diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7952 @@
+{
+  "best_metric": 0.7278521656990051,
+  "best_model_checkpoint": "checkpoints/bart/checkpoint-620000",
+  "epoch": 20.18886356235754,
+  "global_step": 620000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 10.4661,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 10.2325,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5e-06,
+      "loss": 9.8216,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 9.4694,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5e-06,
+      "loss": 9.1591,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3e-06,
+      "loss": 8.9458,
+      "step": 3000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.5e-06,
+      "loss": 8.8337,
+      "step": 3500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 8.7867,
+      "step": 4000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.5e-06,
+      "loss": 8.7671,
+      "step": 4500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-06,
+      "loss": 8.7524,
+      "step": 5000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 8.7339,
+      "step": 5500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6e-06,
+      "loss": 8.6581,
+      "step": 6000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 8.5433,
+      "step": 6500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7e-06,
+      "loss": 8.4347,
+      "step": 7000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 8.2994,
+      "step": 7500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 8.1404,
+      "step": 8000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.5e-06,
+      "loss": 7.9371,
+      "step": 8500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9e-06,
+      "loss": 7.7451,
+      "step": 9000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.5e-06,
+      "loss": 7.5965,
+      "step": 9500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 7.482,
+      "step": 10000
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 7.40041971206665,
+      "eval_runtime": 144.9465,
+      "eval_samples_per_second": 547.871,
+      "eval_steps_per_second": 2.146,
+      "step": 10000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.994949494949497e-06,
+      "loss": 7.385,
+      "step": 10500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.989898989898991e-06,
+      "loss": 7.3015,
+      "step": 11000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.984848484848485e-06,
+      "loss": 7.2289,
+      "step": 11500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.97979797979798e-06,
+      "loss": 7.1624,
+      "step": 12000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.974747474747476e-06,
+      "loss": 7.1031,
+      "step": 12500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.96969696969697e-06,
+      "loss": 7.0448,
+      "step": 13000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.964646464646466e-06,
+      "loss": 6.9899,
+      "step": 13500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.95959595959596e-06,
+      "loss": 6.9354,
+      "step": 14000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.954545454545456e-06,
+      "loss": 6.8847,
+      "step": 14500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.94949494949495e-06,
+      "loss": 6.8338,
+      "step": 15000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.944444444444445e-06,
+      "loss": 6.7857,
+      "step": 15500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.939393939393939e-06,
+      "loss": 6.7382,
+      "step": 16000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.934343434343435e-06,
+      "loss": 6.6942,
+      "step": 16500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.92929292929293e-06,
+      "loss": 6.6523,
+      "step": 17000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.924242424242425e-06,
+      "loss": 6.6065,
+      "step": 17500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.91919191919192e-06,
+      "loss": 6.5534,
+      "step": 18000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.914141414141416e-06,
+      "loss": 6.4647,
+      "step": 18500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.90909090909091e-06,
+      "loss": 6.3421,
+      "step": 19000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.904040404040404e-06,
+      "loss": 6.2265,
+      "step": 19500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.8989898989899e-06,
+      "loss": 6.1141,
+      "step": 20000
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 5.996727466583252,
+      "eval_runtime": 102.6556,
+      "eval_samples_per_second": 773.577,
+      "eval_steps_per_second": 3.03,
+      "step": 20000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.893939393939395e-06,
+      "loss": 6.0157,
+      "step": 20500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.88888888888889e-06,
+      "loss": 5.925,
+      "step": 21000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.883838383838385e-06,
+      "loss": 5.841,
+      "step": 21500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.87878787878788e-06,
+      "loss": 5.7651,
+      "step": 22000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.873737373737373e-06,
+      "loss": 5.6874,
+      "step": 22500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.86868686868687e-06,
+      "loss": 5.6163,
+      "step": 23000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.863636363636364e-06,
+      "loss": 5.5472,
+      "step": 23500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.85858585858586e-06,
+      "loss": 5.4828,
+      "step": 24000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.853535353535354e-06,
+      "loss": 5.4214,
+      "step": 24500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.84848484848485e-06,
+      "loss": 5.3606,
+      "step": 25000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.843434343434344e-06,
+      "loss": 5.302,
+      "step": 25500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.838383838383839e-06,
+      "loss": 5.2467,
+      "step": 26000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.833333333333333e-06,
+      "loss": 5.19,
+      "step": 26500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 9.828282828282829e-06,
+      "loss": 5.1389,
+      "step": 27000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.823232323232325e-06,
+      "loss": 5.0893,
+      "step": 27500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.81818181818182e-06,
+      "loss": 5.036,
+      "step": 28000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 9.813131313131315e-06,
+      "loss": 4.9913,
+      "step": 28500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.80808080808081e-06,
+      "loss": 4.9422,
+      "step": 29000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.803030303030304e-06,
+      "loss": 4.9003,
+      "step": 29500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.797979797979798e-06,
+      "loss": 4.8521,
+      "step": 30000
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 4.616042137145996,
+      "eval_runtime": 89.4823,
+      "eval_samples_per_second": 887.46,
+      "eval_steps_per_second": 3.476,
+      "step": 30000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.792929292929294e-06,
+      "loss": 4.8057,
+      "step": 30500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.787878787878788e-06,
+      "loss": 4.7651,
+      "step": 31000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.782828282828284e-06,
+      "loss": 4.7198,
+      "step": 31500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 4.6846,
+      "step": 32000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.772727272727273e-06,
+      "loss": 4.6463,
+      "step": 32500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.767676767676767e-06,
+      "loss": 4.6111,
+      "step": 33000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.762626262626263e-06,
+      "loss": 4.5757,
+      "step": 33500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.757575757575758e-06,
+      "loss": 4.5376,
+      "step": 34000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.752525252525254e-06,
+      "loss": 4.5036,
+      "step": 34500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.747474747474748e-06,
+      "loss": 4.4697,
+      "step": 35000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.742424242424244e-06,
+      "loss": 4.4346,
+      "step": 35500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 9.737373737373738e-06,
+      "loss": 4.3974,
+      "step": 36000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 9.732323232323232e-06,
+      "loss": 4.3679,
+      "step": 36500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.727272727272728e-06,
+      "loss": 4.3366,
+      "step": 37000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 4.3081,
+      "step": 37500
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.717171717171719e-06,
+      "loss": 4.2805,
+      "step": 38000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.712121212121213e-06,
+      "loss": 4.2505,
+      "step": 38500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.707070707070709e-06,
+      "loss": 4.2206,
+      "step": 39000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.702020202020203e-06,
+      "loss": 4.1958,
+      "step": 39500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 4.1595,
+      "step": 40000
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 4.127486705780029,
+      "eval_runtime": 88.4336,
+      "eval_samples_per_second": 897.985,
+      "eval_steps_per_second": 3.517,
+      "step": 40000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 9.691919191919192e-06,
+      "loss": 4.1409,
+      "step": 40500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 9.686868686868688e-06,
+      "loss": 4.1087,
+      "step": 41000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 9.681818181818182e-06,
+      "loss": 4.0735,
+      "step": 41500
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.676767676767678e-06,
+      "loss": 4.0509,
+      "step": 42000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 9.671717171717172e-06,
+      "loss": 4.0182,
+      "step": 42500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 3.9825,
+      "step": 43000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 9.661616161616163e-06,
+      "loss": 3.9548,
+      "step": 43500
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.656565656565657e-06,
+      "loss": 3.9287,
+      "step": 44000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.651515151515153e-06,
+      "loss": 3.8958,
+      "step": 44500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.646464646464647e-06,
+      "loss": 3.867,
+      "step": 45000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.641414141414143e-06,
+      "loss": 3.8407,
+      "step": 45500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.636363636363638e-06,
+      "loss": 3.813,
+      "step": 46000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.631313131313132e-06,
+      "loss": 3.7928,
+      "step": 46500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.626262626262626e-06,
+      "loss": 3.7673,
+      "step": 47000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.621212121212122e-06,
+      "loss": 3.7444,
+      "step": 47500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.616161616161616e-06,
+      "loss": 3.7203,
+      "step": 48000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.611111111111112e-06,
+      "loss": 3.7023,
+      "step": 48500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.606060606060607e-06,
+      "loss": 3.6724,
+      "step": 49000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.601010101010103e-06,
+      "loss": 3.653,
+      "step": 49500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.595959595959597e-06,
+      "loss": 3.6342,
+      "step": 50000
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 3.4967825412750244,
+      "eval_runtime": 91.2654,
+      "eval_samples_per_second": 870.122,
+      "eval_steps_per_second": 3.408,
+      "step": 50000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.590909090909091e-06,
+      "loss": 3.6095,
+      "step": 50500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.585858585858586e-06,
+      "loss": 3.5816,
+      "step": 51000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.580808080808082e-06,
+      "loss": 3.5653,
+      "step": 51500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.575757575757576e-06,
+      "loss": 3.5455,
+      "step": 52000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.570707070707072e-06,
+      "loss": 3.5238,
+      "step": 52500
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.565656565656566e-06,
+      "loss": 3.4987,
+      "step": 53000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 9.56060606060606e-06,
+      "loss": 3.4826,
+      "step": 53500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 3.4585,
+      "step": 54000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.55050505050505e-06,
+      "loss": 3.442,
+      "step": 54500
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 3.4201,
+      "step": 55000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 9.540404040404041e-06,
+      "loss": 3.4026,
+      "step": 55500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 9.535353535353537e-06,
+      "loss": 3.385,
+      "step": 56000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.530303030303031e-06,
+      "loss": 3.3663,
+      "step": 56500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.525252525252526e-06,
+      "loss": 3.3497,
+      "step": 57000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.52020202020202e-06,
+      "loss": 3.3303,
+      "step": 57500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.515151515151516e-06,
+      "loss": 3.3123,
+      "step": 58000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.51010101010101e-06,
+      "loss": 3.2963,
+      "step": 58500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.505050505050506e-06,
+      "loss": 3.2779,
+      "step": 59000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 9.5e-06,
+      "loss": 3.2636,
+      "step": 59500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.494949494949497e-06,
+      "loss": 3.2404,
+      "step": 60000
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 2.9165635108947754,
+      "eval_runtime": 91.1781,
+      "eval_samples_per_second": 870.955,
+      "eval_steps_per_second": 3.411,
+      "step": 60000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.48989898989899e-06,
+      "loss": 3.2228,
+      "step": 60500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.484848484848485e-06,
+      "loss": 3.2067,
+      "step": 61000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.479797979797981e-06,
+      "loss": 3.1899,
+      "step": 61500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.474747474747475e-06,
+      "loss": 3.1685,
+      "step": 62000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 9.469696969696971e-06,
+      "loss": 3.1507,
+      "step": 62500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 9.464646464646466e-06,
+      "loss": 3.1322,
+      "step": 63000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.45959595959596e-06,
+      "loss": 3.119,
+      "step": 63500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.454545454545456e-06,
+      "loss": 3.0955,
+      "step": 64000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.44949494949495e-06,
+      "loss": 3.0801,
+      "step": 64500
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 3.0589,
+      "step": 65000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.43939393939394e-06,
+      "loss": 3.0396,
+      "step": 65500
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.434343434343435e-06,
+      "loss": 3.0281,
+      "step": 66000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.42929292929293e-06,
+      "loss": 3.003,
+      "step": 66500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.424242424242425e-06,
+      "loss": 2.9854,
+      "step": 67000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.41919191919192e-06,
+      "loss": 2.9694,
+      "step": 67500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 9.414141414141414e-06,
+      "loss": 2.9498,
+      "step": 68000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 9.40909090909091e-06,
+      "loss": 2.9232,
+      "step": 68500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 9.404040404040404e-06,
+      "loss": 2.9021,
+      "step": 69000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 9.3989898989899e-06,
+      "loss": 2.8785,
+      "step": 69500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 2.8581,
+      "step": 70000
+    },
+    {
+      "epoch": 2.28,
+      "eval_loss": 2.5783917903900146,
+      "eval_runtime": 91.27,
+      "eval_samples_per_second": 870.077,
+      "eval_steps_per_second": 3.407,
+      "step": 70000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 9.38888888888889e-06,
+      "loss": 2.8342,
+      "step": 70500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 9.383838383838385e-06,
+      "loss": 2.8058,
+      "step": 71000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 9.378787878787879e-06,
+      "loss": 2.7748,
+      "step": 71500
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 9.373737373737375e-06,
+      "loss": 2.7429,
+      "step": 72000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 9.36868686868687e-06,
+      "loss": 2.7156,
+      "step": 72500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 9.363636363636365e-06,
+      "loss": 2.6877,
+      "step": 73000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 9.35858585858586e-06,
+      "loss": 2.6559,
+      "step": 73500
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.353535353535354e-06,
+      "loss": 2.6292,
+      "step": 74000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.34848484848485e-06,
+      "loss": 2.6008,
+      "step": 74500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.343434343434344e-06,
+      "loss": 2.5744,
+      "step": 75000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.338383838383838e-06,
+      "loss": 2.5463,
+      "step": 75500
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 2.5264,
+      "step": 76000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 9.328282828282829e-06,
+      "loss": 2.4965,
+      "step": 76500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 9.323232323232325e-06,
+      "loss": 2.4692,
+      "step": 77000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 9.318181818181819e-06,
+      "loss": 2.4472,
+      "step": 77500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 9.313131313131313e-06,
+      "loss": 2.426,
+      "step": 78000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 9.30808080808081e-06,
+      "loss": 2.4064,
+      "step": 78500
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.303030303030303e-06,
+      "loss": 2.3843,
+      "step": 79000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.2979797979798e-06,
+      "loss": 2.3624,
+      "step": 79500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.292929292929294e-06,
+      "loss": 2.3427,
+      "step": 80000
+    },
+    {
+      "epoch": 2.61,
+      "eval_loss": 1.944496750831604,
+      "eval_runtime": 90.6503,
+      "eval_samples_per_second": 876.026,
+      "eval_steps_per_second": 3.431,
+      "step": 80000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.28787878787879e-06,
+      "loss": 2.3235,
+      "step": 80500
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.282828282828284e-06,
+      "loss": 2.3018,
+      "step": 81000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 9.277777777777778e-06,
+      "loss": 2.285,
+      "step": 81500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 9.272727272727273e-06,
+      "loss": 2.2638,
+      "step": 82000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 9.267676767676769e-06,
+      "loss": 2.2429,
+      "step": 82500
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.262626262626263e-06,
+      "loss": 2.2254,
+      "step": 83000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.257575757575759e-06,
+      "loss": 2.2084,
+      "step": 83500
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 9.252525252525253e-06,
+      "loss": 2.1904,
+      "step": 84000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 9.24747474747475e-06,
+      "loss": 2.1747,
+      "step": 84500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 9.242424242424244e-06,
+      "loss": 2.1589,
+      "step": 85000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 9.237373737373738e-06,
+      "loss": 2.1424,
+      "step": 85500
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 9.232323232323232e-06,
+      "loss": 2.1265,
+      "step": 86000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 9.227272727272728e-06,
+      "loss": 2.1108,
+      "step": 86500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 9.222222222222224e-06,
+      "loss": 2.0934,
+      "step": 87000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.217171717171718e-06,
+      "loss": 2.0784,
+      "step": 87500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.212121212121213e-06,
+      "loss": 2.0634,
+      "step": 88000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 9.207070707070707e-06,
+      "loss": 2.0508,
+      "step": 88500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 9.202020202020203e-06,
+      "loss": 2.0325,
+      "step": 89000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.196969696969697e-06,
+      "loss": 2.0171,
+      "step": 89500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.191919191919193e-06,
+      "loss": 2.0014,
+      "step": 90000
+    },
+    {
+      "epoch": 2.93,
+      "eval_loss": 1.6067554950714111,
+      "eval_runtime": 95.1881,
+      "eval_samples_per_second": 834.264,
+      "eval_steps_per_second": 3.267,
+      "step": 90000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.186868686868688e-06,
+      "loss": 1.9869,
+      "step": 90500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.181818181818184e-06,
+      "loss": 1.9747,
+      "step": 91000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.176767676767678e-06,
+      "loss": 1.9606,
+      "step": 91500
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 9.171717171717172e-06,
+      "loss": 1.9454,
+      "step": 92000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 1.9341,
+      "step": 92500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 9.161616161616162e-06,
+      "loss": 1.9222,
+      "step": 93000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 9.156565656565657e-06,
+      "loss": 1.9068,
+      "step": 93500
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 9.151515151515153e-06,
+      "loss": 1.8961,
+      "step": 94000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 9.146464646464647e-06,
+      "loss": 1.8837,
+      "step": 94500
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 9.141414141414143e-06,
+      "loss": 1.872,
+      "step": 95000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.136363636363637e-06,
+      "loss": 1.8583,
+      "step": 95500
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.131313131313132e-06,
+      "loss": 1.8481,
+      "step": 96000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 9.126262626262628e-06,
+      "loss": 1.8352,
+      "step": 96500
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 9.121212121212122e-06,
+      "loss": 1.8233,
+      "step": 97000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 9.116161616161618e-06,
+      "loss": 1.814,
+      "step": 97500
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 1.8015,
+      "step": 98000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 9.106060606060606e-06,
+      "loss": 1.7941,
+      "step": 98500
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.1010101010101e-06,
+      "loss": 1.7809,
+      "step": 99000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.095959595959597e-06,
+      "loss": 1.7673,
+      "step": 99500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.7589,
+      "step": 100000
+    },
+    {
+      "epoch": 3.26,
+      "eval_loss": 1.4153616428375244,
+      "eval_runtime": 92.3909,
+      "eval_samples_per_second": 859.522,
+      "eval_steps_per_second": 3.366,
+      "step": 100000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 9.085858585858587e-06,
+      "loss": 1.7465,
+      "step": 100500
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 9.080808080808081e-06,
+      "loss": 1.7378,
+      "step": 101000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 9.075757575757577e-06,
+      "loss": 1.7264,
+      "step": 101500
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 9.070707070707072e-06,
+      "loss": 1.7176,
+      "step": 102000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 9.065656565656566e-06,
+      "loss": 1.7086,
+      "step": 102500
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.06060606060606e-06,
+      "loss": 1.6981,
+      "step": 103000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.055555555555556e-06,
+      "loss": 1.6867,
+      "step": 103500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.050505050505052e-06,
+      "loss": 1.6755,
+      "step": 104000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.045454545454546e-06,
+      "loss": 1.6676,
+      "step": 104500
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.040404040404042e-06,
+      "loss": 1.6582,
+      "step": 105000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 9.035353535353537e-06,
+      "loss": 1.6505,
+      "step": 105500
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.030303030303031e-06,
+      "loss": 1.6398,
+      "step": 106000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.025252525252525e-06,
+      "loss": 1.6315,
+      "step": 106500
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.020202020202021e-06,
+      "loss": 1.6228,
+      "step": 107000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.015151515151516e-06,
+      "loss": 1.6121,
+      "step": 107500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.010101010101012e-06,
+      "loss": 1.606,
+      "step": 108000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.005050505050506e-06,
+      "loss": 1.5963,
+      "step": 108500
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9e-06,
+      "loss": 1.5873,
+      "step": 109000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.994949494949495e-06,
+      "loss": 1.5805,
+      "step": 109500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 8.98989898989899e-06,
+      "loss": 1.5718,
+      "step": 110000
+    },
+    {
+      "epoch": 3.58,
+      "eval_loss": 1.3169385194778442,
+      "eval_runtime": 89.9714,
+      "eval_samples_per_second": 882.636,
+      "eval_steps_per_second": 3.457,
+      "step": 110000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 8.984848484848485e-06,
+      "loss": 1.5651,
+      "step": 110500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.97979797979798e-06,
+      "loss": 1.5574,
+      "step": 111000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.974747474747475e-06,
+      "loss": 1.5482,
+      "step": 111500
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.969696969696971e-06,
+      "loss": 1.5396,
+      "step": 112000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.964646464646465e-06,
+      "loss": 1.5354,
+      "step": 112500
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.95959595959596e-06,
+      "loss": 1.5262,
+      "step": 113000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.954545454545456e-06,
+      "loss": 1.5191,
+      "step": 113500
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.94949494949495e-06,
+      "loss": 1.51,
+      "step": 114000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.944444444444446e-06,
+      "loss": 1.5044,
+      "step": 114500
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.93939393939394e-06,
+      "loss": 1.4982,
+      "step": 115000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.934343434343436e-06,
+      "loss": 1.49,
+      "step": 115500
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.92929292929293e-06,
+      "loss": 1.484,
+      "step": 116000
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.924242424242425e-06,
+      "loss": 1.4765,
+      "step": 116500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 8.919191919191919e-06,
+      "loss": 1.4701,
+      "step": 117000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.914141414141415e-06,
+      "loss": 1.4645,
+      "step": 117500
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.90909090909091e-06,
+      "loss": 1.4558,
+      "step": 118000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 8.904040404040405e-06,
+      "loss": 1.4492,
+      "step": 118500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 8.8989898989899e-06,
+      "loss": 1.4444,
+      "step": 119000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 8.893939393939394e-06,
+      "loss": 1.4393,
+      "step": 119500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.434,
+      "step": 120000
+    },
+    {
+      "epoch": 3.91,
+      "eval_loss": 1.213987946510315,
+      "eval_runtime": 91.2226,
+      "eval_samples_per_second": 870.53,
+      "eval_steps_per_second": 3.409,
+      "step": 120000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 8.883838383838384e-06,
+      "loss": 1.4269,
+      "step": 120500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 8.87878787878788e-06,
+      "loss": 1.42,
+      "step": 121000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 8.873737373737375e-06,
+      "loss": 1.4156,
+      "step": 121500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 8.86868686868687e-06,
+      "loss": 1.4107,
+      "step": 122000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.863636363636365e-06,
+      "loss": 1.4045,
+      "step": 122500
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 8.85858585858586e-06,
+      "loss": 1.3989,
+      "step": 123000
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 8.853535353535353e-06,
+      "loss": 1.3944,
+      "step": 123500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.84848484848485e-06,
+      "loss": 1.3868,
+      "step": 124000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.843434343434344e-06,
+      "loss": 1.383,
+      "step": 124500
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.83838383838384e-06,
+      "loss": 1.3757,
+      "step": 125000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.833333333333334e-06,
+      "loss": 1.3715,
+      "step": 125500
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.82828282828283e-06,
+      "loss": 1.3676,
+      "step": 126000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.823232323232324e-06,
+      "loss": 1.362,
+      "step": 126500
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.818181818181819e-06,
+      "loss": 1.3577,
+      "step": 127000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.813131313131313e-06,
+      "loss": 1.353,
+      "step": 127500
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.808080808080809e-06,
+      "loss": 1.3467,
+      "step": 128000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.803030303030303e-06,
+      "loss": 1.3435,
+      "step": 128500
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.7979797979798e-06,
+      "loss": 1.3383,
+      "step": 129000
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.792929292929293e-06,
+      "loss": 1.3324,
+      "step": 129500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 1.3295,
+      "step": 130000
+    },
+    {
+      "epoch": 4.23,
+      "eval_loss": 1.165347695350647,
+      "eval_runtime": 94.4005,
+      "eval_samples_per_second": 841.225,
+      "eval_steps_per_second": 3.294,
+      "step": 130000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.782828282828284e-06,
+      "loss": 1.3257,
+      "step": 130500
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.777777777777778e-06,
+      "loss": 1.3218,
+      "step": 131000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 8.772727272727274e-06,
+      "loss": 1.317,
+      "step": 131500
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 8.767676767676768e-06,
+      "loss": 1.3128,
+      "step": 132000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 8.762626262626264e-06,
+      "loss": 1.3093,
+      "step": 132500
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 8.757575757575759e-06,
+      "loss": 1.3051,
+      "step": 133000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.752525252525253e-06,
+      "loss": 1.3019,
+      "step": 133500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.747474747474747e-06,
+      "loss": 1.2969,
+      "step": 134000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.742424242424243e-06,
+      "loss": 1.2921,
+      "step": 134500
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.737373737373738e-06,
+      "loss": 1.2898,
+      "step": 135000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.732323232323234e-06,
+      "loss": 1.2858,
+      "step": 135500
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 8.727272727272728e-06,
+      "loss": 1.2819,
+      "step": 136000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.722222222222224e-06,
+      "loss": 1.2785,
+      "step": 136500
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.717171717171718e-06,
+      "loss": 1.2751,
+      "step": 137000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.712121212121212e-06,
+      "loss": 1.272,
+      "step": 137500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 8.707070707070707e-06,
+      "loss": 1.2677,
+      "step": 138000
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.702020202020203e-06,
+      "loss": 1.2648,
+      "step": 138500
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.696969696969699e-06,
+      "loss": 1.2601,
+      "step": 139000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.691919191919193e-06,
+      "loss": 1.2591,
+      "step": 139500
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 8.686868686868687e-06,
+      "loss": 1.2537,
+      "step": 140000
+    },
+    {
+      "epoch": 4.56,
+      "eval_loss": 1.0970981121063232,
+      "eval_runtime": 91.1052,
+      "eval_samples_per_second": 871.651,
+      "eval_steps_per_second": 3.414,
+      "step": 140000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.681818181818182e-06,
+      "loss": 1.2513,
+      "step": 140500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 8.676767676767678e-06,
+      "loss": 1.2472,
+      "step": 141000
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 8.671717171717172e-06,
+      "loss": 1.2447,
+      "step": 141500
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.2418,
+      "step": 142000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 8.661616161616162e-06,
+      "loss": 1.2394,
+      "step": 142500
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 8.656565656565658e-06,
+      "loss": 1.236,
+      "step": 143000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 8.651515151515152e-06,
+      "loss": 1.2334,
+      "step": 143500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 8.646464646464647e-06,
+      "loss": 1.2295,
+      "step": 144000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 8.641414141414141e-06,
+      "loss": 1.2269,
+      "step": 144500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 1.2236,
+      "step": 145000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 8.631313131313131e-06,
+      "loss": 1.22,
+      "step": 145500
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 8.626262626262627e-06,
+      "loss": 1.2184,
+      "step": 146000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 8.621212121212122e-06,
+      "loss": 1.2156,
+      "step": 146500
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 8.616161616161618e-06,
+      "loss": 1.2129,
+      "step": 147000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.209,
+      "step": 147500
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 8.606060606060606e-06,
+      "loss": 1.2071,
+      "step": 148000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 8.601010101010102e-06,
+      "loss": 1.2041,
+      "step": 148500
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 8.595959595959596e-06,
+      "loss": 1.2018,
+      "step": 149000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 8.590909090909092e-06,
+      "loss": 1.1989,
+      "step": 149500
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 8.585858585858587e-06,
+      "loss": 1.1954,
+      "step": 150000
+    },
+    {
+      "epoch": 4.88,
+      "eval_loss": 1.0581802129745483,
+      "eval_runtime": 89.9138,
+      "eval_samples_per_second": 883.201,
+      "eval_steps_per_second": 3.459,
+      "step": 150000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 8.580808080808081e-06,
+      "loss": 1.1939,
+      "step": 150500
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 8.575757575757575e-06,
+      "loss": 1.1918,
+      "step": 151000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 8.570707070707071e-06,
+      "loss": 1.1882,
+      "step": 151500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 8.565656565656566e-06,
+      "loss": 1.1868,
+      "step": 152000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 8.560606060606062e-06,
+      "loss": 1.1855,
+      "step": 152500
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 8.555555555555556e-06,
+      "loss": 1.1829,
+      "step": 153000
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 8.550505050505052e-06,
+      "loss": 1.1791,
+      "step": 153500
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 8.545454545454546e-06,
+      "loss": 1.1762,
+      "step": 154000
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 8.54040404040404e-06,
+      "loss": 1.1743,
+      "step": 154500
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 8.535353535353535e-06,
+      "loss": 1.1741,
+      "step": 155000
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 8.53030303030303e-06,
+      "loss": 1.1709,
+      "step": 155500
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 8.525252525252527e-06,
+      "loss": 1.1675,
+      "step": 156000
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 8.520202020202021e-06,
+      "loss": 1.166,
+      "step": 156500
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 8.515151515151517e-06,
+      "loss": 1.1651,
+      "step": 157000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 8.510101010101011e-06,
+      "loss": 1.1622,
+      "step": 157500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 8.505050505050506e-06,
+      "loss": 1.1614,
+      "step": 158000
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 8.5e-06,
+      "loss": 1.1572,
+      "step": 158500
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 8.494949494949496e-06,
+      "loss": 1.157,
+      "step": 159000
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 8.48989898989899e-06,
+      "loss": 1.1549,
+      "step": 159500
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 1.152,
+      "step": 160000
+    },
+    {
+      "epoch": 5.21,
+      "eval_loss": 1.0367275476455688,
+      "eval_runtime": 89.0095,
+      "eval_samples_per_second": 892.175,
+      "eval_steps_per_second": 3.494,
+      "step": 160000
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 8.47979797979798e-06,
+      "loss": 1.1502,
+      "step": 160500
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 8.474747474747475e-06,
+      "loss": 1.1474,
+      "step": 161000
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 8.46969696969697e-06,
+      "loss": 1.1456,
+      "step": 161500
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 8.464646464646465e-06,
+      "loss": 1.1447,
+      "step": 162000
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 8.45959595959596e-06,
+      "loss": 1.1435,
+      "step": 162500
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 8.454545454545455e-06,
+      "loss": 1.1408,
+      "step": 163000
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 8.44949494949495e-06,
+      "loss": 1.1382,
+      "step": 163500
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 1.1363,
+      "step": 164000
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 8.43939393939394e-06,
+      "loss": 1.1343,
+      "step": 164500
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 8.434343434343434e-06,
+      "loss": 1.1329,
+      "step": 165000
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 8.42929292929293e-06,
+      "loss": 1.1325,
+      "step": 165500
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 8.424242424242425e-06,
+      "loss": 1.1296,
+      "step": 166000
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 8.41919191919192e-06,
+      "loss": 1.1274,
+      "step": 166500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 8.414141414141415e-06,
+      "loss": 1.1253,
+      "step": 167000
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 8.40909090909091e-06,
+      "loss": 1.1251,
+      "step": 167500
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 8.404040404040405e-06,
+      "loss": 1.1218,
+      "step": 168000
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 8.3989898989899e-06,
+      "loss": 1.1217,
+      "step": 168500
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 8.393939393939394e-06,
+      "loss": 1.1199,
+      "step": 169000
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 8.38888888888889e-06,
+      "loss": 1.1171,
+      "step": 169500
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 8.383838383838384e-06,
+      "loss": 1.1158,
+      "step": 170000
+    },
+    {
+      "epoch": 5.54,
+      "eval_loss": 1.0161751508712769,
+      "eval_runtime": 96.0727,
+      "eval_samples_per_second": 826.582,
+      "eval_steps_per_second": 3.237,
+      "step": 170000
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 8.37878787878788e-06,
+      "loss": 1.1143,
+      "step": 170500
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 8.373737373737374e-06,
+      "loss": 1.1134,
+      "step": 171000
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 8.368686868686869e-06,
+      "loss": 1.1119,
+      "step": 171500
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 8.363636363636365e-06,
+      "loss": 1.1093,
+      "step": 172000
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 8.358585858585859e-06,
+      "loss": 1.1083,
+      "step": 172500
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 8.353535353535355e-06,
+      "loss": 1.1076,
+      "step": 173000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 8.348484848484849e-06,
+      "loss": 1.1078,
+      "step": 173500
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 8.343434343434345e-06,
+      "loss": 1.1039,
+      "step": 174000
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 8.33838383838384e-06,
+      "loss": 1.1016,
+      "step": 174500
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.1014,
+      "step": 175000
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 8.328282828282828e-06,
+      "loss": 1.0998,
+      "step": 175500
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 8.323232323232324e-06,
+      "loss": 1.0973,
+      "step": 176000
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 8.318181818181818e-06,
+      "loss": 1.0977,
+      "step": 176500
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 8.313131313131314e-06,
+      "loss": 1.0961,
+      "step": 177000
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 8.308080808080809e-06,
+      "loss": 1.0932,
+      "step": 177500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 8.303030303030305e-06,
+      "loss": 1.091,
+      "step": 178000
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 8.297979797979799e-06,
+      "loss": 1.091,
+      "step": 178500
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 8.292929292929293e-06,
+      "loss": 1.0885,
+      "step": 179000
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 8.287878787878787e-06,
+      "loss": 1.088,
+      "step": 179500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 8.282828282828283e-06,
+      "loss": 1.0873,
+      "step": 180000
+    },
+    {
+      "epoch": 5.86,
+      "eval_loss": 0.9905126094818115,
+      "eval_runtime": 95.2304,
+      "eval_samples_per_second": 833.893,
+      "eval_steps_per_second": 3.266,
+      "step": 180000
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 8.277777777777778e-06,
+      "loss": 1.0862,
+      "step": 180500
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 8.272727272727274e-06,
+      "loss": 1.0836,
+      "step": 181000
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 8.267676767676768e-06,
+      "loss": 1.0828,
+      "step": 181500
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 8.262626262626264e-06,
+      "loss": 1.0814,
+      "step": 182000
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 8.257575757575758e-06,
+      "loss": 1.0798,
+      "step": 182500
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 8.252525252525253e-06,
+      "loss": 1.0786,
+      "step": 183000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 8.247474747474749e-06,
+      "loss": 1.0782,
+      "step": 183500
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 8.242424242424243e-06,
+      "loss": 1.0769,
+      "step": 184000
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 8.237373737373739e-06,
+      "loss": 1.0758,
+      "step": 184500
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 8.232323232323233e-06,
+      "loss": 1.0736,
+      "step": 185000
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 8.227272727272728e-06,
+      "loss": 1.0725,
+      "step": 185500
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 1.0714,
+      "step": 186000
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 8.217171717171718e-06,
+      "loss": 1.0703,
+      "step": 186500
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 8.212121212121212e-06,
+      "loss": 1.0694,
+      "step": 187000
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 8.207070707070708e-06,
+      "loss": 1.0684,
+      "step": 187500
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 8.202020202020202e-06,
+      "loss": 1.0667,
+      "step": 188000
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 8.196969696969698e-06,
+      "loss": 1.0655,
+      "step": 188500
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 8.191919191919193e-06,
+      "loss": 1.0649,
+      "step": 189000
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 8.186868686868687e-06,
+      "loss": 1.0625,
+      "step": 189500
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 1.0615,
+      "step": 190000
+    },
+    {
+      "epoch": 6.19,
+      "eval_loss": 0.9824332594871521,
+      "eval_runtime": 88.8399,
+      "eval_samples_per_second": 893.878,
+      "eval_steps_per_second": 3.501,
+      "step": 190000
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 8.176767676767677e-06,
+      "loss": 1.0608,
+      "step": 190500
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 8.171717171717173e-06,
+      "loss": 1.0598,
+      "step": 191000
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 8.166666666666668e-06,
+      "loss": 1.0585,
+      "step": 191500
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 8.161616161616162e-06,
+      "loss": 1.0579,
+      "step": 192000
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 8.156565656565658e-06,
+      "loss": 1.0568,
+      "step": 192500
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 8.151515151515152e-06,
+      "loss": 1.0562,
+      "step": 193000
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 8.146464646464646e-06,
+      "loss": 1.054,
+      "step": 193500
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 8.141414141414142e-06,
+      "loss": 1.0526,
+      "step": 194000
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 8.136363636363637e-06,
+      "loss": 1.0524,
+      "step": 194500
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 8.131313131313133e-06,
+      "loss": 1.0513,
+      "step": 195000
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 8.126262626262627e-06,
+      "loss": 1.0498,
+      "step": 195500
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 8.121212121212121e-06,
+      "loss": 1.0479,
+      "step": 196000
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 8.116161616161616e-06,
+      "loss": 1.0476,
+      "step": 196500
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 8.111111111111112e-06,
+      "loss": 1.0482,
+      "step": 197000
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 8.106060606060606e-06,
+      "loss": 1.0463,
+      "step": 197500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 8.101010101010102e-06,
+      "loss": 1.0453,
+      "step": 198000
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 8.095959595959598e-06,
+      "loss": 1.0434,
+      "step": 198500
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 8.090909090909092e-06,
+      "loss": 1.0425,
+      "step": 199000
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 8.085858585858586e-06,
+      "loss": 1.041,
+      "step": 199500
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 8.08080808080808e-06,
+      "loss": 1.0406,
+      "step": 200000
+    },
+    {
+      "epoch": 6.51,
+      "eval_loss": 0.9657134413719177,
+      "eval_runtime": 87.9909,
+      "eval_samples_per_second": 902.502,
+      "eval_steps_per_second": 3.534,
+      "step": 200000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 8.075757575757577e-06,
+      "loss": 1.0404,
+      "step": 200500
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 8.070707070707071e-06,
+      "loss": 1.0394,
+      "step": 201000
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 8.065656565656567e-06,
+      "loss": 1.0383,
+      "step": 201500
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 8.060606060606061e-06,
+      "loss": 1.0369,
+      "step": 202000
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 1.0354,
+      "step": 202500
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 8.050505050505052e-06,
+      "loss": 1.0343,
+      "step": 203000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 8.045454545454546e-06,
+      "loss": 1.0338,
+      "step": 203500
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 8.04040404040404e-06,
+      "loss": 1.0321,
+      "step": 204000
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 8.035353535353536e-06,
+      "loss": 1.0315,
+      "step": 204500
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 8.03030303030303e-06,
+      "loss": 1.0313,
+      "step": 205000
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 8.025252525252526e-06,
+      "loss": 1.0303,
+      "step": 205500
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 8.02020202020202e-06,
+      "loss": 1.0292,
+      "step": 206000
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 8.015151515151515e-06,
+      "loss": 1.028,
+      "step": 206500
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 8.010101010101011e-06,
+      "loss": 1.0273,
+      "step": 207000
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 8.005050505050505e-06,
+      "loss": 1.0264,
+      "step": 207500
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0248,
+      "step": 208000
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 7.994949494949496e-06,
+      "loss": 1.0231,
+      "step": 208500
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 7.989898989898992e-06,
+      "loss": 1.0236,
+      "step": 209000
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 7.984848484848486e-06,
+      "loss": 1.0221,
+      "step": 209500
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 7.97979797979798e-06,
+      "loss": 1.0219,
+      "step": 210000
+    },
+    {
+      "epoch": 6.84,
+      "eval_loss": 0.9520924091339111,
+      "eval_runtime": 89.7334,
+      "eval_samples_per_second": 884.977,
+      "eval_steps_per_second": 3.466,
+      "step": 210000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 7.974747474747475e-06,
+      "loss": 1.0195,
+      "step": 210500
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 7.96969696969697e-06,
+      "loss": 1.0197,
+      "step": 211000
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 7.964646464646465e-06,
+      "loss": 1.0178,
+      "step": 211500
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 7.95959595959596e-06,
+      "loss": 1.0177,
+      "step": 212000
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 7.954545454545455e-06,
+      "loss": 1.0166,
+      "step": 212500
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 7.949494949494951e-06,
+      "loss": 1.0153,
+      "step": 213000
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 7.944444444444445e-06,
+      "loss": 1.0145,
+      "step": 213500
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 7.93939393939394e-06,
+      "loss": 1.0142,
+      "step": 214000
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 7.934343434343434e-06,
+      "loss": 1.0136,
+      "step": 214500
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 7.92929292929293e-06,
+      "loss": 1.0124,
+      "step": 215000
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 7.924242424242426e-06,
+      "loss": 1.0116,
+      "step": 215500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 7.91919191919192e-06,
+      "loss": 1.0097,
+      "step": 216000
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 7.914141414141415e-06,
+      "loss": 1.0094,
+      "step": 216500
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 7.909090909090909e-06,
+      "loss": 1.0078,
+      "step": 217000
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 7.904040404040405e-06,
+      "loss": 1.0069,
+      "step": 217500
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 7.898989898989899e-06,
+      "loss": 1.0076,
+      "step": 218000
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 7.893939393939395e-06,
+      "loss": 1.0061,
+      "step": 218500
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 7.88888888888889e-06,
+      "loss": 1.0054,
+      "step": 219000
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 7.883838383838385e-06,
+      "loss": 1.0034,
+      "step": 219500
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 1.0043,
+      "step": 220000
+    },
+    {
+      "epoch": 7.16,
+      "eval_loss": 0.9399166107177734,
+      "eval_runtime": 91.3908,
+      "eval_samples_per_second": 868.928,
+      "eval_steps_per_second": 3.403,
+      "step": 220000
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 7.873737373737374e-06,
+      "loss": 1.0021,
+      "step": 220500
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 7.868686868686868e-06,
+      "loss": 1.0022,
+      "step": 221000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 7.863636363636364e-06,
+      "loss": 1.0009,
+      "step": 221500
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 7.858585858585859e-06,
+      "loss": 1.0004,
+      "step": 222000
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 7.853535353535355e-06,
+      "loss": 0.9988,
+      "step": 222500
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 7.848484848484849e-06,
+      "loss": 0.9988,
+      "step": 223000
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 7.843434343434345e-06,
+      "loss": 0.9983,
+      "step": 223500
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 7.838383838383839e-06,
+      "loss": 0.9971,
+      "step": 224000
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 7.833333333333333e-06,
+      "loss": 0.9956,
+      "step": 224500
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 7.82828282828283e-06,
+      "loss": 0.9954,
+      "step": 225000
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 7.823232323232324e-06,
+      "loss": 0.9942,
+      "step": 225500
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 7.81818181818182e-06,
+      "loss": 0.9932,
+      "step": 226000
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 7.813131313131314e-06,
+      "loss": 0.9927,
+      "step": 226500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 7.808080808080808e-06,
+      "loss": 0.9916,
+      "step": 227000
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 7.803030303030303e-06,
+      "loss": 0.992,
+      "step": 227500
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 7.797979797979799e-06,
+      "loss": 0.9914,
+      "step": 228000
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 7.792929292929293e-06,
+      "loss": 0.9902,
+      "step": 228500
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 7.787878787878789e-06,
+      "loss": 0.989,
+      "step": 229000
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 7.782828282828283e-06,
+      "loss": 0.9896,
+      "step": 229500
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.9876,
+      "step": 230000
+    },
+    {
+      "epoch": 7.49,
+      "eval_loss": 0.9257983565330505,
+      "eval_runtime": 89.899,
+      "eval_samples_per_second": 883.347,
+      "eval_steps_per_second": 3.459,
+      "step": 230000
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 7.772727272727273e-06,
+      "loss": 0.9873,
+      "step": 230500
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 7.767676767676768e-06,
+      "loss": 0.9866,
+      "step": 231000
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 7.762626262626262e-06,
+      "loss": 0.9857,
+      "step": 231500
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 7.757575757575758e-06,
+      "loss": 0.9855,
+      "step": 232000
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 7.752525252525254e-06,
+      "loss": 0.9837,
+      "step": 232500
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 7.747474747474748e-06,
+      "loss": 0.9834,
+      "step": 233000
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 7.742424242424244e-06,
+      "loss": 0.9839,
+      "step": 233500
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 7.737373737373739e-06,
+      "loss": 0.9823,
+      "step": 234000
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 7.732323232323233e-06,
+      "loss": 0.9815,
+      "step": 234500
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.9809,
+      "step": 235000
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 7.722222222222223e-06,
+      "loss": 0.9802,
+      "step": 235500
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 7.717171717171717e-06,
+      "loss": 0.98,
+      "step": 236000
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 7.712121212121213e-06,
+      "loss": 0.9788,
+      "step": 236500
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 7.707070707070708e-06,
+      "loss": 0.9788,
+      "step": 237000
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 7.702020202020202e-06,
+      "loss": 0.9779,
+      "step": 237500
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 7.696969696969696e-06,
+      "loss": 0.9769,
+      "step": 238000
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 7.691919191919192e-06,
+      "loss": 0.9762,
+      "step": 238500
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 7.686868686868687e-06,
+      "loss": 0.9757,
+      "step": 239000
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 7.681818181818183e-06,
+      "loss": 0.9753,
+      "step": 239500
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 7.676767676767677e-06,
+      "loss": 0.9744,
+      "step": 240000
+    },
+    {
+      "epoch": 7.82,
+      "eval_loss": 0.9180858135223389,
+      "eval_runtime": 88.4669,
+      "eval_samples_per_second": 897.647,
+      "eval_steps_per_second": 3.515,
+      "step": 240000
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 7.671717171717173e-06,
+      "loss": 0.9732,
+      "step": 240500
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.9739,
+      "step": 241000
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 7.661616161616162e-06,
+      "loss": 0.9728,
+      "step": 241500
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 7.656565656565658e-06,
+      "loss": 0.9718,
+      "step": 242000
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 7.651515151515152e-06,
+      "loss": 0.971,
+      "step": 242500
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 7.646464646464648e-06,
+      "loss": 0.9707,
+      "step": 243000
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 7.641414141414142e-06,
+      "loss": 0.9698,
+      "step": 243500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 7.636363636363638e-06,
+      "loss": 0.9701,
+      "step": 244000
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 7.631313131313132e-06,
+      "loss": 0.9689,
+      "step": 244500
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 7.6262626262626275e-06,
+      "loss": 0.9683,
+      "step": 245000
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 7.621212121212122e-06,
+      "loss": 0.967,
+      "step": 245500
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 7.616161616161617e-06,
+      "loss": 0.9667,
+      "step": 246000
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 7.611111111111111e-06,
+      "loss": 0.9668,
+      "step": 246500
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 7.606060606060606e-06,
+      "loss": 0.9658,
+      "step": 247000
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 7.6010101010101016e-06,
+      "loss": 0.9655,
+      "step": 247500
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 7.595959595959597e-06,
+      "loss": 0.9648,
+      "step": 248000
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 7.590909090909091e-06,
+      "loss": 0.9642,
+      "step": 248500
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 7.585858585858586e-06,
+      "loss": 0.9638,
+      "step": 249000
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 7.580808080808082e-06,
+      "loss": 0.9635,
+      "step": 249500
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.9624,
+      "step": 250000
+    },
+    {
+      "epoch": 8.14,
+      "eval_loss": 0.9136332869529724,
+      "eval_runtime": 88.8056,
+      "eval_samples_per_second": 894.223,
+      "eval_steps_per_second": 3.502,
+      "step": 250000
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 7.5707070707070716e-06,
+      "loss": 0.9617,
+      "step": 250500
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 7.565656565656566e-06,
+      "loss": 0.9614,
+      "step": 251000
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 7.560606060606062e-06,
+      "loss": 0.9609,
+      "step": 251500
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 0.9604,
+      "step": 252000
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 7.550505050505051e-06,
+      "loss": 0.9603,
+      "step": 252500
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 7.545454545454546e-06,
+      "loss": 0.9592,
+      "step": 253000
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 7.540404040404042e-06,
+      "loss": 0.9587,
+      "step": 253500
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 7.535353535353536e-06,
+      "loss": 0.9578,
+      "step": 254000
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 7.530303030303031e-06,
+      "loss": 0.9575,
+      "step": 254500
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 7.525252525252525e-06,
+      "loss": 0.957,
+      "step": 255000
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 7.520202020202021e-06,
+      "loss": 0.9562,
+      "step": 255500
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 7.515151515151516e-06,
+      "loss": 0.9557,
+      "step": 256000
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 7.510101010101011e-06,
+      "loss": 0.955,
+      "step": 256500
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 7.505050505050505e-06,
+      "loss": 0.9548,
+      "step": 257000
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.9542,
+      "step": 257500
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 7.494949494949496e-06,
+      "loss": 0.9529,
+      "step": 258000
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 7.4898989898989905e-06,
+      "loss": 0.9534,
+      "step": 258500
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 7.484848484848486e-06,
+      "loss": 0.9528,
+      "step": 259000
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 7.47979797979798e-06,
+      "loss": 0.9523,
+      "step": 259500
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 7.474747474747476e-06,
+      "loss": 0.9522,
+      "step": 260000
+    },
+    {
+      "epoch": 8.47,
+      "eval_loss": 0.9024671316146851,
+      "eval_runtime": 87.0285,
+      "eval_samples_per_second": 912.483,
+      "eval_steps_per_second": 3.574,
+      "step": 260000
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 7.46969696969697e-06,
+      "loss": 0.9513,
+      "step": 260500
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 7.464646464646465e-06,
+      "loss": 0.9512,
+      "step": 261000
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 7.45959595959596e-06,
+      "loss": 0.9504,
+      "step": 261500
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 7.454545454545456e-06,
+      "loss": 0.9502,
+      "step": 262000
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 7.44949494949495e-06,
+      "loss": 0.9497,
+      "step": 262500
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 7.444444444444445e-06,
+      "loss": 0.9491,
+      "step": 263000
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 7.439393939393939e-06,
+      "loss": 0.9482,
+      "step": 263500
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 7.434343434343435e-06,
+      "loss": 0.9472,
+      "step": 264000
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 7.42929292929293e-06,
+      "loss": 0.9474,
+      "step": 264500
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 7.424242424242425e-06,
+      "loss": 0.9474,
+      "step": 265000
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 7.419191919191919e-06,
+      "loss": 0.9464,
+      "step": 265500
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 7.414141414141415e-06,
+      "loss": 0.9463,
+      "step": 266000
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 7.40909090909091e-06,
+      "loss": 0.9459,
+      "step": 266500
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 7.4040404040404045e-06,
+      "loss": 0.9462,
+      "step": 267000
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 7.3989898989899e-06,
+      "loss": 0.945,
+      "step": 267500
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 7.393939393939395e-06,
+      "loss": 0.9445,
+      "step": 268000
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 7.38888888888889e-06,
+      "loss": 0.9436,
+      "step": 268500
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 7.383838383838384e-06,
+      "loss": 0.9431,
+      "step": 269000
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 7.378787878787879e-06,
+      "loss": 0.943,
+      "step": 269500
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 7.373737373737374e-06,
+      "loss": 0.9421,
+      "step": 270000
+    },
+    {
+      "epoch": 8.79,
+      "eval_loss": 0.8987886905670166,
+      "eval_runtime": 91.9986,
+      "eval_samples_per_second": 863.187,
+      "eval_steps_per_second": 3.38,
+      "step": 270000
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 7.36868686868687e-06,
+      "loss": 0.9422,
+      "step": 270500
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 7.363636363636364e-06,
+      "loss": 0.9417,
+      "step": 271000
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 7.358585858585859e-06,
+      "loss": 0.9409,
+      "step": 271500
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 7.353535353535353e-06,
+      "loss": 0.9403,
+      "step": 272000
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 7.348484848484849e-06,
+      "loss": 0.9399,
+      "step": 272500
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 7.343434343434344e-06,
+      "loss": 0.9392,
+      "step": 273000
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 7.338383838383839e-06,
+      "loss": 0.9393,
+      "step": 273500
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.9381,
+      "step": 274000
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 7.328282828282829e-06,
+      "loss": 0.9388,
+      "step": 274500
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 7.323232323232324e-06,
+      "loss": 0.9377,
+      "step": 275000
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 7.3181818181818186e-06,
+      "loss": 0.9367,
+      "step": 275500
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 7.3131313131313146e-06,
+      "loss": 0.9372,
+      "step": 276000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 7.308080808080809e-06,
+      "loss": 0.9372,
+      "step": 276500
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 7.303030303030304e-06,
+      "loss": 0.936,
+      "step": 277000
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 7.297979797979798e-06,
+      "loss": 0.9364,
+      "step": 277500
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 7.2929292929292934e-06,
+      "loss": 0.9357,
+      "step": 278000
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 7.287878787878789e-06,
+      "loss": 0.9347,
+      "step": 278500
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 7.282828282828284e-06,
+      "loss": 0.9347,
+      "step": 279000
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 7.277777777777778e-06,
+      "loss": 0.9344,
+      "step": 279500
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.9342,
+      "step": 280000
+    },
+    {
+      "epoch": 9.12,
+      "eval_loss": 0.8905976414680481,
+      "eval_runtime": 91.0407,
+      "eval_samples_per_second": 872.269,
+      "eval_steps_per_second": 3.416,
+      "step": 280000
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 7.2676767676767675e-06,
+      "loss": 0.9337,
+      "step": 280500
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 7.2626262626262635e-06,
+      "loss": 0.933,
+      "step": 281000
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 7.257575757575758e-06,
+      "loss": 0.9325,
+      "step": 281500
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 7.252525252525253e-06,
+      "loss": 0.9321,
+      "step": 282000
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 7.247474747474747e-06,
+      "loss": 0.9319,
+      "step": 282500
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 7.242424242424243e-06,
+      "loss": 0.9319,
+      "step": 283000
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 7.237373737373738e-06,
+      "loss": 0.9306,
+      "step": 283500
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 7.232323232323233e-06,
+      "loss": 0.931,
+      "step": 284000
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 7.227272727272729e-06,
+      "loss": 0.9292,
+      "step": 284500
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.93,
+      "step": 285000
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 7.217171717171718e-06,
+      "loss": 0.9296,
+      "step": 285500
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 7.212121212121212e-06,
+      "loss": 0.9287,
+      "step": 286000
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 7.207070707070708e-06,
+      "loss": 0.9282,
+      "step": 286500
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 7.202020202020203e-06,
+      "loss": 0.9281,
+      "step": 287000
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 7.196969696969698e-06,
+      "loss": 0.9277,
+      "step": 287500
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 7.191919191919192e-06,
+      "loss": 0.9276,
+      "step": 288000
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 7.186868686868688e-06,
+      "loss": 0.9269,
+      "step": 288500
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 7.181818181818182e-06,
+      "loss": 0.9262,
+      "step": 289000
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 7.1767676767676775e-06,
+      "loss": 0.926,
+      "step": 289500
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 7.171717171717172e-06,
+      "loss": 0.926,
+      "step": 290000
+    },
+    {
+      "epoch": 9.44,
+      "eval_loss": 0.8822857737541199,
+      "eval_runtime": 87.516,
+      "eval_samples_per_second": 907.4,
+      "eval_steps_per_second": 3.554,
+      "step": 290000
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 7.166666666666667e-06,
+      "loss": 0.9252,
+      "step": 290500
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 7.161616161616162e-06,
+      "loss": 0.9248,
+      "step": 291000
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 7.156565656565657e-06,
+      "loss": 0.9253,
+      "step": 291500
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 7.151515151515152e-06,
+      "loss": 0.9242,
+      "step": 292000
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 7.146464646464647e-06,
+      "loss": 0.9234,
+      "step": 292500
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 7.141414141414143e-06,
+      "loss": 0.9238,
+      "step": 293000
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 7.136363636363637e-06,
+      "loss": 0.9238,
+      "step": 293500
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 7.131313131313132e-06,
+      "loss": 0.9228,
+      "step": 294000
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 7.126262626262626e-06,
+      "loss": 0.9216,
+      "step": 294500
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 7.121212121212122e-06,
+      "loss": 0.9216,
+      "step": 295000
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 7.116161616161617e-06,
+      "loss": 0.9215,
+      "step": 295500
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 0.9208,
+      "step": 296000
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 7.106060606060606e-06,
+      "loss": 0.9213,
+      "step": 296500
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 7.101010101010102e-06,
+      "loss": 0.9207,
+      "step": 297000
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 7.095959595959596e-06,
+      "loss": 0.9206,
+      "step": 297500
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 7.0909090909090916e-06,
+      "loss": 0.92,
+      "step": 298000
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 7.085858585858586e-06,
+      "loss": 0.9194,
+      "step": 298500
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 7.080808080808082e-06,
+      "loss": 0.9191,
+      "step": 299000
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 7.075757575757576e-06,
+      "loss": 0.9191,
+      "step": 299500
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 7.070707070707071e-06,
+      "loss": 0.9184,
+      "step": 300000
+    },
+    {
+      "epoch": 9.77,
+      "eval_loss": 0.8790136575698853,
+      "eval_runtime": 89.0584,
+      "eval_samples_per_second": 891.684,
+      "eval_steps_per_second": 3.492,
+      "step": 300000
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 7.065656565656566e-06,
+      "loss": 0.9183,
+      "step": 300500
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 7.060606060606061e-06,
+      "loss": 0.918,
+      "step": 301000
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 7.055555555555557e-06,
+      "loss": 0.9174,
+      "step": 301500
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 7.050505050505051e-06,
+      "loss": 0.9162,
+      "step": 302000
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 7.045454545454546e-06,
+      "loss": 0.9162,
+      "step": 302500
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 7.0404040404040404e-06,
+      "loss": 0.9159,
+      "step": 303000
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 7.0353535353535364e-06,
+      "loss": 0.9158,
+      "step": 303500
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 7.030303030303031e-06,
+      "loss": 0.9154,
+      "step": 304000
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 7.025252525252526e-06,
+      "loss": 0.9154,
+      "step": 304500
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 7.02020202020202e-06,
+      "loss": 0.9144,
+      "step": 305000
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 7.015151515151516e-06,
+      "loss": 0.9151,
+      "step": 305500
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 7.0101010101010105e-06,
+      "loss": 0.9142,
+      "step": 306000
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 7.005050505050506e-06,
+      "loss": 0.9145,
+      "step": 306500
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 7e-06,
+      "loss": 0.9138,
+      "step": 307000
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 6.994949494949496e-06,
+      "loss": 0.9128,
+      "step": 307500
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 6.98989898989899e-06,
+      "loss": 0.9138,
+      "step": 308000
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 6.984848484848485e-06,
+      "loss": 0.913,
+      "step": 308500
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 6.979797979797981e-06,
+      "loss": 0.9125,
+      "step": 309000
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 6.974747474747476e-06,
+      "loss": 0.9117,
+      "step": 309500
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.9114,
+      "step": 310000
+    },
+    {
+      "epoch": 10.09,
+      "eval_loss": 0.8732441067695618,
+      "eval_runtime": 91.0609,
+      "eval_samples_per_second": 872.076,
+      "eval_steps_per_second": 3.415,
+      "step": 310000
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 6.964646464646465e-06,
+      "loss": 0.9112,
+      "step": 310500
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 6.95959595959596e-06,
+      "loss": 0.9108,
+      "step": 311000
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 6.954545454545455e-06,
+      "loss": 0.9102,
+      "step": 311500
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 6.9494949494949505e-06,
+      "loss": 0.9105,
+      "step": 312000
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.9103,
+      "step": 312500
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 6.93939393939394e-06,
+      "loss": 0.9101,
+      "step": 313000
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 6.934343434343434e-06,
+      "loss": 0.9097,
+      "step": 313500
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 6.92929292929293e-06,
+      "loss": 0.9086,
+      "step": 314000
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 6.9242424242424245e-06,
+      "loss": 0.9089,
+      "step": 314500
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 6.91919191919192e-06,
+      "loss": 0.9082,
+      "step": 315000
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 6.914141414141414e-06,
+      "loss": 0.9082,
+      "step": 315500
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 6.90909090909091e-06,
+      "loss": 0.9077,
+      "step": 316000
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 6.904040404040404e-06,
+      "loss": 0.9072,
+      "step": 316500
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 6.898989898989899e-06,
+      "loss": 0.907,
+      "step": 317000
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 6.893939393939395e-06,
+      "loss": 0.9067,
+      "step": 317500
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 0.9067,
+      "step": 318000
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 6.883838383838385e-06,
+      "loss": 0.9054,
+      "step": 318500
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 6.878787878787879e-06,
+      "loss": 0.9051,
+      "step": 319000
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 6.873737373737375e-06,
+      "loss": 0.905,
+      "step": 319500
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 6.868686868686869e-06,
+      "loss": 0.9047,
+      "step": 320000
+    },
+    {
+      "epoch": 10.42,
+      "eval_loss": 0.8707190155982971,
+      "eval_runtime": 92.9604,
+      "eval_samples_per_second": 854.256,
+      "eval_steps_per_second": 3.346,
+      "step": 320000
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 6.8636363636363645e-06,
+      "loss": 0.9048,
+      "step": 320500
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 6.858585858585859e-06,
+      "loss": 0.9051,
+      "step": 321000
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 6.853535353535354e-06,
+      "loss": 0.9039,
+      "step": 321500
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 6.848484848484849e-06,
+      "loss": 0.9039,
+      "step": 322000
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 6.843434343434344e-06,
+      "loss": 0.9034,
+      "step": 322500
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 6.8383838383838386e-06,
+      "loss": 0.9028,
+      "step": 323000
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 6.833333333333334e-06,
+      "loss": 0.9025,
+      "step": 323500
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 6.828282828282828e-06,
+      "loss": 0.9027,
+      "step": 324000
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 6.823232323232324e-06,
+      "loss": 0.9017,
+      "step": 324500
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.9026,
+      "step": 325000
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 6.813131313131313e-06,
+      "loss": 0.902,
+      "step": 325500
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 6.808080808080809e-06,
+      "loss": 0.9021,
+      "step": 326000
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 6.803030303030304e-06,
+      "loss": 0.9008,
+      "step": 326500
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 6.797979797979799e-06,
+      "loss": 0.9006,
+      "step": 327000
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 6.792929292929293e-06,
+      "loss": 0.9005,
+      "step": 327500
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 6.787878787878789e-06,
+      "loss": 0.8999,
+      "step": 328000
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 6.7828282828282834e-06,
+      "loss": 0.9,
+      "step": 328500
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 6.777777777777779e-06,
+      "loss": 0.9004,
+      "step": 329000
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 6.772727272727273e-06,
+      "loss": 0.8989,
+      "step": 329500
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 6.767676767676769e-06,
+      "loss": 0.8993,
+      "step": 330000
+    },
+    {
+      "epoch": 10.75,
+      "eval_loss": 0.8648310303688049,
+      "eval_runtime": 90.2528,
+      "eval_samples_per_second": 879.884,
+      "eval_steps_per_second": 3.446,
+      "step": 330000
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 6.762626262626263e-06,
+      "loss": 0.8983,
+      "step": 330500
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 6.757575757575758e-06,
+      "loss": 0.8989,
+      "step": 331000
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 6.752525252525253e-06,
+      "loss": 0.8984,
+      "step": 331500
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 6.747474747474749e-06,
+      "loss": 0.898,
+      "step": 332000
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 6.742424242424243e-06,
+      "loss": 0.8975,
+      "step": 332500
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 6.737373737373738e-06,
+      "loss": 0.8973,
+      "step": 333000
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 6.732323232323232e-06,
+      "loss": 0.8971,
+      "step": 333500
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 6.7272727272727275e-06,
+      "loss": 0.8971,
+      "step": 334000
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 6.7222222222222235e-06,
+      "loss": 0.8962,
+      "step": 334500
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 6.717171717171718e-06,
+      "loss": 0.8963,
+      "step": 335000
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 6.712121212121213e-06,
+      "loss": 0.8961,
+      "step": 335500
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 6.707070707070707e-06,
+      "loss": 0.8961,
+      "step": 336000
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 6.702020202020203e-06,
+      "loss": 0.895,
+      "step": 336500
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 6.6969696969696975e-06,
+      "loss": 0.8949,
+      "step": 337000
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 6.691919191919193e-06,
+      "loss": 0.8944,
+      "step": 337500
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 6.686868686868687e-06,
+      "loss": 0.8942,
+      "step": 338000
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 6.681818181818183e-06,
+      "loss": 0.8942,
+      "step": 338500
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 6.676767676767677e-06,
+      "loss": 0.8944,
+      "step": 339000
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 6.671717171717172e-06,
+      "loss": 0.8938,
+      "step": 339500
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8938,
+      "step": 340000
+    },
+    {
+      "epoch": 11.07,
+      "eval_loss": 0.8565430641174316,
+      "eval_runtime": 91.4844,
+      "eval_samples_per_second": 868.039,
+      "eval_steps_per_second": 3.399,
+      "step": 340000
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 6.661616161616163e-06,
+      "loss": 0.8931,
+      "step": 340500
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 6.656565656565657e-06,
+      "loss": 0.8921,
+      "step": 341000
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 6.651515151515152e-06,
+      "loss": 0.8925,
+      "step": 341500
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 6.646464646464646e-06,
+      "loss": 0.8921,
+      "step": 342000
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 6.641414141414142e-06,
+      "loss": 0.8919,
+      "step": 342500
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 6.6363636363636375e-06,
+      "loss": 0.8913,
+      "step": 343000
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 6.631313131313132e-06,
+      "loss": 0.8913,
+      "step": 343500
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 6.626262626262627e-06,
+      "loss": 0.8904,
+      "step": 344000
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 6.621212121212121e-06,
+      "loss": 0.8906,
+      "step": 344500
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 6.616161616161617e-06,
+      "loss": 0.8901,
+      "step": 345000
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 6.6111111111111115e-06,
+      "loss": 0.8903,
+      "step": 345500
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 6.606060606060607e-06,
+      "loss": 0.8897,
+      "step": 346000
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 6.601010101010101e-06,
+      "loss": 0.8892,
+      "step": 346500
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 6.595959595959597e-06,
+      "loss": 0.8893,
+      "step": 347000
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 6.590909090909091e-06,
+      "loss": 0.8893,
+      "step": 347500
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 6.585858585858586e-06,
+      "loss": 0.8889,
+      "step": 348000
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 6.580808080808081e-06,
+      "loss": 0.888,
+      "step": 348500
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 6.575757575757577e-06,
+      "loss": 0.888,
+      "step": 349000
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 6.570707070707071e-06,
+      "loss": 0.8875,
+      "step": 349500
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 6.565656565656566e-06,
+      "loss": 0.8867,
+      "step": 350000
+    },
+    {
+      "epoch": 11.4,
+      "eval_loss": 0.8537823557853699,
+      "eval_runtime": 91.6919,
+      "eval_samples_per_second": 866.075,
+      "eval_steps_per_second": 3.392,
+      "step": 350000
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 6.56060606060606e-06,
+      "loss": 0.8865,
+      "step": 350500
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 6.555555555555556e-06,
+      "loss": 0.8867,
+      "step": 351000
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 6.5505050505050516e-06,
+      "loss": 0.8868,
+      "step": 351500
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 6.545454545454546e-06,
+      "loss": 0.8861,
+      "step": 352000
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 6.540404040404042e-06,
+      "loss": 0.8854,
+      "step": 352500
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 6.535353535353536e-06,
+      "loss": 0.8856,
+      "step": 353000
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 6.530303030303031e-06,
+      "loss": 0.8856,
+      "step": 353500
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 6.525252525252526e-06,
+      "loss": 0.8851,
+      "step": 354000
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 6.520202020202021e-06,
+      "loss": 0.8844,
+      "step": 354500
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 6.515151515151516e-06,
+      "loss": 0.8842,
+      "step": 355000
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 6.510101010101011e-06,
+      "loss": 0.8838,
+      "step": 355500
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 6.505050505050505e-06,
+      "loss": 0.8834,
+      "step": 356000
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.8828,
+      "step": 356500
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 6.494949494949495e-06,
+      "loss": 0.8829,
+      "step": 357000
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 6.489898989898991e-06,
+      "loss": 0.883,
+      "step": 357500
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 6.484848484848485e-06,
+      "loss": 0.8826,
+      "step": 358000
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 6.47979797979798e-06,
+      "loss": 0.8827,
+      "step": 358500
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 6.4747474747474745e-06,
+      "loss": 0.8824,
+      "step": 359000
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 6.4696969696969705e-06,
+      "loss": 0.8815,
+      "step": 359500
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 6.464646464646466e-06,
+      "loss": 0.881,
+      "step": 360000
+    },
+    {
+      "epoch": 11.72,
+      "eval_loss": 0.8482265472412109,
+      "eval_runtime": 90.1088,
+      "eval_samples_per_second": 881.29,
+      "eval_steps_per_second": 3.451,
+      "step": 360000
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 6.45959595959596e-06,
+      "loss": 0.8809,
+      "step": 360500
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 6.454545454545456e-06,
+      "loss": 0.8808,
+      "step": 361000
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 6.44949494949495e-06,
+      "loss": 0.8804,
+      "step": 361500
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 0.8803,
+      "step": 362000
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 6.43939393939394e-06,
+      "loss": 0.8799,
+      "step": 362500
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 6.434343434343436e-06,
+      "loss": 0.8795,
+      "step": 363000
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 6.42929292929293e-06,
+      "loss": 0.88,
+      "step": 363500
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 6.424242424242425e-06,
+      "loss": 0.8782,
+      "step": 364000
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 6.419191919191919e-06,
+      "loss": 0.8792,
+      "step": 364500
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 6.4141414141414145e-06,
+      "loss": 0.8784,
+      "step": 365000
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 6.40909090909091e-06,
+      "loss": 0.8781,
+      "step": 365500
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 6.404040404040405e-06,
+      "loss": 0.8786,
+      "step": 366000
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 6.398989898989899e-06,
+      "loss": 0.8776,
+      "step": 366500
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 6.393939393939394e-06,
+      "loss": 0.8779,
+      "step": 367000
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.877,
+      "step": 367500
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 6.3838383838383845e-06,
+      "loss": 0.8773,
+      "step": 368000
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 6.37878787878788e-06,
+      "loss": 0.8764,
+      "step": 368500
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 6.373737373737374e-06,
+      "loss": 0.8766,
+      "step": 369000
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 6.36868686868687e-06,
+      "loss": 0.8764,
+      "step": 369500
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.8756,
+      "step": 370000
+    },
+    {
+      "epoch": 12.05,
+      "eval_loss": 0.842664361000061,
+      "eval_runtime": 89.679,
+      "eval_samples_per_second": 885.514,
+      "eval_steps_per_second": 3.468,
+      "step": 370000
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 6.358585858585859e-06,
+      "loss": 0.8757,
+      "step": 370500
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 6.353535353535354e-06,
+      "loss": 0.8754,
+      "step": 371000
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 6.34848484848485e-06,
+      "loss": 0.8751,
+      "step": 371500
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 6.343434343434344e-06,
+      "loss": 0.8745,
+      "step": 372000
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 6.338383838383839e-06,
+      "loss": 0.8755,
+      "step": 372500
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.8746,
+      "step": 373000
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 6.328282828282829e-06,
+      "loss": 0.8745,
+      "step": 373500
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 6.323232323232324e-06,
+      "loss": 0.8742,
+      "step": 374000
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 6.318181818181819e-06,
+      "loss": 0.8738,
+      "step": 374500
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 6.313131313131313e-06,
+      "loss": 0.8734,
+      "step": 375000
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 6.308080808080809e-06,
+      "loss": 0.873,
+      "step": 375500
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 6.303030303030303e-06,
+      "loss": 0.873,
+      "step": 376000
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 6.2979797979797986e-06,
+      "loss": 0.8729,
+      "step": 376500
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 6.292929292929294e-06,
+      "loss": 0.8722,
+      "step": 377000
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 6.287878787878788e-06,
+      "loss": 0.872,
+      "step": 377500
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 6.282828282828284e-06,
+      "loss": 0.872,
+      "step": 378000
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 6.277777777777778e-06,
+      "loss": 0.8715,
+      "step": 378500
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 6.2727272727272734e-06,
+      "loss": 0.8714,
+      "step": 379000
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 6.267676767676768e-06,
+      "loss": 0.8718,
+      "step": 379500
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 6.262626262626264e-06,
+      "loss": 0.8713,
+      "step": 380000
+    },
+    {
+      "epoch": 12.37,
+      "eval_loss": 0.8392786383628845,
+      "eval_runtime": 91.281,
+      "eval_samples_per_second": 869.973,
+      "eval_steps_per_second": 3.407,
+      "step": 380000
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 6.257575757575758e-06,
+      "loss": 0.8704,
+      "step": 380500
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 6.252525252525253e-06,
+      "loss": 0.871,
+      "step": 381000
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 6.2474747474747474e-06,
+      "loss": 0.8705,
+      "step": 381500
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 6.2424242424242434e-06,
+      "loss": 0.8699,
+      "step": 382000
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 6.237373737373738e-06,
+      "loss": 0.8698,
+      "step": 382500
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 6.232323232323233e-06,
+      "loss": 0.87,
+      "step": 383000
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 6.227272727272727e-06,
+      "loss": 0.8696,
+      "step": 383500
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 0.8691,
+      "step": 384000
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 6.2171717171717175e-06,
+      "loss": 0.8687,
+      "step": 384500
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 6.212121212121213e-06,
+      "loss": 0.8683,
+      "step": 385000
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 6.207070707070707e-06,
+      "loss": 0.8683,
+      "step": 385500
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 6.202020202020203e-06,
+      "loss": 0.8683,
+      "step": 386000
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 6.196969696969698e-06,
+      "loss": 0.8675,
+      "step": 386500
+    },
+    {
+      "epoch": 12.6,
+      "learning_rate": 6.191919191919192e-06,
+      "loss": 0.8675,
+      "step": 387000
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 6.1868686868686875e-06,
+      "loss": 0.8671,
+      "step": 387500
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 6.181818181818182e-06,
+      "loss": 0.8675,
+      "step": 388000
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 6.176767676767678e-06,
+      "loss": 0.8675,
+      "step": 388500
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 6.171717171717172e-06,
+      "loss": 0.8672,
+      "step": 389000
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 6.166666666666667e-06,
+      "loss": 0.8665,
+      "step": 389500
+    },
+    {
+      "epoch": 12.7,
+      "learning_rate": 6.1616161616161615e-06,
+      "loss": 0.8666,
+      "step": 390000
+    },
+    {
+      "epoch": 12.7,
+      "eval_loss": 0.8361514806747437,
+      "eval_runtime": 88.2176,
+      "eval_samples_per_second": 900.183,
+      "eval_steps_per_second": 3.525,
+      "step": 390000
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 6.1565656565656575e-06,
+      "loss": 0.8666,
+      "step": 390500
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 6.151515151515152e-06,
+      "loss": 0.8662,
+      "step": 391000
+    },
+    {
+      "epoch": 12.75,
+      "learning_rate": 6.146464646464647e-06,
+      "loss": 0.8659,
+      "step": 391500
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 6.141414141414141e-06,
+      "loss": 0.8662,
+      "step": 392000
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 6.136363636363637e-06,
+      "loss": 0.866,
+      "step": 392500
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 6.1313131313131315e-06,
+      "loss": 0.8651,
+      "step": 393000
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 6.126262626262627e-06,
+      "loss": 0.8647,
+      "step": 393500
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 6.121212121212121e-06,
+      "loss": 0.8647,
+      "step": 394000
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 6.116161616161617e-06,
+      "loss": 0.8649,
+      "step": 394500
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.8641,
+      "step": 395000
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 6.106060606060606e-06,
+      "loss": 0.8637,
+      "step": 395500
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 6.1010101010101015e-06,
+      "loss": 0.8637,
+      "step": 396000
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 6.095959595959597e-06,
+      "loss": 0.8636,
+      "step": 396500
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 6.090909090909092e-06,
+      "loss": 0.8634,
+      "step": 397000
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 6.085858585858586e-06,
+      "loss": 0.8637,
+      "step": 397500
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 6.080808080808081e-06,
+      "loss": 0.863,
+      "step": 398000
+    },
+    {
+      "epoch": 12.98,
+      "learning_rate": 6.0757575757575755e-06,
+      "loss": 0.8627,
+      "step": 398500
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 6.0707070707070715e-06,
+      "loss": 0.8623,
+      "step": 399000
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 6.065656565656566e-06,
+      "loss": 0.8623,
+      "step": 399500
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.8623,
+      "step": 400000
+    },
+    {
+      "epoch": 13.03,
+      "eval_loss": 0.832267165184021,
+      "eval_runtime": 90.9314,
+      "eval_samples_per_second": 873.318,
+      "eval_steps_per_second": 3.42,
+      "step": 400000
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 6.055555555555555e-06,
+      "loss": 0.8615,
+      "step": 400500
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 6.050505050505051e-06,
+      "loss": 0.8617,
+      "step": 401000
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 6.0454545454545456e-06,
+      "loss": 0.8615,
+      "step": 401500
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 6.040404040404041e-06,
+      "loss": 0.8618,
+      "step": 402000
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 6.035353535353535e-06,
+      "loss": 0.8611,
+      "step": 402500
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 6.030303030303031e-06,
+      "loss": 0.8611,
+      "step": 403000
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 6.025252525252526e-06,
+      "loss": 0.86,
+      "step": 403500
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 6.0202020202020204e-06,
+      "loss": 0.8604,
+      "step": 404000
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 6.015151515151516e-06,
+      "loss": 0.8607,
+      "step": 404500
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 6.010101010101011e-06,
+      "loss": 0.8601,
+      "step": 405000
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 6.005050505050506e-06,
+      "loss": 0.8596,
+      "step": 405500
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 6e-06,
+      "loss": 0.8596,
+      "step": 406000
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 5.994949494949496e-06,
+      "loss": 0.8595,
+      "step": 406500
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 5.9898989898989904e-06,
+      "loss": 0.8591,
+      "step": 407000
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 5.984848484848486e-06,
+      "loss": 0.8594,
+      "step": 407500
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 5.97979797979798e-06,
+      "loss": 0.8587,
+      "step": 408000
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 5.974747474747475e-06,
+      "loss": 0.8585,
+      "step": 408500
+    },
+    {
+      "epoch": 13.32,
+      "learning_rate": 5.96969696969697e-06,
+      "loss": 0.8586,
+      "step": 409000
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 5.964646464646465e-06,
+      "loss": 0.8582,
+      "step": 409500
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 5.95959595959596e-06,
+      "loss": 0.8578,
+      "step": 410000
+    },
+    {
+      "epoch": 13.35,
+      "eval_loss": 0.8293071389198303,
+      "eval_runtime": 90.0727,
+      "eval_samples_per_second": 881.643,
+      "eval_steps_per_second": 3.453,
+      "step": 410000
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 5.954545454545455e-06,
+      "loss": 0.8582,
+      "step": 410500
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 5.949494949494949e-06,
+      "loss": 0.858,
+      "step": 411000
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 5.944444444444445e-06,
+      "loss": 0.8582,
+      "step": 411500
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 5.93939393939394e-06,
+      "loss": 0.8575,
+      "step": 412000
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 5.9343434343434345e-06,
+      "loss": 0.8575,
+      "step": 412500
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 5.9292929292929305e-06,
+      "loss": 0.8566,
+      "step": 413000
+    },
+    {
+      "epoch": 13.46,
+      "learning_rate": 5.924242424242425e-06,
+      "loss": 0.8575,
+      "step": 413500
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 5.91919191919192e-06,
+      "loss": 0.8563,
+      "step": 414000
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 5.914141414141414e-06,
+      "loss": 0.8565,
+      "step": 414500
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.8561,
+      "step": 415000
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 5.9040404040404045e-06,
+      "loss": 0.8554,
+      "step": 415500
+    },
+    {
+      "epoch": 13.55,
+      "learning_rate": 5.8989898989899e-06,
+      "loss": 0.8558,
+      "step": 416000
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 5.893939393939394e-06,
+      "loss": 0.8555,
+      "step": 416500
+    },
+    {
+      "epoch": 13.58,
+      "learning_rate": 5.88888888888889e-06,
+      "loss": 0.8558,
+      "step": 417000
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 5.883838383838384e-06,
+      "loss": 0.8552,
+      "step": 417500
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 5.878787878787879e-06,
+      "loss": 0.8546,
+      "step": 418000
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 5.873737373737374e-06,
+      "loss": 0.8551,
+      "step": 418500
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 5.868686868686869e-06,
+      "loss": 0.8545,
+      "step": 419000
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 5.863636363636364e-06,
+      "loss": 0.8549,
+      "step": 419500
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 5.858585858585859e-06,
+      "loss": 0.8536,
+      "step": 420000
+    },
+    {
+      "epoch": 13.68,
+      "eval_loss": 0.8251619935035706,
+      "eval_runtime": 88.5613,
+      "eval_samples_per_second": 896.69,
+      "eval_steps_per_second": 3.512,
+      "step": 420000
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 5.853535353535354e-06,
+      "loss": 0.8547,
+      "step": 420500
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 5.8484848484848485e-06,
+      "loss": 0.8536,
+      "step": 421000
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 5.8434343434343445e-06,
+      "loss": 0.8538,
+      "step": 421500
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 5.838383838383839e-06,
+      "loss": 0.8532,
+      "step": 422000
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.8535,
+      "step": 422500
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 5.828282828282828e-06,
+      "loss": 0.8528,
+      "step": 423000
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 5.823232323232324e-06,
+      "loss": 0.8527,
+      "step": 423500
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 5.8181818181818185e-06,
+      "loss": 0.8533,
+      "step": 424000
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 5.813131313131314e-06,
+      "loss": 0.8523,
+      "step": 424500
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 5.808080808080808e-06,
+      "loss": 0.8524,
+      "step": 425000
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 5.803030303030304e-06,
+      "loss": 0.8525,
+      "step": 425500
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 5.797979797979798e-06,
+      "loss": 0.8525,
+      "step": 426000
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 5.792929292929293e-06,
+      "loss": 0.8518,
+      "step": 426500
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 5.787878787878788e-06,
+      "loss": 0.8514,
+      "step": 427000
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 5.782828282828284e-06,
+      "loss": 0.8515,
+      "step": 427500
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 0.8508,
+      "step": 428000
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 5.772727272727273e-06,
+      "loss": 0.8512,
+      "step": 428500
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 5.767676767676768e-06,
+      "loss": 0.8506,
+      "step": 429000
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 5.762626262626263e-06,
+      "loss": 0.8511,
+      "step": 429500
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.8505,
+      "step": 430000
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 0.8214622139930725,
+      "eval_runtime": 91.6492,
+      "eval_samples_per_second": 866.478,
+      "eval_steps_per_second": 3.393,
+      "step": 430000
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 5.752525252525253e-06,
+      "loss": 0.8508,
+      "step": 430500
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 5.747474747474748e-06,
+      "loss": 0.8502,
+      "step": 431000
+    },
+    {
+      "epoch": 14.05,
+      "learning_rate": 5.742424242424242e-06,
+      "loss": 0.8499,
+      "step": 431500
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 5.737373737373738e-06,
+      "loss": 0.8496,
+      "step": 432000
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 5.732323232323233e-06,
+      "loss": 0.8497,
+      "step": 432500
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 5.727272727272728e-06,
+      "loss": 0.8494,
+      "step": 433000
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 5.722222222222222e-06,
+      "loss": 0.8488,
+      "step": 433500
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 5.717171717171718e-06,
+      "loss": 0.8482,
+      "step": 434000
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 5.712121212121212e-06,
+      "loss": 0.8489,
+      "step": 434500
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 5.7070707070707075e-06,
+      "loss": 0.8489,
+      "step": 435000
+    },
+    {
+      "epoch": 14.18,
+      "learning_rate": 5.702020202020202e-06,
+      "loss": 0.8484,
+      "step": 435500
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 5.696969696969698e-06,
+      "loss": 0.8482,
+      "step": 436000
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 5.691919191919192e-06,
+      "loss": 0.848,
+      "step": 436500
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 5.686868686868687e-06,
+      "loss": 0.8477,
+      "step": 437000
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 5.681818181818183e-06,
+      "loss": 0.8482,
+      "step": 437500
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 5.6767676767676775e-06,
+      "loss": 0.8478,
+      "step": 438000
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 5.671717171717173e-06,
+      "loss": 0.8472,
+      "step": 438500
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 0.8471,
+      "step": 439000
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 5.661616161616162e-06,
+      "loss": 0.8474,
+      "step": 439500
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 5.656565656565657e-06,
+      "loss": 0.8468,
+      "step": 440000
+    },
+    {
+      "epoch": 14.33,
+      "eval_loss": 0.8191090226173401,
+      "eval_runtime": 90.6354,
+      "eval_samples_per_second": 876.169,
+      "eval_steps_per_second": 3.431,
+      "step": 440000
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 5.651515151515152e-06,
+      "loss": 0.847,
+      "step": 440500
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 5.646464646464647e-06,
+      "loss": 0.8464,
+      "step": 441000
+    },
+    {
+      "epoch": 14.38,
+      "learning_rate": 5.641414141414142e-06,
+      "loss": 0.8464,
+      "step": 441500
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 5.636363636363636e-06,
+      "loss": 0.8463,
+      "step": 442000
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 5.631313131313132e-06,
+      "loss": 0.8457,
+      "step": 442500
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 5.626262626262626e-06,
+      "loss": 0.846,
+      "step": 443000
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 5.6212121212121215e-06,
+      "loss": 0.8456,
+      "step": 443500
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 5.616161616161616e-06,
+      "loss": 0.8448,
+      "step": 444000
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 5.611111111111112e-06,
+      "loss": 0.8457,
+      "step": 444500
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 5.606060606060606e-06,
+      "loss": 0.8455,
+      "step": 445000
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 5.601010101010101e-06,
+      "loss": 0.845,
+      "step": 445500
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 5.595959595959597e-06,
+      "loss": 0.8447,
+      "step": 446000
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 5.5909090909090915e-06,
+      "loss": 0.8451,
+      "step": 446500
+    },
+    {
+      "epoch": 14.56,
+      "learning_rate": 5.585858585858587e-06,
+      "loss": 0.8442,
+      "step": 447000
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 5.580808080808081e-06,
+      "loss": 0.8443,
+      "step": 447500
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 5.575757575757577e-06,
+      "loss": 0.8442,
+      "step": 448000
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 5.570707070707071e-06,
+      "loss": 0.844,
+      "step": 448500
+    },
+    {
+      "epoch": 14.62,
+      "learning_rate": 5.565656565656566e-06,
+      "loss": 0.8441,
+      "step": 449000
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 5.560606060606061e-06,
+      "loss": 0.8434,
+      "step": 449500
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.8437,
+      "step": 450000
+    },
+    {
+      "epoch": 14.65,
+      "eval_loss": 0.8145061135292053,
+      "eval_runtime": 90.5256,
+      "eval_samples_per_second": 877.233,
+      "eval_steps_per_second": 3.435,
+      "step": 450000
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 5.550505050505051e-06,
+      "loss": 0.8433,
+      "step": 450500
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 5.545454545454546e-06,
+      "loss": 0.8435,
+      "step": 451000
+    },
+    {
+      "epoch": 14.7,
+      "learning_rate": 5.54040404040404e-06,
+      "loss": 0.8427,
+      "step": 451500
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 5.5353535353535355e-06,
+      "loss": 0.8424,
+      "step": 452000
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 5.530303030303031e-06,
+      "loss": 0.8433,
+      "step": 452500
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 5.525252525252526e-06,
+      "loss": 0.8424,
+      "step": 453000
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 5.52020202020202e-06,
+      "loss": 0.8425,
+      "step": 453500
+    },
+    {
+      "epoch": 14.78,
+      "learning_rate": 5.515151515151515e-06,
+      "loss": 0.8418,
+      "step": 454000
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 5.510101010101011e-06,
+      "loss": 0.8421,
+      "step": 454500
+    },
+    {
+      "epoch": 14.82,
+      "learning_rate": 5.5050505050505056e-06,
+      "loss": 0.8417,
+      "step": 455000
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.8412,
+      "step": 455500
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 5.494949494949495e-06,
+      "loss": 0.8412,
+      "step": 456000
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 5.489898989898991e-06,
+      "loss": 0.8417,
+      "step": 456500
+    },
+    {
+      "epoch": 14.88,
+      "learning_rate": 5.484848484848485e-06,
+      "loss": 0.8412,
+      "step": 457000
+    },
+    {
+      "epoch": 14.9,
+      "learning_rate": 5.4797979797979804e-06,
+      "loss": 0.8415,
+      "step": 457500
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 5.474747474747475e-06,
+      "loss": 0.8407,
+      "step": 458000
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 5.469696969696971e-06,
+      "loss": 0.8402,
+      "step": 458500
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 5.464646464646465e-06,
+      "loss": 0.8405,
+      "step": 459000
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 5.45959595959596e-06,
+      "loss": 0.8401,
+      "step": 459500
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.8396,
+      "step": 460000
+    },
+    {
+      "epoch": 14.98,
+      "eval_loss": 0.8129719495773315,
+      "eval_runtime": 88.2583,
+      "eval_samples_per_second": 899.768,
+      "eval_steps_per_second": 3.524,
+      "step": 460000
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 5.4494949494949504e-06,
+      "loss": 0.8402,
+      "step": 460500
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 5.444444444444445e-06,
+      "loss": 0.8397,
+      "step": 461000
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 5.43939393939394e-06,
+      "loss": 0.8401,
+      "step": 461500
+    },
+    {
+      "epoch": 15.04,
+      "learning_rate": 5.434343434343434e-06,
+      "loss": 0.8389,
+      "step": 462000
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 5.429292929292929e-06,
+      "loss": 0.8388,
+      "step": 462500
+    },
+    {
+      "epoch": 15.08,
+      "learning_rate": 5.424242424242425e-06,
+      "loss": 0.839,
+      "step": 463000
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 5.41919191919192e-06,
+      "loss": 0.8382,
+      "step": 463500
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 5.414141414141415e-06,
+      "loss": 0.8387,
+      "step": 464000
+    },
+    {
+      "epoch": 15.13,
+      "learning_rate": 5.409090909090909e-06,
+      "loss": 0.8384,
+      "step": 464500
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 5.404040404040405e-06,
+      "loss": 0.8385,
+      "step": 465000
+    },
+    {
+      "epoch": 15.16,
+      "learning_rate": 5.398989898989899e-06,
+      "loss": 0.8386,
+      "step": 465500
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 5.3939393939393945e-06,
+      "loss": 0.8383,
+      "step": 466000
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 5.388888888888889e-06,
+      "loss": 0.8376,
+      "step": 466500
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 5.383838383838385e-06,
+      "loss": 0.8381,
+      "step": 467000
+    },
+    {
+      "epoch": 15.22,
+      "learning_rate": 5.378787878787879e-06,
+      "loss": 0.8379,
+      "step": 467500
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 5.373737373737374e-06,
+      "loss": 0.8377,
+      "step": 468000
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 5.3686868686868685e-06,
+      "loss": 0.8374,
+      "step": 468500
+    },
+    {
+      "epoch": 15.27,
+      "learning_rate": 5.3636363636363645e-06,
+      "loss": 0.837,
+      "step": 469000
+    },
+    {
+      "epoch": 15.29,
+      "learning_rate": 5.358585858585859e-06,
+      "loss": 0.8364,
+      "step": 469500
+    },
+    {
+      "epoch": 15.3,
+      "learning_rate": 5.353535353535354e-06,
+      "loss": 0.8371,
+      "step": 470000
+    },
+    {
+      "epoch": 15.3,
+      "eval_loss": 0.8082245588302612,
+      "eval_runtime": 89.0044,
+      "eval_samples_per_second": 892.226,
+      "eval_steps_per_second": 3.494,
+      "step": 470000
+    },
+    {
+      "epoch": 15.32,
+      "learning_rate": 5.348484848484848e-06,
+      "loss": 0.8364,
+      "step": 470500
+    },
+    {
+      "epoch": 15.34,
+      "learning_rate": 5.343434343434344e-06,
+      "loss": 0.8365,
+      "step": 471000
+    },
+    {
+      "epoch": 15.35,
+      "learning_rate": 5.338383838383839e-06,
+      "loss": 0.836,
+      "step": 471500
+    },
+    {
+      "epoch": 15.37,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.8365,
+      "step": 472000
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 5.328282828282829e-06,
+      "loss": 0.8355,
+      "step": 472500
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 5.323232323232324e-06,
+      "loss": 0.8358,
+      "step": 473000
+    },
+    {
+      "epoch": 15.42,
+      "learning_rate": 5.318181818181819e-06,
+      "loss": 0.8357,
+      "step": 473500
+    },
+    {
+      "epoch": 15.43,
+      "learning_rate": 5.313131313131313e-06,
+      "loss": 0.8348,
+      "step": 474000
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 5.3080808080808085e-06,
+      "loss": 0.8353,
+      "step": 474500
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 5.303030303030303e-06,
+      "loss": 0.8347,
+      "step": 475000
+    },
+    {
+      "epoch": 15.48,
+      "learning_rate": 5.297979797979799e-06,
+      "loss": 0.8347,
+      "step": 475500
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 5.292929292929293e-06,
+      "loss": 0.8347,
+      "step": 476000
+    },
+    {
+      "epoch": 15.52,
+      "learning_rate": 5.287878787878788e-06,
+      "loss": 0.8348,
+      "step": 476500
+    },
+    {
+      "epoch": 15.53,
+      "learning_rate": 5.2828282828282825e-06,
+      "loss": 0.834,
+      "step": 477000
+    },
+    {
+      "epoch": 15.55,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 0.8344,
+      "step": 477500
+    },
+    {
+      "epoch": 15.56,
+      "learning_rate": 5.272727272727273e-06,
+      "loss": 0.8339,
+      "step": 478000
+    },
+    {
+      "epoch": 15.58,
+      "learning_rate": 5.267676767676768e-06,
+      "loss": 0.834,
+      "step": 478500
+    },
+    {
+      "epoch": 15.6,
+      "learning_rate": 5.262626262626262e-06,
+      "loss": 0.8343,
+      "step": 479000
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 5.257575757575758e-06,
+      "loss": 0.8337,
+      "step": 479500
+    },
+    {
+      "epoch": 15.63,
+      "learning_rate": 5.252525252525253e-06,
+      "loss": 0.8329,
+      "step": 480000
+    },
+    {
+      "epoch": 15.63,
+      "eval_loss": 0.8069521188735962,
+      "eval_runtime": 92.7218,
+      "eval_samples_per_second": 856.454,
+      "eval_steps_per_second": 3.354,
+      "step": 480000
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 5.247474747474748e-06,
+      "loss": 0.833,
+      "step": 480500
+    },
+    {
+      "epoch": 15.66,
+      "learning_rate": 5.242424242424244e-06,
+      "loss": 0.8334,
+      "step": 481000
+    },
+    {
+      "epoch": 15.68,
+      "learning_rate": 5.237373737373738e-06,
+      "loss": 0.8331,
+      "step": 481500
+    },
+    {
+      "epoch": 15.7,
+      "learning_rate": 5.232323232323233e-06,
+      "loss": 0.8327,
+      "step": 482000
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 5.2272727272727274e-06,
+      "loss": 0.833,
+      "step": 482500
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 5.2222222222222226e-06,
+      "loss": 0.8325,
+      "step": 483000
+    },
+    {
+      "epoch": 15.74,
+      "learning_rate": 5.217171717171718e-06,
+      "loss": 0.8324,
+      "step": 483500
+    },
+    {
+      "epoch": 15.76,
+      "learning_rate": 5.212121212121213e-06,
+      "loss": 0.8318,
+      "step": 484000
+    },
+    {
+      "epoch": 15.78,
+      "learning_rate": 5.207070707070707e-06,
+      "loss": 0.8318,
+      "step": 484500
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 5.202020202020202e-06,
+      "loss": 0.8325,
+      "step": 485000
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 5.196969696969697e-06,
+      "loss": 0.8318,
+      "step": 485500
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 5.191919191919193e-06,
+      "loss": 0.8316,
+      "step": 486000
+    },
+    {
+      "epoch": 15.84,
+      "learning_rate": 5.186868686868687e-06,
+      "loss": 0.8312,
+      "step": 486500
+    },
+    {
+      "epoch": 15.86,
+      "learning_rate": 5.181818181818182e-06,
+      "loss": 0.8309,
+      "step": 487000
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 5.176767676767676e-06,
+      "loss": 0.8314,
+      "step": 487500
+    },
+    {
+      "epoch": 15.89,
+      "learning_rate": 5.171717171717172e-06,
+      "loss": 0.8309,
+      "step": 488000
+    },
+    {
+      "epoch": 15.91,
+      "learning_rate": 5.1666666666666675e-06,
+      "loss": 0.8307,
+      "step": 488500
+    },
+    {
+      "epoch": 15.92,
+      "learning_rate": 5.161616161616162e-06,
+      "loss": 0.8306,
+      "step": 489000
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 5.156565656565658e-06,
+      "loss": 0.8302,
+      "step": 489500
+    },
+    {
+      "epoch": 15.96,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.8301,
+      "step": 490000
+    },
+    {
+      "epoch": 15.96,
+      "eval_loss": 0.8032618761062622,
+      "eval_runtime": 92.9139,
+      "eval_samples_per_second": 854.684,
+      "eval_steps_per_second": 3.347,
+      "step": 490000
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 5.146464646464647e-06,
+      "loss": 0.8298,
+      "step": 490500
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 5.1414141414141415e-06,
+      "loss": 0.8299,
+      "step": 491000
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 5.1363636363636375e-06,
+      "loss": 0.8295,
+      "step": 491500
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 5.131313131313132e-06,
+      "loss": 0.8295,
+      "step": 492000
+    },
+    {
+      "epoch": 16.04,
+      "learning_rate": 5.126262626262627e-06,
+      "loss": 0.8287,
+      "step": 492500
+    },
+    {
+      "epoch": 16.05,
+      "learning_rate": 5.121212121212121e-06,
+      "loss": 0.8289,
+      "step": 493000
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 5.116161616161617e-06,
+      "loss": 0.8283,
+      "step": 493500
+    },
+    {
+      "epoch": 16.09,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 0.829,
+      "step": 494000
+    },
+    {
+      "epoch": 16.1,
+      "learning_rate": 5.106060606060607e-06,
+      "loss": 0.828,
+      "step": 494500
+    },
+    {
+      "epoch": 16.12,
+      "learning_rate": 5.101010101010101e-06,
+      "loss": 0.8285,
+      "step": 495000
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 5.095959595959596e-06,
+      "loss": 0.8283,
+      "step": 495500
+    },
+    {
+      "epoch": 16.15,
+      "learning_rate": 5.090909090909091e-06,
+      "loss": 0.8276,
+      "step": 496000
+    },
+    {
+      "epoch": 16.17,
+      "learning_rate": 5.085858585858586e-06,
+      "loss": 0.8281,
+      "step": 496500
+    },
+    {
+      "epoch": 16.18,
+      "learning_rate": 5.0808080808080815e-06,
+      "loss": 0.8273,
+      "step": 497000
+    },
+    {
+      "epoch": 16.2,
+      "learning_rate": 5.075757575757576e-06,
+      "loss": 0.8277,
+      "step": 497500
+    },
+    {
+      "epoch": 16.22,
+      "learning_rate": 5.070707070707072e-06,
+      "loss": 0.8271,
+      "step": 498000
+    },
+    {
+      "epoch": 16.23,
+      "learning_rate": 5.065656565656566e-06,
+      "loss": 0.827,
+      "step": 498500
+    },
+    {
+      "epoch": 16.25,
+      "learning_rate": 5.060606060606061e-06,
+      "loss": 0.8269,
+      "step": 499000
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 5.0555555555555555e-06,
+      "loss": 0.8271,
+      "step": 499500
+    },
+    {
+      "epoch": 16.28,
+      "learning_rate": 5.0505050505050515e-06,
+      "loss": 0.8264,
+      "step": 500000
+    },
+    {
+      "epoch": 16.28,
+      "eval_loss": 0.7977773547172546,
+      "eval_runtime": 89.9319,
+      "eval_samples_per_second": 883.023,
+      "eval_steps_per_second": 3.458,
+      "step": 500000
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 5.045454545454546e-06,
+      "loss": 0.8265,
+      "step": 500500
+    },
+    {
+      "epoch": 16.31,
+      "learning_rate": 5.040404040404041e-06,
+      "loss": 0.8264,
+      "step": 501000
+    },
+    {
+      "epoch": 16.33,
+      "learning_rate": 5.035353535353535e-06,
+      "loss": 0.8255,
+      "step": 501500
+    },
+    {
+      "epoch": 16.35,
+      "learning_rate": 5.030303030303031e-06,
+      "loss": 0.8261,
+      "step": 502000
+    },
+    {
+      "epoch": 16.36,
+      "learning_rate": 5.0252525252525255e-06,
+      "loss": 0.8252,
+      "step": 502500
+    },
+    {
+      "epoch": 16.38,
+      "learning_rate": 5.020202020202021e-06,
+      "loss": 0.8253,
+      "step": 503000
+    },
+    {
+      "epoch": 16.4,
+      "learning_rate": 5.015151515151515e-06,
+      "loss": 0.8254,
+      "step": 503500
+    },
+    {
+      "epoch": 16.41,
+      "learning_rate": 5.010101010101011e-06,
+      "loss": 0.8252,
+      "step": 504000
+    },
+    {
+      "epoch": 16.43,
+      "learning_rate": 5.005050505050505e-06,
+      "loss": 0.8253,
+      "step": 504500
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 5e-06,
+      "loss": 0.8254,
+      "step": 505000
+    },
+    {
+      "epoch": 16.46,
+      "learning_rate": 4.9949494949494956e-06,
+      "loss": 0.8245,
+      "step": 505500
+    },
+    {
+      "epoch": 16.48,
+      "learning_rate": 4.98989898989899e-06,
+      "loss": 0.8238,
+      "step": 506000
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 4.984848484848485e-06,
+      "loss": 0.8241,
+      "step": 506500
+    },
+    {
+      "epoch": 16.51,
+      "learning_rate": 4.97979797979798e-06,
+      "loss": 0.8239,
+      "step": 507000
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 4.974747474747475e-06,
+      "loss": 0.8242,
+      "step": 507500
+    },
+    {
+      "epoch": 16.54,
+      "learning_rate": 4.9696969696969696e-06,
+      "loss": 0.8239,
+      "step": 508000
+    },
+    {
+      "epoch": 16.56,
+      "learning_rate": 4.964646464646465e-06,
+      "loss": 0.8235,
+      "step": 508500
+    },
+    {
+      "epoch": 16.57,
+      "learning_rate": 4.95959595959596e-06,
+      "loss": 0.8235,
+      "step": 509000
+    },
+    {
+      "epoch": 16.59,
+      "learning_rate": 4.954545454545455e-06,
+      "loss": 0.8233,
+      "step": 509500
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 4.94949494949495e-06,
+      "loss": 0.8233,
+      "step": 510000
+    },
+    {
+      "epoch": 16.61,
+      "eval_loss": 0.7961931824684143,
+      "eval_runtime": 87.3644,
+      "eval_samples_per_second": 908.975,
+      "eval_steps_per_second": 3.56,
+      "step": 510000
+    },
+    {
+      "epoch": 16.62,
+      "learning_rate": 4.944444444444445e-06,
+      "loss": 0.8221,
+      "step": 510500
+    },
+    {
+      "epoch": 16.64,
+      "learning_rate": 4.93939393939394e-06,
+      "loss": 0.8231,
+      "step": 511000
+    },
+    {
+      "epoch": 16.66,
+      "learning_rate": 4.934343434343435e-06,
+      "loss": 0.8224,
+      "step": 511500
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 4.92929292929293e-06,
+      "loss": 0.8228,
+      "step": 512000
+    },
+    {
+      "epoch": 16.69,
+      "learning_rate": 4.924242424242425e-06,
+      "loss": 0.822,
+      "step": 512500
+    },
+    {
+      "epoch": 16.7,
+      "learning_rate": 4.919191919191919e-06,
+      "loss": 0.822,
+      "step": 513000
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 4.9141414141414145e-06,
+      "loss": 0.8223,
+      "step": 513500
+    },
+    {
+      "epoch": 16.74,
+      "learning_rate": 4.90909090909091e-06,
+      "loss": 0.8216,
+      "step": 514000
+    },
+    {
+      "epoch": 16.75,
+      "learning_rate": 4.904040404040405e-06,
+      "loss": 0.8216,
+      "step": 514500
+    },
+    {
+      "epoch": 16.77,
+      "learning_rate": 4.898989898989899e-06,
+      "loss": 0.8214,
+      "step": 515000
+    },
+    {
+      "epoch": 16.79,
+      "learning_rate": 4.893939393939394e-06,
+      "loss": 0.8216,
+      "step": 515500
+    },
+    {
+      "epoch": 16.8,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 0.8211,
+      "step": 516000
+    },
+    {
+      "epoch": 16.82,
+      "learning_rate": 4.883838383838384e-06,
+      "loss": 0.8207,
+      "step": 516500
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 4.878787878787879e-06,
+      "loss": 0.8211,
+      "step": 517000
+    },
+    {
+      "epoch": 16.85,
+      "learning_rate": 4.873737373737374e-06,
+      "loss": 0.8205,
+      "step": 517500
+    },
+    {
+      "epoch": 16.87,
+      "learning_rate": 4.868686868686869e-06,
+      "loss": 0.8203,
+      "step": 518000
+    },
+    {
+      "epoch": 16.88,
+      "learning_rate": 4.863636363636364e-06,
+      "loss": 0.82,
+      "step": 518500
+    },
+    {
+      "epoch": 16.9,
+      "learning_rate": 4.858585858585859e-06,
+      "loss": 0.8198,
+      "step": 519000
+    },
+    {
+      "epoch": 16.92,
+      "learning_rate": 4.8535353535353545e-06,
+      "loss": 0.8195,
+      "step": 519500
+    },
+    {
+      "epoch": 16.93,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.8196,
+      "step": 520000
+    },
+    {
+      "epoch": 16.93,
+      "eval_loss": 0.7930386662483215,
+      "eval_runtime": 89.4487,
+      "eval_samples_per_second": 887.794,
+      "eval_steps_per_second": 3.477,
+      "step": 520000
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 4.843434343434344e-06,
+      "loss": 0.8193,
+      "step": 520500
+    },
+    {
+      "epoch": 16.97,
+      "learning_rate": 4.838383838383839e-06,
+      "loss": 0.8193,
+      "step": 521000
+    },
+    {
+      "epoch": 16.98,
+      "learning_rate": 4.833333333333333e-06,
+      "loss": 0.819,
+      "step": 521500
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 4.8282828282828285e-06,
+      "loss": 0.8188,
+      "step": 522000
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 4.823232323232324e-06,
+      "loss": 0.8185,
+      "step": 522500
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 4.818181818181819e-06,
+      "loss": 0.8185,
+      "step": 523000
+    },
+    {
+      "epoch": 17.05,
+      "learning_rate": 4.813131313131313e-06,
+      "loss": 0.8178,
+      "step": 523500
+    },
+    {
+      "epoch": 17.06,
+      "learning_rate": 4.808080808080808e-06,
+      "loss": 0.8177,
+      "step": 524000
+    },
+    {
+      "epoch": 17.08,
+      "learning_rate": 4.803030303030303e-06,
+      "loss": 0.8182,
+      "step": 524500
+    },
+    {
+      "epoch": 17.1,
+      "learning_rate": 4.7979797979797985e-06,
+      "loss": 0.8173,
+      "step": 525000
+    },
+    {
+      "epoch": 17.11,
+      "learning_rate": 4.792929292929293e-06,
+      "loss": 0.8181,
+      "step": 525500
+    },
+    {
+      "epoch": 17.13,
+      "learning_rate": 4.787878787878788e-06,
+      "loss": 0.8172,
+      "step": 526000
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 4.782828282828283e-06,
+      "loss": 0.8174,
+      "step": 526500
+    },
+    {
+      "epoch": 17.16,
+      "learning_rate": 4.777777777777778e-06,
+      "loss": 0.817,
+      "step": 527000
+    },
+    {
+      "epoch": 17.18,
+      "learning_rate": 4.772727272727273e-06,
+      "loss": 0.8164,
+      "step": 527500
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 4.7676767676767685e-06,
+      "loss": 0.8169,
+      "step": 528000
+    },
+    {
+      "epoch": 17.21,
+      "learning_rate": 4.762626262626263e-06,
+      "loss": 0.8162,
+      "step": 528500
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 4.757575757575758e-06,
+      "loss": 0.8166,
+      "step": 529000
+    },
+    {
+      "epoch": 17.24,
+      "learning_rate": 4.752525252525253e-06,
+      "loss": 0.8156,
+      "step": 529500
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 4.747474747474748e-06,
+      "loss": 0.8154,
+      "step": 530000
+    },
+    {
+      "epoch": 17.26,
+      "eval_loss": 0.7897705435752869,
+      "eval_runtime": 89.4546,
+      "eval_samples_per_second": 887.735,
+      "eval_steps_per_second": 3.477,
+      "step": 530000
+    },
+    {
+      "epoch": 17.27,
+      "learning_rate": 4.7424242424242426e-06,
+      "loss": 0.815,
+      "step": 530500
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 4.737373737373738e-06,
+      "loss": 0.8146,
+      "step": 531000
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 4.732323232323233e-06,
+      "loss": 0.8152,
+      "step": 531500
+    },
+    {
+      "epoch": 17.32,
+      "learning_rate": 4.727272727272728e-06,
+      "loss": 0.815,
+      "step": 532000
+    },
+    {
+      "epoch": 17.34,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.8143,
+      "step": 532500
+    },
+    {
+      "epoch": 17.36,
+      "learning_rate": 4.717171717171717e-06,
+      "loss": 0.8143,
+      "step": 533000
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 4.7121212121212126e-06,
+      "loss": 0.814,
+      "step": 533500
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 4.707070707070707e-06,
+      "loss": 0.8141,
+      "step": 534000
+    },
+    {
+      "epoch": 17.4,
+      "learning_rate": 4.702020202020202e-06,
+      "loss": 0.8139,
+      "step": 534500
+    },
+    {
+      "epoch": 17.42,
+      "learning_rate": 4.696969696969698e-06,
+      "loss": 0.8132,
+      "step": 535000
+    },
+    {
+      "epoch": 17.44,
+      "learning_rate": 4.691919191919192e-06,
+      "loss": 0.8127,
+      "step": 535500
+    },
+    {
+      "epoch": 17.45,
+      "learning_rate": 4.6868686868686874e-06,
+      "loss": 0.8138,
+      "step": 536000
+    },
+    {
+      "epoch": 17.47,
+      "learning_rate": 4.681818181818183e-06,
+      "loss": 0.8128,
+      "step": 536500
+    },
+    {
+      "epoch": 17.49,
+      "learning_rate": 4.676767676767677e-06,
+      "loss": 0.8125,
+      "step": 537000
+    },
+    {
+      "epoch": 17.5,
+      "learning_rate": 4.671717171717172e-06,
+      "loss": 0.8125,
+      "step": 537500
+    },
+    {
+      "epoch": 17.52,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.8122,
+      "step": 538000
+    },
+    {
+      "epoch": 17.54,
+      "learning_rate": 4.661616161616162e-06,
+      "loss": 0.812,
+      "step": 538500
+    },
+    {
+      "epoch": 17.55,
+      "learning_rate": 4.656565656565657e-06,
+      "loss": 0.8116,
+      "step": 539000
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 4.651515151515152e-06,
+      "loss": 0.8121,
+      "step": 539500
+    },
+    {
+      "epoch": 17.58,
+      "learning_rate": 4.646464646464647e-06,
+      "loss": 0.8114,
+      "step": 540000
+    },
+    {
+      "epoch": 17.58,
+      "eval_loss": 0.7859430313110352,
+      "eval_runtime": 87.9327,
+      "eval_samples_per_second": 903.1,
+      "eval_steps_per_second": 3.537,
+      "step": 540000
+    },
+    {
+      "epoch": 17.6,
+      "learning_rate": 4.641414141414142e-06,
+      "loss": 0.8116,
+      "step": 540500
+    },
+    {
+      "epoch": 17.62,
+      "learning_rate": 4.636363636363636e-06,
+      "loss": 0.8106,
+      "step": 541000
+    },
+    {
+      "epoch": 17.63,
+      "learning_rate": 4.6313131313131315e-06,
+      "loss": 0.8103,
+      "step": 541500
+    },
+    {
+      "epoch": 17.65,
+      "learning_rate": 4.626262626262627e-06,
+      "loss": 0.81,
+      "step": 542000
+    },
+    {
+      "epoch": 17.67,
+      "learning_rate": 4.621212121212122e-06,
+      "loss": 0.8101,
+      "step": 542500
+    },
+    {
+      "epoch": 17.68,
+      "learning_rate": 4.616161616161616e-06,
+      "loss": 0.8094,
+      "step": 543000
+    },
+    {
+      "epoch": 17.7,
+      "learning_rate": 4.611111111111112e-06,
+      "loss": 0.8095,
+      "step": 543500
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 4.606060606060606e-06,
+      "loss": 0.8092,
+      "step": 544000
+    },
+    {
+      "epoch": 17.73,
+      "learning_rate": 4.6010101010101015e-06,
+      "loss": 0.8087,
+      "step": 544500
+    },
+    {
+      "epoch": 17.75,
+      "learning_rate": 4.595959595959597e-06,
+      "loss": 0.8085,
+      "step": 545000
+    },
+    {
+      "epoch": 17.76,
+      "learning_rate": 4.590909090909092e-06,
+      "loss": 0.8088,
+      "step": 545500
+    },
+    {
+      "epoch": 17.78,
+      "learning_rate": 4.585858585858586e-06,
+      "loss": 0.8081,
+      "step": 546000
+    },
+    {
+      "epoch": 17.8,
+      "learning_rate": 4.580808080808081e-06,
+      "loss": 0.8078,
+      "step": 546500
+    },
+    {
+      "epoch": 17.81,
+      "learning_rate": 4.575757575757576e-06,
+      "loss": 0.8073,
+      "step": 547000
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 4.5707070707070715e-06,
+      "loss": 0.8075,
+      "step": 547500
+    },
+    {
+      "epoch": 17.84,
+      "learning_rate": 4.565656565656566e-06,
+      "loss": 0.8066,
+      "step": 548000
+    },
+    {
+      "epoch": 17.86,
+      "learning_rate": 4.560606060606061e-06,
+      "loss": 0.8064,
+      "step": 548500
+    },
+    {
+      "epoch": 17.88,
+      "learning_rate": 4.555555555555556e-06,
+      "loss": 0.806,
+      "step": 549000
+    },
+    {
+      "epoch": 17.89,
+      "learning_rate": 4.55050505050505e-06,
+      "loss": 0.8062,
+      "step": 549500
+    },
+    {
+      "epoch": 17.91,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.8052,
+      "step": 550000
+    },
+    {
+      "epoch": 17.91,
+      "eval_loss": 0.7798945903778076,
+      "eval_runtime": 88.3824,
+      "eval_samples_per_second": 898.505,
+      "eval_steps_per_second": 3.519,
+      "step": 550000
+    },
+    {
+      "epoch": 17.93,
+      "learning_rate": 4.540404040404041e-06,
+      "loss": 0.8054,
+      "step": 550500
+    },
+    {
+      "epoch": 17.94,
+      "learning_rate": 4.535353535353536e-06,
+      "loss": 0.8051,
+      "step": 551000
+    },
+    {
+      "epoch": 17.96,
+      "learning_rate": 4.53030303030303e-06,
+      "loss": 0.8043,
+      "step": 551500
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 4.525252525252526e-06,
+      "loss": 0.8051,
+      "step": 552000
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 4.520202020202021e-06,
+      "loss": 0.8036,
+      "step": 552500
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 4.5151515151515155e-06,
+      "loss": 0.8038,
+      "step": 553000
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 4.510101010101011e-06,
+      "loss": 0.803,
+      "step": 553500
+    },
+    {
+      "epoch": 18.04,
+      "learning_rate": 4.505050505050506e-06,
+      "loss": 0.8026,
+      "step": 554000
+    },
+    {
+      "epoch": 18.06,
+      "learning_rate": 4.5e-06,
+      "loss": 0.8019,
+      "step": 554500
+    },
+    {
+      "epoch": 18.07,
+      "learning_rate": 4.494949494949495e-06,
+      "loss": 0.8017,
+      "step": 555000
+    },
+    {
+      "epoch": 18.09,
+      "learning_rate": 4.48989898989899e-06,
+      "loss": 0.801,
+      "step": 555500
+    },
+    {
+      "epoch": 18.1,
+      "learning_rate": 4.4848484848484855e-06,
+      "loss": 0.8016,
+      "step": 556000
+    },
+    {
+      "epoch": 18.12,
+      "learning_rate": 4.47979797979798e-06,
+      "loss": 0.8005,
+      "step": 556500
+    },
+    {
+      "epoch": 18.14,
+      "learning_rate": 4.474747474747475e-06,
+      "loss": 0.7995,
+      "step": 557000
+    },
+    {
+      "epoch": 18.15,
+      "learning_rate": 4.46969696969697e-06,
+      "loss": 0.7997,
+      "step": 557500
+    },
+    {
+      "epoch": 18.17,
+      "learning_rate": 4.464646464646465e-06,
+      "loss": 0.7994,
+      "step": 558000
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 4.4595959595959596e-06,
+      "loss": 0.7986,
+      "step": 558500
+    },
+    {
+      "epoch": 18.2,
+      "learning_rate": 4.454545454545455e-06,
+      "loss": 0.7985,
+      "step": 559000
+    },
+    {
+      "epoch": 18.22,
+      "learning_rate": 4.44949494949495e-06,
+      "loss": 0.7975,
+      "step": 559500
+    },
+    {
+      "epoch": 18.24,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.7969,
+      "step": 560000
+    },
+    {
+      "epoch": 18.24,
+      "eval_loss": 0.7699345350265503,
+      "eval_runtime": 91.5062,
+      "eval_samples_per_second": 867.832,
+      "eval_steps_per_second": 3.399,
+      "step": 560000
+    },
+    {
+      "epoch": 18.25,
+      "learning_rate": 4.43939393939394e-06,
+      "loss": 0.7963,
+      "step": 560500
+    },
+    {
+      "epoch": 18.27,
+      "learning_rate": 4.434343434343435e-06,
+      "loss": 0.7959,
+      "step": 561000
+    },
+    {
+      "epoch": 18.28,
+      "learning_rate": 4.42929292929293e-06,
+      "loss": 0.7952,
+      "step": 561500
+    },
+    {
+      "epoch": 18.3,
+      "learning_rate": 4.424242424242425e-06,
+      "loss": 0.7947,
+      "step": 562000
+    },
+    {
+      "epoch": 18.32,
+      "learning_rate": 4.41919191919192e-06,
+      "loss": 0.794,
+      "step": 562500
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 4.414141414141415e-06,
+      "loss": 0.7937,
+      "step": 563000
+    },
+    {
+      "epoch": 18.35,
+      "learning_rate": 4.409090909090909e-06,
+      "loss": 0.7929,
+      "step": 563500
+    },
+    {
+      "epoch": 18.37,
+      "learning_rate": 4.4040404040404044e-06,
+      "loss": 0.7926,
+      "step": 564000
+    },
+    {
+      "epoch": 18.38,
+      "learning_rate": 4.3989898989899e-06,
+      "loss": 0.7924,
+      "step": 564500
+    },
+    {
+      "epoch": 18.4,
+      "learning_rate": 4.393939393939394e-06,
+      "loss": 0.7908,
+      "step": 565000
+    },
+    {
+      "epoch": 18.41,
+      "learning_rate": 4.388888888888889e-06,
+      "loss": 0.7906,
+      "step": 565500
+    },
+    {
+      "epoch": 18.43,
+      "learning_rate": 4.383838383838384e-06,
+      "loss": 0.7902,
+      "step": 566000
+    },
+    {
+      "epoch": 18.45,
+      "learning_rate": 4.378787878787879e-06,
+      "loss": 0.7894,
+      "step": 566500
+    },
+    {
+      "epoch": 18.46,
+      "learning_rate": 4.373737373737374e-06,
+      "loss": 0.7884,
+      "step": 567000
+    },
+    {
+      "epoch": 18.48,
+      "learning_rate": 4.368686868686869e-06,
+      "loss": 0.7881,
+      "step": 567500
+    },
+    {
+      "epoch": 18.5,
+      "learning_rate": 4.363636363636364e-06,
+      "loss": 0.7869,
+      "step": 568000
+    },
+    {
+      "epoch": 18.51,
+      "learning_rate": 4.358585858585859e-06,
+      "loss": 0.7861,
+      "step": 568500
+    },
+    {
+      "epoch": 18.53,
+      "learning_rate": 4.353535353535353e-06,
+      "loss": 0.7858,
+      "step": 569000
+    },
+    {
+      "epoch": 18.54,
+      "learning_rate": 4.348484848484849e-06,
+      "loss": 0.7854,
+      "step": 569500
+    },
+    {
+      "epoch": 18.56,
+      "learning_rate": 4.343434343434344e-06,
+      "loss": 0.7846,
+      "step": 570000
+    },
+    {
+      "epoch": 18.56,
+      "eval_loss": 0.7586176991462708,
+      "eval_runtime": 88.5769,
+      "eval_samples_per_second": 896.531,
+      "eval_steps_per_second": 3.511,
+      "step": 570000
+    },
+    {
+      "epoch": 18.58,
+      "learning_rate": 4.338383838383839e-06,
+      "loss": 0.7837,
+      "step": 570500
+    },
+    {
+      "epoch": 18.59,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.7832,
+      "step": 571000
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 4.328282828282829e-06,
+      "loss": 0.7826,
+      "step": 571500
+    },
+    {
+      "epoch": 18.63,
+      "learning_rate": 4.323232323232323e-06,
+      "loss": 0.7824,
+      "step": 572000
+    },
+    {
+      "epoch": 18.64,
+      "learning_rate": 4.3181818181818185e-06,
+      "loss": 0.7817,
+      "step": 572500
+    },
+    {
+      "epoch": 18.66,
+      "learning_rate": 4.313131313131314e-06,
+      "loss": 0.7812,
+      "step": 573000
+    },
+    {
+      "epoch": 18.67,
+      "learning_rate": 4.308080808080809e-06,
+      "loss": 0.7805,
+      "step": 573500
+    },
+    {
+      "epoch": 18.69,
+      "learning_rate": 4.303030303030303e-06,
+      "loss": 0.78,
+      "step": 574000
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 4.297979797979798e-06,
+      "loss": 0.7789,
+      "step": 574500
+    },
+    {
+      "epoch": 18.72,
+      "learning_rate": 4.292929292929293e-06,
+      "loss": 0.7789,
+      "step": 575000
+    },
+    {
+      "epoch": 18.74,
+      "learning_rate": 4.287878787878788e-06,
+      "loss": 0.7782,
+      "step": 575500
+    },
+    {
+      "epoch": 18.76,
+      "learning_rate": 4.282828282828283e-06,
+      "loss": 0.778,
+      "step": 576000
+    },
+    {
+      "epoch": 18.77,
+      "learning_rate": 4.277777777777778e-06,
+      "loss": 0.777,
+      "step": 576500
+    },
+    {
+      "epoch": 18.79,
+      "learning_rate": 4.272727272727273e-06,
+      "loss": 0.7762,
+      "step": 577000
+    },
+    {
+      "epoch": 18.8,
+      "learning_rate": 4.267676767676767e-06,
+      "loss": 0.776,
+      "step": 577500
+    },
+    {
+      "epoch": 18.82,
+      "learning_rate": 4.262626262626263e-06,
+      "loss": 0.7755,
+      "step": 578000
+    },
+    {
+      "epoch": 18.84,
+      "learning_rate": 4.2575757575757585e-06,
+      "loss": 0.7748,
+      "step": 578500
+    },
+    {
+      "epoch": 18.85,
+      "learning_rate": 4.252525252525253e-06,
+      "loss": 0.7744,
+      "step": 579000
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 4.247474747474748e-06,
+      "loss": 0.7742,
+      "step": 579500
+    },
+    {
+      "epoch": 18.89,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.7735,
+      "step": 580000
+    },
+    {
+      "epoch": 18.89,
+      "eval_loss": 0.7486483454704285,
+      "eval_runtime": 89.3916,
+      "eval_samples_per_second": 888.361,
+      "eval_steps_per_second": 3.479,
+      "step": 580000
+    },
+    {
+      "epoch": 18.9,
+      "learning_rate": 4.237373737373737e-06,
+      "loss": 0.7729,
+      "step": 580500
+    },
+    {
+      "epoch": 18.92,
+      "learning_rate": 4.2323232323232325e-06,
+      "loss": 0.7723,
+      "step": 581000
+    },
+    {
+      "epoch": 18.94,
+      "learning_rate": 4.227272727272728e-06,
+      "loss": 0.7725,
+      "step": 581500
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.7715,
+      "step": 582000
+    },
+    {
+      "epoch": 18.97,
+      "learning_rate": 4.217171717171717e-06,
+      "loss": 0.7712,
+      "step": 582500
+    },
+    {
+      "epoch": 18.98,
+      "learning_rate": 4.212121212121212e-06,
+      "loss": 0.7706,
+      "step": 583000
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 4.207070707070707e-06,
+      "loss": 0.7705,
+      "step": 583500
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 4.2020202020202026e-06,
+      "loss": 0.7698,
+      "step": 584000
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 4.196969696969697e-06,
+      "loss": 0.769,
+      "step": 584500
+    },
+    {
+      "epoch": 19.05,
+      "learning_rate": 4.191919191919192e-06,
+      "loss": 0.7683,
+      "step": 585000
+    },
+    {
+      "epoch": 19.07,
+      "learning_rate": 4.186868686868687e-06,
+      "loss": 0.7689,
+      "step": 585500
+    },
+    {
+      "epoch": 19.08,
+      "learning_rate": 4.181818181818182e-06,
+      "loss": 0.7683,
+      "step": 586000
+    },
+    {
+      "epoch": 19.1,
+      "learning_rate": 4.1767676767676774e-06,
+      "loss": 0.767,
+      "step": 586500
+    },
+    {
+      "epoch": 19.11,
+      "learning_rate": 4.1717171717171726e-06,
+      "loss": 0.7677,
+      "step": 587000
+    },
+    {
+      "epoch": 19.13,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.7673,
+      "step": 587500
+    },
+    {
+      "epoch": 19.15,
+      "learning_rate": 4.161616161616162e-06,
+      "loss": 0.7668,
+      "step": 588000
+    },
+    {
+      "epoch": 19.16,
+      "learning_rate": 4.156565656565657e-06,
+      "loss": 0.7665,
+      "step": 588500
+    },
+    {
+      "epoch": 19.18,
+      "learning_rate": 4.151515151515152e-06,
+      "loss": 0.7663,
+      "step": 589000
+    },
+    {
+      "epoch": 19.2,
+      "learning_rate": 4.146464646464647e-06,
+      "loss": 0.7653,
+      "step": 589500
+    },
+    {
+      "epoch": 19.21,
+      "learning_rate": 4.141414141414142e-06,
+      "loss": 0.7648,
+      "step": 590000
+    },
+    {
+      "epoch": 19.21,
+      "eval_loss": 0.7404940724372864,
+      "eval_runtime": 90.2196,
+      "eval_samples_per_second": 880.208,
+      "eval_steps_per_second": 3.447,
+      "step": 590000
+    },
+    {
+      "epoch": 19.23,
+      "learning_rate": 4.136363636363637e-06,
+      "loss": 0.7649,
+      "step": 590500
+    },
+    {
+      "epoch": 19.24,
+      "learning_rate": 4.131313131313132e-06,
+      "loss": 0.7642,
+      "step": 591000
+    },
+    {
+      "epoch": 19.26,
+      "learning_rate": 4.126262626262626e-06,
+      "loss": 0.764,
+      "step": 591500
+    },
+    {
+      "epoch": 19.28,
+      "learning_rate": 4.1212121212121215e-06,
+      "loss": 0.7635,
+      "step": 592000
+    },
+    {
+      "epoch": 19.29,
+      "learning_rate": 4.116161616161617e-06,
+      "loss": 0.7635,
+      "step": 592500
+    },
+    {
+      "epoch": 19.31,
+      "learning_rate": 4.111111111111111e-06,
+      "loss": 0.7627,
+      "step": 593000
+    },
+    {
+      "epoch": 19.33,
+      "learning_rate": 4.106060606060606e-06,
+      "loss": 0.7625,
+      "step": 593500
+    },
+    {
+      "epoch": 19.34,
+      "learning_rate": 4.101010101010101e-06,
+      "loss": 0.7623,
+      "step": 594000
+    },
+    {
+      "epoch": 19.36,
+      "learning_rate": 4.095959595959596e-06,
+      "loss": 0.7613,
+      "step": 594500
+    },
+    {
+      "epoch": 19.37,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.762,
+      "step": 595000
+    },
+    {
+      "epoch": 19.39,
+      "learning_rate": 4.085858585858587e-06,
+      "loss": 0.7613,
+      "step": 595500
+    },
+    {
+      "epoch": 19.41,
+      "learning_rate": 4.080808080808081e-06,
+      "loss": 0.7611,
+      "step": 596000
+    },
+    {
+      "epoch": 19.42,
+      "learning_rate": 4.075757575757576e-06,
+      "loss": 0.7599,
+      "step": 596500
+    },
+    {
+      "epoch": 19.44,
+      "learning_rate": 4.070707070707071e-06,
+      "loss": 0.7601,
+      "step": 597000
+    },
+    {
+      "epoch": 19.46,
+      "learning_rate": 4.065656565656566e-06,
+      "loss": 0.7604,
+      "step": 597500
+    },
+    {
+      "epoch": 19.47,
+      "learning_rate": 4.060606060606061e-06,
+      "loss": 0.7596,
+      "step": 598000
+    },
+    {
+      "epoch": 19.49,
+      "learning_rate": 4.055555555555556e-06,
+      "loss": 0.7597,
+      "step": 598500
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 4.050505050505051e-06,
+      "loss": 0.7588,
+      "step": 599000
+    },
+    {
+      "epoch": 19.52,
+      "learning_rate": 4.045454545454546e-06,
+      "loss": 0.7588,
+      "step": 599500
+    },
+    {
+      "epoch": 19.54,
+      "learning_rate": 4.04040404040404e-06,
+      "loss": 0.7586,
+      "step": 600000
+    },
+    {
+      "epoch": 19.54,
+      "eval_loss": 0.7349938750267029,
+      "eval_runtime": 88.7424,
+      "eval_samples_per_second": 894.86,
+      "eval_steps_per_second": 3.505,
+      "step": 600000
+    },
+    {
+      "epoch": 19.55,
+      "learning_rate": 4.0353535353535355e-06,
+      "loss": 0.7584,
+      "step": 600500
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 4.030303030303031e-06,
+      "loss": 0.7582,
+      "step": 601000
+    },
+    {
+      "epoch": 19.59,
+      "learning_rate": 4.025252525252526e-06,
+      "loss": 0.7574,
+      "step": 601500
+    },
+    {
+      "epoch": 19.6,
+      "learning_rate": 4.02020202020202e-06,
+      "loss": 0.7573,
+      "step": 602000
+    },
+    {
+      "epoch": 19.62,
+      "learning_rate": 4.015151515151515e-06,
+      "loss": 0.7572,
+      "step": 602500
+    },
+    {
+      "epoch": 19.64,
+      "learning_rate": 4.01010101010101e-06,
+      "loss": 0.7571,
+      "step": 603000
+    },
+    {
+      "epoch": 19.65,
+      "learning_rate": 4.0050505050505055e-06,
+      "loss": 0.7563,
+      "step": 603500
+    },
+    {
+      "epoch": 19.67,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7561,
+      "step": 604000
+    },
+    {
+      "epoch": 19.68,
+      "learning_rate": 3.994949494949496e-06,
+      "loss": 0.7559,
+      "step": 604500
+    },
+    {
+      "epoch": 19.7,
+      "learning_rate": 3.98989898989899e-06,
+      "loss": 0.7559,
+      "step": 605000
+    },
+    {
+      "epoch": 19.72,
+      "learning_rate": 3.984848484848485e-06,
+      "loss": 0.7557,
+      "step": 605500
+    },
+    {
+      "epoch": 19.73,
+      "learning_rate": 3.97979797979798e-06,
+      "loss": 0.7557,
+      "step": 606000
+    },
+    {
+      "epoch": 19.75,
+      "learning_rate": 3.9747474747474755e-06,
+      "loss": 0.7552,
+      "step": 606500
+    },
+    {
+      "epoch": 19.77,
+      "learning_rate": 3.96969696969697e-06,
+      "loss": 0.7546,
+      "step": 607000
+    },
+    {
+      "epoch": 19.78,
+      "learning_rate": 3.964646464646465e-06,
+      "loss": 0.755,
+      "step": 607500
+    },
+    {
+      "epoch": 19.8,
+      "learning_rate": 3.95959595959596e-06,
+      "loss": 0.7547,
+      "step": 608000
+    },
+    {
+      "epoch": 19.81,
+      "learning_rate": 3.954545454545454e-06,
+      "loss": 0.7539,
+      "step": 608500
+    },
+    {
+      "epoch": 19.83,
+      "learning_rate": 3.9494949494949496e-06,
+      "loss": 0.7541,
+      "step": 609000
+    },
+    {
+      "epoch": 19.85,
+      "learning_rate": 3.944444444444445e-06,
+      "loss": 0.7541,
+      "step": 609500
+    },
+    {
+      "epoch": 19.86,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.7531,
+      "step": 610000
+    },
+    {
+      "epoch": 19.86,
+      "eval_loss": 0.7310336828231812,
+      "eval_runtime": 88.6702,
+      "eval_samples_per_second": 895.588,
+      "eval_steps_per_second": 3.507,
+      "step": 610000
+    },
+    {
+      "epoch": 19.88,
+      "learning_rate": 3.934343434343434e-06,
+      "loss": 0.7535,
+      "step": 610500
+    },
+    {
+      "epoch": 19.9,
+      "learning_rate": 3.929292929292929e-06,
+      "loss": 0.7533,
+      "step": 611000
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 3.9242424242424244e-06,
+      "loss": 0.7529,
+      "step": 611500
+    },
+    {
+      "epoch": 19.93,
+      "learning_rate": 3.9191919191919196e-06,
+      "loss": 0.7529,
+      "step": 612000
+    },
+    {
+      "epoch": 19.94,
+      "learning_rate": 3.914141414141415e-06,
+      "loss": 0.7525,
+      "step": 612500
+    },
+    {
+      "epoch": 19.96,
+      "learning_rate": 3.90909090909091e-06,
+      "loss": 0.7522,
+      "step": 613000
+    },
+    {
+      "epoch": 19.98,
+      "learning_rate": 3.904040404040404e-06,
+      "loss": 0.7523,
+      "step": 613500
+    },
+    {
+      "epoch": 19.99,
+      "learning_rate": 3.898989898989899e-06,
+      "loss": 0.7517,
+      "step": 614000
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 3.8939393939393944e-06,
+      "loss": 0.7513,
+      "step": 614500
+    },
+    {
+      "epoch": 20.03,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.7514,
+      "step": 615000
+    },
+    {
+      "epoch": 20.04,
+      "learning_rate": 3.883838383838384e-06,
+      "loss": 0.7512,
+      "step": 615500
+    },
+    {
+      "epoch": 20.06,
+      "learning_rate": 3.878787878787879e-06,
+      "loss": 0.7506,
+      "step": 616000
+    },
+    {
+      "epoch": 20.07,
+      "learning_rate": 3.873737373737374e-06,
+      "loss": 0.7507,
+      "step": 616500
+    },
+    {
+      "epoch": 20.09,
+      "learning_rate": 3.868686868686869e-06,
+      "loss": 0.7504,
+      "step": 617000
+    },
+    {
+      "epoch": 20.11,
+      "learning_rate": 3.863636363636364e-06,
+      "loss": 0.7504,
+      "step": 617500
+    },
+    {
+      "epoch": 20.12,
+      "learning_rate": 3.858585858585859e-06,
+      "loss": 0.7503,
+      "step": 618000
+    },
+    {
+      "epoch": 20.14,
+      "learning_rate": 3.853535353535354e-06,
+      "loss": 0.75,
+      "step": 618500
+    },
+    {
+      "epoch": 20.16,
+      "learning_rate": 3.848484848484848e-06,
+      "loss": 0.7501,
+      "step": 619000
+    },
+    {
+      "epoch": 20.17,
+      "learning_rate": 3.843434343434343e-06,
+      "loss": 0.7496,
+      "step": 619500
+    },
+    {
+      "epoch": 20.19,
+      "learning_rate": 3.8383838383838385e-06,
+      "loss": 0.7494,
+      "step": 620000
+    },
+    {
+      "epoch": 20.19,
+      "eval_loss": 0.7278521656990051,
+      "eval_runtime": 91.3349,
+      "eval_samples_per_second": 869.459,
+      "eval_steps_per_second": 3.405,
+      "step": 620000
+    }
+  ],
+  "max_steps": 1000000,
+  "num_train_epochs": 33,
+  "total_flos": 4.911250179876717e+18,
+  "trial_name": null,
+  "trial_params": null
+}