.99,
-      "learning_rate": 6.52952656302784e-05,
-      "loss": 0.2556,
-      "step": 589000
-    },
-    {
-      "epoch": 9.01,
-      "learning_rate": 6.506902584793773e-05,
-      "loss": 0.2553,
-      "step": 590000
-    },
-    {
-      "epoch": 9.01,
-      "eval_runtime": 0.9015,
-      "eval_samples_per_second": 1109.201,
-      "eval_steps_per_second": 17.747,
-      "step": 590000
-    },
-    {
-      "epoch": 9.02,
-      "learning_rate": 6.484294934833822e-05,
-      "loss": 0.2552,
-      "step": 591000
-    },
-    {
-      "epoch": 9.04,
-      "learning_rate": 6.461703860381628e-05,
-      "loss": 0.2551,
-      "step": 592000
-    },
-    {
-      "epoch": 9.06,
-      "learning_rate": 6.439129608489559e-05,
-      "loss": 0.2555,
-      "step": 593000
-    },
-    {
-      "epoch": 9.07,
-      "learning_rate": 6.41657242602602e-05,
-      "loss": 0.2549,
-      "step": 594000
-    },
-    {
-      "epoch": 9.09,
-      "learning_rate": 6.39403255967274e-05,
-      "loss": 0.255,
-      "step": 595000
-    },
-    {
-      "epoch": 9.09,
-      "eval_runtime": 1.1107,
-      "eval_samples_per_second": 900.319,
-      "eval_steps_per_second": 14.405,
-      "step": 595000
-    },
-    {
-      "epoch": 9.1,
-      "learning_rate": 6.371510255922088e-05,
-      "loss": 0.2545,
-      "step": 596000
-    },
-    {
-      "epoch": 9.12,
-      "learning_rate": 6.349005761074372e-05,
-      "loss": 0.2547,
-      "step": 597000
-    },
-    {
-      "epoch": 9.13,
-      "learning_rate": 6.326519321235139e-05,
-      "loss": 0.2546,
-      "step": 598000
-    },
-    {
-      "epoch": 9.15,
-      "learning_rate": 6.304051182312496e-05,
-      "loss": 0.2549,
-      "step": 599000
-    },
-    {
-      "epoch": 9.16,
-      "learning_rate": 6.281601590014407e-05,
-      "loss": 0.2546,
-      "step": 600000
-    },
-    {
-      "epoch": 9.16,
-      "eval_runtime": 1.0772,
-      "eval_samples_per_second": 928.316,
-      "eval_steps_per_second": 14.853,
-      "step": 600000
-    },
-    {
-      "epoch": 9.18,
-      "learning_rate": 6.259170789846017e-05,
-      "loss": 0.2546,
-      "step": 601000
-    },
-    {
-      "epoch": 9.19,
-      "learning_rate": 6.236759027106965e-05,
-      "loss": 0.2542,
-      "step": 602000
-    },
-    {
-      "epoch": 9.21,
-      "learning_rate": 6.214366546888694e-05,
-      "loss": 0.2541,
-      "step": 603000
-    },
-    {
-      "epoch": 9.22,
-      "learning_rate": 6.191993594071785e-05,
-      "loss": 0.2541,
-      "step": 604000
-    },
-    {
-      "epoch": 9.24,
-      "learning_rate": 6.169640413323262e-05,
-      "loss": 0.254,
-      "step": 605000
-    },
-    {
-      "epoch": 9.24,
-      "eval_runtime": 1.0913,
-      "eval_samples_per_second": 916.334,
-      "eval_steps_per_second": 14.661,
-      "step": 605000
-    },
-    {
-      "epoch": 9.25,
-      "learning_rate": 6.147307249093929e-05,
-      "loss": 0.2537,
-      "step": 606000
-    },
-    {
-      "epoch": 9.27,
-      "learning_rate": 6.124994345615693e-05,
-      "loss": 0.2532,
-      "step": 607000
-    },
-    {
-      "epoch": 9.28,
-      "learning_rate": 6.102701946898891e-05,
-      "loss": 0.2536,
-      "step": 608000
-    },
-    {
-      "epoch": 9.3,
-      "learning_rate": 6.0804302967296225e-05,
-      "loss": 0.2545,
-      "step": 609000
-    },
-    {
-      "epoch": 9.31,
-      "learning_rate": 6.058179638667089e-05,
-      "loss": 0.2536,
-      "step": 610000
-    },
-    {
-      "epoch": 9.31,
-      "eval_runtime": 1.0284,
-      "eval_samples_per_second": 972.365,
-      "eval_steps_per_second": 15.558,
-      "step": 610000
-    },
-    {
-      "epoch": 9.33,
-      "learning_rate": 6.035950216040917e-05,
-      "loss": 0.2533,
-      "step": 611000
-    },
-    {
-      "epoch": 9.35,
-      "learning_rate": 6.0137422719485145e-05,
-      "loss": 0.2531,
-      "step": 612000
-    },
-    {
-      "epoch": 9.36,
-      "learning_rate": 5.991556049252401e-05,
-      "loss": 0.2532,
-      "step": 613000
-    },
-    {
-      "epoch": 9.38,
-      "learning_rate": 5.969391790577551e-05,
-      "loss": 0.2532,
-      "step": 614000
-    },
-    {
-      "epoch": 9.39,
-      "learning_rate": 5.947249738308747e-05,
-      "loss": 0.2529,
-      "step": 615000
-    },
-    {
-      "epoch": 9.39,
-      "eval_runtime": 1.014,
-      "eval_samples_per_second": 986.174,
-      "eval_steps_per_second": 15.779,
-      "step": 615000
-    },
-    {
-      "epoch": 9.41,
-      "learning_rate": 5.925130134587924e-05,
-      "loss": 0.2527,
-      "step": 616000
-    },
-    {
-      "epoch": 9.42,
-      "learning_rate": 5.903033221311528e-05,
-      "loss": 0.2525,
-      "step": 617000
-    },
-    {
-      "epoch": 9.44,
-      "learning_rate": 5.880959240127858e-05,
-      "loss": 0.2524,
-      "step": 618000
-    },
-    {
-      "epoch": 9.45,
-      "learning_rate": 5.858908432434438e-05,
-      "loss": 0.2525,
-      "step": 619000
-    },
-    {
-      "epoch": 9.47,
-      "learning_rate": 5.8368810393753684e-05,
-      "loss": 0.2524,
-      "step": 620000
-    },
-    {
-      "epoch": 9.47,
-      "eval_runtime": 1.0588,
-      "eval_samples_per_second": 944.48,
-      "eval_steps_per_second": 15.112,
-      "step": 620000
-    },
-    {
-      "epoch": 9.48,
-      "learning_rate": 5.814877301838688e-05,
-      "loss": 0.2523,
-      "step": 621000
-    },
-    {
-      "epoch": 9.5,
-      "learning_rate": 5.7928974604537494e-05,
-      "loss": 0.2522,
-      "step": 622000
-    },
-    {
-      "epoch": 9.51,
-      "learning_rate": 5.770941755588573e-05,
-      "loss": 0.2537,
-      "step": 623000
-    },
-    {
-      "epoch": 9.53,
-      "learning_rate": 5.749010427347233e-05,
-      "loss": 0.254,
-      "step": 624000
-    },
-    {
-      "epoch": 9.54,
-      "learning_rate": 5.7271037155672156e-05,
-      "loss": 0.2522,
-      "step": 625000
-    },
-    {
-      "epoch": 9.54,
-      "eval_runtime": 1.0707,
-      "eval_samples_per_second": 934.001,
-      "eval_steps_per_second": 14.944,
-      "step": 625000
-    },
-    {
-      "epoch": 9.56,
-      "learning_rate": 5.7052218598168154e-05,
-      "loss": 0.2524,
-      "step": 626000
-    },
-    {
-      "epoch": 9.57,
-      "learning_rate": 5.6833650993925016e-05,
-      "loss": 0.2522,
-      "step": 627000
-    },
-    {
-      "epoch": 9.59,
-      "learning_rate": 5.661533673316303e-05,
-      "loss": 0.2522,
-      "step": 628000
-    },
-    {
-      "epoch": 9.6,
-      "learning_rate": 5.639727820333198e-05,
-      "loss": 0.2518,
-      "step": 629000
-    },
-    {
-      "epoch": 9.62,
-      "learning_rate": 5.617947778908498e-05,
-      "loss": 0.2517,
-      "step": 630000
-    },
-    {
-      "epoch": 9.62,
-      "eval_runtime": 1.1949,
-      "eval_samples_per_second": 836.899,
-      "eval_steps_per_second": 13.39,
-      "step": 630000
-    },
-    {
-      "epoch": 9.64,
-      "learning_rate": 5.596193787225254e-05,
-      "loss": 0.2514,
-      "step": 631000
-    },
-    {
-      "epoch": 9.65,
-      "learning_rate": 5.574466083181624e-05,
-      "loss": 0.2512,
-      "step": 632000
-    },
-    {
-      "epoch": 9.67,
-      "learning_rate": 5.552764904388305e-05,
-      "loss": 0.2511,
-      "step": 633000
-    },
-    {
-      "epoch": 9.68,
-      "learning_rate": 5.5310904881659116e-05,
-      "loss": 0.2511,
-      "step": 634000
-    },
-    {
-      "epoch": 9.7,
-      "learning_rate": 5.5094430715423835e-05,
-      "loss": 0.2509,
-      "step": 635000
-    },
-    {
-      "epoch": 9.7,
-      "eval_runtime": 1.0102,
-      "eval_samples_per_second": 989.889,
-      "eval_steps_per_second": 15.838,
-      "step": 635000
-    },
-    {
-      "epoch": 9.71,
-      "learning_rate": 5.487822891250406e-05,
-      "loss": 0.2511,
-      "step": 636000
-    },
-    {
-      "epoch": 9.73,
-      "learning_rate": 5.4662301837247985e-05,
-      "loss": 0.2508,
-      "step": 637000
-    },
-    {
-      "epoch": 9.74,
-      "learning_rate": 5.4446651850999604e-05,
-      "loss": 0.2506,
-      "step": 638000
-    },
-    {
-      "epoch": 9.76,
-      "learning_rate": 5.4231281312072544e-05,
-      "loss": 0.2505,
-      "step": 639000
-    },
-    {
-      "epoch": 9.77,
-      "learning_rate": 5.401619257572453e-05,
-      "loss": 0.2502,
-      "step": 640000
-    },
-    {
-      "epoch": 9.77,
-      "eval_runtime": 1.0069,
-      "eval_samples_per_second": 993.184,
-      "eval_steps_per_second": 15.891,
-      "step": 640000
-    },
-    {
-      "epoch": 9.79,
-      "learning_rate": 5.3801387994131576e-05,
-      "loss": 0.2501,
-      "step": 641000
-    },
-    {
-      "epoch": 9.8,
-      "learning_rate": 5.358686991636209e-05,
-      "loss": 0.2503,
-      "step": 642000
-    },
-    {
-      "epoch": 9.82,
-      "learning_rate": 5.3372640688351476e-05,
-      "loss": 0.2505,
-      "step": 643000
-    },
-    {
-      "epoch": 9.83,
-      "learning_rate": 5.315870265287618e-05,
-      "loss": 0.2502,
-      "step": 644000
-    },
-    {
-      "epoch": 9.85,
-      "learning_rate": 5.294505814952835e-05,
-      "loss": 0.2501,
-      "step": 645000
-    },
-    {
-      "epoch": 9.85,
-      "eval_runtime": 1.0688,
-      "eval_samples_per_second": 935.652,
-      "eval_steps_per_second": 14.97,
-      "step": 645000
-    },
-    {
-      "epoch": 9.86,
-      "learning_rate": 5.2731709514689995e-05,
-      "loss": 0.2502,
-      "step": 646000
-    },
-    {
-      "epoch": 9.88,
-      "learning_rate": 5.25186590815076e-05,
-      "loss": 0.2501,
-      "step": 647000
-    },
-    {
-      "epoch": 9.9,
-      "learning_rate": 5.2305909179866635e-05,
-      "loss": 0.2495,
-      "step": 648000
-    },
-    {
-      "epoch": 9.91,
-      "learning_rate": 5.209346213636584e-05,
-      "loss": 0.2498,
-      "step": 649000
-    },
-    {
-      "epoch": 9.93,
-      "learning_rate": 5.188132027429215e-05,
-      "loss": 0.2495,
-      "step": 650000
-    },
-    {
-      "epoch": 9.93,
-      "eval_runtime": 1.0361,
-      "eval_samples_per_second": 965.164,
-      "eval_steps_per_second": 15.443,
-      "step": 650000
-    },
-    {
-      "epoch": 9.94,
-      "learning_rate": 5.166948591359489e-05,
-      "loss": 0.2493,
-      "step": 651000
-    },
-    {
-      "epoch": 9.96,
-      "learning_rate": 5.145796137086076e-05,
-      "loss": 0.2493,
-      "step": 652000
-    },
-    {
-      "epoch": 9.97,
-      "learning_rate": 5.124674895928823e-05,
-      "loss": 0.2493,
-      "step": 653000
-    },
-    {
-      "epoch": 9.99,
-      "learning_rate": 5.103585098866237e-05,
-      "loss": 0.2491,
-      "step": 654000
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 5.082526976532968e-05,
-      "loss": 0.249,
-      "step": 655000
-    },
-    {
-      "epoch": 10.0,
-      "eval_runtime": 1.0267,
-      "eval_samples_per_second": 974.027,
-      "eval_steps_per_second": 15.584,
-      "step": 655000
-    },
-    {
-      "epoch": 10.02,
-      "learning_rate": 5.061500759217261e-05,
-      "loss": 0.2494,
-      "step": 656000
-    },
-    {
-      "epoch": 10.03,
-      "learning_rate": 5.04050667685846e-05,
-      "loss": 0.2487,
-      "step": 657000
-    },
-    {
-      "epoch": 10.05,
-      "learning_rate": 5.01954495904449e-05,
-      "loss": 0.2485,
-      "step": 658000
-    },
-    {
-      "epoch": 10.06,
-      "learning_rate": 4.998615835009339e-05,
-      "loss": 0.2488,
-      "step": 659000
-    },
-    {
-      "epoch": 10.08,
-      "learning_rate": 4.97771953363055e-05,
-      "loss": 0.2489,
-      "step": 660000
-    },
-    {
-      "epoch": 10.08,
-      "eval_runtime": 1.0445,
-      "eval_samples_per_second": 957.361,
-      "eval_steps_per_second": 15.318,
-      "step": 660000
-    },
-    {
-      "epoch": 10.09,
-      "learning_rate": 4.956856283426728e-05,
-      "loss": 0.2487,
-      "step": 661000
-    },
-    {
-      "epoch": 10.11,
-      "learning_rate": 4.936026312555037e-05,
-      "loss": 0.248,
-      "step": 662000
-    },
-    {
-      "epoch": 10.12,
-      "learning_rate": 4.915229848808698e-05,
-      "loss": 0.2478,
-      "step": 663000
-    },
-    {
-      "epoch": 10.14,
-      "learning_rate": 4.8944671196145136e-05,
-      "loss": 0.2484,
-      "step": 664000
-    },
-    {
-      "epoch": 10.15,
-      "learning_rate": 4.8737383520303546e-05,
-      "loss": 0.2485,
-      "step": 665000
-    },
-    {
-      "epoch": 10.15,
-      "eval_runtime": 1.1085,
-      "eval_samples_per_second": 902.106,
-      "eval_steps_per_second": 14.434,
-      "step": 665000
-    },
-    {
-      "epoch": 10.17,
-      "learning_rate": 4.853043772742709e-05,
-      "loss": 0.248,
-      "step": 666000
-    },
-    {
-      "epoch": 10.19,
-      "learning_rate": 4.832383608064172e-05,
-      "loss": 0.2476,
-      "step": 667000
-    },
-    {
-      "epoch": 10.2,
-      "learning_rate": 4.811758083931005e-05,
-      "loss": 0.2478,
-      "step": 668000
-    },
-    {
-      "epoch": 10.22,
-      "learning_rate": 4.791167425900632e-05,
-      "loss": 0.2481,
-      "step": 669000
-    },
-    {
-      "epoch": 10.23,
-      "learning_rate": 4.770611859149185e-05,
-      "loss": 0.2508,
-      "step": 670000
-    },
-    {
-      "epoch": 10.23,
-      "eval_runtime": 1.1412,
-      "eval_samples_per_second": 876.243,
-      "eval_steps_per_second": 14.02,
-      "step": 670000
-    },
-    {
-      "epoch": 10.25,
-      "learning_rate": 4.7500916084690564e-05,
-      "loss": 0.2542,
-      "step": 671000
-    },
-    {
-      "epoch": 10.26,
-      "learning_rate": 4.729606898266411e-05,
-      "loss": 0.2507,
-      "step": 672000
-    },
-    {
-      "epoch": 10.28,
-      "learning_rate": 4.709157952558768e-05,
-      "loss": 0.2478,
-      "step": 673000
-    },
-    {
-      "epoch": 10.29,
-      "learning_rate": 4.688744994972514e-05,
-      "loss": 0.2482,
-      "step": 674000
-    },
-    {
-      "epoch": 10.31,
-      "learning_rate": 4.668368248740485e-05,
-      "loss": 0.247,
-      "step": 675000
-    },
-    {
-      "epoch": 10.31,
-      "eval_runtime": 0.9224,
-      "eval_samples_per_second": 1084.145,
-      "eval_steps_per_second": 17.346,
-      "step": 675000
-    },
-    {
-      "epoch": 10.32,
-      "learning_rate": 4.6480279366995116e-05,
-      "loss": 0.2472,
-      "step": 676000
-    },
-    {
-      "epoch": 10.34,
-      "learning_rate": 4.6277242812879914e-05,
-      "loss": 0.2473,
-      "step": 677000
-    },
-    {
-      "epoch": 10.35,
-      "learning_rate": 4.607457504543447e-05,
-      "loss": 0.2471,
-      "step": 678000
-    },
-    {
-      "epoch": 10.37,
-      "learning_rate": 4.5872278281000955e-05,
-      "loss": 0.2469,
-      "step": 679000
-    },
-    {
-      "epoch": 10.38,
-      "learning_rate": 4.567035473186444e-05,
-      "loss": 0.2469,
-      "step": 680000
-    },
-    {
-      "epoch": 10.38,
-      "eval_runtime": 0.7393,
-      "eval_samples_per_second": 1352.617,
-      "eval_steps_per_second": 21.642,
-      "step": 680000
-    },
-    {
-      "epoch": 10.4,
-      "learning_rate": 4.546880660622845e-05,
-      "loss": 0.2463,
-      "step": 681000
-    },
-    {
-      "epoch": 10.41,
-      "learning_rate": 4.5267636108191036e-05,
-      "loss": 0.2466,
-      "step": 682000
-    },
-    {
-      "epoch": 10.43,
-      "learning_rate": 4.5066845437720555e-05,
-      "loss": 0.2462,
-      "step": 683000
-    },
-    {
-      "epoch": 10.44,
-      "learning_rate": 4.4866436790631564e-05,
-      "loss": 0.2463,
-      "step": 684000
-    },
-    {
-      "epoch": 10.46,
-      "learning_rate": 4.4666412358560955e-05,
-      "loss": 0.2461,
-      "step": 685000
-    },
-    {
-      "epoch": 10.46,
-      "eval_runtime": 0.7931,
-      "eval_samples_per_second": 1260.883,
-      "eval_steps_per_second": 20.174,
-      "step": 685000
-    },
-    {
-      "epoch": 10.48,
-      "learning_rate": 4.4466774328943796e-05,
-      "loss": 0.2462,
-      "step": 686000
-    },
-    {
-      "epoch": 10.49,
-      "learning_rate": 4.426752488498972e-05,
-      "loss": 0.2462,
-      "step": 687000
-    },
-    {
-      "epoch": 10.51,
-      "learning_rate": 4.406866620565862e-05,
-      "loss": 0.2459,
-      "step": 688000
-    },
-    {
-      "epoch": 10.52,
-      "learning_rate": 4.3870200465637164e-05,
-      "loss": 0.2471,
-      "step": 689000
-    },
-    {
-      "epoch": 10.54,
-      "learning_rate": 4.3672129835314955e-05,
-      "loss": 0.2481,
-      "step": 690000
-    },
-    {
-      "epoch": 10.54,
-      "eval_runtime": 0.8338,
-      "eval_samples_per_second": 1199.26,
-      "eval_steps_per_second": 19.188,
-      "step": 690000
-    },
-    {
-      "epoch": 10.55,
-      "learning_rate": 4.347445648076057e-05,
-      "loss": 0.2463,
-      "step": 691000
-    },
-    {
-      "epoch": 10.57,
-      "learning_rate": 4.327718256369826e-05,
-      "loss": 0.2458,
-      "step": 692000
-    },
-    {
-      "epoch": 10.58,
-      "learning_rate": 4.3080310241483885e-05,
-      "loss": 0.2451,
-      "step": 693000
-    },
-    {
-      "epoch": 10.6,
-      "learning_rate": 4.2883841667081675e-05,
-      "loss": 0.2454,
-      "step": 694000
-    },
-    {
-      "epoch": 10.61,
-      "learning_rate": 4.268777898904044e-05,
-      "loss": 0.2455,
-      "step": 695000
-    },
-    {
-      "epoch": 10.61,
-      "eval_runtime": 0.794,
-      "eval_samples_per_second": 1259.505,
-      "eval_steps_per_second": 20.152,
-      "step": 695000
-    },
-    {
-      "epoch": 10.63,
-      "learning_rate": 4.2492124351470214e-05,
-      "loss": 0.2453,
-      "step": 696000
-    },
-    {
-      "epoch": 10.64,
-      "learning_rate": 4.2296879894018835e-05,
-      "loss": 0.2449,
-      "step": 697000
-    },
-    {
-      "epoch": 10.66,
-      "learning_rate": 4.210204775184834e-05,
-      "loss": 0.245,
-      "step": 698000
-    },
-    {
-      "epoch": 10.67,
-      "learning_rate": 4.190763005561186e-05,
-      "loss": 0.2447,
-      "step": 699000
-    },
-    {
-      "epoch": 10.69,
-      "learning_rate": 4.171362893143013e-05,
-      "loss": 0.2444,
-      "step": 700000
-    },
-    {
-      "epoch": 10.69,
-      "eval_runtime": 0.7798,
-      "eval_samples_per_second": 1282.444,
-      "eval_steps_per_second": 20.519,
-      "step": 700000
-    },
-    {
-      "epoch": 10.7,
-      "learning_rate": 4.1520046500868384e-05,
-      "loss": 0.2442,
-      "step": 701000
-    },
-    {
-      "epoch": 10.72,
-      "learning_rate": 4.1326884880913074e-05,
-      "loss": 0.2454,
-      "step": 702000
-    },
-    {
-      "epoch": 10.73,
-      "learning_rate": 4.1134146183948724e-05,
-      "loss": 0.2445,
-      "step": 703000
-    },
-    {
-      "epoch": 10.75,
-      "learning_rate": 4.0941832517734885e-05,
-      "loss": 0.2448,
-      "step": 704000
-    },
-    {
-      "epoch": 10.77,
-      "learning_rate": 4.0749945985382915e-05,
-      "loss": 0.2445,
-      "step": 705000
-    },
-    {
-      "epoch": 10.77,
-      "eval_runtime": 0.7458,
-      "eval_samples_per_second": 1340.853,
-      "eval_steps_per_second": 21.454,
-      "step": 705000
-    },
-    {
-      "epoch": 10.78,
-      "learning_rate": 4.0558488685333235e-05,
-      "loss": 0.253,
-      "step": 706000
-    },
-    {
-      "epoch": 10.8,
-      "learning_rate": 4.036746271133223e-05,
-      "loss": 0.2533,
-      "step": 707000
-    },
-    {
-      "epoch": 10.81,
-      "learning_rate": 4.0176870152409324e-05,
-      "loss": 0.2547,
-      "step": 708000
-    },
-    {
-      "epoch": 10.83,
-      "learning_rate": 3.998671309285417e-05,
-      "loss": 0.2529,
-      "step": 709000
-    },
-    {
-      "epoch": 10.84,
-      "learning_rate": 3.979699361219395e-05,
-      "loss": 0.2457,
-      "step": 710000
-    },
-    {
-      "epoch": 10.84,
-      "eval_runtime": 0.7472,
-      "eval_samples_per_second": 1338.326,
-      "eval_steps_per_second": 21.413,
-      "step": 710000
-    },
-    {
-      "epoch": 10.86,
-      "learning_rate": 3.960771378517049e-05,
-      "loss": 0.2438,
-      "step": 711000
-    },
-    {
-      "epoch": 10.87,
-      "learning_rate": 3.941887568171766e-05,
-      "loss": 0.2464,
-      "step": 712000
-    },
-    {
-      "epoch": 10.89,
-      "learning_rate": 3.923048136693873e-05,
-      "loss": 0.2445,
-      "step": 713000
-    },
-    {
-      "epoch": 10.9,
-      "learning_rate": 3.904253290108369e-05,
-      "loss": 0.2435,
-      "step": 714000
-    },
-    {
-      "epoch": 10.92,
-      "learning_rate": 3.885503233952689e-05,
-      "loss": 0.2446,
-      "step": 715000
-    },
-    {
-      "epoch": 10.92,
-      "eval_runtime": 0.8432,
-      "eval_samples_per_second": 1186.017,
-      "eval_steps_per_second": 18.976,
-      "step": 715000
-    },
-    {
-      "epoch": 10.93,
-      "learning_rate": 3.86679817327444e-05,
-      "loss": 0.2432,
-      "step": 716000
-    },
-    {
-      "epoch": 10.95,
-      "learning_rate": 3.848138312629171e-05,
-      "loss": 0.2433,
-      "step": 717000
-    },
-    {
-      "epoch": 10.96,
-      "learning_rate": 3.8295238560781317e-05,
-      "loss": 0.2436,
-      "step": 718000
-    },
-    {
-      "epoch": 10.98,
-      "learning_rate": 3.810955007186029e-05,
-      "loss": 0.2433,
-      "step": 719000
-    },
-    {
-      "epoch": 10.99,
-      "learning_rate": 3.792431969018824e-05,
-      "loss": 0.243,
-      "step": 720000
-    },
-    {
-      "epoch": 10.99,
-      "eval_runtime": 0.7755,
-      "eval_samples_per_second": 1289.466,
-      "eval_steps_per_second": 20.631,
-      "step": 720000
-    },
-    {
-      "epoch": 11.01,
-      "learning_rate": 3.7739549441414945e-05,
-      "loss": 0.2427,
-      "step": 721000
-    },
-    {
-      "epoch": 11.03,
-      "learning_rate": 3.755524134615825e-05,
-      "loss": 0.2429,
-      "step": 722000
-    },
-    {
-      "epoch": 11.04,
-      "learning_rate": 3.7371397419981925e-05,
-      "loss": 0.2428,
-      "step": 723000
-    },
-    {
-      "epoch": 11.06,
-      "learning_rate": 3.7188019673373706e-05,
-      "loss": 0.2431,
-      "step": 724000
-    },
-    {
-      "epoch": 11.07,
-      "learning_rate": 3.700511011172325e-05,
-      "loss": 0.2436,
-      "step": 725000
-    },
-    {
-      "epoch": 11.07,
-      "eval_runtime": 0.7297,
-      "eval_samples_per_second": 1370.472,
-      "eval_steps_per_second": 21.928,
-      "step": 725000
-    },
-    {
-      "epoch": 11.09,
-      "learning_rate": 3.682267073530023e-05,
-      "loss": 0.243,
-      "step": 726000
-    },
-    {
-      "epoch": 11.1,
-      "learning_rate": 3.664070353923245e-05,
-      "loss": 0.2424,
-      "step": 727000
-    },
-    {
-      "epoch": 11.12,
-      "learning_rate": 3.645921051348396e-05,
-      "loss": 0.2423,
-      "step": 728000
-    },
-    {
-      "epoch": 11.13,
-      "learning_rate": 3.627819364283345e-05,
-      "loss": 0.2456,
-      "step": 729000
-    },
-    {
-      "epoch": 11.15,
-      "learning_rate": 3.6097654906852405e-05,
-      "loss": 0.2431,
-      "step": 730000
-    },
-    {
-      "epoch": 11.15,
-      "eval_runtime": 0.7906,
-      "eval_samples_per_second": 1264.795,
-      "eval_steps_per_second": 20.237,
-      "step": 730000
-    },
-    {
-      "epoch": 11.16,
-      "learning_rate": 3.591759627988353e-05,
-      "loss": 0.242,
-      "step": 731000
-    },
-    {
-      "epoch": 11.18,
-      "learning_rate": 3.573801973101913e-05,
-      "loss": 0.2418,
-      "step": 732000
-    },
-    {
-      "epoch": 11.19,
-      "learning_rate": 3.5558927224079534e-05,
-      "loss": 0.2418,
-      "step": 733000
-    },
-    {
-      "epoch": 11.21,
-      "learning_rate": 3.5380320717591716e-05,
-      "loss": 0.2419,
-      "step": 734000
-    },
-    {
-      "epoch": 11.22,
-      "learning_rate": 3.5202202164767836e-05,
-      "loss": 0.2418,
-      "step": 735000
-    },
-    {
-      "epoch": 11.22,
-      "eval_runtime": 0.8971,
-      "eval_samples_per_second": 1114.723,
-      "eval_steps_per_second": 17.836,
-      "step": 735000
-    },
-    {
-      "epoch": 11.24,
-      "learning_rate": 3.5024573513483864e-05,
-      "loss": 0.2415,
-      "step": 736000
-    },
-    {
-      "epoch": 11.25,
-      "learning_rate": 3.484743670625822e-05,
-      "loss": 0.2414,
-      "step": 737000
-    },
-    {
-      "epoch": 11.27,
-      "learning_rate": 3.467079368023068e-05,
-      "loss": 0.2413,
-      "step": 738000
-    },
-    {
-      "epoch": 11.28,
-      "learning_rate": 3.449464636714107e-05,
-      "loss": 0.2415,
-      "step": 739000
-    },
-    {
-      "epoch": 11.3,
-      "learning_rate": 3.431899669330819e-05,
-      "loss": 0.2414,
-      "step": 740000
-    },
-    {
-      "epoch": 11.3,
-      "eval_runtime": 0.7754,
-      "eval_samples_per_second": 1289.598,
-      "eval_steps_per_second": 20.634,
-      "step": 740000
-    },
-    {
-      "epoch": 11.32,
-      "learning_rate": 3.4143846579608744e-05,
-      "loss": 0.2411,
-      "step": 741000
-    },
-    {
-      "epoch": 11.33,
-      "learning_rate": 3.396919794145629e-05,
-      "loss": 0.2412,
-      "step": 742000
-    },
-    {
-      "epoch": 11.35,
-      "learning_rate": 3.3795052688780345e-05,
-      "loss": 0.241,
-      "step": 743000
-    },
-    {
-      "epoch": 11.36,
-      "learning_rate": 3.362141272600552e-05,
-      "loss": 0.2413,
-      "step": 744000
-    },
-    {
-      "epoch": 11.38,
-      "learning_rate": 3.3448279952030615e-05,
-      "loss": 0.241,
-      "step": 745000
-    },
-    {
-      "epoch": 11.38,
-      "eval_runtime": 0.937,
-      "eval_samples_per_second": 1067.221,
-      "eval_steps_per_second": 17.076,
-      "step": 745000
-    },
-    {
-      "epoch": 11.39,
-      "learning_rate": 3.327565626020793e-05,
-      "loss": 0.2408,
-      "step": 746000
-    },
-    {
-      "epoch": 11.41,
-      "learning_rate": 3.3103543538322455e-05,
-      "loss": 0.2408,
-      "step": 747000
-    },
-    {
-      "epoch": 11.42,
-      "learning_rate": 3.293194366857137e-05,
-      "loss": 0.2407,
-      "step": 748000
-    },
-    {
-      "epoch": 11.44,
-      "learning_rate": 3.276085852754336e-05,
-      "loss": 0.2409,
-      "step": 749000
-    },
-    {
-      "epoch": 11.45,
-      "learning_rate": 3.259028998619814e-05,
-      "loss": 0.2405,
-      "step": 750000
-    },
-    {
-      "epoch": 11.45,
-      "eval_runtime": 0.7243,
-      "eval_samples_per_second": 1380.717,
-      "eval_steps_per_second": 22.091,
-      "step": 750000
-    },
-    {
-      "epoch": 11.47,
-      "learning_rate": 3.2420239909845894e-05,
-      "loss": 0.2403,
-      "step": 751000
-    },
-    {
-      "epoch": 11.48,
-      "learning_rate": 3.2250710158127045e-05,
-      "loss": 0.2402,
-      "step": 752000
-    },
-    {
-      "epoch": 11.5,
-      "learning_rate": 3.2081702584991786e-05,
-      "loss": 0.2398,
-      "step": 753000
-    },
-    {
-      "epoch": 11.51,
-      "learning_rate": 3.191321903867988e-05,
-      "loss": 0.2401,
-      "step": 754000
-    },
-    {
-      "epoch": 11.53,
-      "learning_rate": 3.174526136170039e-05,
-      "loss": 0.2403,
-      "step": 755000
-    },
-    {
-      "epoch": 11.53,
-      "eval_runtime": 0.695,
-      "eval_samples_per_second": 1438.835,
-      "eval_steps_per_second": 23.021,
-      "step": 755000
-    },
-    {
-      "epoch": 11.54,
-      "learning_rate": 3.157783139081155e-05,
-      "loss": 0.24,
-      "step": 756000
-    },
-    {
-      "epoch": 11.56,
-      "learning_rate": 3.141093095700072e-05,
-      "loss": 0.2401,
-      "step": 757000
-    },
-    {
-      "epoch": 11.57,
-      "learning_rate": 3.1244561885464244e-05,
-      "loss": 0.252,
-      "step": 758000
-    },
-    {
-      "epoch": 11.59,
-      "learning_rate": 3.107872599558769e-05,
-      "loss": 0.24,
-      "step": 759000
-    },
-    {
-      "epoch": 11.61,
-      "learning_rate": 3.0913425100925795e-05,
-      "loss": 0.2396,
-      "step": 760000
-    },
-    {
-      "epoch": 11.61,
-      "eval_runtime": 0.7192,
-      "eval_samples_per_second": 1390.499,
-      "eval_steps_per_second": 22.248,
-      "step": 760000
-    },
-    {
-      "epoch": 11.62,
-      "learning_rate": 3.0748661009182616e-05,
-      "loss": 0.2396,
-      "step": 761000
-    },
-    {
-      "epoch": 11.64,
-      "learning_rate": 3.0584435522191896e-05,
-      "loss": 0.2395,
-      "step": 762000
-    },
-    {
-      "epoch": 11.65,
-      "learning_rate": 3.0420750435897183e-05,
-      "loss": 0.2393,
-      "step": 763000
-    },
-    {
-      "epoch": 11.67,
-      "learning_rate": 3.025760754033246e-05,
-      "loss": 0.239,
-      "step": 764000
-    },
-    {
-      "epoch": 11.68,
-      "learning_rate": 3.0095008619602206e-05,
-      "loss": 0.2392,
-      "step": 765000
-    },
-    {
-      "epoch": 11.68,
-      "eval_runtime": 0.7905,
-      "eval_samples_per_second": 1264.968,
-      "eval_steps_per_second": 20.239,
-      "step": 765000
-    },
-    {
-      "epoch": 11.7,
-      "learning_rate": 2.993295545186223e-05,
-      "loss": 0.2393,
-      "step": 766000
-    },
-    {
-      "epoch": 11.71,
-      "learning_rate": 2.977144980929996e-05,
-      "loss": 0.2392,
-      "step": 767000
-    },
-    {
-      "epoch": 11.73,
-      "learning_rate": 2.961049345811523e-05,
-      "loss": 0.2388,
-      "step": 768000
-    },
-    {
-      "epoch": 11.74,
-      "learning_rate": 2.945008815850097e-05,
-      "loss": 0.2392,
-      "step": 769000
-    },
-    {
-      "epoch": 11.76,
-      "learning_rate": 2.929023566462377e-05,
-      "loss": 0.2391,
-      "step": 770000
-    },
-    {
-      "epoch": 11.76,
-      "eval_runtime": 0.8418,
-      "eval_samples_per_second": 1187.898,
-      "eval_steps_per_second": 19.006,
-      "step": 770000
-    },
-    {
-      "epoch": 11.77,
-      "learning_rate": 2.9130937724604947e-05,
-      "loss": 0.2401,
-      "step": 771000
-    },
-    {
-      "epoch": 11.79,
-      "learning_rate": 2.8972196080501208e-05,
-      "loss": 0.2392,
-      "step": 772000
-    },
-    {
-      "epoch": 11.8,
-      "learning_rate": 2.8814012468285748e-05,
-      "loss": 0.2395,
-      "step": 773000
-    },
-    {
-      "epoch": 11.82,
-      "learning_rate": 2.865638861782922e-05,
-      "loss": 0.2387,
-      "step": 774000
-    },
-    {
-      "epoch": 11.83,
-      "learning_rate": 2.849932625288079e-05,
-      "loss": 0.2383,
-      "step": 775000
-    },
-    {
-      "epoch": 11.83,
-      "eval_runtime": 0.7561,
-      "eval_samples_per_second": 1322.637,
-      "eval_steps_per_second": 21.162,
-      "step": 775000
-    },
-    {
-      "epoch": 11.85,
-      "learning_rate": 2.8342827091049336e-05,
-      "loss": 0.2383,
-      "step": 776000
-    },
-    {
-      "epoch": 11.86,
-      "learning_rate": 2.8186892843784587e-05,
-      "loss": 0.2384,
-      "step": 777000
-    },
-    {
-      "epoch": 11.88,
-      "learning_rate": 2.803152521635851e-05,
-      "loss": 0.2382,
-      "step": 778000
-    },
-    {
-      "epoch": 11.9,
-      "learning_rate": 2.7876725907846578e-05,
-      "loss": 0.2378,
-      "step": 779000
-    },
-    {
-      "epoch": 11.91,
-      "learning_rate": 2.7722496611109243e-05,
-      "loss": 0.2378,
-      "step": 780000
-    },
-    {
-      "epoch": 11.91,
-      "eval_runtime": 0.7835,
-      "eval_samples_per_second": 1276.363,
-      "eval_steps_per_second": 20.422,
-      "step": 780000
-    },
-    {
-      "epoch": 11.93,
-      "learning_rate": 2.7568839012773365e-05,
-      "loss": 0.238,
-      "step": 781000
-    },
-    {
-      "epoch": 11.94,
-      "learning_rate": 2.7415754793213826e-05,
-      "loss": 0.2375,
-      "step": 782000
-    },
-    {
-      "epoch": 11.96,
-      "learning_rate": 2.7263245626535116e-05,
-      "loss": 0.2377,
-      "step": 783000
-    },
-    {
-      "epoch": 11.97,
-      "learning_rate": 2.7111313180553077e-05,
-      "loss": 0.2378,
-      "step": 784000
-    },
-    {
-      "epoch": 11.99,
-      "learning_rate": 2.6959959116776587e-05,
-      "loss": 0.2376,
-      "step": 785000
-    },
-    {
-      "epoch": 11.99,
-      "eval_runtime": 0.7664,
-      "eval_samples_per_second": 1304.853,
-      "eval_steps_per_second": 20.878,
-      "step": 785000
-    },
-    {
-      "epoch": 12.0,
-      "learning_rate": 2.6809185090389406e-05,
-      "loss": 0.2371,
-      "step": 786000
-    },
-    {
-      "epoch": 12.02,
-      "learning_rate": 2.6658992750232167e-05,
-      "loss": 0.2373,
-      "step": 787000
-    },
-    {
-      "epoch": 12.03,
-      "learning_rate": 2.6509383738784218e-05,
-      "loss": 0.2374,
-      "step": 788000
-    },
-    {
-      "epoch": 12.05,
-      "learning_rate": 2.6360359692145757e-05,
-      "loss": 0.237,
-      "step": 789000
-    },
-    {
-      "epoch": 12.06,
-      "learning_rate": 2.6211922240019883e-05,
-      "loss": 0.2368,
-      "step": 790000
-    },
-    {
-      "epoch": 12.06,
-      "eval_runtime": 0.7543,
-      "eval_samples_per_second": 1325.719,
-      "eval_steps_per_second": 21.212,
-      "step": 790000
-    },
-    {
-      "epoch": 12.08,
-      "learning_rate": 2.6064073005694758e-05,
-      "loss": 0.2381,
-      "step": 791000
-    },
-    {
-      "epoch": 12.09,
-      "learning_rate": 2.591681360602595e-05,
-      "loss": 0.2373,
-      "step": 792000
-    },
-    {
-      "epoch": 12.11,
-      "learning_rate": 2.577014565141866e-05,
-      "loss": 0.2377,
-      "step": 793000
-    },
-    {
-      "epoch": 12.12,
-      "learning_rate": 2.562407074581014e-05,
-      "loss": 0.2382,
-      "step": 794000
-    },
-    {
-      "epoch": 12.14,
-      "learning_rate": 2.5478590486652137e-05,
-      "loss": 0.2374,
-      "step": 795000
-    },
-    {
-      "epoch": 12.14,
-      "eval_runtime": 0.8227,
-      "eval_samples_per_second": 1215.581,
-      "eval_steps_per_second": 19.449,
-      "step": 795000
-    },
-    {
-      "epoch": 12.16,
-      "learning_rate": 2.533370646489347e-05,
-      "loss": 0.237,
-      "step": 796000
-    },
-    {
-      "epoch": 12.17,
-      "learning_rate": 2.5189420264962586e-05,
-      "loss": 0.2367,
-      "step": 797000
-    },
-    {
-      "epoch": 12.19,
-      "learning_rate": 2.504573346475026e-05,
-      "loss": 0.2371,
-      "step": 798000
-    },
-    {
-      "epoch": 12.2,
-      "learning_rate": 2.4902647635592324e-05,
-      "loss": 0.2372,
-      "step": 799000
-    },
-    {
-      "epoch": 12.22,
-      "learning_rate": 2.476016434225246e-05,
-      "loss": 0.2372,
-      "step": 800000
-    },
-    {
-      "epoch": 12.22,
-      "eval_runtime": 0.741,
-      "eval_samples_per_second": 1349.61,
-      "eval_steps_per_second": 21.594,
-      "step": 800000
-    },
-    {
-      "epoch": 12.23,
-      "learning_rate": 2.461828514290513e-05,
-      "loss": 0.2364,
-      "step": 801000
-    },
-    {
-      "epoch": 12.25,
-      "learning_rate": 2.447701158911855e-05,
-      "loss": 0.2373,
-      "step": 802000
-    },
-    {
-      "epoch": 12.26,
-      "learning_rate": 2.4336345225837658e-05,
-      "loss": 0.2369,
-      "step": 803000
-    },
-    {
-      "epoch": 12.28,
-      "learning_rate": 2.4196287591367296e-05,
-      "loss": 0.2363,
-      "step": 804000
-    },
-    {
-      "epoch": 12.29,
-      "learning_rate": 2.405684021735527e-05,
-      "loss": 0.2366,
-      "step": 805000
-    },
-    {
-      "epoch": 12.29,
-      "eval_runtime": 0.7797,
-      "eval_samples_per_second": 1282.575,
-      "eval_steps_per_second": 20.521,
-      "step": 805000
-    },
-    {
-      "epoch": 12.31,
-      "learning_rate": 2.3918004628775736e-05,
-      "loss": 0.2366,
-      "step": 806000
-    },
-    {
-      "epoch": 12.32,
-      "learning_rate": 2.3779782343912463e-05,
-      "loss": 0.2367,
-      "step": 807000
-    },
-    {
-      "epoch": 12.34,
-      "learning_rate": 2.364217487434221e-05,
-      "loss": 0.24,
-      "step": 808000
-    },
-    {
-      "epoch": 12.35,
-      "learning_rate": 2.3505183724918196e-05,
-      "loss": 0.2369,
-      "step": 809000
-    },
-    {
-      "epoch": 12.37,
-      "learning_rate": 2.3368810393753687e-05,
-      "loss": 0.2365,
-      "step": 810000
-    },
-    {
-      "epoch": 12.37,
-      "eval_runtime": 0.7457,
-      "eval_samples_per_second": 1341.078,
-      "eval_steps_per_second": 21.457,
-      "step": 810000
-    },
-    {
-      "epoch": 12.38,
-      "learning_rate": 2.32330563722056e-05,
-      "loss": 0.2357,
-      "step": 811000
-    },
-    {
-      "epoch": 12.4,
-      "learning_rate": 2.309792314485815e-05,
-      "loss": 0.2356,
-      "step": 812000
-    },
-    {
-      "epoch": 12.41,
-      "learning_rate": 2.2963412189506695e-05,
-      "loss": 0.2358,
-      "step": 813000
-    },
-    {
-      "epoch": 12.43,
-      "learning_rate": 2.282952497714145e-05,
-      "loss": 0.2356,
-      "step": 814000
-    },
-    {
-      "epoch": 12.45,
-      "learning_rate": 2.2696262971931538e-05,
-      "loss": 0.2357,
-      "step": 815000
-    },
-    {
-      "epoch": 12.45,
-      "eval_runtime": 0.7163,
-      "eval_samples_per_second": 1396.105,
-      "eval_steps_per_second": 22.338,
-      "step": 815000
-    },
-    {
-      "epoch": 12.46,
-      "learning_rate": 2.2563627631208887e-05,
-      "loss": 0.2355,
-      "step": 816000
-    },
-    {
-      "epoch": 12.48,
-      "learning_rate": 2.2431620405452336e-05,
-      "loss": 0.2351,
-      "step": 817000
-    },
-    {
-      "epoch": 12.49,
-      "learning_rate": 2.230024273827179e-05,
-      "loss": 0.2357,
-      "step": 818000
-    },
-    {
-      "epoch": 12.51,
-      "learning_rate": 2.216949606639231e-05,
-      "loss": 0.2353,
-      "step": 819000
-    },
-    {
-      "epoch": 12.52,
-      "learning_rate": 2.2039381819638596e-05,
-      "loss": 0.2351,
-      "step": 820000
-    },
-    {
-      "epoch": 12.52,
-      "eval_runtime": 0.6211,
-      "eval_samples_per_second": 1609.99,
-      "eval_steps_per_second": 25.76,
-      "step": 820000
-    },
-    {
-      "epoch": 12.54,
-      "learning_rate": 2.1909901420919184e-05,
-      "loss": 0.2351,
-      "step": 821000
-    },
-    {
-      "epoch": 12.55,
-      "learning_rate": 2.1781056286210997e-05,
-      "loss": 0.235,
-      "step": 822000
-    },
-    {
-      "epoch": 12.57,
-      "learning_rate": 2.1652847824543744e-05,
-      "loss": 0.2347,
-      "step": 823000
-    },
-    {
-      "epoch": 12.58,
-      "learning_rate": 2.1525277437984636e-05,
-      "loss": 0.2348,
-      "step": 824000
-    },
-    {
-      "epoch": 12.6,
-      "learning_rate": 2.1398346521623e-05,
-      "loss": 0.2345,
-      "step": 825000
-    },
-    {
-      "epoch": 12.6,
-      "eval_runtime": 0.7605,
-      "eval_samples_per_second": 1314.934,
-      "eval_steps_per_second": 21.039,
-      "step": 825000
-    },
-    {
-      "epoch": 12.61,
-      "learning_rate": 2.1272056463554978e-05,
-      "loss": 0.2343,
-      "step": 826000
-    },
-    {
-      "epoch": 12.63,
-      "learning_rate": 2.114640864486845e-05,
-      "loss": 0.2346,
-      "step": 827000
-    },
-    {
-      "epoch": 12.64,
-      "learning_rate": 2.1021404439627775e-05,
-      "loss": 0.2344,
-      "step": 828000
-    },
-    {
-      "epoch": 12.66,
-      "learning_rate": 2.089704521485896e-05,
-      "loss": 0.2344,
-      "step": 829000
-    },
-    {
-      "epoch": 12.67,
-      "learning_rate": 2.0773332330534513e-05,
-      "loss": 0.2343,
-      "step": 830000
-    },
-    {
-      "epoch": 12.67,
-      "eval_runtime": 0.7327,
-      "eval_samples_per_second": 1364.889,
-      "eval_steps_per_second": 21.838,
-      "step": 830000
-    },
-    {
-      "epoch": 12.69,
-      "learning_rate": 2.0650267139558772e-05,
-      "loss": 0.2339,
-      "step": 831000
-    },
-    {
-      "epoch": 12.7,
-      "learning_rate": 2.052785098775293e-05,
-      "loss": 0.2339,
-      "step": 832000
-    },
-    {
-      "epoch": 12.72,
-      "learning_rate": 2.04060852138404e-05,
-      "loss": 0.234,
-      "step": 833000
-    },
-    {
-      "epoch": 12.74,
-      "learning_rate": 2.028497114943219e-05,
-      "loss": 0.234,
-      "step": 834000
-    },
-    {
-      "epoch": 12.75,
-      "learning_rate": 2.0164510119012263e-05,
-      "loss": 0.2338,
-      "step": 835000
-    },
-    {
-      "epoch": 12.75,
-      "eval_runtime": 0.7099,
-      "eval_samples_per_second": 1408.578,
-      "eval_steps_per_second": 22.537,
-      "step": 835000
-    },
-    {
-      "epoch": 12.77,
-      "learning_rate": 2.0044703439923217e-05,
-      "loss": 0.2336,
-      "step": 836000
-    },
-    {
-      "epoch": 12.78,
-      "learning_rate": 1.9925552422351654e-05,
-      "loss": 0.2338,
-      "step": 837000
-    },
-    {
-      "epoch": 12.8,
-      "learning_rate": 1.9807058369314016e-05,
-      "loss": 0.2335,
-      "step": 838000
-    },
-    {
-      "epoch": 12.81,
-      "learning_rate": 1.968922257664231e-05,
-      "loss": 0.2337,
-      "step": 839000
-    },
-    {
-      "epoch": 12.83,
-      "learning_rate": 1.9572046332969825e-05,
-      "loss": 0.2335,
-      "step": 840000
-    },
-    {
-      "epoch": 12.83,
-      "eval_runtime": 0.7491,
-      "eval_samples_per_second": 1334.897,
-      "eval_steps_per_second": 21.358,
-      "step": 840000
-    },
-    {
-      "epoch": 12.84,
-      "learning_rate": 1.945553091971727e-05,
-      "loss": 0.2334,
-      "step": 841000
-    },
-    {
-      "epoch": 12.86,
-      "learning_rate": 1.933967761107847e-05,
-      "loss": 0.234,
-      "step": 842000
-    },
-    {
-      "epoch": 12.87,
-      "learning_rate": 1.9224487674006694e-05,
-      "loss": 0.234,
-      "step": 843000
-    },
-    {
-      "epoch": 12.89,
-      "learning_rate": 1.9109962368200602e-05,
-      "loss": 0.2379,
-      "step": 844000
-    },
-    {
-      "epoch": 12.9,
-      "learning_rate": 1.8996102946090586e-05,
-      "loss": 0.2335,
-      "step": 845000
-    },
-    {
-      "epoch": 12.9,
-      "eval_runtime": 0.7039,
-      "eval_samples_per_second": 1420.612,
-      "eval_steps_per_second": 22.73,
-      "step": 845000
-    },
-    {
-      "epoch": 12.92,
-      "learning_rate": 1.888291065282509e-05,
-      "loss": 0.2338,
-      "step": 846000
-    },
-    {
-      "epoch": 12.93,
-      "learning_rate": 1.8770386726256865e-05,
-      "loss": 0.2329,
-      "step": 847000
-    },
-    {
-      "epoch": 12.95,
-      "learning_rate": 1.8658532396929565e-05,
-      "loss": 0.2334,
-      "step": 848000
-    },
-    {
-      "epoch": 12.96,
-      "learning_rate": 1.8547348888064178e-05,
-      "loss": 0.2341,
-      "step": 849000
-    },
-    {
-      "epoch": 12.98,
-      "learning_rate": 1.8436837415545772e-05,
-      "loss": 0.2356,
-      "step": 850000
-    },
-    {
-      "epoch": 12.98,
-      "eval_runtime": 0.8308,
-      "eval_samples_per_second": 1203.685,
-      "eval_steps_per_second": 19.259,
-      "step": 850000
-    },
-    {
-      "epoch": 12.99,
-      "learning_rate": 1.8326999187910095e-05,
-      "loss": 0.2342,
-      "step": 851000
-    },
-    {
-      "epoch": 13.01,
-      "learning_rate": 1.8217835406330415e-05,
-      "loss": 0.2344,
-      "step": 852000
-    },
-    {
-      "epoch": 13.03,
-      "learning_rate": 1.810934726460436e-05,
-      "loss": 0.2328,
-      "step": 853000
-    },
-    {
-      "epoch": 13.04,
-      "learning_rate": 1.800153594914084e-05,
-      "loss": 0.2326,
-      "step": 854000
-    },
-    {
-      "epoch": 13.06,
-      "learning_rate": 1.7894402638947176e-05,
-      "loss": 0.2325,
-      "step": 855000
-    },
-    {
-      "epoch": 13.06,
-      "eval_runtime": 0.7234,
-      "eval_samples_per_second": 1382.419,
-      "eval_steps_per_second": 22.119,
-      "step": 855000
-    },
-    {
-      "epoch": 13.07,
-      "learning_rate": 1.778794850561604e-05,
-      "loss": 0.2327,
-      "step": 856000
-    },
-    {
-      "epoch": 13.09,
-      "learning_rate": 1.7682174713312805e-05,
-      "loss": 0.2326,
-      "step": 857000
-    },
-    {
-      "epoch": 13.1,
-      "learning_rate": 1.75770824187627e-05,
-      "loss": 0.2325,
-      "step": 858000
-    },
-    {
-      "epoch": 13.12,
-      "learning_rate": 1.747267277123821e-05,
-      "loss": 0.2327,
-      "step": 859000
-    },
-    {
-      "epoch": 13.13,
-      "learning_rate": 1.7368946912546556e-05,
-      "loss": 0.2329,
-      "step": 860000
-    },
-    {
-      "epoch": 13.13,
-      "eval_runtime": 0.7568,
-      "eval_samples_per_second": 1321.327,
-      "eval_steps_per_second": 21.141,
-      "step": 860000
-    },
-    {
-      "epoch": 13.15,
-      "learning_rate": 1.726590597701708e-05,
-      "loss": 0.2322,
-      "step": 861000
-    },
-    {
-      "epoch": 13.16,
-      "learning_rate": 1.7163551091488952e-05,
-      "loss": 0.2375,
-      "step": 862000
-    },
-    {
-      "epoch": 13.18,
-      "learning_rate": 1.7061883375298788e-05,
-      "loss": 0.2328,
-      "step": 863000
-    },
-    {
-      "epoch": 13.19,
-      "learning_rate": 1.6960903940268456e-05,
-      "loss": 0.2323,
-      "step": 864000
-    },
-    {
-      "epoch": 13.21,
-      "learning_rate": 1.6860613890692876e-05,
-      "loss": 0.2334,
-      "step": 865000
-    },
-    {
-      "epoch": 13.21,
-      "eval_runtime": 0.7389,
-      "eval_samples_per_second": 1353.416,
-      "eval_steps_per_second": 21.655,
-      "step": 865000
-    },
-    {
-      "epoch": 13.22,
-      "learning_rate": 1.6761014323327962e-05,
-      "loss": 0.233,
-      "step": 866000
-    },
-    {
-      "epoch": 13.24,
-      "learning_rate": 1.6662106327378645e-05,
-      "loss": 0.2334,
-      "step": 867000
-    },
-    {
-      "epoch": 13.25,
-      "learning_rate": 1.6563890984486884e-05,
-      "loss": 0.2333,
-      "step": 868000
-    },
-    {
-      "epoch": 13.27,
-      "learning_rate": 1.6466369368719955e-05,
-      "loss": 0.2324,
-      "step": 869000
-    },
-    {
-      "epoch": 13.29,
-      "learning_rate": 1.6369542546558626e-05,
-      "loss": 0.2324,
-      "step": 870000
-    },
-    {
-      "epoch": 13.29,
-      "eval_runtime": 0.8823,
-      "eval_samples_per_second": 1133.455,
-      "eval_steps_per_second": 18.135,
-      "step": 870000
-    },
-    {
-      "epoch": 13.3,
-      "learning_rate": 1.6273411576885517e-05,
-      "loss": 0.2323,
-      "step": 871000
-    },
-    {
-      "epoch": 13.32,
-      "learning_rate": 1.617797751097349e-05,
-      "loss": 0.2322,
-      "step": 872000
-    },
-    {
-      "epoch": 13.33,
-      "learning_rate": 1.608324139247421e-05,
-      "loss": 0.2335,
-      "step": 873000
-    },
-    {
-      "epoch": 13.35,
-      "learning_rate": 1.5989204257406693e-05,
-      "loss": 0.2329,
-      "step": 874000
-    },
-    {
-      "epoch": 13.36,
-      "learning_rate": 1.5895867134145974e-05,
-      "loss": 0.2325,
-      "step": 875000
-    },
-    {
-      "epoch": 13.36,
-      "eval_runtime": 0.8114,
-      "eval_samples_per_second": 1232.442,
-      "eval_steps_per_second": 19.719,
-      "step": 875000
-    },
-    {
-      "epoch": 13.38,
-      "learning_rate": 1.5803231043411912e-05,
-      "loss": 0.2322,
-      "step": 876000
-    },
-    {
-      "epoch": 13.39,
-      "learning_rate": 1.5711296998257902e-05,
-      "loss": 0.232,
-      "step": 877000
-    },
-    {
-      "epoch": 13.41,
-      "learning_rate": 1.562006600405996e-05,
-      "loss": 0.2322,
-      "step": 878000
-    },
-    {
-      "epoch": 13.42,
-      "learning_rate": 1.5529539058505624e-05,
-      "loss": 0.2317,
-      "step": 879000
-    },
-    {
-      "epoch": 13.44,
-      "learning_rate": 1.543971715158307e-05,
-      "loss": 0.2318,
-      "step": 880000
-    },
-    {
-      "epoch": 13.44,
-      "eval_runtime": 0.8591,
-      "eval_samples_per_second": 1164.009,
-      "eval_steps_per_second": 18.624,
-      "step": 880000
-    },
-    {
-      "epoch": 13.45,
-      "learning_rate": 1.535060126557028e-05,
-      "loss": 0.2322,
-      "step": 881000
-    },
-    {
-      "epoch": 13.47,
-      "learning_rate": 1.5262192375024284e-05,
-      "loss": 0.232,
-      "step": 882000
-    },
-    {
-      "epoch": 13.48,
-      "learning_rate": 1.5174491446770566e-05,
-      "loss": 0.2314,
-      "step": 883000
-    },
-    {
-      "epoch": 13.5,
-      "learning_rate": 1.508749943989242e-05,
-      "loss": 0.2312,
-      "step": 884000
-    },
-    {
-      "epoch": 13.51,
-      "learning_rate": 1.500121730572051e-05,
-      "loss": 0.2314,
-      "step": 885000
-    },
-    {
-      "epoch": 13.51,
-      "eval_runtime": 0.7508,
-      "eval_samples_per_second": 1331.906,
-      "eval_steps_per_second": 21.31,
-      "step": 885000
-    },
-    {
-      "epoch": 13.53,
-      "learning_rate": 1.4915645987822406e-05,
-      "loss": 0.2314,
-      "step": 886000
-    },
-    {
-      "epoch": 13.54,
-      "learning_rate": 1.4830786421992347e-05,
-      "loss": 0.2316,
-      "step": 887000
-    },
-    {
-      "epoch": 13.56,
-      "learning_rate": 1.4746639536240942e-05,
-      "loss": 0.2312,
-      "step": 888000
-    },
-    {
-      "epoch": 13.58,
-      "learning_rate": 1.4663206250785055e-05,
-      "loss": 0.2315,
-      "step": 889000
-    },
-    {
-      "epoch": 13.59,
-      "learning_rate": 1.4580487478037748e-05,
-      "loss": 0.2311,
-      "step": 890000
-    },
-    {
-      "epoch": 13.59,
-      "eval_runtime": 0.7331,
-      "eval_samples_per_second": 1364.004,
-      "eval_steps_per_second": 21.824,
-      "step": 890000
-    },
-    {
-      "epoch": 13.61,
-      "learning_rate": 1.4498484122598232e-05,
-      "loss": 0.2308,
-      "step": 891000
-    },
-    {
-      "epoch": 13.62,
-      "learning_rate": 1.4417197081242083e-05,
-      "loss": 0.2305,
-      "step": 892000
-    },
-    {
-      "epoch": 13.64,
-      "learning_rate": 1.433662724291136e-05,
-      "loss": 0.2313,
-      "step": 893000
-    },
-    {
-      "epoch": 13.65,
-      "learning_rate": 1.4256775488704904e-05,
-      "loss": 0.2311,
-      "step": 894000
-    },
-    {
-      "epoch": 13.67,
-      "learning_rate": 1.4177642691868717e-05,
-      "loss": 0.231,
-      "step": 895000
-    },
-    {
-      "epoch": 13.67,
-      "eval_runtime": 0.7236,
-      "eval_samples_per_second": 1382.053,
-      "eval_steps_per_second": 22.113,
-      "step": 895000
-    },
-    {
-      "epoch": 13.68,
-      "learning_rate": 1.4099229717786368e-05,
-      "loss": 0.231,
-      "step": 896000
-    },
-    {
-      "epoch": 13.7,
-      "learning_rate": 1.4021537423969588e-05,
-      "loss": 0.2317,
-      "step": 897000
-    },
-    {
-      "epoch": 13.71,
-      "learning_rate": 1.3944566660048863e-05,
-      "loss": 0.2308,
-      "step": 898000
-    },
-    {
-      "epoch": 13.73,
-      "learning_rate": 1.3868318267764128e-05,
-      "loss": 0.2309,
-      "step": 899000
-    },
-    {
-      "epoch": 13.74,
-      "learning_rate": 1.3792793080955574e-05,
-      "loss": 0.2308,
-      "step": 900000
-    },
-    {
-      "epoch": 13.74,
-      "eval_runtime": 0.7542,
-      "eval_samples_per_second": 1325.982,
-      "eval_steps_per_second": 21.216,
-      "step": 900000
-    },
-    {
-      "epoch": 13.76,
-      "learning_rate": 1.3717991925554562e-05,
-      "loss": 0.2308,
-      "step": 901000
-    },
-    {
-      "epoch": 13.77,
-      "learning_rate": 1.3643915619574529e-05,
-      "loss": 0.2305,
-      "step": 902000
-    },
-    {
-      "epoch": 13.79,
-      "learning_rate": 1.35705649731021e-05,
-      "loss": 0.2304,
-      "step": 903000
-    },
-    {
-      "epoch": 13.8,
-      "learning_rate": 1.3497940788288195e-05,
-      "loss": 0.2301,
-      "step": 904000
-    },
-    {
-      "epoch": 13.82,
-      "learning_rate": 1.3426043859339253e-05,
-      "loss": 0.2304,
-      "step": 905000
-    },
-    {
-      "epoch": 13.82,
-      "eval_runtime": 0.9802,
-      "eval_samples_per_second": 1020.243,
-      "eval_steps_per_second": 16.324,
-      "step": 905000
-    },
-    {
-      "epoch": 13.83,
-      "learning_rate": 1.3354874972508582e-05,
-      "loss": 0.2302,
-      "step": 906000
-    },
-    {
-      "epoch": 13.85,
-      "learning_rate": 1.3284434906087695e-05,
-      "loss": 0.2303,
-      "step": 907000
-    },
-    {
-      "epoch": 13.87,
-      "learning_rate": 1.3214724430397915e-05,
-      "loss": 0.2304,
-      "step": 908000
-    },
-    {
-      "epoch": 13.88,
-      "learning_rate": 1.314574430778182e-05,
-      "loss": 0.2302,
-      "step": 909000
-    },
-    {
-      "epoch": 13.9,
-      "learning_rate": 1.3077495292594966e-05,
-      "loss": 0.2305,
-      "step": 910000
-    },
-    {
-      "epoch": 13.9,
-      "eval_runtime": 0.7262,
-      "eval_samples_per_second": 1377.03,
-      "eval_steps_per_second": 22.032,
-      "step": 910000
-    },
-    {
-      "epoch": 13.91,
-      "learning_rate": 1.3009978131197669e-05,
-      "loss": 0.2314,
-      "step": 911000
-    },
-    {
-      "epoch": 13.93,
-      "learning_rate": 1.2943193561946762e-05,
-      "loss": 0.2304,
-      "step": 912000
-    },
-    {
-      "epoch": 13.94,
-      "learning_rate": 1.2877142315187628e-05,
-      "loss": 0.2299,
-      "step": 913000
-    },
-    {
-      "epoch": 13.96,
-      "learning_rate": 1.28118251132461e-05,
-      "loss": 0.23,
-      "step": 914000
-    },
-    {
-      "epoch": 13.97,
-      "learning_rate": 1.274724267042063e-05,
-      "loss": 0.2299,
-      "step": 915000
-    },
-    {
-      "epoch": 13.97,
-      "eval_runtime": 0.795,
-      "eval_samples_per_second": 1257.794,
-      "eval_steps_per_second": 20.125,
-      "step": 915000
-    },
-    {
-      "epoch": 13.99,
-      "learning_rate": 1.2683395692974472e-05,
-      "loss": 0.23,
-      "step": 916000
-    },
-    {
-      "epoch": 14.0,
-      "learning_rate": 1.2620284879127947e-05,
-      "loss": 0.23,
-      "step": 917000
-    },
-    {
-      "epoch": 14.02,
-      "learning_rate": 1.2557910919050803e-05,
-      "loss": 0.2295,
-      "step": 918000
-    },
-    {
-      "epoch": 14.03,
-      "learning_rate": 1.2496274494854666e-05,
-      "loss": 0.2296,
-      "step": 919000
-    },
-    {
-      "epoch": 14.05,
-      "learning_rate": 1.24353762805856e-05,
-      "loss": 0.2297,
-      "step": 920000
-    },
-    {
-      "epoch": 14.05,
-      "eval_runtime": 0.7692,
-      "eval_samples_per_second": 1300.053,
-      "eval_steps_per_second": 20.801,
-      "step": 920000
-    },
-    {
-      "epoch": 14.06,
-      "learning_rate": 1.2375216942216713e-05,
-      "loss": 0.2306,
-      "step": 921000
-    },
-    {
-      "epoch": 14.08,
-      "learning_rate": 1.2315797137640906e-05,
-      "loss": 0.2298,
-      "step": 922000
-    },
-    {
-      "epoch": 14.09,
-      "learning_rate": 1.225711751666363e-05,
-      "loss": 0.2295,
-      "step": 923000
-    },
-    {
-      "epoch": 14.11,
-      "learning_rate": 1.2199178720995825e-05,
-      "loss": 0.2299,
-      "step": 924000
-    },
-    {
-      "epoch": 14.12,
-      "learning_rate": 1.2141981384246874e-05,
-      "loss": 0.23,
-      "step": 925000
-    },
-    {
-      "epoch": 14.12,
-      "eval_runtime": 0.827,
-      "eval_samples_per_second": 1209.23,
-      "eval_steps_per_second": 19.348,
-      "step": 925000
-    },
-    {
-      "epoch": 14.14,
-      "learning_rate": 1.2085526131917685e-05,
-      "loss": 0.2294,
-      "step": 926000
-    },
-    {
-      "epoch": 14.16,
-      "learning_rate": 1.2029813581393866e-05,
-      "loss": 0.2289,
-      "step": 927000
-    },
-    {
-      "epoch": 14.17,
-      "learning_rate": 1.197484434193893e-05,
-      "loss": 0.2295,
-      "step": 928000
-    },
-    {
-      "epoch": 14.19,
-      "learning_rate": 1.192061901468768e-05,
-      "loss": 0.2293,
-      "step": 929000
-    },
-    {
-      "epoch": 14.2,
-      "learning_rate": 1.1867138192639601e-05,
-      "loss": 0.2293,
-      "step": 930000
-    },
-    {
-      "epoch": 14.2,
-      "eval_runtime": 0.9644,
-      "eval_samples_per_second": 1036.936,
-      "eval_steps_per_second": 16.591,
-      "step": 930000
-    },
-    {
-      "epoch": 14.22,
-      "learning_rate": 1.1814402460652382e-05,
-      "loss": 0.2291,
-      "step": 931000
-    },
-    {
-      "epoch": 14.23,
-      "learning_rate": 1.176241239543558e-05,
-      "loss": 0.229,
-      "step": 932000
-    },
-    {
-      "epoch": 14.25,
-      "learning_rate": 1.171116856554418e-05,
-      "loss": 0.2291,
-      "step": 933000
-    },
-    {
-      "epoch": 14.26,
-      "learning_rate": 1.1660671531372517e-05,
-      "loss": 0.2301,
-      "step": 934000
-    },
-    {
-      "epoch": 14.28,
-      "learning_rate": 1.1610921845148052e-05,
-      "loss": 0.2295,
-      "step": 935000
-    },
-    {
-      "epoch": 14.28,
-      "eval_runtime": 0.8534,
-      "eval_samples_per_second": 1171.832,
-      "eval_steps_per_second": 18.749,
-      "step": 935000
-    },
-    {
-      "epoch": 14.29,
-      "learning_rate": 1.156192005092539e-05,
-      "loss": 0.2301,
-      "step": 936000
-    },
-    {
-      "epoch": 14.31,
-      "learning_rate": 1.1513666684580308e-05,
-      "loss": 0.2291,
-      "step": 937000
-    },
-    {
-      "epoch": 14.32,
-      "learning_rate": 1.1466162273803876e-05,
-      "loss": 0.2292,
-      "step": 938000
-    },
-    {
-      "epoch": 14.34,
-      "learning_rate": 1.1419407338096732e-05,
-      "loss": 0.2287,
-      "step": 939000
-    },
-    {
-      "epoch": 14.35,
-      "learning_rate": 1.1373402388763346e-05,
-      "loss": 0.2286,
-      "step": 940000
-    },
-    {
-      "epoch": 14.35,
-      "eval_runtime": 0.7875,
-      "eval_samples_per_second": 1269.803,
-      "eval_steps_per_second": 20.317,
-      "step": 940000
-    },
-    {
-      "epoch": 14.37,
-      "learning_rate": 1.1328147928906494e-05,
-      "loss": 0.2287,
-      "step": 941000
-    },
-    {
-      "epoch": 14.38,
-      "learning_rate": 1.1283644453421678e-05,
-      "loss": 0.2289,
-      "step": 942000
-    },
-    {
-      "epoch": 14.4,
-      "learning_rate": 1.1239892448991798e-05,
-      "loss": 0.2284,
-      "step": 943000
-    },
-    {
-      "epoch": 14.42,
-      "learning_rate": 1.1196892394081743e-05,
-      "loss": 0.2287,
-      "step": 944000
-    },
-    {
-      "epoch": 14.43,
-      "learning_rate": 1.1154644758933235e-05,
-      "loss": 0.2285,
-      "step": 945000
-    },
-    {
-      "epoch": 14.43,
-      "eval_runtime": 0.7294,
-      "eval_samples_per_second": 1370.909,
-      "eval_steps_per_second": 21.935,
-      "step": 945000
-    },
-    {
-      "epoch": 14.45,
-      "learning_rate": 1.1113150005559644e-05,
-      "loss": 0.2283,
-      "step": 946000
-    },
-    {
-      "epoch": 14.46,
-      "learning_rate": 1.1072408587740942e-05,
-      "loss": 0.2282,
-      "step": 947000
-    },
-    {
-      "epoch": 14.48,
-      "learning_rate": 1.1032420951018755e-05,
-      "loss": 0.228,
-      "step": 948000
-    },
-    {
-      "epoch": 14.49,
-      "learning_rate": 1.0993187532691458e-05,
-      "loss": 0.2281,
-      "step": 949000
-    },
-    {
-      "epoch": 14.51,
-      "learning_rate": 1.0954708761809438e-05,
-      "loss": 0.2281,
-      "step": 950000
-    },
-    {
-      "epoch": 14.51,
-      "eval_runtime": 0.7692,
-      "eval_samples_per_second": 1300.007,
-      "eval_steps_per_second": 20.8,
-      "step": 950000
-    },
-    {
-      "epoch": 14.52,
-      "learning_rate": 1.091698505917036e-05,
-      "loss": 0.2281,
-      "step": 951000
-    },
-    {
-      "epoch": 14.54,
-      "learning_rate": 1.0880016837314599e-05,
-      "loss": 0.2283,
-      "step": 952000
-    },
-    {
-      "epoch": 14.55,
-      "learning_rate": 1.084380450052071e-05,
-      "loss": 0.2281,
-      "step": 953000
-    },
-    {
-      "epoch": 14.57,
-      "learning_rate": 1.0808348444801e-05,
-      "loss": 0.2278,
-      "step": 954000
-    },
-    {
-      "epoch": 14.58,
-      "learning_rate": 1.0773649057897206e-05,
-      "loss": 0.2283,
-      "step": 955000
-    },
-    {
-      "epoch": 14.58,
-      "eval_runtime": 0.7689,
-      "eval_samples_per_second": 1300.511,
-      "eval_steps_per_second": 20.808,
-      "step": 955000
-    },
-    {
-      "epoch": 14.6,
-      "learning_rate": 1.073970671927628e-05,
-      "loss": 0.2277,
-      "step": 956000
-    },
-    {
-      "epoch": 14.61,
-      "learning_rate": 1.0706521800126198e-05,
-      "loss": 0.2279,
-      "step": 957000
-    },
-    {
-      "epoch": 14.63,
-      "learning_rate": 1.0674094663351906e-05,
-      "loss": 0.2278,
-      "step": 958000
-    },
-    {
-      "epoch": 14.64,
-      "learning_rate": 1.0642425663571383e-05,
-      "loss": 0.2279,
-      "step": 959000
-    },
-    {
-      "epoch": 14.66,
-      "learning_rate": 1.0611515147111736e-05,
-      "loss": 0.2279,
-      "step": 960000
-    },
-    {
-      "epoch": 14.66,
-      "eval_runtime": 0.8148,
-      "eval_samples_per_second": 1227.238,
-      "eval_steps_per_second": 19.636,
-      "step": 960000
-    },
-    {
-      "epoch": 14.67,
-      "learning_rate": 1.0581363452005424e-05,
-      "loss": 0.2279,
-      "step": 961000
-    },
-    {
-      "epoch": 14.69,
-      "learning_rate": 1.0551970907986557e-05,
-      "loss": 0.2277,
-      "step": 962000
-    },
-    {
-      "epoch": 14.71,
-      "learning_rate": 1.0523337836487271e-05,
-      "loss": 0.2276,
-      "step": 963000
-    },
-    {
-      "epoch": 14.72,
-      "learning_rate": 1.0495464550634267e-05,
-      "loss": 0.2278,
-      "step": 964000
-    },
-    {
-      "epoch": 14.74,
-      "learning_rate": 1.046835135524533e-05,
-      "loss": 0.2277,
-      "step": 965000
-    },
-    {
-      "epoch": 14.74,
-      "eval_runtime": 0.7884,
-      "eval_samples_per_second": 1268.404,
-      "eval_steps_per_second": 20.294,
-      "step": 965000
-    },
-    {
-      "epoch": 14.75,
-      "learning_rate": 1.044199854682601e-05,
-      "loss": 0.2278,
-      "step": 966000
-    },
-    {
-      "epoch": 14.77,
-      "learning_rate": 1.0416406413566414e-05,
-      "loss": 0.2279,
-      "step": 967000
-    },
-    {
-      "epoch": 14.78,
-      "learning_rate": 1.0391575235337991e-05,
-      "loss": 0.2278,
-      "step": 968000
-    },
-    {
-      "epoch": 14.8,
-      "learning_rate": 1.0367505283690547e-05,
-      "loss": 0.2276,
-      "step": 969000
-    },
-    {
-      "epoch": 14.81,
-      "learning_rate": 1.0344196821849202e-05,
-      "loss": 0.2279,
-      "step": 970000
-    },
-    {
-      "epoch": 14.81,
-      "eval_runtime": 0.7534,
-      "eval_samples_per_second": 1327.252,
-      "eval_steps_per_second": 21.236,
-      "step": 970000
-    },
-    {
-      "epoch": 14.83,
-      "learning_rate": 1.032165010471157e-05,
-      "loss": 0.2277,
-      "step": 971000
-    },
-    {
-      "epoch": 14.84,
-      "learning_rate": 1.0299865378844936e-05,
-      "loss": 0.2275,
-      "step": 972000
-    },
-    {
-      "epoch": 14.86,
-      "learning_rate": 1.0278842882483569e-05,
-      "loss": 0.2275,
-      "step": 973000
-    },
-    {
-      "epoch": 14.87,
-      "learning_rate": 1.025858284552612e-05,
-      "loss": 0.2276,
-      "step": 974000
-    },
-    {
-      "epoch": 14.89,
-      "learning_rate": 1.023908548953311e-05,
-      "loss": 0.2275,
-      "step": 975000
-    },
-    {
-      "epoch": 14.89,
-      "eval_runtime": 0.7861,
-      "eval_samples_per_second": 1272.066,
-      "eval_steps_per_second": 20.353,
-      "step": 975000
-    },
-    {
-      "epoch": 14.9,
-      "learning_rate": 1.02203510277245e-05,
-      "loss": 0.2276,
-      "step": 976000
-    },
-    {
-      "epoch": 14.92,
-      "learning_rate": 1.0202379664977364e-05,
-      "loss": 0.2272,
-      "step": 977000
-    },
-    {
-      "epoch": 14.93,
-      "learning_rate": 1.018517159782365e-05,
-      "loss": 0.2274,
-      "step": 978000
-    },
-    {
-      "epoch": 14.95,
-      "learning_rate": 1.0168727014448004e-05,
-      "loss": 0.2272,
-      "step": 979000
-    },
-    {
-      "epoch": 14.96,
-      "learning_rate": 1.0153046094685783e-05,
-      "loss": 0.227,
-      "step": 980000
-    },
-    {
-      "epoch": 14.96,
-      "eval_runtime": 0.7489,
-      "eval_samples_per_second": 1335.226,
-      "eval_steps_per_second": 21.364,
-      "step": 980000
-    },
-    {
-      "epoch": 14.98,
-      "learning_rate": 1.0138129010020992e-05,
-      "loss": 0.2272,
-      "step": 981000
-    },
-    {
-      "epoch": 15.0,
-      "learning_rate": 1.0123975923584488e-05,
-      "loss": 0.2273,
-      "step": 982000
-    },
-    {
-      "epoch": 15.01,
-      "learning_rate": 1.0110586990152152e-05,
-      "loss": 0.227,
-      "step": 983000
-    },
-    {
-      "epoch": 15.03,
-      "learning_rate": 1.0097962356143219e-05,
-      "loss": 0.2273,
-      "step": 984000
-    },
-    {
-      "epoch": 15.04,
-      "learning_rate": 1.0086102159618668e-05,
-      "loss": 0.227,
-      "step": 985000
-    },
-    {
-      "epoch": 15.04,
-      "eval_runtime": 0.7868,
-      "eval_samples_per_second": 1271.022,
-      "eval_steps_per_second": 20.336,
-      "step": 985000
-    },
-    {
-      "epoch": 15.06,
-      "learning_rate": 1.0075006530279694e-05,
-      "loss": 0.2271,
-      "step": 986000
-    },
-    {
-      "epoch": 15.07,
-      "learning_rate": 1.0064675589466339e-05,
-      "loss": 0.2268,
-      "step": 987000
-    },
-    {
-      "epoch": 15.09,
-      "learning_rate": 1.0055109450156098e-05,
-      "loss": 0.2272,
-      "step": 988000
-    },
-    {
-      "epoch": 15.1,
-      "learning_rate": 1.0046308216962759e-05,
-      "loss": 0.2269,
-      "step": 989000
-    },
-    {
-      "epoch": 15.12,
-      "learning_rate": 1.0038271986135177e-05,
-      "loss": 0.2272,
-      "step": 990000
-    },
-    {
-      "epoch": 15.12,
-      "eval_runtime": 0.7713,
-      "eval_samples_per_second": 1296.438,
-      "eval_steps_per_second": 20.743,
-      "step": 990000
-    },
-    {
-      "epoch": 15.13,
-      "learning_rate": 1.0031000845556304e-05,
-      "loss": 0.2272,
-      "step": 991000
-    },
-    {
-      "epoch": 15.15,
-      "learning_rate": 1.0024494874742152e-05,
-      "loss": 0.2272,
-      "step": 992000
-    },
-    {
-      "epoch": 15.16,
-      "learning_rate": 1.0018754144840986e-05,
-      "loss": 0.2272,
-      "step": 993000
-    },
-    {
-      "epoch": 15.18,
-      "learning_rate": 1.0013778718632507e-05,
-      "loss": 0.227,
-      "step": 994000
-    },
-    {
-      "epoch": 15.19,
-      "learning_rate": 1.000956865052717e-05,
-      "loss": 0.2269,
-      "step": 995000
-    },
-    {
-      "epoch": 15.19,
-      "eval_runtime": 0.7194,
-      "eval_samples_per_second": 1390.055,
-      "eval_steps_per_second": 22.241,
-      "step": 995000
-    },
-    {
-      "epoch": 15.21,
-      "learning_rate": 1.0006123986565623e-05,
-      "loss": 0.2267,
-      "step": 996000
-    },
-    {
-      "epoch": 15.22,
-      "learning_rate": 1.0003444764418138e-05,
-      "loss": 0.2265,
-      "step": 997000
-    },
-    {
-      "epoch": 15.24,
-      "learning_rate": 1.000153101338428e-05,
-      "loss": 0.2268,
-      "step": 998000
-    },
-    {
-      "epoch": 15.25,
-      "learning_rate": 1.00003827543925e-05,
-      "loss": 0.2269,
-      "step": 999000
-    },
-    {
-      "epoch": 15.27,
-      "learning_rate": 1e-05,
-      "loss": 0.2268,
-      "step": 1000000
-    },
-    {
-      "epoch": 15.27,
-      "eval_runtime": 0.8245,
-      "eval_samples_per_second": 1212.903,
-      "eval_steps_per_second": 19.406,
-      "step": 1000000
-    }
-  ],
-  "max_steps": 1000000,
-  "num_train_epochs": 16,
-  "total_flos": 7.010016247012483e+22,
-  "trial_name": null,
-  "trial_params": null
-}