diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
deleted file mode 100644--- "a/last-checkpoint/trainer_state.json"
+++ /dev/null
@@ -1,15216 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 0.1,
-  "global_step": 2000000,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0,
-      "learning_rate": 7.499999999999999e-07,
-      "loss": 1.0632,
-      "step": 1000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4999999999999998e-06,
-      "loss": 0.9867,
-      "step": 2000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 2.2499999999999996e-06,
-      "loss": 0.9811,
-      "step": 3000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 0.9787,
-      "step": 4000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.7499999999999997e-06,
-      "loss": 0.9729,
-      "step": 5000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.9619815349578857,
-      "eval_runtime": 82.8842,
-      "eval_samples_per_second": 77.216,
-      "eval_steps_per_second": 0.603,
-      "step": 5000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 4.499999999999999e-06,
-      "loss": 0.9491,
-      "step": 6000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 5.25e-06,
-      "loss": 0.8618,
-      "step": 7000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 5.999999999999999e-06,
-      "loss": 0.8256,
-      "step": 8000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 6.749999999999999e-06,
-      "loss": 0.8148,
-      "step": 9000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 7.499999999999999e-06,
-      "loss": 0.8019,
-      "step": 10000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.8153918385505676,
-      "eval_runtime": 78.5433,
-      "eval_samples_per_second": 81.484,
-      "eval_steps_per_second": 0.637,
-      "step": 10000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.249999999999999e-06,
-      "loss": 0.7961,
-      "step": 11000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.999999999999999e-06,
-      "loss": 0.7914,
-      "step": 12000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 9.75e-06,
-      "loss": 0.7849,
-      "step": 13000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.05e-05,
-      "loss": 0.7817,
-      "step": 14000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.1249999999999999e-05,
-      "loss": 0.7777,
-      "step": 15000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.7774372696876526,
-      "eval_runtime": 79.7934,
-      "eval_samples_per_second": 80.207,
-      "eval_steps_per_second": 0.627,
-      "step": 15000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.1999999999999999e-05,
-      "loss": 0.7751,
-      "step": 16000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.2749249999999998e-05,
-      "loss": 0.7714,
-      "step": 17000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.34985e-05,
-      "loss": 0.7673,
-      "step": 18000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.42485e-05,
-      "loss": 0.7649,
-      "step": 19000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.4998499999999999e-05,
-      "loss": 0.7629,
-      "step": 20000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.7626805305480957,
-      "eval_runtime": 78.8878,
-      "eval_samples_per_second": 81.128,
-      "eval_steps_per_second": 0.634,
-      "step": 20000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.57485e-05,
-      "loss": 0.7616,
-      "step": 21000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.649775e-05,
-      "loss": 0.7592,
-      "step": 22000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.7247749999999998e-05,
-      "loss": 0.7577,
-      "step": 23000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.799775e-05,
-      "loss": 0.7557,
-      "step": 24000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.8747749999999997e-05,
-      "loss": 0.7542,
-      "step": 25000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.7557830810546875,
-      "eval_runtime": 80.9075,
-      "eval_samples_per_second": 79.103,
-      "eval_steps_per_second": 0.618,
-      "step": 25000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.9497749999999998e-05,
-      "loss": 0.7529,
-      "step": 26000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 2.0246999999999998e-05,
-      "loss": 0.7526,
-      "step": 27000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 2.0996999999999996e-05,
-      "loss": 0.7518,
-      "step": 28000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 2.1746999999999997e-05,
-      "loss": 0.7508,
-      "step": 29000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 2.24955e-05,
-      "loss": 0.7496,
-      "step": 30000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.7457370162010193,
-      "eval_runtime": 76.4686,
-      "eval_samples_per_second": 83.694,
-      "eval_steps_per_second": 0.654,
-      "step": 30000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.3245499999999996e-05,
-      "loss": 0.748,
-      "step": 31000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.3995499999999998e-05,
-      "loss": 0.7472,
-      "step": 32000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.4744749999999997e-05,
-      "loss": 0.7467,
-      "step": 33000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.549475e-05,
-      "loss": 0.7465,
-      "step": 34000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.6243999999999998e-05,
-      "loss": 0.7447,
-      "step": 35000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.7436805963516235,
-      "eval_runtime": 80.6483,
-      "eval_samples_per_second": 79.357,
-      "eval_steps_per_second": 0.62,
-      "step": 35000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.6994e-05,
-      "loss": 0.7446,
-      "step": 36000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.774325e-05,
-      "loss": 0.7432,
-      "step": 37000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.849325e-05,
-      "loss": 0.7435,
-      "step": 38000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.9243249999999995e-05,
-      "loss": 0.7398,
-      "step": 39000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.9992499999999994e-05,
-      "loss": 0.7365,
-      "step": 40000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.7301322817802429,
-      "eval_runtime": 81.961,
-      "eval_samples_per_second": 78.086,
-      "eval_steps_per_second": 0.61,
-      "step": 40000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.07425e-05,
-      "loss": 0.7345,
-      "step": 41000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.149175e-05,
-      "loss": 0.7318,
-      "step": 42000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.224175e-05,
-      "loss": 0.7293,
-      "step": 43000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.2990999999999996e-05,
-      "loss": 0.7263,
-      "step": 44000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.3740999999999994e-05,
-      "loss": 0.719,
-      "step": 45000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.70879727602005,
-      "eval_runtime": 80.7536,
-      "eval_samples_per_second": 79.253,
-      "eval_steps_per_second": 0.619,
-      "step": 45000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.4489499999999996e-05,
-      "loss": 0.7129,
-      "step": 46000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.5239499999999994e-05,
-      "loss": 0.7101,
-      "step": 47000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.59895e-05,
-      "loss": 0.7054,
-      "step": 48000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.6739499999999996e-05,
-      "loss": 0.7032,
-      "step": 49000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.74895e-05,
-      "loss": 0.6997,
-      "step": 50000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.6915566921234131,
-      "eval_runtime": 80.6444,
-      "eval_samples_per_second": 79.361,
-      "eval_steps_per_second": 0.62,
-      "step": 50000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.823875e-05,
-      "loss": 0.6956,
-      "step": 51000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.898875e-05,
-      "loss": 0.6912,
-      "step": 52000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.9738e-05,
-      "loss": 0.6889,
-      "step": 53000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.0487999999999996e-05,
-      "loss": 0.6873,
-      "step": 54000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.1237999999999994e-05,
-      "loss": 0.6834,
-      "step": 55000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.6724392771720886,
-      "eval_runtime": 82.1061,
-      "eval_samples_per_second": 77.948,
-      "eval_steps_per_second": 0.609,
-      "step": 55000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.198725e-05,
-      "loss": 0.6817,
-      "step": 56000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.273724999999999e-05,
-      "loss": 0.6811,
-      "step": 57000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.34865e-05,
-      "loss": 0.677,
-      "step": 58000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.4236499999999996e-05,
-      "loss": 0.6756,
-      "step": 59000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.49865e-05,
-      "loss": 0.6735,
-      "step": 60000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.6614792346954346,
-      "eval_runtime": 80.4754,
-      "eval_samples_per_second": 79.527,
-      "eval_steps_per_second": 0.621,
-      "step": 60000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.573574999999999e-05,
-      "loss": 0.6712,
-      "step": 61000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.648574999999999e-05,
-      "loss": 0.6687,
-      "step": 62000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.7235e-05,
-      "loss": 0.6685,
-      "step": 63000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.7984999999999995e-05,
-      "loss": 0.6658,
-      "step": 64000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.8735e-05,
-      "loss": 0.6629,
-      "step": 65000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.6503811478614807,
-      "eval_runtime": 79.7102,
-      "eval_samples_per_second": 80.291,
-      "eval_steps_per_second": 0.627,
-      "step": 65000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.948424999999999e-05,
-      "loss": 0.661,
-      "step": 66000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 5.023425e-05,
-      "loss": 0.6599,
-      "step": 67000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 5.09835e-05,
-      "loss": 0.657,
-      "step": 68000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 5.173349999999999e-05,
-      "loss": 0.6563,
-      "step": 69000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.248349999999999e-05,
-      "loss": 0.6541,
-      "step": 70000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.6402961015701294,
-      "eval_runtime": 81.1292,
-      "eval_samples_per_second": 78.887,
-      "eval_steps_per_second": 0.616,
-      "step": 70000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.323275e-05,
-      "loss": 0.6524,
-      "step": 71000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.398275e-05,
-      "loss": 0.6514,
-      "step": 72000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.4732749999999995e-05,
-      "loss": 0.6488,
-      "step": 73000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.5481999999999995e-05,
-      "loss": 0.6476,
-      "step": 74000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.623199999999999e-05,
-      "loss": 0.6474,
-      "step": 75000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.6383292078971863,
-      "eval_runtime": 80.3652,
-      "eval_samples_per_second": 79.636,
-      "eval_steps_per_second": 0.622,
-      "step": 75000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.698125e-05,
-      "loss": 0.6443,
-      "step": 76000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.773125e-05,
-      "loss": 0.6436,
-      "step": 77000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.84805e-05,
-      "loss": 0.6408,
-      "step": 78000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.9230499999999995e-05,
-      "loss": 0.6409,
-      "step": 79000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 5.998049999999999e-05,
-      "loss": 0.6399,
-      "step": 80000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.630017876625061,
-      "eval_runtime": 81.3873,
-      "eval_samples_per_second": 78.636,
-      "eval_steps_per_second": 0.614,
-      "step": 80000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.072974999999999e-05,
-      "loss": 0.6372,
-      "step": 81000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.147975e-05,
-      "loss": 0.6375,
-      "step": 82000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.222974999999999e-05,
-      "loss": 0.6351,
-      "step": 83000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.297974999999999e-05,
-      "loss": 0.6345,
-      "step": 84000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.372975e-05,
-      "loss": 0.6331,
-      "step": 85000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.6252214908599854,
-      "eval_runtime": 81.0014,
-      "eval_samples_per_second": 79.011,
-      "eval_steps_per_second": 0.617,
-      "step": 85000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.4479e-05,
-      "loss": 0.6321,
-      "step": 86000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.5229e-05,
-      "loss": 0.6313,
-      "step": 87000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.597899999999999e-05,
-      "loss": 0.6299,
-      "step": 88000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.672824999999999e-05,
-      "loss": 0.6284,
-      "step": 89000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 6.74775e-05,
-      "loss": 0.6264,
-      "step": 90000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.6188452839851379,
-      "eval_runtime": 79.754,
-      "eval_samples_per_second": 80.247,
-      "eval_steps_per_second": 0.627,
-      "step": 90000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 6.822749999999999e-05,
-      "loss": 0.6273,
-      "step": 91000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 6.89775e-05,
-      "loss": 0.6241,
-      "step": 92000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 6.972675e-05,
-      "loss": 0.625,
-      "step": 93000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.047674999999999e-05,
-      "loss": 0.6227,
-      "step": 94000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.122675e-05,
-      "loss": 0.6223,
-      "step": 95000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.6109737753868103,
-      "eval_runtime": 80.7059,
-      "eval_samples_per_second": 79.3,
-      "eval_steps_per_second": 0.62,
-      "step": 95000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.197599999999999e-05,
-      "loss": 0.6206,
-      "step": 96000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.2726e-05,
-      "loss": 0.6193,
-      "step": 97000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.3476e-05,
-      "loss": 0.6178,
-      "step": 98000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.422524999999999e-05,
-      "loss": 0.6179,
-      "step": 99000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.497524999999998e-05,
-      "loss": 0.6168,
-      "step": 100000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.6086182594299316,
-      "eval_runtime": 81.3282,
-      "eval_samples_per_second": 78.693,
-      "eval_steps_per_second": 0.615,
-      "step": 100000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.572449999999999e-05,
-      "loss": 0.6163,
-      "step": 101000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.64745e-05,
-      "loss": 0.6152,
-      "step": 102000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.72245e-05,
-      "loss": 0.6134,
-      "step": 103000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.797374999999999e-05,
-      "loss": 0.612,
-      "step": 104000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.872375e-05,
-      "loss": 0.6118,
-      "step": 105000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.6021054983139038,
-      "eval_runtime": 80.8303,
-      "eval_samples_per_second": 79.178,
-      "eval_steps_per_second": 0.619,
-      "step": 105000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.947225e-05,
-      "loss": 0.6106,
-      "step": 106000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 8.022225e-05,
-      "loss": 0.6099,
-      "step": 107000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 8.097225e-05,
-      "loss": 0.609,
-      "step": 108000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 8.172225e-05,
-      "loss": 0.6088,
-      "step": 109000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.247225e-05,
-      "loss": 0.6063,
-      "step": 110000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.5952876806259155,
-      "eval_runtime": 80.8626,
-      "eval_samples_per_second": 79.147,
-      "eval_steps_per_second": 0.618,
-      "step": 110000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.322224999999999e-05,
-      "loss": 0.6055,
-      "step": 111000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.39715e-05,
-      "loss": 0.6059,
-      "step": 112000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.471999999999999e-05,
-      "loss": 0.6036,
-      "step": 113000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.546999999999999e-05,
-      "loss": 0.6035,
-      "step": 114000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.621999999999998e-05,
-      "loss": 0.6017,
-      "step": 115000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.5951609015464783,
-      "eval_runtime": 80.5699,
-      "eval_samples_per_second": 79.434,
-      "eval_steps_per_second": 0.621,
-      "step": 115000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.696999999999999e-05,
-      "loss": 0.5997,
-      "step": 116000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.771924999999998e-05,
-      "loss": 0.6007,
-      "step": 117000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.846924999999999e-05,
-      "loss": 0.5999,
-      "step": 118000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.921924999999999e-05,
-      "loss": 0.5991,
-      "step": 119000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 8.996924999999998e-05,
-      "loss": 0.5988,
-      "step": 120000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.5855764746665955,
-      "eval_runtime": 79.6751,
-      "eval_samples_per_second": 80.326,
-      "eval_steps_per_second": 0.628,
-      "step": 120000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.07185e-05,
-      "loss": 0.5964,
-      "step": 121000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.14685e-05,
-      "loss": 0.5964,
-      "step": 122000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.221775e-05,
-      "loss": 0.596,
-      "step": 123000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.296774999999999e-05,
-      "loss": 0.5956,
-      "step": 124000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.3717e-05,
-      "loss": 0.5941,
-      "step": 125000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.5847644209861755,
-      "eval_runtime": 79.7033,
-      "eval_samples_per_second": 80.298,
-      "eval_steps_per_second": 0.627,
-      "step": 125000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.446624999999999e-05,
-      "loss": 0.5936,
-      "step": 126000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.521625e-05,
-      "loss": 0.5935,
-      "step": 127000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.596624999999999e-05,
-      "loss": 0.5927,
-      "step": 128000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 9.671624999999999e-05,
-      "loss": 0.5925,
-      "step": 129000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 9.746549999999998e-05,
-      "loss": 0.5914,
-      "step": 130000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.5818326473236084,
-      "eval_runtime": 80.9583,
-      "eval_samples_per_second": 79.053,
-      "eval_steps_per_second": 0.618,
-      "step": 130000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 9.821549999999999e-05,
-      "loss": 0.5887,
-      "step": 131000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 9.896549999999999e-05,
-      "loss": 0.5892,
-      "step": 132000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 9.971475e-05,
-      "loss": 0.5884,
-      "step": 133000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010046475,
-      "loss": 0.5882,
-      "step": 134000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010121249999999999,
-      "loss": 0.5879,
-      "step": 135000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.5765495300292969,
-      "eval_runtime": 78.6913,
-      "eval_samples_per_second": 81.33,
-      "eval_steps_per_second": 0.635,
-      "step": 135000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010196249999999999,
-      "loss": 0.5872,
-      "step": 136000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010271249999999998,
-      "loss": 0.5854,
-      "step": 137000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010346249999999999,
-      "loss": 0.5846,
-      "step": 138000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010421249999999999,
-      "loss": 0.5845,
-      "step": 139000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010496174999999998,
-      "loss": 0.5847,
-      "step": 140000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.5738531351089478,
-      "eval_runtime": 79.0795,
-      "eval_samples_per_second": 80.931,
-      "eval_steps_per_second": 0.632,
-      "step": 140000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010571174999999999,
-      "loss": 0.5829,
-      "step": 141000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010646174999999998,
-      "loss": 0.5832,
-      "step": 142000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010721174999999999,
-      "loss": 0.5826,
-      "step": 143000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010796099999999999,
-      "loss": 0.581,
-      "step": 144000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.000108711,
-      "loss": 0.5812,
-      "step": 145000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.570769190788269,
-      "eval_runtime": 78.7666,
-      "eval_samples_per_second": 81.253,
-      "eval_steps_per_second": 0.635,
-      "step": 145000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010946024999999999,
-      "loss": 0.5804,
-      "step": 146000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00011021025,
-      "loss": 0.5807,
-      "step": 147000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00011096025,
-      "loss": 0.5797,
-      "step": 148000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00011170949999999999,
-      "loss": 0.5804,
-      "step": 149000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001124595,
-      "loss": 0.5785,
-      "step": 150000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.5669087767601013,
-      "eval_runtime": 78.7624,
-      "eval_samples_per_second": 81.257,
-      "eval_steps_per_second": 0.635,
-      "step": 150000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011320874999999999,
-      "loss": 0.5777,
-      "step": 151000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011395875,
-      "loss": 0.5778,
-      "step": 152000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011470799999999999,
-      "loss": 0.5772,
-      "step": 153000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011545799999999998,
-      "loss": 0.5776,
-      "step": 154000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011620799999999998,
-      "loss": 0.5745,
-      "step": 155000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.5640676021575928,
-      "eval_runtime": 79.5789,
-      "eval_samples_per_second": 80.423,
-      "eval_steps_per_second": 0.628,
-      "step": 155000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011695799999999999,
-      "loss": 0.5758,
-      "step": 156000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011770725,
-      "loss": 0.5747,
-      "step": 157000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011845724999999998,
-      "loss": 0.5743,
-      "step": 158000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00011920649999999999,
-      "loss": 0.5736,
-      "step": 159000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001199565,
-      "loss": 0.5738,
-      "step": 160000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.5617207884788513,
-      "eval_runtime": 79.215,
-      "eval_samples_per_second": 80.793,
-      "eval_steps_per_second": 0.631,
-      "step": 160000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00012070574999999999,
-      "loss": 0.5723,
-      "step": 161000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00012145575,
-      "loss": 0.5704,
-      "step": 162000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00012220575,
-      "loss": 0.5703,
-      "step": 163000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00012295499999999998,
-      "loss": 0.5708,
-      "step": 164000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.000123705,
-      "loss": 0.5712,
-      "step": 165000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.5598272085189819,
-      "eval_runtime": 78.3832,
-      "eval_samples_per_second": 81.65,
-      "eval_steps_per_second": 0.638,
-      "step": 165000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00012445424999999997,
-      "loss": 0.57,
-      "step": 166000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00012520425,
-      "loss": 0.5686,
-      "step": 167000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001259535,
-      "loss": 0.5678,
-      "step": 168000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001267035,
-      "loss": 0.5695,
-      "step": 169000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00012745349999999998,
-      "loss": 0.5689,
-      "step": 170000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.5588927865028381,
-      "eval_runtime": 79.0237,
-      "eval_samples_per_second": 80.988,
-      "eval_steps_per_second": 0.633,
-      "step": 170000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00012820275,
-      "loss": 0.5689,
-      "step": 171000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00012895275,
-      "loss": 0.5668,
-      "step": 172000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00012970199999999999,
-      "loss": 0.5663,
-      "step": 173000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013045125,
-      "loss": 0.5655,
-      "step": 174000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013120124999999998,
-      "loss": 0.566,
-      "step": 175000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.5537316203117371,
-      "eval_runtime": 78.8358,
-      "eval_samples_per_second": 81.181,
-      "eval_steps_per_second": 0.634,
-      "step": 175000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013195124999999997,
-      "loss": 0.5653,
-      "step": 176000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013270049999999998,
-      "loss": 0.5651,
-      "step": 177000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013345049999999997,
-      "loss": 0.5649,
-      "step": 178000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001342005,
-      "loss": 0.5624,
-      "step": 179000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013495049999999998,
-      "loss": 0.5645,
-      "step": 180000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.5552637577056885,
-      "eval_runtime": 77.1932,
-      "eval_samples_per_second": 82.909,
-      "eval_steps_per_second": 0.648,
-      "step": 180000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013569975,
-      "loss": 0.5634,
-      "step": 181000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013644974999999998,
-      "loss": 0.5635,
-      "step": 182000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013719974999999997,
-      "loss": 0.5625,
-      "step": 183000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.000137949,
-      "loss": 0.5614,
-      "step": 184000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.000138699,
-      "loss": 0.5606,
-      "step": 185000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.5475028157234192,
-      "eval_runtime": 79.3213,
-      "eval_samples_per_second": 80.684,
-      "eval_steps_per_second": 0.63,
-      "step": 185000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013944899999999999,
-      "loss": 0.561,
-      "step": 186000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00014019749999999998,
-      "loss": 0.5607,
-      "step": 187000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001409475,
-      "loss": 0.5589,
-      "step": 188000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001416975,
-      "loss": 0.5586,
-      "step": 189000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014244749999999998,
-      "loss": 0.5586,
-      "step": 190000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.5459640026092529,
-      "eval_runtime": 79.4666,
-      "eval_samples_per_second": 80.537,
-      "eval_steps_per_second": 0.629,
-      "step": 190000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014319675,
-      "loss": 0.5576,
-      "step": 191000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014394674999999998,
-      "loss": 0.5587,
-      "step": 192000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014469599999999998,
-      "loss": 0.5584,
-      "step": 193000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014544599999999998,
-      "loss": 0.5583,
-      "step": 194000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014619524999999998,
-      "loss": 0.5585,
-      "step": 195000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.5435135960578918,
-      "eval_runtime": 80.1194,
-      "eval_samples_per_second": 79.881,
-      "eval_steps_per_second": 0.624,
-      "step": 195000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001469445,
-      "loss": 0.5572,
-      "step": 196000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014769449999999998,
-      "loss": 0.5556,
-      "step": 197000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014844374999999999,
-      "loss": 0.5552,
-      "step": 198000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014919374999999998,
-      "loss": 0.5557,
-      "step": 199000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014994375,
-      "loss": 0.554,
-      "step": 200000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.5443071126937866,
-      "eval_runtime": 78.383,
-      "eval_samples_per_second": 81.65,
-      "eval_steps_per_second": 0.638,
-      "step": 200000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014999990877662188,
-      "loss": 0.5531,
-      "step": 201000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001499996057416134,
-      "loss": 0.5546,
-      "step": 202000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014999908908161323,
-      "loss": 0.5529,
-      "step": 203000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001499983591922482,
-      "loss": 0.5527,
-      "step": 204000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001499974171253642,
-      "loss": 0.5518,
-      "step": 205000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.5403967499732971,
-      "eval_runtime": 79.4655,
-      "eval_samples_per_second": 80.538,
-      "eval_steps_per_second": 0.629,
-      "step": 205000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014999626099781158,
-      "loss": 0.5512,
-      "step": 206000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014999489312535964,
-      "loss": 0.5513,
-      "step": 207000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014999331077369244,
-      "loss": 0.5519,
-      "step": 208000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014999151521026205,
-      "loss": 0.5513,
-      "step": 209000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00014998950644053808,
-      "loss": 0.5499,
-      "step": 210000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.5364459753036499,
-      "eval_runtime": 78.5073,
-      "eval_samples_per_second": 81.521,
-      "eval_steps_per_second": 0.637,
-      "step": 210000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001499872844706396,
-      "loss": 0.5499,
-      "step": 211000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001499848493073351,
-      "loss": 0.548,
-      "step": 212000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001499822037128756,
-      "loss": 0.5483,
-      "step": 213000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014997934536662948,
-      "loss": 0.5483,
-      "step": 214000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014997627427728645,
-      "loss": 0.5483,
-      "step": 215000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.5351453423500061,
-      "eval_runtime": 79.4482,
-      "eval_samples_per_second": 80.556,
-      "eval_steps_per_second": 0.629,
-      "step": 215000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014997298706049593,
-      "loss": 0.5467,
-      "step": 216000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014996948669385182,
-      "loss": 0.5479,
-      "step": 217000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014996577318801676,
-      "loss": 0.5476,
-      "step": 218000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014996185058739,
-      "loss": 0.5451,
-      "step": 219000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014995771105086807,
-      "loss": 0.5458,
-      "step": 220000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.5317646861076355,
-      "eval_runtime": 79.21,
-      "eval_samples_per_second": 80.798,
-      "eval_steps_per_second": 0.631,
-      "step": 220000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014995335841102588,
-      "loss": 0.5443,
-      "step": 221000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014994879268112236,
-      "loss": 0.5443,
-      "step": 222000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014994401876029827,
-      "loss": 0.5445,
-      "step": 223000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014993902710569922,
-      "loss": 0.5446,
-      "step": 224000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014993382240469446,
-      "loss": 0.5442,
-      "step": 225000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.5309335589408875,
-      "eval_runtime": 78.9413,
-      "eval_samples_per_second": 81.073,
-      "eval_steps_per_second": 0.633,
-      "step": 225000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014992841019727328,
-      "loss": 0.5452,
-      "step": 226000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001499227796646749,
-      "loss": 0.5426,
-      "step": 227000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001499169361351633,
-      "loss": 0.5419,
-      "step": 228000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00014991087962653893,
-      "loss": 0.542,
-      "step": 229000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014990461653308776,
-      "loss": 0.5432,
-      "step": 230000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.5244734883308411,
-      "eval_runtime": 78.2913,
-      "eval_samples_per_second": 81.746,
-      "eval_steps_per_second": 0.639,
-      "step": 230000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014989813433516588,
-      "loss": 0.5417,
-      "step": 231000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014989144601687223,
-      "loss": 0.5414,
-      "step": 232000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014988454522270304,
-      "loss": 0.5408,
-      "step": 233000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014987742474683482,
-      "loss": 0.5409,
-      "step": 234000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014987009141221238,
-      "loss": 0.5391,
-      "step": 235000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.5243217349052429,
-      "eval_runtime": 78.6423,
-      "eval_samples_per_second": 81.381,
-      "eval_steps_per_second": 0.636,
-      "step": 235000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001498625528936495,
-      "loss": 0.5386,
-      "step": 236000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014985479412198444,
-      "loss": 0.5398,
-      "step": 237000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014984682256050195,
-      "loss": 0.5385,
-      "step": 238000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001498386382334848,
-      "loss": 0.5394,
-      "step": 239000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001498302496691869,
-      "loss": 0.5377,
-      "step": 240000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.5230529308319092,
-      "eval_runtime": 77.6998,
-      "eval_samples_per_second": 82.368,
-      "eval_steps_per_second": 0.644,
-      "step": 240000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014982164881505554,
-      "loss": 0.5377,
-      "step": 241000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001498128267689627,
-      "loss": 0.5373,
-      "step": 242000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014980379206089227,
-      "loss": 0.5354,
-      "step": 243000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014979455407190963,
-      "loss": 0.5354,
-      "step": 244000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001497850943356876,
-      "loss": 0.5348,
-      "step": 245000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.5176389217376709,
-      "eval_runtime": 78.4635,
-      "eval_samples_per_second": 81.567,
-      "eval_steps_per_second": 0.637,
-      "step": 245000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014977544157872605,
-      "loss": 0.5361,
-      "step": 246000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014976555714203422,
-      "loss": 0.5355,
-      "step": 247000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00014975546018735622,
-      "loss": 0.5346,
-      "step": 248000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001497451507454492,
-      "loss": 0.5336,
-      "step": 249000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001497346288477175,
-      "loss": 0.5338,
-      "step": 250000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.5214279294013977,
-      "eval_runtime": 79.6078,
-      "eval_samples_per_second": 80.394,
-      "eval_steps_per_second": 0.628,
-      "step": 250000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014972390536662896,
-      "loss": 0.5344,
-      "step": 251000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014971296991900254,
-      "loss": 0.5342,
-      "step": 252000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014970181127337587,
-      "loss": 0.5347,
-      "step": 253000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.000149690440303944,
-      "loss": 0.5323,
-      "step": 254000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.000149678857045345,
-      "loss": 0.5328,
-      "step": 255000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.5172922611236572,
-      "eval_runtime": 79.325,
-      "eval_samples_per_second": 80.681,
-      "eval_steps_per_second": 0.63,
-      "step": 255000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014966706153286336,
-      "loss": 0.5313,
-      "step": 256000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.0001496550659161515,
-      "loss": 0.5312,
-      "step": 257000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014964284621650758,
-      "loss": 0.5321,
-      "step": 258000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.0001496304143726763,
-      "loss": 0.531,
-      "step": 259000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014961777042252718,
-      "loss": 0.5295,
-      "step": 260000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.5135677456855774,
-      "eval_runtime": 77.9166,
-      "eval_samples_per_second": 82.139,
-      "eval_steps_per_second": 0.642,
-      "step": 260000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.0001496049403282296,
-      "loss": 0.5307,
-      "step": 261000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014959187270565565,
-      "loss": 0.53,
-      "step": 262000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014957860647965505,
-      "loss": 0.5301,
-      "step": 263000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014956511513163472,
-      "loss": 0.5297,
-      "step": 264000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014955143949417623,
-      "loss": 0.5295,
-      "step": 265000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.513587236404419,
-      "eval_runtime": 79.2411,
-      "eval_samples_per_second": 80.766,
-      "eval_steps_per_second": 0.631,
-      "step": 265000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014953752479677766,
-      "loss": 0.5289,
-      "step": 266000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014952339827601903,
-      "loss": 0.5306,
-      "step": 267000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014950905997493217,
-      "loss": 0.5292,
-      "step": 268000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00014949450993719408,
-      "loss": 0.5277,
-      "step": 269000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014947976307458214,
-      "loss": 0.5274,
-      "step": 270000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.5130758881568909,
-      "eval_runtime": 82.5777,
-      "eval_samples_per_second": 77.503,
-      "eval_steps_per_second": 0.605,
-      "step": 270000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014946478990877685,
-      "loss": 0.5284,
-      "step": 271000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014944960514117462,
-      "loss": 0.5273,
-      "step": 272000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014943420881803084,
-      "loss": 0.5271,
-      "step": 273000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014941861669970997,
-      "loss": 0.526,
-      "step": 274000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014940281354295447,
-      "loss": 0.5265,
-      "step": 275000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.510474681854248,
-      "eval_runtime": 79.0154,
-      "eval_samples_per_second": 80.997,
-      "eval_steps_per_second": 0.633,
-      "step": 275000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014938678325993965,
-      "loss": 0.5256,
-      "step": 276000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014937054161274965,
-      "loss": 0.5268,
-      "step": 277000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014935408865085934,
-      "loss": 0.5247,
-      "step": 278000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014933742442438727,
-      "loss": 0.5247,
-      "step": 279000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014932056596502014,
-      "loss": 0.5239,
-      "step": 280000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.5078325867652893,
-      "eval_runtime": 79.4774,
-      "eval_samples_per_second": 80.526,
-      "eval_steps_per_second": 0.629,
-      "step": 280000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014930347957345093,
-      "loss": 0.5242,
-      "step": 281000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014928619947439815,
-      "loss": 0.5243,
-      "step": 282000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014926869112571586,
-      "loss": 0.5242,
-      "step": 283000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014925098959731945,
-      "loss": 0.524,
-      "step": 284000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001492330595046446,
-      "loss": 0.525,
-      "step": 285000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.508979320526123,
-      "eval_runtime": 78.5336,
-      "eval_samples_per_second": 81.494,
-      "eval_steps_per_second": 0.637,
-      "step": 285000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014921491851606374,
-      "loss": 0.5229,
-      "step": 286000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014919658514396308,
-      "loss": 0.5233,
-      "step": 287000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014917802274075096,
-      "loss": 0.5229,
-      "step": 288000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014915926848765564,
-      "loss": 0.5233,
-      "step": 289000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00014914028489576315,
-      "loss": 0.5225,
-      "step": 290000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.5040385723114014,
-      "eval_runtime": 82.1473,
-      "eval_samples_per_second": 77.909,
-      "eval_steps_per_second": 0.609,
-      "step": 290000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.0001491210906905722,
-      "loss": 0.5212,
-      "step": 291000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00014910168593055176,
-      "loss": 0.5221,
-      "step": 292000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.000149082070674812,
-      "loss": 0.522,
-      "step": 293000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.0001490622649138889,
-      "loss": 0.521,
-      "step": 294000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00014904224919787054,
-      "loss": 0.5206,
-      "step": 295000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.5053508877754211,
-      "eval_runtime": 82.4285,
-      "eval_samples_per_second": 77.643,
-      "eval_steps_per_second": 0.607,
-      "step": 295000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00014902200323661346,
-      "loss": 0.5199,
-      "step": 296000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.0001490015470223644,
-      "loss": 0.5213,
-      "step": 297000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00014898090138881081,
-      "loss": 0.5202,
-      "step": 298000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00014896004604751396,
-      "loss": 0.5201,
-      "step": 299000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00014893895987079335,
-      "loss": 0.5199,
-      "step": 300000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.5044429898262024,
-      "eval_runtime": 81.8484,
-      "eval_samples_per_second": 78.193,
-      "eval_steps_per_second": 0.611,
-      "step": 300000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014891766369404524,
-      "loss": 0.519,
-      "step": 301000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.0001488961575821415,
-      "loss": 0.519,
-      "step": 302000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014887444160059348,
-      "loss": 0.5188,
-      "step": 303000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014885251581555183,
-      "loss": 0.5185,
-      "step": 304000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014883040253406912,
-      "loss": 0.5184,
-      "step": 305000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.5007572174072266,
-      "eval_runtime": 81.9944,
-      "eval_samples_per_second": 78.054,
-      "eval_steps_per_second": 0.61,
-      "step": 305000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014880805755268368,
-      "loss": 0.5178,
-      "step": 306000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014878550297002193,
-      "loss": 0.518,
-      "step": 307000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014876276172354279,
-      "loss": 0.5176,
-      "step": 308000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014873978835451058,
-      "loss": 0.5193,
-      "step": 309000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014871662887949272,
-      "loss": 0.5163,
-      "step": 310000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.5015575289726257,
-      "eval_runtime": 79.7243,
-      "eval_samples_per_second": 80.277,
-      "eval_steps_per_second": 0.627,
-      "step": 310000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014869323700373343,
-      "loss": 0.518,
-      "step": 311000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014866965958208565,
-      "loss": 0.5179,
-      "step": 312000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001486458733988959,
-      "loss": 0.5176,
-      "step": 313000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014862185440305553,
-      "loss": 0.5176,
-      "step": 314000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014859762637314596,
-      "loss": 0.516,
-      "step": 315000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.5005845427513123,
-      "eval_runtime": 78.6119,
-      "eval_samples_per_second": 81.413,
-      "eval_steps_per_second": 0.636,
-      "step": 315000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001485731893829699,
-      "loss": 0.5173,
-      "step": 316000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014854856825715626,
-      "loss": 0.5169,
-      "step": 317000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001485237137791746,
-      "loss": 0.5151,
-      "step": 318000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014849867573352797,
-      "loss": 0.5154,
-      "step": 319000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014847340407028189,
-      "loss": 0.516,
-      "step": 320000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.500928521156311,
-      "eval_runtime": 79.795,
-      "eval_samples_per_second": 80.206,
-      "eval_steps_per_second": 0.627,
-      "step": 320000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014844792382517202,
-      "loss": 0.5163,
-      "step": 321000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014842223507581563,
-      "loss": 0.5155,
-      "step": 322000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014839633790046512,
-      "loss": 0.515,
-      "step": 323000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001483702847969303,
-      "loss": 0.514,
-      "step": 324000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001483439714233434,
-      "loss": 0.5138,
-      "step": 325000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.5003533363342285,
-      "eval_runtime": 78.4102,
-      "eval_samples_per_second": 81.622,
-      "eval_steps_per_second": 0.638,
-      "step": 325000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014831744986216702,
-      "loss": 0.5143,
-      "step": 326000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014829072019419042,
-      "loss": 0.5139,
-      "step": 327000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001482638095424116,
-      "loss": 0.5143,
-      "step": 328000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014823666411364003,
-      "loss": 0.513,
-      "step": 329000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014820931082415564,
-      "loss": 0.5136,
-      "step": 330000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.49581122398376465,
-      "eval_runtime": 78.2187,
-      "eval_samples_per_second": 81.822,
-      "eval_steps_per_second": 0.639,
-      "step": 330000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001481817497572813,
-      "loss": 0.5129,
-      "step": 331000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014815398099697283,
-      "loss": 0.5129,
-      "step": 332000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014812603270786007,
-      "loss": 0.5123,
-      "step": 333000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014809784902256195,
-      "loss": 0.512,
-      "step": 334000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014806948639411544,
-      "loss": 0.5112,
-      "step": 335000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.4955292046070099,
-      "eval_runtime": 79.8311,
-      "eval_samples_per_second": 80.169,
-      "eval_steps_per_second": 0.626,
-      "step": 335000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001480408881268936,
-      "loss": 0.5122,
-      "step": 336000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014801211150435752,
-      "loss": 0.5114,
-      "step": 337000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014798312811928091,
-      "loss": 0.5126,
-      "step": 338000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014795390873698235,
-      "loss": 0.5112,
-      "step": 339000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014792451188504096,
-      "loss": 0.5114,
-      "step": 340000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.4929465353488922,
-      "eval_runtime": 78.1965,
-      "eval_samples_per_second": 81.845,
-      "eval_steps_per_second": 0.639,
-      "step": 340000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014789487880046323,
-      "loss": 0.5116,
-      "step": 341000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014786503889631,
-      "loss": 0.5108,
-      "step": 342000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014783499226347888,
-      "loss": 0.5108,
-      "step": 343000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014780473899349703,
-      "loss": 0.5101,
-      "step": 344000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001477743403042215,
-      "loss": 0.5103,
-      "step": 345000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.49353981018066406,
-      "eval_runtime": 79.8609,
-      "eval_samples_per_second": 80.139,
-      "eval_steps_per_second": 0.626,
-      "step": 345000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014774367444984895,
-      "loss": 0.51,
-      "step": 346000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014771280223649564,
-      "loss": 0.5104,
-      "step": 347000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014768172375820364,
-      "loss": 0.5083,
-      "step": 348000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014765047049724233,
-      "loss": 0.5092,
-      "step": 349000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001476190115731596,
-      "loss": 0.5097,
-      "step": 350000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.49557891488075256,
-      "eval_runtime": 80.7409,
-      "eval_samples_per_second": 79.266,
-      "eval_steps_per_second": 0.619,
-      "step": 350000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001475873152824479,
-      "loss": 0.5086,
-      "step": 351000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001475554131090526,
-      "loss": 0.508,
-      "step": 352000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014752330515015318,
-      "loss": 0.509,
-      "step": 353000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014749102391991098,
-      "loss": 0.5094,
-      "step": 354000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.000147458504889589,
-      "loss": 0.5079,
-      "step": 355000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.49167051911354065,
-      "eval_runtime": 79.1142,
-      "eval_samples_per_second": 80.896,
-      "eval_steps_per_second": 0.632,
-      "step": 355000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014742578036896223,
-      "loss": 0.5081,
-      "step": 356000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001473929165224509,
-      "loss": 0.51,
-      "step": 357000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014735978173137384,
-      "loss": 0.507,
-      "step": 358000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014732644175071978,
-      "loss": 0.5079,
-      "step": 359000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014729289668204807,
-      "loss": 0.5079,
-      "step": 360000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.48893874883651733,
-      "eval_runtime": 79.0429,
-      "eval_samples_per_second": 80.969,
-      "eval_steps_per_second": 0.633,
-      "step": 360000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014725914662754285,
-      "loss": 0.5076,
-      "step": 361000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014722522574725483,
-      "loss": 0.5066,
-      "step": 362000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001471910662348599,
-      "loss": 0.5085,
-      "step": 363000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014715670204682463,
-      "loss": 0.5072,
-      "step": 364000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014712220262943823,
-      "loss": 0.5059,
-      "step": 365000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4867703914642334,
-      "eval_runtime": 74.6947,
-      "eval_samples_per_second": 85.682,
-      "eval_steps_per_second": 0.669,
-      "step": 365000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001470874298136091,
-      "loss": 0.506,
-      "step": 366000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014705245263783414,
-      "loss": 0.5073,
-      "step": 367000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014701727120865993,
-      "loss": 0.5052,
-      "step": 368000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014698188563325515,
-      "loss": 0.5066,
-      "step": 369000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014694636740219654,
-      "loss": 0.5055,
-      "step": 370000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4855530560016632,
-      "eval_runtime": 79.7083,
-      "eval_samples_per_second": 80.293,
-      "eval_steps_per_second": 0.627,
-      "step": 370000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014691057426607544,
-      "loss": 0.5063,
-      "step": 371000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001468745773087412,
-      "loss": 0.5064,
-      "step": 372000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014683841294223294,
-      "loss": 0.5051,
-      "step": 373000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014680200887559787,
-      "loss": 0.5059,
-      "step": 374000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001467654380276222,
-      "loss": 0.5066,
-      "step": 375000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4906100034713745,
-      "eval_runtime": 80.9132,
-      "eval_samples_per_second": 79.097,
-      "eval_steps_per_second": 0.618,
-      "step": 375000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014672862729481488,
-      "loss": 0.5051,
-      "step": 376000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014669161329503643,
-      "loss": 0.505,
-      "step": 377000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001466543961410379,
-      "loss": 0.5034,
-      "step": 378000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001466170134677651,
-      "loss": 0.5048,
-      "step": 379000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001465793905489244,
-      "loss": 0.5054,
-      "step": 380000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4870378375053406,
-      "eval_runtime": 80.5642,
-      "eval_samples_per_second": 79.44,
-      "eval_steps_per_second": 0.621,
-      "step": 380000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014654160274471124,
-      "loss": 0.5036,
-      "step": 381000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014650357451899344,
-      "loss": 0.5046,
-      "step": 382000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014646538204381214,
-      "loss": 0.5035,
-      "step": 383000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014642694897411588,
-      "loss": 0.504,
-      "step": 384000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001463883522928644,
-      "loss": 0.504,
-      "step": 385000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.48908066749572754,
-      "eval_runtime": 80.3298,
-      "eval_samples_per_second": 79.672,
-      "eval_steps_per_second": 0.622,
-      "step": 385000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014634951484701638,
-      "loss": 0.5041,
-      "step": 386000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014631047528903952,
-      "loss": 0.504,
-      "step": 387000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014627123373785504,
-      "loss": 0.5036,
-      "step": 388000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00014623182985722033,
-      "loss": 0.5033,
-      "step": 389000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001461922246262511,
-      "loss": 0.502,
-      "step": 390000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4866912066936493,
-      "eval_runtime": 80.948,
-      "eval_samples_per_second": 79.063,
-      "eval_steps_per_second": 0.618,
-      "step": 390000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001461523782182672,
-      "loss": 0.5041,
-      "step": 391000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014611233029866684,
-      "loss": 0.5011,
-      "step": 392000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014607208098944285,
-      "loss": 0.5036,
-      "step": 393000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014603167096426982,
-      "loss": 0.5021,
-      "step": 394000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014599101944531292,
-      "loss": 0.5024,
-      "step": 395000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4846630394458771,
-      "eval_runtime": 80.3433,
-      "eval_samples_per_second": 79.658,
-      "eval_steps_per_second": 0.622,
-      "step": 395000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014595020785917355,
-      "loss": 0.5013,
-      "step": 396000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014590915462531445,
-      "loss": 0.5013,
-      "step": 397000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001458679006207401,
-      "loss": 0.5011,
-      "step": 398000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001458264875259472,
-      "loss": 0.5018,
-      "step": 399000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001457848325580096,
-      "loss": 0.5021,
-      "step": 400000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.486747145652771,
-      "eval_runtime": 75.8258,
-      "eval_samples_per_second": 84.404,
-      "eval_steps_per_second": 0.659,
-      "step": 400000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001457429771980633,
-      "loss": 0.5019,
-      "step": 401000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.000145700963729221,
-      "loss": 0.502,
-      "step": 402000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014565870816843566,
-      "loss": 0.5006,
-      "step": 403000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001456162951552679,
-      "loss": 0.5002,
-      "step": 404000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014557363990799942,
-      "loss": 0.5011,
-      "step": 405000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4839228689670563,
-      "eval_runtime": 81.4745,
-      "eval_samples_per_second": 78.552,
-      "eval_steps_per_second": 0.614,
-      "step": 405000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014553078491205172,
-      "loss": 0.5007,
-      "step": 406000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00014548777345224824,
-      "loss": 0.5004,
-      "step": 407000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001454445629041043,
-      "loss": 0.5006,
-      "step": 408000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001454011098463918,
-      "loss": 0.5018,
-      "step": 409000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014535745756555653,
-      "loss": 0.4996,
-      "step": 410000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4806668162345886,
-      "eval_runtime": 80.4842,
-      "eval_samples_per_second": 79.519,
-      "eval_steps_per_second": 0.621,
-      "step": 410000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014531360619457094,
-      "loss": 0.4996,
-      "step": 411000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014526955586701388,
-      "loss": 0.5014,
-      "step": 412000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014522530671707033,
-      "loss": 0.4991,
-      "step": 413000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001451809479734062,
-      "loss": 0.4997,
-      "step": 414000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014513630198063546,
-      "loss": 0.4985,
-      "step": 415000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4842430055141449,
-      "eval_runtime": 79.3255,
-      "eval_samples_per_second": 80.68,
-      "eval_steps_per_second": 0.63,
-      "step": 415000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001450915025148655,
-      "loss": 0.4976,
-      "step": 416000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001450464600239659,
-      "loss": 0.4988,
-      "step": 417000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014500126472982865,
-      "loss": 0.4992,
-      "step": 418000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014495587182637829,
-      "loss": 0.4989,
-      "step": 419000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014491023571671524,
-      "loss": 0.5,
-      "step": 420000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4820484519004822,
-      "eval_runtime": 78.241,
-      "eval_samples_per_second": 81.799,
-      "eval_steps_per_second": 0.639,
-      "step": 420000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014486440187921716,
-      "loss": 0.498,
-      "step": 421000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001448183704535019,
-      "loss": 0.4985,
-      "step": 422000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014477214157978911,
-      "loss": 0.4985,
-      "step": 423000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014472576192358876,
-      "loss": 0.4983,
-      "step": 424000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014467913877404042,
-      "loss": 0.4979,
-      "step": 425000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4826773405075073,
-      "eval_runtime": 78.9952,
-      "eval_samples_per_second": 81.018,
-      "eval_steps_per_second": 0.633,
-      "step": 425000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014463236551915786,
-      "loss": 0.4984,
-      "step": 426000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014458534866129436,
-      "loss": 0.4991,
-      "step": 427000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014453813506525818,
-      "loss": 0.4982,
-      "step": 428000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00014449072487487006,
-      "loss": 0.4987,
-      "step": 429000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001444431182345497,
-      "loss": 0.4986,
-      "step": 430000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4816123843193054,
-      "eval_runtime": 79.868,
-      "eval_samples_per_second": 80.132,
-      "eval_steps_per_second": 0.626,
-      "step": 430000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014439536319026632,
-      "loss": 0.4959,
-      "step": 431000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014434736428182006,
-      "loss": 0.4972,
-      "step": 432000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001442991693601428,
-      "loss": 0.4971,
-      "step": 433000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014425082706061895,
-      "loss": 0.4973,
-      "step": 434000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001442022894331757,
-      "loss": 0.4971,
-      "step": 435000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.47952431440353394,
-      "eval_runtime": 79.6918,
-      "eval_samples_per_second": 80.309,
-      "eval_steps_per_second": 0.627,
-      "step": 435000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014415350774604595,
-      "loss": 0.4975,
-      "step": 436000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001441045306362058,
-      "loss": 0.4959,
-      "step": 437000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014405535825284804,
-      "loss": 0.4956,
-      "step": 438000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014400604021068167,
-      "loss": 0.4963,
-      "step": 439000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001439564779251439,
-      "loss": 0.4968,
-      "step": 440000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4796030819416046,
-      "eval_runtime": 79.6283,
-      "eval_samples_per_second": 80.373,
-      "eval_steps_per_second": 0.628,
-      "step": 440000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014390672081708274,
-      "loss": 0.4956,
-      "step": 441000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014385681908703354,
-      "loss": 0.4963,
-      "step": 442000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014380667298366756,
-      "loss": 0.496,
-      "step": 443000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014375633251411,
-      "loss": 0.4955,
-      "step": 444000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001437057978317067,
-      "loss": 0.4955,
-      "step": 445000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.47917410731315613,
-      "eval_runtime": 79.9877,
-      "eval_samples_per_second": 80.012,
-      "eval_steps_per_second": 0.625,
-      "step": 445000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001436551199160174,
-      "loss": 0.4956,
-      "step": 446000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014360419746415294,
-      "loss": 0.4965,
-      "step": 447000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00014355308126287247,
-      "loss": 0.4953,
-      "step": 448000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001435018228743279,
-      "loss": 0.4965,
-      "step": 449000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001434503198352907,
-      "loss": 0.4951,
-      "step": 450000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.47704780101776123,
-      "eval_runtime": 78.6754,
-      "eval_samples_per_second": 81.347,
-      "eval_steps_per_second": 0.636,
-      "step": 450000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014339862351557547,
-      "loss": 0.4934,
-      "step": 451000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.0001433467860585135,
-      "loss": 0.4934,
-      "step": 452000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014329470384334429,
-      "loss": 0.4932,
-      "step": 453000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014324242882153025,
-      "loss": 0.4935,
-      "step": 454000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014318996115231028,
-      "loss": 0.4931,
-      "step": 455000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.4748893678188324,
-      "eval_runtime": 82.5143,
-      "eval_samples_per_second": 77.562,
-      "eval_steps_per_second": 0.606,
-      "step": 455000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014313730099550992,
-      "loss": 0.4927,
-      "step": 456000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014308455440834473,
-      "loss": 0.4931,
-      "step": 457000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.0001430315101423767,
-      "loss": 0.4928,
-      "step": 458000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00014297827387149757,
-      "loss": 0.4945,
-      "step": 459000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014292484575787425,
-      "loss": 0.4917,
-      "step": 460000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4765481948852539,
-      "eval_runtime": 80.7233,
-      "eval_samples_per_second": 79.283,
-      "eval_steps_per_second": 0.619,
-      "step": 460000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014287133339503343,
-      "loss": 0.4935,
-      "step": 461000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001428175224676294,
-      "loss": 0.4942,
-      "step": 462000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014276352018715785,
-      "loss": 0.4933,
-      "step": 463000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001427093267181191,
-      "loss": 0.4933,
-      "step": 464000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014265494222559578,
-      "loss": 0.4927,
-      "step": 465000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4773631989955902,
-      "eval_runtime": 80.561,
-      "eval_samples_per_second": 79.443,
-      "eval_steps_per_second": 0.621,
-      "step": 465000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014260036687525253,
-      "loss": 0.4932,
-      "step": 466000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001425457105556186,
-      "loss": 0.4931,
-      "step": 467000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014249075436983697,
-      "loss": 0.4911,
-      "step": 468000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014243560782638008,
-      "loss": 0.4917,
-      "step": 469000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014238027109323388,
-      "loss": 0.4925,
-      "step": 470000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4760352373123169,
-      "eval_runtime": 79.1679,
-      "eval_samples_per_second": 80.841,
-      "eval_steps_per_second": 0.632,
-      "step": 470000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014232479996057723,
-      "loss": 0.4928,
-      "step": 471000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001422690835440944,
-      "loss": 0.4921,
-      "step": 472000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014221317744518412,
-      "loss": 0.4919,
-      "step": 473000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014215708183414603,
-      "loss": 0.4908,
-      "step": 474000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014210085326132836,
-      "loss": 0.4908,
-      "step": 475000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.47689467668533325,
-      "eval_runtime": 79.0304,
-      "eval_samples_per_second": 80.981,
-      "eval_steps_per_second": 0.633,
-      "step": 475000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001420443793283263,
-      "loss": 0.4908,
-      "step": 476000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001419877731546628,
-      "loss": 0.4906,
-      "step": 477000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014193092158712586,
-      "loss": 0.4914,
-      "step": 478000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00014187393850165534,
-      "loss": 0.4906,
-      "step": 479000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001418167099916094,
-      "loss": 0.4912,
-      "step": 480000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4743531346321106,
-      "eval_runtime": 80.9996,
-      "eval_samples_per_second": 79.013,
-      "eval_steps_per_second": 0.617,
-      "step": 480000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001417592931771443,
-      "loss": 0.49,
-      "step": 481000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001417016882331616,
-      "loss": 0.4902,
-      "step": 482000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014164395322185843,
-      "loss": 0.4899,
-      "step": 483000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001415859727335269,
-      "loss": 0.4897,
-      "step": 484000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014152786290537982,
-      "loss": 0.4893,
-      "step": 485000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.473435640335083,
-      "eval_runtime": 81.6676,
-      "eval_samples_per_second": 78.366,
-      "eval_steps_per_second": 0.612,
-      "step": 485000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014146950757846707,
-      "loss": 0.49,
-      "step": 486000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001414110236408181,
-      "loss": 0.4894,
-      "step": 487000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014135229418565457,
-      "loss": 0.4909,
-      "step": 488000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014129337784076295,
-      "loss": 0.4892,
-      "step": 489000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014123427478561265,
-      "loss": 0.4899,
-      "step": 490000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.47095221281051636,
-      "eval_runtime": 81.6054,
-      "eval_samples_per_second": 78.426,
-      "eval_steps_per_second": 0.613,
-      "step": 490000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.000141175044582939,
-      "loss": 0.49,
-      "step": 491000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001411155688342132,
-      "loss": 0.4894,
-      "step": 492000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001410559069168658,
-      "loss": 0.4898,
-      "step": 493000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014099605901263734,
-      "loss": 0.4885,
-      "step": 494000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001409360854302923,
-      "loss": 0.4894,
-      "step": 495000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.47246918082237244,
-      "eval_runtime": 79.94,
-      "eval_samples_per_second": 80.06,
-      "eval_steps_per_second": 0.625,
-      "step": 495000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014087592659712124,
-      "loss": 0.4882,
-      "step": 496000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014081552219904862,
-      "loss": 0.4895,
-      "step": 497000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014075493254634853,
-      "loss": 0.4874,
-      "step": 498000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00014069421869069117,
-      "loss": 0.4883,
-      "step": 499000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014063325926779179,
-      "loss": 0.4889,
-      "step": 500000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4746360778808594,
-      "eval_runtime": 80.6889,
-      "eval_samples_per_second": 79.317,
-      "eval_steps_per_second": 0.62,
-      "step": 500000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014057211514546938,
-      "loss": 0.4878,
-      "step": 501000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014051084793071725,
-      "loss": 0.4872,
-      "step": 502000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001404493351531105,
-      "loss": 0.4881,
-      "step": 503000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014038763823634546,
-      "loss": 0.4891,
-      "step": 504000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014032575736836167,
-      "loss": 0.4875,
-      "step": 505000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.47102564573287964,
-      "eval_runtime": 79.1525,
-      "eval_samples_per_second": 80.857,
-      "eval_steps_per_second": 0.632,
-      "step": 505000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014026381705018997,
-      "loss": 0.4878,
-      "step": 506000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014020156921278623,
-      "loss": 0.4892,
-      "step": 507000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014013913799096225,
-      "loss": 0.4868,
-      "step": 508000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001400765862807525,
-      "loss": 0.4881,
-      "step": 509000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00014001378904408309,
-      "loss": 0.4885,
-      "step": 510000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4696248769760132,
-      "eval_runtime": 80.7581,
-      "eval_samples_per_second": 79.249,
-      "eval_steps_per_second": 0.619,
-      "step": 510000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013995093513742188,
-      "loss": 0.4875,
-      "step": 511000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001398877728328313,
-      "loss": 0.4867,
-      "step": 512000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013982442809969868,
-      "loss": 0.4888,
-      "step": 513000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013976090113098301,
-      "loss": 0.4873,
-      "step": 514000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.0001396971921201985,
-      "loss": 0.4863,
-      "step": 515000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.46735358238220215,
-      "eval_runtime": 82.0128,
-      "eval_samples_per_second": 78.037,
-      "eval_steps_per_second": 0.61,
-      "step": 515000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013963336524304095,
-      "loss": 0.4864,
-      "step": 516000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013956929291243476,
-      "loss": 0.4878,
-      "step": 517000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013950510346779393,
-      "loss": 0.4858,
-      "step": 518000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00013944066859728662,
-      "loss": 0.4871,
-      "step": 519000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001393760526601943,
-      "loss": 0.4861,
-      "step": 520000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4665950834751129,
-      "eval_runtime": 80.8998,
-      "eval_samples_per_second": 79.11,
-      "eval_steps_per_second": 0.618,
-      "step": 520000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013931132074043395,
-      "loss": 0.487,
-      "step": 521000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001392463434417902,
-      "loss": 0.4869,
-      "step": 522000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013918118566850956,
-      "loss": 0.4861,
-      "step": 523000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013911591304710506,
-      "loss": 0.4867,
-      "step": 524000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013905039510052293,
-      "loss": 0.4878,
-      "step": 525000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.47149112820625305,
-      "eval_runtime": 79.7289,
-      "eval_samples_per_second": 80.272,
-      "eval_steps_per_second": 0.627,
-      "step": 525000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013898469727619654,
-      "loss": 0.4862,
-      "step": 526000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.000138918885741437,
-      "loss": 0.4859,
-      "step": 527000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001388528289419268,
-      "loss": 0.4846,
-      "step": 528000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013878659286649316,
-      "loss": 0.4859,
-      "step": 529000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001387202442214324,
-      "loss": 0.4864,
-      "step": 530000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.46929115056991577,
-      "eval_runtime": 81.2258,
-      "eval_samples_per_second": 78.793,
-      "eval_steps_per_second": 0.616,
-      "step": 530000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013865365037876758,
-      "loss": 0.485,
-      "step": 531000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013858694472860118,
-      "loss": 0.4861,
-      "step": 532000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013851999392931896,
-      "loss": 0.4848,
-      "step": 533000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001384529320860216,
-      "loss": 0.4852,
-      "step": 534000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013838562514508073,
-      "loss": 0.4854,
-      "step": 535000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.47047823667526245,
-      "eval_runtime": 80.5052,
-      "eval_samples_per_second": 79.498,
-      "eval_steps_per_second": 0.621,
-      "step": 535000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013831814035140017,
-      "loss": 0.4855,
-      "step": 536000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001382504779105503,
-      "loss": 0.4856,
-      "step": 537000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013818277388535536,
-      "loss": 0.4855,
-      "step": 538000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001381148252285261,
-      "loss": 0.4849,
-      "step": 539000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00013804663161577017,
-      "loss": 0.4838,
-      "step": 540000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4682275354862213,
-      "eval_runtime": 81.8185,
-      "eval_samples_per_second": 78.222,
-      "eval_steps_per_second": 0.611,
-      "step": 540000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013797826118290763,
-      "loss": 0.4858,
-      "step": 541000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013790971413820663,
-      "loss": 0.4831,
-      "step": 542000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013784099069047342,
-      "loss": 0.4858,
-      "step": 543000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013777216003663188,
-      "loss": 0.4845,
-      "step": 544000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013770308458728033,
-      "loss": 0.4846,
-      "step": 545000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.46753159165382385,
-      "eval_runtime": 78.6329,
-      "eval_samples_per_second": 81.391,
-      "eval_steps_per_second": 0.636,
-      "step": 545000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013763390270327768,
-      "loss": 0.4835,
-      "step": 546000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001375644760931289,
-      "loss": 0.4832,
-      "step": 547000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013749494382108183,
-      "loss": 0.4834,
-      "step": 548000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013742516689523143,
-      "loss": 0.4843,
-      "step": 549000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013735521504236372,
-      "loss": 0.4847,
-      "step": 550000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4699910283088684,
-      "eval_runtime": 77.0998,
-      "eval_samples_per_second": 83.009,
-      "eval_steps_per_second": 0.649,
-      "step": 550000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013728515868932954,
-      "loss": 0.4843,
-      "step": 551000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013721485779660927,
-      "loss": 0.485,
-      "step": 552000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001371443826175097,
-      "loss": 0.4833,
-      "step": 553000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013707380410283847,
-      "loss": 0.4829,
-      "step": 554000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001370029811692982,
-      "loss": 0.4838,
-      "step": 555000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.46684029698371887,
-      "eval_runtime": 76.4066,
-      "eval_samples_per_second": 83.762,
-      "eval_steps_per_second": 0.654,
-      "step": 555000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001369319845947916,
-      "loss": 0.4838,
-      "step": 556000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013686081459558647,
-      "loss": 0.4836,
-      "step": 557000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013678961424761038,
-      "loss": 0.4837,
-      "step": 558000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00013671809839568835,
-      "loss": 0.4844,
-      "step": 559000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013664640977060228,
-      "loss": 0.4825,
-      "step": 560000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4646618962287903,
-      "eval_runtime": 77.1272,
-      "eval_samples_per_second": 82.98,
-      "eval_steps_per_second": 0.648,
-      "step": 560000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013657462053802633,
-      "loss": 0.4829,
-      "step": 561000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013650258719449216,
-      "loss": 0.4844,
-      "step": 562000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013643045402563777,
-      "loss": 0.4823,
-      "step": 563000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013635807684048157,
-      "loss": 0.4832,
-      "step": 564000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013628560061338996,
-      "loss": 0.4822,
-      "step": 565000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4679878354072571,
-      "eval_runtime": 77.1624,
-      "eval_samples_per_second": 82.942,
-      "eval_steps_per_second": 0.648,
-      "step": 565000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001362128804676232,
-      "loss": 0.4811,
-      "step": 566000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013614006206451144,
-      "loss": 0.4821,
-      "step": 567000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013606699984332037,
-      "loss": 0.4824,
-      "step": 568000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013599376683232436,
-      "loss": 0.4827,
-      "step": 569000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013592036325460387,
-      "loss": 0.4828,
-      "step": 570000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4641859531402588,
-      "eval_runtime": 77.5968,
-      "eval_samples_per_second": 82.478,
-      "eval_steps_per_second": 0.644,
-      "step": 570000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013584686299269162,
-      "loss": 0.4812,
-      "step": 571000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013577319295161774,
-      "loss": 0.4817,
-      "step": 572000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013569927935696313,
-      "loss": 0.4822,
-      "step": 573000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.000135625196092644,
-      "loss": 0.4822,
-      "step": 574000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013555094338433074,
-      "loss": 0.4816,
-      "step": 575000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.46690383553504944,
-      "eval_runtime": 77.9328,
-      "eval_samples_per_second": 82.122,
-      "eval_steps_per_second": 0.642,
-      "step": 575000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013547652145821007,
-      "loss": 0.4822,
-      "step": 576000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00013540200521623672,
-      "loss": 0.4829,
-      "step": 577000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001353272457037724,
-      "loss": 0.4822,
-      "step": 578000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001352523176549226,
-      "loss": 0.4825,
-      "step": 579000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013517729647828186,
-      "loss": 0.4824,
-      "step": 580000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.46679091453552246,
-      "eval_runtime": 78.4477,
-      "eval_samples_per_second": 81.583,
-      "eval_steps_per_second": 0.637,
-      "step": 580000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013510203220987017,
-      "loss": 0.4801,
-      "step": 581000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013502667560699584,
-      "loss": 0.4814,
-      "step": 582000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013495107603517676,
-      "loss": 0.4813,
-      "step": 583000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013487530907285017,
-      "loss": 0.4814,
-      "step": 584000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001347994509683568,
-      "loss": 0.4812,
-      "step": 585000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.465150386095047,
-      "eval_runtime": 77.7438,
-      "eval_samples_per_second": 82.322,
-      "eval_steps_per_second": 0.643,
-      "step": 585000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001347233500847351,
-      "loss": 0.4818,
-      "step": 586000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013464708250429864,
-      "loss": 0.4804,
-      "step": 587000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013457064845937174,
-      "loss": 0.4798,
-      "step": 588000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013449404818278575,
-      "loss": 0.4807,
-      "step": 589000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001344174356059392,
-      "loss": 0.4811,
-      "step": 590000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.46463653445243835,
-      "eval_runtime": 77.8511,
-      "eval_samples_per_second": 82.208,
-      "eval_steps_per_second": 0.642,
-      "step": 590000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013434050389784906,
-      "loss": 0.4798,
-      "step": 591000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001342634066591601,
-      "loss": 0.4808,
-      "step": 592000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001341862214697452,
-      "loss": 0.4804,
-      "step": 593000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013410887154965859,
-      "loss": 0.48,
-      "step": 594000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013403127945970912,
-      "loss": 0.4801,
-      "step": 595000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4652685523033142,
-      "eval_runtime": 81.8053,
-      "eval_samples_per_second": 78.235,
-      "eval_steps_per_second": 0.611,
-      "step": 595000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001339535227811121,
-      "loss": 0.4801,
-      "step": 596000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013387560175072777,
-      "loss": 0.4811,
-      "step": 597000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001337975166059172,
-      "loss": 0.4802,
-      "step": 598000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00013371934591533964,
-      "loss": 0.4815,
-      "step": 599000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001336409334192773,
-      "loss": 0.4802,
-      "step": 600000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.46285489201545715,
-      "eval_runtime": 79.5065,
-      "eval_samples_per_second": 80.497,
-      "eval_steps_per_second": 0.629,
-      "step": 600000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013356235752362913,
-      "loss": 0.4814,
-      "step": 601000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013348369728822555,
-      "loss": 0.482,
-      "step": 602000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013340479547477053,
-      "loss": 0.4818,
-      "step": 603000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013332573098104674,
-      "loss": 0.4825,
-      "step": 604000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013324650404789842,
-      "loss": 0.4801,
-      "step": 605000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.4614011347293854,
-      "eval_runtime": 79.077,
-      "eval_samples_per_second": 80.934,
-      "eval_steps_per_second": 0.632,
-      "step": 605000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013316719438673325,
-      "loss": 0.482,
-      "step": 606000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013308764346108213,
-      "loss": 0.4808,
-      "step": 607000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013300801061459758,
-      "loss": 0.4817,
-      "step": 608000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001329281366647775,
-      "loss": 0.4809,
-      "step": 609000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001328481014866741,
-      "loss": 0.4822,
-      "step": 610000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.4625004827976227,
-      "eval_runtime": 82.853,
-      "eval_samples_per_second": 77.245,
-      "eval_steps_per_second": 0.603,
-      "step": 610000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013276798560058145,
-      "loss": 0.4807,
-      "step": 611000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013268770929537345,
-      "loss": 0.4793,
-      "step": 612000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001326071922179333,
-      "loss": 0.4794,
-      "step": 613000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013252651488986224,
-      "loss": 0.4804,
-      "step": 614000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013244567755691747,
-      "loss": 0.4814,
-      "step": 615000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.46313124895095825,
-      "eval_runtime": 83.0415,
-      "eval_samples_per_second": 77.07,
-      "eval_steps_per_second": 0.602,
-      "step": 615000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013236468046534346,
-      "loss": 0.4802,
-      "step": 616000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013228352386187144,
-      "loss": 0.4791,
-      "step": 617000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013220228938905693,
-      "loss": 0.4795,
-      "step": 618000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00013212081466281838,
-      "loss": 0.481,
-      "step": 619000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001320393445928162,
-      "loss": 0.4812,
-      "step": 620000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4640228748321533,
-      "eval_runtime": 83.7482,
-      "eval_samples_per_second": 76.42,
-      "eval_steps_per_second": 0.597,
-      "step": 620000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013195755289395868,
-      "loss": 0.4815,
-      "step": 621000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013187560292338433,
-      "loss": 0.4794,
-      "step": 622000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013179357711756845,
-      "loss": 0.4798,
-      "step": 623000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013171131151059012,
-      "loss": 0.4789,
-      "step": 624000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013162888838198894,
-      "loss": 0.4796,
-      "step": 625000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.46150654554367065,
-      "eval_runtime": 77.9007,
-      "eval_samples_per_second": 82.156,
-      "eval_steps_per_second": 0.642,
-      "step": 625000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013154630798284004,
-      "loss": 0.4796,
-      "step": 626000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001314636533804629,
-      "loss": 0.4797,
-      "step": 627000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013138084232425004,
-      "loss": 0.48,
-      "step": 628000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001312977919374723,
-      "loss": 0.4779,
-      "step": 629000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013121458528872397,
-      "loss": 0.481,
-      "step": 630000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4608282148838043,
-      "eval_runtime": 78.9952,
-      "eval_samples_per_second": 81.018,
-      "eval_steps_per_second": 0.633,
-      "step": 630000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013113122263146702,
-      "loss": 0.4799,
-      "step": 631000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001310477878157651,
-      "loss": 0.4802,
-      "step": 632000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013096411405914954,
-      "loss": 0.4795,
-      "step": 633000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013088028505700412,
-      "loss": 0.4789,
-      "step": 634000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.0001307963010646865,
-      "loss": 0.478,
-      "step": 635000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.458748996257782,
-      "eval_runtime": 79.5488,
-      "eval_samples_per_second": 80.454,
-      "eval_steps_per_second": 0.629,
-      "step": 635000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013071224655395776,
-      "loss": 0.4792,
-      "step": 636000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013062803787373327,
-      "loss": 0.4787,
-      "step": 637000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013054359075594853,
-      "loss": 0.4779,
-      "step": 638000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00013045907435152275,
-      "loss": 0.4776,
-      "step": 639000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00013037440454726602,
-      "loss": 0.4786,
-      "step": 640000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4617770314216614,
-      "eval_runtime": 77.6784,
-      "eval_samples_per_second": 82.391,
-      "eval_steps_per_second": 0.644,
-      "step": 640000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00013028949661608873,
-      "loss": 0.4772,
-      "step": 641000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00013020443549438331,
-      "loss": 0.4787,
-      "step": 642000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001301192214412607,
-      "loss": 0.4789,
-      "step": 643000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001300339401591957,
-      "loss": 0.4792,
-      "step": 644000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012994842117471585,
-      "loss": 0.4802,
-      "step": 645000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4589656591415405,
-      "eval_runtime": 78.0666,
-      "eval_samples_per_second": 81.981,
-      "eval_steps_per_second": 0.64,
-      "step": 645000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001298627500386828,
-      "loss": 0.4785,
-      "step": 646000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001297770129108747,
-      "loss": 0.4776,
-      "step": 647000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012969112445675987,
-      "loss": 0.4777,
-      "step": 648000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001296049987362049,
-      "loss": 0.4782,
-      "step": 649000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012951872191022104,
-      "loss": 0.4767,
-      "step": 650000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4610587954521179,
-      "eval_runtime": 80.9253,
-      "eval_samples_per_second": 79.085,
-      "eval_steps_per_second": 0.618,
-      "step": 650000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012943229424162249,
-      "loss": 0.4763,
-      "step": 651000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012934588930028175,
-      "loss": 0.4782,
-      "step": 652000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012925916103710117,
-      "loss": 0.4765,
-      "step": 653000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012917228272197337,
-      "loss": 0.4771,
-      "step": 654000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012908525461954466,
-      "loss": 0.4766,
-      "step": 655000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4596790373325348,
-      "eval_runtime": 78.9864,
-      "eval_samples_per_second": 81.027,
-      "eval_steps_per_second": 0.633,
-      "step": 655000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012899816424714043,
-      "loss": 0.4765,
-      "step": 656000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012891083751499744,
-      "loss": 0.4771,
-      "step": 657000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012882336179196102,
-      "loss": 0.4784,
-      "step": 658000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00012873582504314393,
-      "loss": 0.478,
-      "step": 659000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001286480522864964,
-      "loss": 0.477,
-      "step": 660000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.45967376232147217,
-      "eval_runtime": 76.8819,
-      "eval_samples_per_second": 83.245,
-      "eval_steps_per_second": 0.65,
-      "step": 660000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012856021933432393,
-      "loss": 0.4775,
-      "step": 661000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001284721506124795,
-      "loss": 0.4759,
-      "step": 662000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.000128383934236058,
-      "loss": 0.4762,
-      "step": 663000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001282956589110738,
-      "loss": 0.4768,
-      "step": 664000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012820723676322364,
-      "loss": 0.4772,
-      "step": 665000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.45860305428504944,
-      "eval_runtime": 78.1325,
-      "eval_samples_per_second": 81.912,
-      "eval_steps_per_second": 0.64,
-      "step": 665000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012811857933062512,
-      "loss": 0.4762,
-      "step": 666000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012802977532048903,
-      "loss": 0.4779,
-      "step": 667000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012794082500332778,
-      "loss": 0.4779,
-      "step": 668000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001278517286500994,
-      "loss": 0.4755,
-      "step": 669000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012776257584704384,
-      "loss": 0.4766,
-      "step": 670000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.45803070068359375,
-      "eval_runtime": 78.0714,
-      "eval_samples_per_second": 81.976,
-      "eval_steps_per_second": 0.64,
-      "step": 670000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012767327784173963,
-      "loss": 0.4757,
-      "step": 671000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.000127583745300672,
-      "loss": 0.4757,
-      "step": 672000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012749406781126398,
-      "loss": 0.4767,
-      "step": 673000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012740424564668877,
-      "loss": 0.4768,
-      "step": 674000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012731427908056026,
-      "loss": 0.4764,
-      "step": 675000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.4579757750034332,
-      "eval_runtime": 78.9147,
-      "eval_samples_per_second": 81.1,
-      "eval_steps_per_second": 0.634,
-      "step": 675000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012722416838693222,
-      "loss": 0.4778,
-      "step": 676000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012713400416660716,
-      "loss": 0.4759,
-      "step": 677000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00012704360618533725,
-      "loss": 0.4755,
-      "step": 678000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.0001269531555138567,
-      "loss": 0.4759,
-      "step": 679000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012686247134531458,
-      "loss": 0.4775,
-      "step": 680000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.45441490411758423,
-      "eval_runtime": 78.5989,
-      "eval_samples_per_second": 81.426,
-      "eval_steps_per_second": 0.636,
-      "step": 680000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012677173532368958,
-      "loss": 0.4763,
-      "step": 681000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012668076607172854,
-      "loss": 0.4762,
-      "step": 682000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012658965462205847,
-      "loss": 0.4764,
-      "step": 683000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012649840125222064,
-      "loss": 0.4761,
-      "step": 684000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012640700624018863,
-      "loss": 0.4751,
-      "step": 685000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4569858908653259,
-      "eval_runtime": 78.7381,
-      "eval_samples_per_second": 81.282,
-      "eval_steps_per_second": 0.635,
-      "step": 685000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012631546986436747,
-      "loss": 0.4751,
-      "step": 686000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001262238841514327,
-      "loss": 0.4765,
-      "step": 687000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012613206602563615,
-      "loss": 0.4756,
-      "step": 688000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012604029143092928,
-      "loss": 0.4767,
-      "step": 689000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.000125948284981498,
-      "loss": 0.4752,
-      "step": 690000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.456228107213974,
-      "eval_runtime": 80.1461,
-      "eval_samples_per_second": 79.854,
-      "eval_steps_per_second": 0.624,
-      "step": 690000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012585604653715085,
-      "loss": 0.4758,
-      "step": 691000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001257636684073334,
-      "loss": 0.4757,
-      "step": 692000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012567124346051786,
-      "loss": 0.4738,
-      "step": 693000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012557858694336514,
-      "loss": 0.4739,
-      "step": 694000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012548579158593213,
-      "loss": 0.4736,
-      "step": 695000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.45279574394226074,
-      "eval_runtime": 78.8366,
-      "eval_samples_per_second": 81.181,
-      "eval_steps_per_second": 0.634,
-      "step": 695000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012539295067392003,
-      "loss": 0.4758,
-      "step": 696000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012529987862249377,
-      "loss": 0.475,
-      "step": 697000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.0001252066685797811,
-      "loss": 0.4745,
-      "step": 698000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012511332082971603,
-      "loss": 0.4738,
-      "step": 699000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00012501983565665193,
-      "loss": 0.4737,
-      "step": 700000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4576612412929535,
-      "eval_runtime": 79.2297,
-      "eval_samples_per_second": 80.778,
-      "eval_steps_per_second": 0.631,
-      "step": 700000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012492621334536086,
-      "loss": 0.475,
-      "step": 701000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012483254800846005,
-      "loss": 0.4748,
-      "step": 702000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.000124738652413126,
-      "loss": 0.476,
-      "step": 703000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.0001246447146359523,
-      "loss": 0.4738,
-      "step": 704000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012455054689951964,
-      "loss": 0.4741,
-      "step": 705000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4546070992946625,
-      "eval_runtime": 79.5809,
-      "eval_samples_per_second": 80.421,
-      "eval_steps_per_second": 0.628,
-      "step": 705000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012445624345439494,
-      "loss": 0.4732,
-      "step": 706000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012436180458784285,
-      "loss": 0.4735,
-      "step": 707000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012426741987021326,
-      "loss": 0.4746,
-      "step": 708000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.0001241727112936531,
-      "loss": 0.4737,
-      "step": 709000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012407786815935373,
-      "loss": 0.4727,
-      "step": 710000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4537406861782074,
-      "eval_runtime": 78.7373,
-      "eval_samples_per_second": 81.283,
-      "eval_steps_per_second": 0.635,
-      "step": 710000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012398289075622377,
-      "loss": 0.4743,
-      "step": 711000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012388787455178986,
-      "loss": 0.4736,
-      "step": 712000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012379262961290468,
-      "loss": 0.4737,
-      "step": 713000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012369725127407496,
-      "loss": 0.4736,
-      "step": 714000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012360173982583968,
-      "loss": 0.4736,
-      "step": 715000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4546468257904053,
-      "eval_runtime": 78.1145,
-      "eval_samples_per_second": 81.931,
-      "eval_steps_per_second": 0.64,
-      "step": 715000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012350619126965588,
-      "loss": 0.4734,
-      "step": 716000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012341051045099063,
-      "loss": 0.4748,
-      "step": 717000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00012331460168600204,
-      "loss": 0.4745,
-      "step": 718000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.0001232185609772249,
-      "loss": 0.4729,
-      "step": 719000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012312238861721586,
-      "loss": 0.4736,
-      "step": 720000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4559849202632904,
-      "eval_runtime": 80.1667,
-      "eval_samples_per_second": 79.834,
-      "eval_steps_per_second": 0.624,
-      "step": 720000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001230260848989326,
-      "loss": 0.4745,
-      "step": 721000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012292984311590698,
-      "loss": 0.4735,
-      "step": 722000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012283327782279692,
-      "loss": 0.4731,
-      "step": 723000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012273658205209426,
-      "loss": 0.4728,
-      "step": 724000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012263985298923314,
-      "loss": 0.4735,
-      "step": 725000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4538413882255554,
-      "eval_runtime": 78.3414,
-      "eval_samples_per_second": 81.694,
-      "eval_steps_per_second": 0.638,
-      "step": 725000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012254289727713877,
-      "loss": 0.4721,
-      "step": 726000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001224458119720013,
-      "loss": 0.4738,
-      "step": 727000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001223486946486473,
-      "loss": 0.4716,
-      "step": 728000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001222513511738867,
-      "loss": 0.4722,
-      "step": 729000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012215387899418292,
-      "loss": 0.4714,
-      "step": 730000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.45341652631759644,
-      "eval_runtime": 81.7464,
-      "eval_samples_per_second": 78.291,
-      "eval_steps_per_second": 0.612,
-      "step": 730000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012205637607108188,
-      "loss": 0.4721,
-      "step": 731000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012195864750059665,
-      "loss": 0.4724,
-      "step": 732000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012186088903692187,
-      "loss": 0.4726,
-      "step": 733000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012176290526525648,
-      "loss": 0.4721,
-      "step": 734000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012166479427653891,
-      "loss": 0.4736,
-      "step": 735000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4538614749908447,
-      "eval_runtime": 76.948,
-      "eval_samples_per_second": 83.173,
-      "eval_steps_per_second": 0.65,
-      "step": 735000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001215665563696323,
-      "loss": 0.4717,
-      "step": 736000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012146819184378635,
-      "loss": 0.4721,
-      "step": 737000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.00012136989810619402,
-      "loss": 0.472,
-      "step": 738000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001212712814934992,
-      "loss": 0.4725,
-      "step": 739000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001211725391613235,
-      "loss": 0.4721,
-      "step": 740000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.45002052187919617,
-      "eval_runtime": 81.7464,
-      "eval_samples_per_second": 78.291,
-      "eval_steps_per_second": 0.612,
-      "step": 740000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012107367141045319,
-      "loss": 0.4713,
-      "step": 741000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012097477759732085,
-      "loss": 0.4721,
-      "step": 742000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012087566003761258,
-      "loss": 0.4723,
-      "step": 743000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012077641796355505,
-      "loss": 0.4713,
-      "step": 744000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012067715110568603,
-      "loss": 0.4708,
-      "step": 745000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.45279091596603394,
-      "eval_runtime": 81.252,
-      "eval_samples_per_second": 78.767,
-      "eval_steps_per_second": 0.615,
-      "step": 745000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012057766103399194,
-      "loss": 0.4705,
-      "step": 746000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.0001204780473557052,
-      "loss": 0.4721,
-      "step": 747000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012037831037426628,
-      "loss": 0.4716,
-      "step": 748000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012027865023600415,
-      "loss": 0.4714,
-      "step": 749000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00012017866780517054,
-      "loss": 0.4716,
-      "step": 750000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.45302364230155945,
-      "eval_runtime": 80.5157,
-      "eval_samples_per_second": 79.488,
-      "eval_steps_per_second": 0.621,
-      "step": 750000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00012007856298314685,
-      "loss": 0.4718,
-      "step": 751000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011997833607486966,
-      "loss": 0.4709,
-      "step": 752000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011987808779506468,
-      "loss": 0.472,
-      "step": 753000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011977771828251805,
-      "loss": 0.472,
-      "step": 754000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011967712719084747,
-      "loss": 0.4715,
-      "step": 755000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.4534043073654175,
-      "eval_runtime": 81.188,
-      "eval_samples_per_second": 78.829,
-      "eval_steps_per_second": 0.616,
-      "step": 755000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011957641523576647,
-      "loss": 0.4711,
-      "step": 756000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011947558272406116,
-      "loss": 0.4698,
-      "step": 757000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.0001193746299628848,
-      "loss": 0.4694,
-      "step": 758000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00011927365839226875,
-      "loss": 0.4721,
-      "step": 759000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011917246617455475,
-      "loss": 0.4685,
-      "step": 760000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4503335952758789,
-      "eval_runtime": 78.7744,
-      "eval_samples_per_second": 81.245,
-      "eval_steps_per_second": 0.635,
-      "step": 760000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011907125600175987,
-      "loss": 0.4697,
-      "step": 761000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011896992704905985,
-      "loss": 0.4715,
-      "step": 762000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011886837801627817,
-      "loss": 0.4708,
-      "step": 763000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001187667105836763,
-      "loss": 0.472,
-      "step": 764000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011866502690535649,
-      "loss": 0.4707,
-      "step": 765000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4522061049938202,
-      "eval_runtime": 81.4742,
-      "eval_samples_per_second": 78.552,
-      "eval_steps_per_second": 0.614,
-      "step": 765000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.000118563123720189,
-      "loss": 0.47,
-      "step": 766000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001184612051442868,
-      "loss": 0.4708,
-      "step": 767000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011835906744697045,
-      "loss": 0.4705,
-      "step": 768000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011825681290128225,
-      "loss": 0.4706,
-      "step": 769000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011815444181870728,
-      "loss": 0.4695,
-      "step": 770000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4502740204334259,
-      "eval_runtime": 80.2036,
-      "eval_samples_per_second": 79.797,
-      "eval_steps_per_second": 0.623,
-      "step": 770000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.000118052159601486,
-      "loss": 0.4714,
-      "step": 771000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011794955661252567,
-      "loss": 0.4696,
-      "step": 772000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011784683802263448,
-      "loss": 0.4709,
-      "step": 773000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001177440041447111,
-      "loss": 0.4702,
-      "step": 774000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011764105529200537,
-      "loss": 0.4718,
-      "step": 775000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4481835961341858,
-      "eval_runtime": 81.6948,
-      "eval_samples_per_second": 78.34,
-      "eval_steps_per_second": 0.612,
-      "step": 775000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011753809489880001,
-      "loss": 0.4715,
-      "step": 776000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011743502038662826,
-      "loss": 0.4699,
-      "step": 777000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011733172872032392,
-      "loss": 0.4693,
-      "step": 778000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00011722842679718266,
-      "loss": 0.4696,
-      "step": 779000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001171249081212757,
-      "loss": 0.4686,
-      "step": 780000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.45043304562568665,
-      "eval_runtime": 83.0198,
-      "eval_samples_per_second": 77.09,
-      "eval_steps_per_second": 0.602,
-      "step": 780000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011702127635646021,
-      "loss": 0.4694,
-      "step": 781000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011691753181841654,
-      "loss": 0.4696,
-      "step": 782000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011681367482316854,
-      "loss": 0.4702,
-      "step": 783000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011670970568708263,
-      "loss": 0.4699,
-      "step": 784000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011660572886357773,
-      "loss": 0.471,
-      "step": 785000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.4487108588218689,
-      "eval_runtime": 81.1941,
-      "eval_samples_per_second": 78.823,
-      "eval_steps_per_second": 0.616,
-      "step": 785000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011650153650762927,
-      "loss": 0.4691,
-      "step": 786000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011639723296167041,
-      "loss": 0.469,
-      "step": 787000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011629292301312084,
-      "loss": 0.4703,
-      "step": 788000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011618839815105406,
-      "loss": 0.4698,
-      "step": 789000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011608376305285055,
-      "loss": 0.469,
-      "step": 790000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.45146042108535767,
-      "eval_runtime": 80.5783,
-      "eval_samples_per_second": 79.426,
-      "eval_steps_per_second": 0.621,
-      "step": 790000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011597901803724696,
-      "loss": 0.4701,
-      "step": 791000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011587426833256681,
-      "loss": 0.4702,
-      "step": 792000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011576930454883066,
-      "loss": 0.4692,
-      "step": 793000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011566423180558957,
-      "loss": 0.4688,
-      "step": 794000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011555905042291336,
-      "loss": 0.4701,
-      "step": 795000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.44838276505470276,
-      "eval_runtime": 79.5446,
-      "eval_samples_per_second": 80.458,
-      "eval_steps_per_second": 0.629,
-      "step": 795000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011545386606490308,
-      "loss": 0.4692,
-      "step": 796000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011534857392415738,
-      "loss": 0.4686,
-      "step": 797000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011524306876193287,
-      "loss": 0.4683,
-      "step": 798000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00011513756190924513,
-      "loss": 0.4699,
-      "step": 799000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011503184246260971,
-      "loss": 0.4694,
-      "step": 800000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4521300494670868,
-      "eval_runtime": 81.6914,
-      "eval_samples_per_second": 78.344,
-      "eval_steps_per_second": 0.612,
-      "step": 800000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011492601630290677,
-      "loss": 0.4693,
-      "step": 801000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011482008375250114,
-      "loss": 0.4687,
-      "step": 802000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011471415122557365,
-      "loss": 0.4671,
-      "step": 803000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011460800696773604,
-      "loss": 0.4689,
-      "step": 804000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011450186359013775,
-      "loss": 0.467,
-      "step": 805000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4480891525745392,
-      "eval_runtime": 82.1197,
-      "eval_samples_per_second": 77.935,
-      "eval_steps_per_second": 0.609,
-      "step": 805000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011439550891690946,
-      "loss": 0.4671,
-      "step": 806000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011428904946899599,
-      "loss": 0.4684,
-      "step": 807000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011418248557069125,
-      "loss": 0.4687,
-      "step": 808000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011407592426653418,
-      "loss": 0.4681,
-      "step": 809000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011396915254523895,
-      "loss": 0.4671,
-      "step": 810000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4491584897041321,
-      "eval_runtime": 80.5568,
-      "eval_samples_per_second": 79.447,
-      "eval_steps_per_second": 0.621,
-      "step": 810000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011386227734801394,
-      "loss": 0.469,
-      "step": 811000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011375529900041961,
-      "loss": 0.4685,
-      "step": 812000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011364843209307678,
-      "loss": 0.4686,
-      "step": 813000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011354124862735163,
-      "loss": 0.466,
-      "step": 814000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011343407032573083,
-      "loss": 0.4682,
-      "step": 815000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4499417841434479,
-      "eval_runtime": 82.5085,
-      "eval_samples_per_second": 77.568,
-      "eval_steps_per_second": 0.606,
-      "step": 815000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011332668294357827,
-      "loss": 0.4678,
-      "step": 816000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011321919404256975,
-      "loss": 0.4682,
-      "step": 817000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011311160395013516,
-      "loss": 0.467,
-      "step": 818000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00011300391299401257,
-      "loss": 0.4667,
-      "step": 819000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011289633718534724,
-      "loss": 0.467,
-      "step": 820000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4481469690799713,
-      "eval_runtime": 82.9256,
-      "eval_samples_per_second": 77.178,
-      "eval_steps_per_second": 0.603,
-      "step": 820000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011278844568637758,
-      "loss": 0.4666,
-      "step": 821000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011268045430811603,
-      "loss": 0.4677,
-      "step": 822000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011257236337952306,
-      "loss": 0.4687,
-      "step": 823000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011246428146946332,
-      "loss": 0.4687,
-      "step": 824000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011235610086525684,
-      "loss": 0.4675,
-      "step": 825000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.449359655380249,
-      "eval_runtime": 81.428,
-      "eval_samples_per_second": 78.597,
-      "eval_steps_per_second": 0.614,
-      "step": 825000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011224771345925375,
-      "loss": 0.467,
-      "step": 826000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001121392278211229,
-      "loss": 0.4688,
-      "step": 827000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011203064428133039,
-      "loss": 0.4676,
-      "step": 828000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001119220719003777,
-      "loss": 0.4664,
-      "step": 829000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001118132936469263,
-      "loss": 0.4676,
-      "step": 830000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.44925883412361145,
-      "eval_runtime": 83.2246,
-      "eval_samples_per_second": 76.9,
-      "eval_steps_per_second": 0.601,
-      "step": 830000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001117045274081231,
-      "loss": 0.4689,
-      "step": 831000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011159566478488941,
-      "loss": 0.4686,
-      "step": 832000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011148659699242007,
-      "loss": 0.4672,
-      "step": 833000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011137743328631572,
-      "loss": 0.4678,
-      "step": 834000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011126817399910799,
-      "loss": 0.4658,
-      "step": 835000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.450771689414978,
-      "eval_runtime": 83.2311,
-      "eval_samples_per_second": 76.894,
-      "eval_steps_per_second": 0.601,
-      "step": 835000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011115892886562073,
-      "loss": 0.4664,
-      "step": 836000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001110494795097132,
-      "loss": 0.4651,
-      "step": 837000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011094004516277705,
-      "loss": 0.4672,
-      "step": 838000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011083040707044233,
-      "loss": 0.4674,
-      "step": 839000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00011072078484214722,
-      "loss": 0.4676,
-      "step": 840000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4472251832485199,
-      "eval_runtime": 80.5148,
-      "eval_samples_per_second": 79.488,
-      "eval_steps_per_second": 0.621,
-      "step": 840000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00011061095934795527,
-      "loss": 0.4667,
-      "step": 841000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00011050115057267808,
-      "loss": 0.4662,
-      "step": 842000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00011039113901348006,
-      "loss": 0.4667,
-      "step": 843000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00011028103487760205,
-      "loss": 0.4679,
-      "step": 844000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001101709487427803,
-      "loss": 0.4665,
-      "step": 845000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.44985419511795044,
-      "eval_runtime": 82.384,
-      "eval_samples_per_second": 77.685,
-      "eval_steps_per_second": 0.607,
-      "step": 845000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00011006066055174999,
-      "loss": 0.4655,
-      "step": 846000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001099503912161227,
-      "loss": 0.4666,
-      "step": 847000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001098399203124234,
-      "loss": 0.4657,
-      "step": 848000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010972946911824438,
-      "loss": 0.4672,
-      "step": 849000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010961881684658344,
-      "loss": 0.4675,
-      "step": 850000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4436103105545044,
-      "eval_runtime": 81.2881,
-      "eval_samples_per_second": 78.732,
-      "eval_steps_per_second": 0.615,
-      "step": 850000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010950818513832014,
-      "loss": 0.4659,
-      "step": 851000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001093973528456127,
-      "loss": 0.4661,
-      "step": 852000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010928643100347024,
-      "loss": 0.4664,
-      "step": 853000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010917541994977997,
-      "loss": 0.4647,
-      "step": 854000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010906443116690753,
-      "loss": 0.4666,
-      "step": 855000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.44435253739356995,
-      "eval_runtime": 82.1977,
-      "eval_samples_per_second": 77.861,
-      "eval_steps_per_second": 0.608,
-      "step": 855000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010895324279323492,
-      "loss": 0.467,
-      "step": 856000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010884196622296353,
-      "loss": 0.4652,
-      "step": 857000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00010873071320326107,
-      "loss": 0.4658,
-      "step": 858000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.0001086192613443119,
-      "loss": 0.4657,
-      "step": 859000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010850772230612937,
-      "loss": 0.4656,
-      "step": 860000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.44788363575935364,
-      "eval_runtime": 81.6254,
-      "eval_samples_per_second": 78.407,
-      "eval_steps_per_second": 0.613,
-      "step": 860000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010839609642848077,
-      "loss": 0.4663,
-      "step": 861000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010828449580686817,
-      "loss": 0.4649,
-      "step": 862000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010817269735663531,
-      "loss": 0.4656,
-      "step": 863000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010806081308748046,
-      "loss": 0.4656,
-      "step": 864000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001079490673647975,
-      "loss": 0.4664,
-      "step": 865000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4458025395870209,
-      "eval_runtime": 83.4659,
-      "eval_samples_per_second": 76.678,
-      "eval_steps_per_second": 0.599,
-      "step": 865000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001078370126504492,
-      "loss": 0.4645,
-      "step": 866000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010772487313973276,
-      "loss": 0.4664,
-      "step": 867000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010761276144028143,
-      "loss": 0.4661,
-      "step": 868000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010750056579571238,
-      "loss": 0.4647,
-      "step": 869000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010738817411369786,
-      "loss": 0.4648,
-      "step": 870000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.44627639651298523,
-      "eval_runtime": 78.7772,
-      "eval_samples_per_second": 81.242,
-      "eval_steps_per_second": 0.635,
-      "step": 870000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010727569900255524,
-      "loss": 0.464,
-      "step": 871000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010716314080490315,
-      "loss": 0.4656,
-      "step": 872000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010705061254577097,
-      "loss": 0.4645,
-      "step": 873000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010693788928619646,
-      "loss": 0.467,
-      "step": 874000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010682508396913951,
-      "loss": 0.4644,
-      "step": 875000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4447041451931,
-      "eval_runtime": 84.2151,
-      "eval_samples_per_second": 75.996,
-      "eval_steps_per_second": 0.594,
-      "step": 875000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010671219693822467,
-      "loss": 0.4653,
-      "step": 876000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010659934154625607,
-      "loss": 0.4652,
-      "step": 877000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001064862922003476,
-      "loss": 0.4654,
-      "step": 878000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00010637327534281335,
-      "loss": 0.4641,
-      "step": 879000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010626006505800474,
-      "loss": 0.4642,
-      "step": 880000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.44686806201934814,
-      "eval_runtime": 79.0811,
-      "eval_samples_per_second": 80.93,
-      "eval_steps_per_second": 0.632,
-      "step": 880000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010614688811060288,
-      "loss": 0.4646,
-      "step": 881000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001060336316744615,
-      "loss": 0.4652,
-      "step": 882000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010592018260532381,
-      "loss": 0.4655,
-      "step": 883000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010580665457881469,
-      "loss": 0.4649,
-      "step": 884000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001056930479407602,
-      "loss": 0.4663,
-      "step": 885000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.44526371359825134,
-      "eval_runtime": 78.5736,
-      "eval_samples_per_second": 81.452,
-      "eval_steps_per_second": 0.636,
-      "step": 885000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001055793630372259,
-      "loss": 0.4655,
-      "step": 886000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010546560021451571,
-      "loss": 0.4629,
-      "step": 887000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010535187369819837,
-      "loss": 0.4643,
-      "step": 888000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010523795615404908,
-      "loss": 0.4637,
-      "step": 889000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010512396173070846,
-      "loss": 0.4653,
-      "step": 890000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.44463416934013367,
-      "eval_runtime": 78.8452,
-      "eval_samples_per_second": 81.172,
-      "eval_steps_per_second": 0.634,
-      "step": 890000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010500989077542328,
-      "loss": 0.4651,
-      "step": 891000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010489597200575402,
-      "loss": 0.4635,
-      "step": 892000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010478174918057791,
-      "loss": 0.4635,
-      "step": 893000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010466745086589606,
-      "loss": 0.4639,
-      "step": 894000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010455319182075405,
-      "loss": 0.4637,
-      "step": 895000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4454698860645294,
-      "eval_runtime": 81.883,
-      "eval_samples_per_second": 78.16,
-      "eval_steps_per_second": 0.611,
-      "step": 895000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010443874364642582,
-      "loss": 0.4642,
-      "step": 896000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010432433558713417,
-      "loss": 0.4642,
-      "step": 897000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010420973894628034,
-      "loss": 0.4635,
-      "step": 898000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010409518326547141,
-      "loss": 0.4638,
-      "step": 899000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00010398043955301988,
-      "loss": 0.4637,
-      "step": 900000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.44443508982658386,
-      "eval_runtime": 80.8881,
-      "eval_samples_per_second": 79.122,
-      "eval_steps_per_second": 0.618,
-      "step": 900000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010386562279199064,
-      "loss": 0.4619,
-      "step": 901000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010375084825779196,
-      "loss": 0.4642,
-      "step": 902000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001036358865212581,
-      "loss": 0.4644,
-      "step": 903000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010352096785529729,
-      "loss": 0.4642,
-      "step": 904000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010340586254263087,
-      "loss": 0.4625,
-      "step": 905000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.44563567638397217,
-      "eval_runtime": 78.8321,
-      "eval_samples_per_second": 81.185,
-      "eval_steps_per_second": 0.634,
-      "step": 905000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010329080114375571,
-      "loss": 0.4652,
-      "step": 906000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010317555365607644,
-      "loss": 0.4641,
-      "step": 907000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010306023557164153,
-      "loss": 0.4634,
-      "step": 908000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010294484724172987,
-      "loss": 0.4623,
-      "step": 909000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010282938901783426,
-      "loss": 0.4621,
-      "step": 910000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.4458223581314087,
-      "eval_runtime": 80.4596,
-      "eval_samples_per_second": 79.543,
-      "eval_steps_per_second": 0.621,
-      "step": 910000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.000102713976814046,
-      "loss": 0.4627,
-      "step": 911000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010259849555785737,
-      "loss": 0.4638,
-      "step": 912000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010248282990041591,
-      "loss": 0.4626,
-      "step": 913000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.0001023670957563755,
-      "loss": 0.4615,
-      "step": 914000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010225129347828241,
-      "loss": 0.4636,
-      "step": 915000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.44487082958221436,
-      "eval_runtime": 81.5786,
-      "eval_samples_per_second": 78.452,
-      "eval_steps_per_second": 0.613,
-      "step": 915000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010213542341889048,
-      "loss": 0.4642,
-      "step": 916000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010201960190221062,
-      "loss": 0.4627,
-      "step": 917000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010190359740620597,
-      "loss": 0.4622,
-      "step": 918000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00010178752618804497,
-      "loss": 0.4633,
-      "step": 919000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010167150477192083,
-      "loss": 0.4605,
-      "step": 920000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.44223639369010925,
-      "eval_runtime": 78.8789,
-      "eval_samples_per_second": 81.137,
-      "eval_steps_per_second": 0.634,
-      "step": 920000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010155530123620656,
-      "loss": 0.4632,
-      "step": 921000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010143914834118332,
-      "loss": 0.4628,
-      "step": 922000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010132281390240366,
-      "loss": 0.4628,
-      "step": 923000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010120653094235274,
-      "loss": 0.4634,
-      "step": 924000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010109006701658896,
-      "loss": 0.4619,
-      "step": 925000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.44582226872444153,
-      "eval_runtime": 79.59,
-      "eval_samples_per_second": 80.412,
-      "eval_steps_per_second": 0.628,
-      "step": 925000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010097353884682351,
-      "loss": 0.4634,
-      "step": 926000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010085706341187459,
-      "loss": 0.4616,
-      "step": 927000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010074040788255187,
-      "loss": 0.4636,
-      "step": 928000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010062368917434986,
-      "loss": 0.4636,
-      "step": 929000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010050702445560737,
-      "loss": 0.4621,
-      "step": 930000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4443116784095764,
-      "eval_runtime": 81.0459,
-      "eval_samples_per_second": 78.968,
-      "eval_steps_per_second": 0.617,
-      "step": 930000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010039018051876424,
-      "loss": 0.4629,
-      "step": 931000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010027327446989505,
-      "loss": 0.4637,
-      "step": 932000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010015642366364926,
-      "loss": 0.4613,
-      "step": 933000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00010003939452048603,
-      "loss": 0.4623,
-      "step": 934000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 9.992242145441094e-05,
-      "loss": 0.4611,
-      "step": 935000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.44253501296043396,
-      "eval_runtime": 80.3792,
-      "eval_samples_per_second": 79.623,
-      "eval_steps_per_second": 0.622,
-      "step": 935000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 9.980527064149124e-05,
-      "loss": 0.4616,
-      "step": 936000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 9.968817673944183e-05,
-      "loss": 0.4624,
-      "step": 937000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 9.957090568278425e-05,
-      "loss": 0.4624,
-      "step": 938000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 9.94535750097539e-05,
-      "loss": 0.4625,
-      "step": 939000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.933618507776033e-05,
-      "loss": 0.4631,
-      "step": 940000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4430947005748749,
-      "eval_runtime": 79.9849,
-      "eval_samples_per_second": 80.015,
-      "eval_steps_per_second": 0.625,
-      "step": 940000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.921885372252918e-05,
-      "loss": 0.4624,
-      "step": 941000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.910134640392372e-05,
-      "loss": 0.4617,
-      "step": 942000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.898401608814534e-05,
-      "loss": 0.4625,
-      "step": 943000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.886639287093549e-05,
-      "loss": 0.4625,
-      "step": 944000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.874871218342105e-05,
-      "loss": 0.4623,
-      "step": 945000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4420580267906189,
-      "eval_runtime": 78.2862,
-      "eval_samples_per_second": 81.751,
-      "eval_steps_per_second": 0.639,
-      "step": 945000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.863097438407781e-05,
-      "loss": 0.4612,
-      "step": 946000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.851317983155551e-05,
-      "loss": 0.4621,
-      "step": 947000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.839544676367323e-05,
-      "loss": 0.4615,
-      "step": 948000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.827753983728851e-05,
-      "loss": 0.4621,
-      "step": 949000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.815957723434772e-05,
-      "loss": 0.4625,
-      "step": 950000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.44188880920410156,
-      "eval_runtime": 79.7091,
-      "eval_samples_per_second": 80.292,
-      "eval_steps_per_second": 0.627,
-      "step": 950000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.804167735961679e-05,
-      "loss": 0.4612,
-      "step": 951000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.7923604536514e-05,
-      "loss": 0.4611,
-      "step": 952000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.780559526957703e-05,
-      "loss": 0.4621,
-      "step": 953000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.768741366355399e-05,
-      "loss": 0.4617,
-      "step": 954000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.756941470310317e-05,
-      "loss": 0.4623,
-      "step": 955000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4422100782394409,
-      "eval_runtime": 78.3534,
-      "eval_samples_per_second": 81.681,
-      "eval_steps_per_second": 0.638,
-      "step": 955000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.7451125806065e-05,
-      "loss": 0.4611,
-      "step": 956000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.733278374987115e-05,
-      "loss": 0.4614,
-      "step": 957000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.721438889501207e-05,
-      "loss": 0.4627,
-      "step": 958000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.709594160213901e-05,
-      "loss": 0.4614,
-      "step": 959000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 9.697756075732548e-05,
-      "loss": 0.4612,
-      "step": 960000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4385935962200165,
-      "eval_runtime": 79.7647,
-      "eval_samples_per_second": 80.236,
-      "eval_steps_per_second": 0.627,
-      "step": 960000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.685912829929902e-05,
-      "loss": 0.4614,
-      "step": 961000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.674052596023316e-05,
-      "loss": 0.4619,
-      "step": 962000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.662187262662203e-05,
-      "loss": 0.4627,
-      "step": 963000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.650316865990422e-05,
-      "loss": 0.4612,
-      "step": 964000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.638453320090109e-05,
-      "loss": 0.4601,
-      "step": 965000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.44188645482063293,
-      "eval_runtime": 78.5733,
-      "eval_samples_per_second": 81.453,
-      "eval_steps_per_second": 0.636,
-      "step": 965000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.626572910263071e-05,
-      "loss": 0.4598,
-      "step": 966000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.6146994334404e-05,
-      "loss": 0.4605,
-      "step": 967000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.60282104779561e-05,
-      "loss": 0.4619,
-      "step": 968000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.590925891857327e-05,
-      "loss": 0.4605,
-      "step": 969000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.579025889682872e-05,
-      "loss": 0.4614,
-      "step": 970000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.4392482042312622,
-      "eval_runtime": 78.2208,
-      "eval_samples_per_second": 81.82,
-      "eval_steps_per_second": 0.639,
-      "step": 970000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.56712107752171e-05,
-      "loss": 0.459,
-      "step": 971000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.555223403596256e-05,
-      "loss": 0.4609,
-      "step": 972000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.543309084987918e-05,
-      "loss": 0.4614,
-      "step": 973000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.531401986548419e-05,
-      "loss": 0.4608,
-      "step": 974000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.51947830651998e-05,
-      "loss": 0.46,
-      "step": 975000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.4403558373451233,
-      "eval_runtime": 78.7273,
-      "eval_samples_per_second": 81.293,
-      "eval_steps_per_second": 0.635,
-      "step": 975000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.507561928505619e-05,
-      "loss": 0.4601,
-      "step": 976000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.495629032198109e-05,
-      "loss": 0.4609,
-      "step": 977000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.48369157994519e-05,
-      "loss": 0.4606,
-      "step": 978000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 9.471749608110409e-05,
-      "loss": 0.4596,
-      "step": 979000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.459803153071084e-05,
-      "loss": 0.4597,
-      "step": 980000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.44013381004333496,
-      "eval_runtime": 77.6815,
-      "eval_samples_per_second": 82.388,
-      "eval_steps_per_second": 0.644,
-      "step": 980000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.447852251218189e-05,
-      "loss": 0.46,
-      "step": 981000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.435908896459387e-05,
-      "loss": 0.4592,
-      "step": 982000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.42394921456214e-05,
-      "loss": 0.4604,
-      "step": 983000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.411997161242634e-05,
-      "loss": 0.4601,
-      "step": 984000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.400028844880299e-05,
-      "loss": 0.4593,
-      "step": 985000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4422694146633148,
-      "eval_runtime": 78.8261,
-      "eval_samples_per_second": 81.191,
-      "eval_steps_per_second": 0.634,
-      "step": 985000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.388068238486491e-05,
-      "loss": 0.4605,
-      "step": 986000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.376091433343432e-05,
-      "loss": 0.4603,
-      "step": 987000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.364110436387217e-05,
-      "loss": 0.4614,
-      "step": 988000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.352137271329748e-05,
-      "loss": 0.46,
-      "step": 989000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.340148004349036e-05,
-      "loss": 0.4605,
-      "step": 990000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.44141891598701477,
-      "eval_runtime": 80.7223,
-      "eval_samples_per_second": 79.284,
-      "eval_steps_per_second": 0.619,
-      "step": 990000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.328166650421353e-05,
-      "loss": 0.4585,
-      "step": 991000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.316169259355529e-05,
-      "loss": 0.4598,
-      "step": 992000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.30416785900988e-05,
-      "loss": 0.4601,
-      "step": 993000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.292174493288034e-05,
-      "loss": 0.4596,
-      "step": 994000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.28016518798777e-05,
-      "loss": 0.4601,
-      "step": 995000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4383966028690338,
-      "eval_runtime": 82.6078,
-      "eval_samples_per_second": 77.475,
-      "eval_steps_per_second": 0.605,
-      "step": 995000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.26816399822291e-05,
-      "loss": 0.4592,
-      "step": 996000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.256146934151332e-05,
-      "loss": 0.4586,
-      "step": 997000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.244138066427978e-05,
-      "loss": 0.4607,
-      "step": 998000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.232113389862827e-05,
-      "loss": 0.4598,
-      "step": 999000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 9.220084960066614e-05,
-      "loss": 0.459,
-      "step": 1000000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.43742096424102783,
-      "eval_runtime": 80.287,
-      "eval_samples_per_second": 79.714,
-      "eval_steps_per_second": 0.623,
-      "step": 1000000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.208052813680018e-05,
-      "loss": 0.4597,
-      "step": 1001000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.196029025007298e-05,
-      "loss": 0.4589,
-      "step": 1002000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.183989559032122e-05,
-      "loss": 0.4593,
-      "step": 1003000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.171946486419417e-05,
-      "loss": 0.4588,
-      "step": 1004000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.159911892268011e-05,
-      "loss": 0.4592,
-      "step": 1005000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.44004783034324646,
-      "eval_runtime": 78.0539,
-      "eval_samples_per_second": 81.995,
-      "eval_steps_per_second": 0.641,
-      "step": 1005000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.14786171996191e-05,
-      "loss": 0.459,
-      "step": 1006000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.135808051069939e-05,
-      "loss": 0.4581,
-      "step": 1007000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.123762981154407e-05,
-      "loss": 0.4588,
-      "step": 1008000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.11170243265873e-05,
-      "loss": 0.4606,
-      "step": 1009000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.099650563338879e-05,
-      "loss": 0.4595,
-      "step": 1010000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4407427906990051,
-      "eval_runtime": 80.6794,
-      "eval_samples_per_second": 79.326,
-      "eval_steps_per_second": 0.62,
-      "step": 1010000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.087595350988804e-05,
-      "loss": 0.46,
-      "step": 1011000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.075524760023795e-05,
-      "loss": 0.4585,
-      "step": 1012000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.063450892823849e-05,
-      "loss": 0.4594,
-      "step": 1013000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.051373786168048e-05,
-      "loss": 0.4591,
-      "step": 1014000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.039293476845361e-05,
-      "loss": 0.4586,
-      "step": 1015000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.438150554895401,
-      "eval_runtime": 80.3486,
-      "eval_samples_per_second": 79.653,
-      "eval_steps_per_second": 0.622,
-      "step": 1015000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.0272220866988e-05,
-      "loss": 0.4582,
-      "step": 1016000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.01513548555881e-05,
-      "loss": 0.4578,
-      "step": 1017000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.00304579214008e-05,
-      "loss": 0.4578,
-      "step": 1018000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 8.990965137532721e-05,
-      "loss": 0.4595,
-      "step": 1019000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.978869373047941e-05,
-      "loss": 0.4578,
-      "step": 1020000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4372638761997223,
-      "eval_runtime": 79.8699,
-      "eval_samples_per_second": 80.13,
-      "eval_steps_per_second": 0.626,
-      "step": 1020000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.966782726980726e-05,
-      "loss": 0.4595,
-      "step": 1021000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.954681038665626e-05,
-      "loss": 0.4581,
-      "step": 1022000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.942588548263289e-05,
-      "loss": 0.4566,
-      "step": 1023000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.930493192312473e-05,
-      "loss": 0.4596,
-      "step": 1024000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.918382895884115e-05,
-      "loss": 0.458,
-      "step": 1025000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4378396272659302,
-      "eval_runtime": 80.7357,
-      "eval_samples_per_second": 79.271,
-      "eval_steps_per_second": 0.619,
-      "step": 1025000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.906281916381266e-05,
-      "loss": 0.4577,
-      "step": 1026000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.894166064486576e-05,
-      "loss": 0.4586,
-      "step": 1027000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.882059608730904e-05,
-      "loss": 0.4586,
-      "step": 1028000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.869938348850115e-05,
-      "loss": 0.4591,
-      "step": 1029000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.85781443898625e-05,
-      "loss": 0.4592,
-      "step": 1030000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.43954479694366455,
-      "eval_runtime": 78.7138,
-      "eval_samples_per_second": 81.307,
-      "eval_steps_per_second": 0.635,
-      "step": 1030000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.845700043886677e-05,
-      "loss": 0.4591,
-      "step": 1031000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.833570947416866e-05,
-      "loss": 0.4569,
-      "step": 1032000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.82143931174539e-05,
-      "loss": 0.4581,
-      "step": 1033000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.809305173827309e-05,
-      "loss": 0.4579,
-      "step": 1034000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.797168570625316e-05,
-      "loss": 0.4576,
-      "step": 1035000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.44192150235176086,
-      "eval_runtime": 82.0458,
-      "eval_samples_per_second": 78.005,
-      "eval_steps_per_second": 0.609,
-      "step": 1035000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.785041679341762e-05,
-      "loss": 0.4594,
-      "step": 1036000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.772900258862799e-05,
-      "loss": 0.4585,
-      "step": 1037000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.760768628934177e-05,
-      "loss": 0.4576,
-      "step": 1038000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.748622538969719e-05,
-      "loss": 0.4581,
-      "step": 1039000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 8.736486318068334e-05,
-      "loss": 0.4595,
-      "step": 1040000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4380738139152527,
-      "eval_runtime": 79.7233,
-      "eval_samples_per_second": 80.278,
-      "eval_steps_per_second": 0.627,
-      "step": 1040000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.724347858186586e-05,
-      "loss": 0.4577,
-      "step": 1041000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.712195042320727e-05,
-      "loss": 0.4582,
-      "step": 1042000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.700040056985257e-05,
-      "loss": 0.4583,
-      "step": 1043000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.687882939206366e-05,
-      "loss": 0.4585,
-      "step": 1044000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.67573588626427e-05,
-      "loss": 0.4577,
-      "step": 1045000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.437910795211792,
-      "eval_runtime": 79.8686,
-      "eval_samples_per_second": 80.132,
-      "eval_steps_per_second": 0.626,
-      "step": 1045000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.663586779028231e-05,
-      "loss": 0.4578,
-      "step": 1046000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.651435654432562e-05,
-      "loss": 0.4583,
-      "step": 1047000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.639270383167568e-05,
-      "loss": 0.4583,
-      "step": 1048000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.62710316457412e-05,
-      "loss": 0.4579,
-      "step": 1049000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 8.614934035715678e-05,
-      "loss": 0.4578,
-      "step": 1050000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.4384676218032837,
-      "eval_runtime": 84.4358,
-      "eval_samples_per_second": 75.797,
-      "eval_steps_per_second": 0.592,
-      "step": 1050000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.602763033661519e-05,
-      "loss": 0.4575,
-      "step": 1051000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.590602369229602e-05,
-      "loss": 0.4576,
-      "step": 1052000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.578427733795066e-05,
-      "loss": 0.4565,
-      "step": 1053000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.566251336369341e-05,
-      "loss": 0.4565,
-      "step": 1054000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.554073214043843e-05,
-      "loss": 0.4574,
-      "step": 1055000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.4384410083293915,
-      "eval_runtime": 82.1023,
-      "eval_samples_per_second": 77.951,
-      "eval_steps_per_second": 0.609,
-      "step": 1055000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.541893403915243e-05,
-      "loss": 0.4573,
-      "step": 1056000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.529724125358366e-05,
-      "loss": 0.4578,
-      "step": 1057000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.517541052529095e-05,
-      "loss": 0.4594,
-      "step": 1058000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 8.505368588604491e-05,
-      "loss": 0.4575,
-      "step": 1059000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.493182401373038e-05,
-      "loss": 0.456,
-      "step": 1060000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4374844431877136,
-      "eval_runtime": 81.8521,
-      "eval_samples_per_second": 78.19,
-      "eval_steps_per_second": 0.611,
-      "step": 1060000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.480994711822458e-05,
-      "loss": 0.4569,
-      "step": 1061000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.468805557078571e-05,
-      "loss": 0.4569,
-      "step": 1062000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.456614974271656e-05,
-      "loss": 0.4574,
-      "step": 1063000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.444435193192483e-05,
-      "loss": 0.4561,
-      "step": 1064000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.432254060992934e-05,
-      "loss": 0.4574,
-      "step": 1065000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.43869927525520325,
-      "eval_runtime": 85.7182,
-      "eval_samples_per_second": 74.663,
-      "eval_steps_per_second": 0.583,
-      "step": 1065000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.420059419417763e-05,
-      "loss": 0.4574,
-      "step": 1066000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.407875694816683e-05,
-      "loss": 0.4565,
-      "step": 1067000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.395678532468588e-05,
-      "loss": 0.4557,
-      "step": 1068000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.383480164815062e-05,
-      "loss": 0.4571,
-      "step": 1069000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.371280629014447e-05,
-      "loss": 0.4583,
-      "step": 1070000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.43844690918922424,
-      "eval_runtime": 78.8816,
-      "eval_samples_per_second": 81.134,
-      "eval_steps_per_second": 0.634,
-      "step": 1070000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.359079962228652e-05,
-      "loss": 0.4568,
-      "step": 1071000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.346890403917613e-05,
-      "loss": 0.4551,
-      "step": 1072000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.334687587698898e-05,
-      "loss": 0.4569,
-      "step": 1073000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.322495956296366e-05,
-      "loss": 0.4567,
-      "step": 1074000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.310315549773483e-05,
-      "loss": 0.4558,
-      "step": 1075000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.43670228123664856,
-      "eval_runtime": 92.0013,
-      "eval_samples_per_second": 69.564,
-      "eval_steps_per_second": 0.543,
-      "step": 1075000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.298109789313348e-05,
-      "loss": 0.4578,
-      "step": 1076000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.285903120759091e-05,
-      "loss": 0.4559,
-      "step": 1077000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.273695581294347e-05,
-      "loss": 0.456,
-      "step": 1078000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.2614872081054e-05,
-      "loss": 0.4573,
-      "step": 1079000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 8.249290247936282e-05,
-      "loss": 0.457,
-      "step": 1080000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.43457353115081787,
-      "eval_runtime": 80.6194,
-      "eval_samples_per_second": 79.385,
-      "eval_steps_per_second": 0.62,
-      "step": 1080000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.237080319608577e-05,
-      "loss": 0.4561,
-      "step": 1081000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.224869669093106e-05,
-      "loss": 0.4572,
-      "step": 1082000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.212658333585626e-05,
-      "loss": 0.4564,
-      "step": 1083000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.200446350283978e-05,
-      "loss": 0.4561,
-      "step": 1084000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.188258182160425e-05,
-      "loss": 0.4564,
-      "step": 1085000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.4339691400527954,
-      "eval_runtime": 80.6306,
-      "eval_samples_per_second": 79.374,
-      "eval_steps_per_second": 0.62,
-      "step": 1085000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.176045015981425e-05,
-      "loss": 0.4559,
-      "step": 1086000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.163831313538776e-05,
-      "loss": 0.4571,
-      "step": 1087000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.151617112037531e-05,
-      "loss": 0.4568,
-      "step": 1088000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.139414663565934e-05,
-      "loss": 0.4562,
-      "step": 1089000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.127211791262032e-05,
-      "loss": 0.4564,
-      "step": 1090000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.435046911239624,
-      "eval_runtime": 80.4772,
-      "eval_samples_per_second": 79.526,
-      "eval_steps_per_second": 0.621,
-      "step": 1090000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.114996316567713e-05,
-      "loss": 0.4563,
-      "step": 1091000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.10278049157467e-05,
-      "loss": 0.4557,
-      "step": 1092000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.090564353494425e-05,
-      "loss": 0.4561,
-      "step": 1093000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.078347939539452e-05,
-      "loss": 0.4562,
-      "step": 1094000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.06615572044175e-05,
-      "loss": 0.4572,
-      "step": 1095000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.43538710474967957,
-      "eval_runtime": 79.2403,
-      "eval_samples_per_second": 80.767,
-      "eval_steps_per_second": 0.631,
-      "step": 1095000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.053938866743753e-05,
-      "loss": 0.4573,
-      "step": 1096000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.041721848738611e-05,
-      "loss": 0.4551,
-      "step": 1097000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.029504703641487e-05,
-      "loss": 0.4563,
-      "step": 1098000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.017287468667916e-05,
-      "loss": 0.4561,
-      "step": 1099000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 8.005082398335268e-05,
-      "loss": 0.4559,
-      "step": 1100000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4374575912952423,
-      "eval_runtime": 81.4835,
-      "eval_samples_per_second": 78.543,
-      "eval_steps_per_second": 0.614,
-      "step": 1100000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.992865095253273e-05,
-      "loss": 0.4551,
-      "step": 1101000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.980647813905439e-05,
-      "loss": 0.4569,
-      "step": 1102000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.968430591507717e-05,
-      "loss": 0.4554,
-      "step": 1103000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.95623789940764e-05,
-      "loss": 0.4542,
-      "step": 1104000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.94402090625329e-05,
-      "loss": 0.455,
-      "step": 1105000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.43497970700263977,
-      "eval_runtime": 80.3312,
-      "eval_samples_per_second": 79.67,
-      "eval_steps_per_second": 0.622,
-      "step": 1105000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.931804083620962e-05,
-      "loss": 0.4547,
-      "step": 1106000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.919599685223971e-05,
-      "loss": 0.4568,
-      "step": 1107000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.90738331501521e-05,
-      "loss": 0.4551,
-      "step": 1108000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.895179442867707e-05,
-      "loss": 0.4551,
-      "step": 1109000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.882975889383518e-05,
-      "loss": 0.4551,
-      "step": 1110000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.43510711193084717,
-      "eval_runtime": 80.3235,
-      "eval_samples_per_second": 79.678,
-      "eval_steps_per_second": 0.622,
-      "step": 1110000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.87076047643997e-05,
-      "loss": 0.4553,
-      "step": 1111000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.858545457182424e-05,
-      "loss": 0.4551,
-      "step": 1112000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.846330868819947e-05,
-      "loss": 0.4547,
-      "step": 1113000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.834128962434351e-05,
-      "loss": 0.4547,
-      "step": 1114000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.821915346959958e-05,
-      "loss": 0.4549,
-      "step": 1115000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.436400443315506,
-      "eval_runtime": 81.1367,
-      "eval_samples_per_second": 78.879,
-      "eval_steps_per_second": 0.616,
-      "step": 1115000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.809702273962302e-05,
-      "loss": 0.4564,
-      "step": 1116000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.797501992835905e-05,
-      "loss": 0.4543,
-      "step": 1117000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.78529011576391e-05,
-      "loss": 0.4551,
-      "step": 1118000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 7.773091103619413e-05,
-      "loss": 0.4555,
-      "step": 1119000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.760880571122187e-05,
-      "loss": 0.4542,
-      "step": 1120000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.43270257115364075,
-      "eval_runtime": 81.0478,
-      "eval_samples_per_second": 78.966,
-      "eval_steps_per_second": 0.617,
-      "step": 1120000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.74867076702421e-05,
-      "loss": 0.454,
-      "step": 1121000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.73646172851866e-05,
-      "loss": 0.4548,
-      "step": 1122000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.724265700618751e-05,
-      "loss": 0.4545,
-      "step": 1123000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.712070510974305e-05,
-      "loss": 0.4547,
-      "step": 1124000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.699863990589822e-05,
-      "loss": 0.4557,
-      "step": 1125000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4347490966320038,
-      "eval_runtime": 82.2635,
-      "eval_samples_per_second": 77.799,
-      "eval_steps_per_second": 0.608,
-      "step": 1125000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.687658384471673e-05,
-      "loss": 0.4557,
-      "step": 1126000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.675465933967303e-05,
-      "loss": 0.456,
-      "step": 1127000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.663262266912916e-05,
-      "loss": 0.4559,
-      "step": 1128000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.651071827736612e-05,
-      "loss": 0.455,
-      "step": 1129000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.638870248295153e-05,
-      "loss": 0.4552,
-      "step": 1130000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.43748027086257935,
-      "eval_runtime": 81.7683,
-      "eval_samples_per_second": 78.27,
-      "eval_steps_per_second": 0.611,
-      "step": 1130000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.626669768917545e-05,
-      "loss": 0.4546,
-      "step": 1131000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.614470426768568e-05,
-      "loss": 0.454,
-      "step": 1132000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.602284456578315e-05,
-      "loss": 0.4538,
-      "step": 1133000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.590087499136845e-05,
-      "loss": 0.4543,
-      "step": 1134000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.577903985432644e-05,
-      "loss": 0.4535,
-      "step": 1135000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4324733018875122,
-      "eval_runtime": 79.5142,
-      "eval_samples_per_second": 80.489,
-      "eval_steps_per_second": 0.629,
-      "step": 1135000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.565709561166183e-05,
-      "loss": 0.4542,
-      "step": 1136000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.553528652251894e-05,
-      "loss": 0.4539,
-      "step": 1137000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.541336909597102e-05,
-      "loss": 0.4544,
-      "step": 1138000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.529158753744867e-05,
-      "loss": 0.4546,
-      "step": 1139000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 7.516969841105768e-05,
-      "loss": 0.4552,
-      "step": 1140000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.43759095668792725,
-      "eval_runtime": 81.8138,
-      "eval_samples_per_second": 78.226,
-      "eval_steps_per_second": 0.611,
-      "step": 1140000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.504782399860302e-05,
-      "loss": 0.453,
-      "step": 1141000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.492596467133529e-05,
-      "loss": 0.454,
-      "step": 1142000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.480424263648597e-05,
-      "loss": 0.4547,
-      "step": 1143000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.468253639717616e-05,
-      "loss": 0.4541,
-      "step": 1144000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.456072451973997e-05,
-      "loss": 0.4548,
-      "step": 1145000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.435017853975296,
-      "eval_runtime": 81.3392,
-      "eval_samples_per_second": 78.683,
-      "eval_steps_per_second": 0.615,
-      "step": 1145000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.443905099818896e-05,
-      "loss": 0.4538,
-      "step": 1146000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.431727261258993e-05,
-      "loss": 0.4538,
-      "step": 1147000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.419551153756405e-05,
-      "loss": 0.4537,
-      "step": 1148000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.407376814401668e-05,
-      "loss": 0.4538,
-      "step": 1149000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.395204280279928e-05,
-      "loss": 0.4538,
-      "step": 1150000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4336395263671875,
-      "eval_runtime": 80.9118,
-      "eval_samples_per_second": 79.098,
-      "eval_steps_per_second": 0.618,
-      "step": 1150000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.383045758230073e-05,
-      "loss": 0.4541,
-      "step": 1151000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.370889111773011e-05,
-      "loss": 0.4531,
-      "step": 1152000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.358722211935709e-05,
-      "loss": 0.4531,
-      "step": 1153000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.346557265541761e-05,
-      "loss": 0.4528,
-      "step": 1154000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.334394309647706e-05,
-      "loss": 0.4528,
-      "step": 1155000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.43530333042144775,
-      "eval_runtime": 80.885,
-      "eval_samples_per_second": 79.125,
-      "eval_steps_per_second": 0.618,
-      "step": 1155000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.322233381304011e-05,
-      "loss": 0.4533,
-      "step": 1156000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.31008667537514e-05,
-      "loss": 0.4547,
-      "step": 1157000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.29792991113865e-05,
-      "loss": 0.4541,
-      "step": 1158000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 7.285775285529407e-05,
-      "loss": 0.4538,
-      "step": 1159000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.273634986923404e-05,
-      "loss": 0.454,
-      "step": 1160000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.43437331914901733,
-      "eval_runtime": 80.8329,
-      "eval_samples_per_second": 79.176,
-      "eval_steps_per_second": 0.619,
-      "step": 1160000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.26148474740616e-05,
-      "loss": 0.4538,
-      "step": 1161000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.249348904388309e-05,
-      "loss": 0.4549,
-      "step": 1162000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.237203198861888e-05,
-      "loss": 0.4542,
-      "step": 1163000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.225059816946814e-05,
-      "loss": 0.4523,
-      "step": 1164000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.21293093546381e-05,
-      "loss": 0.4528,
-      "step": 1165000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.43269097805023193,
-      "eval_runtime": 81.0783,
-      "eval_samples_per_second": 78.936,
-      "eval_steps_per_second": 0.617,
-      "step": 1165000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.200792309320717e-05,
-      "loss": 0.4529,
-      "step": 1166000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.188656117702852e-05,
-      "loss": 0.4529,
-      "step": 1167000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.17652239757916e-05,
-      "loss": 0.4545,
-      "step": 1168000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.164391185911055e-05,
-      "loss": 0.4525,
-      "step": 1169000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.15227464703484e-05,
-      "loss": 0.453,
-      "step": 1170000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.43528246879577637,
-      "eval_runtime": 77.7536,
-      "eval_samples_per_second": 82.311,
-      "eval_steps_per_second": 0.643,
-      "step": 1170000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.140148560530676e-05,
-      "loss": 0.454,
-      "step": 1171000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.128049337579044e-05,
-      "loss": 0.452,
-      "step": 1172000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.115940640646439e-05,
-      "loss": 0.4537,
-      "step": 1173000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.103822514560007e-05,
-      "loss": 0.4533,
-      "step": 1174000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.09170711838562e-05,
-      "loss": 0.4536,
-      "step": 1175000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.43259185552597046,
-      "eval_runtime": 78.8646,
-      "eval_samples_per_second": 81.152,
-      "eval_steps_per_second": 0.634,
-      "step": 1175000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.079594489028868e-05,
-      "loss": 0.4528,
-      "step": 1176000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.067484663386926e-05,
-      "loss": 0.4547,
-      "step": 1177000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.055377678348416e-05,
-      "loss": 0.4527,
-      "step": 1178000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.043285673451301e-05,
-      "loss": 0.4529,
-      "step": 1179000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 7.031184477318059e-05,
-      "loss": 0.4531,
-      "step": 1180000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.43487441539764404,
-      "eval_runtime": 80.3946,
-      "eval_samples_per_second": 79.607,
-      "eval_steps_per_second": 0.622,
-      "step": 1180000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 7.019098329123491e-05,
-      "loss": 0.4525,
-      "step": 1181000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 7.007003069197359e-05,
-      "loss": 0.4535,
-      "step": 1182000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.99491083411209e-05,
-      "loss": 0.4525,
-      "step": 1183000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.982833748334574e-05,
-      "loss": 0.4529,
-      "step": 1184000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.970747670309944e-05,
-      "loss": 0.4523,
-      "step": 1185000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.43467041850090027,
-      "eval_runtime": 79.4202,
-      "eval_samples_per_second": 80.584,
-      "eval_steps_per_second": 0.63,
-      "step": 1185000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.95868889029842e-05,
-      "loss": 0.4527,
-      "step": 1186000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.946609113261812e-05,
-      "loss": 0.4522,
-      "step": 1187000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.934532545036533e-05,
-      "loss": 0.4523,
-      "step": 1188000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.922471294099106e-05,
-      "loss": 0.4532,
-      "step": 1189000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.91040125054781e-05,
-      "loss": 0.4527,
-      "step": 1190000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4338874816894531,
-      "eval_runtime": 78.8008,
-      "eval_samples_per_second": 81.217,
-      "eval_steps_per_second": 0.635,
-      "step": 1190000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.898346591157581e-05,
-      "loss": 0.4524,
-      "step": 1191000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.886283219202908e-05,
-      "loss": 0.4502,
-      "step": 1192000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.874235298095256e-05,
-      "loss": 0.4521,
-      "step": 1193000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.862178744578633e-05,
-      "loss": 0.4528,
-      "step": 1194000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.850125657062878e-05,
-      "loss": 0.4527,
-      "step": 1195000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4356515109539032,
-      "eval_runtime": 78.1346,
-      "eval_samples_per_second": 81.91,
-      "eval_steps_per_second": 0.64,
-      "step": 1195000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.838088120086767e-05,
-      "loss": 0.4522,
-      "step": 1196000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.826042071151715e-05,
-      "loss": 0.4505,
-      "step": 1197000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.814011638970504e-05,
-      "loss": 0.4519,
-      "step": 1198000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.801972775246722e-05,
-      "loss": 0.452,
-      "step": 1199000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 6.789937560921629e-05,
-      "loss": 0.4534,
-      "step": 1200000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4337690472602844,
-      "eval_runtime": 78.2694,
-      "eval_samples_per_second": 81.769,
-      "eval_steps_per_second": 0.639,
-      "step": 1200000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.777906032656586e-05,
-      "loss": 0.4514,
-      "step": 1201000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.765878227101707e-05,
-      "loss": 0.4509,
-      "step": 1202000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.753866203051995e-05,
-      "loss": 0.452,
-      "step": 1203000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.741857967353855e-05,
-      "loss": 0.4519,
-      "step": 1204000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.729841542014526e-05,
-      "loss": 0.4516,
-      "step": 1205000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.43374085426330566,
-      "eval_runtime": 78.9278,
-      "eval_samples_per_second": 81.087,
-      "eval_steps_per_second": 0.633,
-      "step": 1205000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.717828985798137e-05,
-      "loss": 0.451,
-      "step": 1206000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.705820335297017e-05,
-      "loss": 0.4515,
-      "step": 1207000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.693827629818452e-05,
-      "loss": 0.4499,
-      "step": 1208000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.681838895213777e-05,
-      "loss": 0.452,
-      "step": 1209000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.669842173225596e-05,
-      "loss": 0.4505,
-      "step": 1210000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.4336735010147095,
-      "eval_runtime": 78.5787,
-      "eval_samples_per_second": 81.447,
-      "eval_steps_per_second": 0.636,
-      "step": 1210000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.657849503128835e-05,
-      "loss": 0.4514,
-      "step": 1211000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.645860921455248e-05,
-      "loss": 0.452,
-      "step": 1212000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.633912411889133e-05,
-      "loss": 0.4525,
-      "step": 1213000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.621932104068297e-05,
-      "loss": 0.4506,
-      "step": 1214000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.609955994081265e-05,
-      "loss": 0.4523,
-      "step": 1215000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.43103519082069397,
-      "eval_runtime": 79.501,
-      "eval_samples_per_second": 80.502,
-      "eval_steps_per_second": 0.629,
-      "step": 1215000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.597984118409344e-05,
-      "loss": 0.4525,
-      "step": 1216000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.586016513520938e-05,
-      "loss": 0.4517,
-      "step": 1217000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.574065177005509e-05,
-      "loss": 0.4514,
-      "step": 1218000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 6.562106218675344e-05,
-      "loss": 0.4521,
-      "step": 1219000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.550163592797291e-05,
-      "loss": 0.4509,
-      "step": 1220000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4309111535549164,
-      "eval_runtime": 78.8684,
-      "eval_samples_per_second": 81.148,
-      "eval_steps_per_second": 0.634,
-      "step": 1220000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.538213426595817e-05,
-      "loss": 0.4515,
-      "step": 1221000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.526267713249757e-05,
-      "loss": 0.4507,
-      "step": 1222000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.51433842811744e-05,
-      "loss": 0.4505,
-      "step": 1223000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.50240172509089e-05,
-      "loss": 0.4511,
-      "step": 1224000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.490481513858697e-05,
-      "loss": 0.4503,
-      "step": 1225000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.43227553367614746,
-      "eval_runtime": 78.1714,
-      "eval_samples_per_second": 81.871,
-      "eval_steps_per_second": 0.64,
-      "step": 1225000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.478553966451205e-05,
-      "loss": 0.4517,
-      "step": 1226000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.466642974217942e-05,
-      "loss": 0.4505,
-      "step": 1227000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.454724727617616e-05,
-      "loss": 0.4515,
-      "step": 1228000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.442811188193757e-05,
-      "loss": 0.4497,
-      "step": 1229000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.430902392237073e-05,
-      "loss": 0.4507,
-      "step": 1230000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4315786361694336,
-      "eval_runtime": 79.653,
-      "eval_samples_per_second": 80.348,
-      "eval_steps_per_second": 0.628,
-      "step": 1230000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.419010277640487e-05,
-      "loss": 0.4516,
-      "step": 1231000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.407134866177907e-05,
-      "loss": 0.4508,
-      "step": 1232000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.395240503611e-05,
-      "loss": 0.4506,
-      "step": 1233000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.38335102941982e-05,
-      "loss": 0.4513,
-      "step": 1234000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.371466479821773e-05,
-      "loss": 0.452,
-      "step": 1235000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.43573638796806335,
-      "eval_runtime": 81.2864,
-      "eval_samples_per_second": 78.734,
-      "eval_steps_per_second": 0.615,
-      "step": 1235000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.359598768118097e-05,
-      "loss": 0.4505,
-      "step": 1236000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.347724171283347e-05,
-      "loss": 0.4511,
-      "step": 1237000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.335854607567315e-05,
-      "loss": 0.4513,
-      "step": 1238000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.32399011312675e-05,
-      "loss": 0.4507,
-      "step": 1239000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.312130724102964e-05,
-      "loss": 0.4512,
-      "step": 1240000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4293386936187744,
-      "eval_runtime": 78.7202,
-      "eval_samples_per_second": 81.301,
-      "eval_steps_per_second": 0.635,
-      "step": 1240000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.300276476621711e-05,
-      "loss": 0.4493,
-      "step": 1241000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.288439253264661e-05,
-      "loss": 0.451,
-      "step": 1242000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.276595391951207e-05,
-      "loss": 0.4505,
-      "step": 1243000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.264768616404324e-05,
-      "loss": 0.45,
-      "step": 1244000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.252935285428112e-05,
-      "loss": 0.4496,
-      "step": 1245000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.42953261733055115,
-      "eval_runtime": 80.8366,
-      "eval_samples_per_second": 79.172,
-      "eval_steps_per_second": 0.619,
-      "step": 1245000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.241107276314003e-05,
-      "loss": 0.4495,
-      "step": 1246000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.229284625092165e-05,
-      "loss": 0.45,
-      "step": 1247000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.217479182327523e-05,
-      "loss": 0.4507,
-      "step": 1248000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.205679158576122e-05,
-      "loss": 0.4498,
-      "step": 1249000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.193872786080677e-05,
-      "loss": 0.4505,
-      "step": 1250000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4315950870513916,
-      "eval_runtime": 78.3108,
-      "eval_samples_per_second": 81.726,
-      "eval_steps_per_second": 0.638,
-      "step": 1250000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.182071915361952e-05,
-      "loss": 0.4497,
-      "step": 1251000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.170276582367444e-05,
-      "loss": 0.4511,
-      "step": 1252000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.158498609991133e-05,
-      "loss": 0.4506,
-      "step": 1253000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.146714454592476e-05,
-      "loss": 0.4503,
-      "step": 1254000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.13494772030113e-05,
-      "loss": 0.4502,
-      "step": 1255000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.43262848258018494,
-      "eval_runtime": 79.5141,
-      "eval_samples_per_second": 80.489,
-      "eval_steps_per_second": 0.629,
-      "step": 1255000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.123174885940984e-05,
-      "loss": 0.4513,
-      "step": 1256000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.111407768715625e-05,
-      "loss": 0.4491,
-      "step": 1257000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.099658162948436e-05,
-      "loss": 0.4495,
-      "step": 1258000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.087902581702466e-05,
-      "loss": 0.4498,
-      "step": 1259000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 6.076164571872208e-05,
-      "loss": 0.4501,
-      "step": 1260000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4304799735546112,
-      "eval_runtime": 80.0574,
-      "eval_samples_per_second": 79.943,
-      "eval_steps_per_second": 0.625,
-      "step": 1260000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 6.064420669700471e-05,
-      "loss": 0.4504,
-      "step": 1261000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 6.05268266363905e-05,
-      "loss": 0.4509,
-      "step": 1262000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 6.0409505894439516e-05,
-      "loss": 0.4498,
-      "step": 1263000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 6.029236205966984e-05,
-      "loss": 0.4514,
-      "step": 1264000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 6.01751609667898e-05,
-      "loss": 0.4504,
-      "step": 1265000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.4325391352176666,
-      "eval_runtime": 78.7597,
-      "eval_samples_per_second": 81.26,
-      "eval_steps_per_second": 0.635,
-      "step": 1265000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 6.005802026380759e-05,
-      "loss": 0.4504,
-      "step": 1266000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.994094030755408e-05,
-      "loss": 0.4508,
-      "step": 1267000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.982403844288822e-05,
-      "loss": 0.4489,
-      "step": 1268000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.9707080988205694e-05,
-      "loss": 0.4486,
-      "step": 1269000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.959018534927381e-05,
-      "loss": 0.4496,
-      "step": 1270000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.4292341470718384,
-      "eval_runtime": 81.0604,
-      "eval_samples_per_second": 78.953,
-      "eval_steps_per_second": 0.617,
-      "step": 1270000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.947346868447071e-05,
-      "loss": 0.4495,
-      "step": 1271000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.9356697682398364e-05,
-      "loss": 0.4497,
-      "step": 1272000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.9240222916649246e-05,
-      "loss": 0.4497,
-      "step": 1273000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.9123577909416044e-05,
-      "loss": 0.4494,
-      "step": 1274000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.9006996495385025e-05,
-      "loss": 0.4515,
-      "step": 1275000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.4315658211708069,
-      "eval_runtime": 79.1549,
-      "eval_samples_per_second": 80.854,
-      "eval_steps_per_second": 0.632,
-      "step": 1275000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.889047902968337e-05,
-      "loss": 0.4494,
-      "step": 1276000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.877414228816838e-05,
-      "loss": 0.4492,
-      "step": 1277000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.8657753718891686e-05,
-      "loss": 0.4496,
-      "step": 1278000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 5.8541546452764014e-05,
-      "loss": 0.4512,
-      "step": 1279000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.84252881966525e-05,
-      "loss": 0.4492,
-      "step": 1280000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.427651971578598,
-      "eval_runtime": 79.7313,
-      "eval_samples_per_second": 80.27,
-      "eval_steps_per_second": 0.627,
-      "step": 1280000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.830909566085471e-05,
-      "loss": 0.4483,
-      "step": 1281000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.819308529265298e-05,
-      "loss": 0.4501,
-      "step": 1282000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.807702519250463e-05,
-      "loss": 0.45,
-      "step": 1283000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.796103187353951e-05,
-      "loss": 0.4498,
-      "step": 1284000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.784510568909343e-05,
-      "loss": 0.4488,
-      "step": 1285000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.43311890959739685,
-      "eval_runtime": 78.3828,
-      "eval_samples_per_second": 81.651,
-      "eval_steps_per_second": 0.638,
-      "step": 1285000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.7729362817166816e-05,
-      "loss": 0.4488,
-      "step": 1286000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.7613571892930174e-05,
-      "loss": 0.4495,
-      "step": 1287000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.7497964850187466e-05,
-      "loss": 0.4489,
-      "step": 1288000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.73823105956256e-05,
-      "loss": 0.4498,
-      "step": 1289000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.7266725238468133e-05,
-      "loss": 0.4491,
-      "step": 1290000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.43072545528411865,
-      "eval_runtime": 79.3715,
-      "eval_samples_per_second": 80.633,
-      "eval_steps_per_second": 0.63,
-      "step": 1290000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.715132461220847e-05,
-      "loss": 0.4488,
-      "step": 1291000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.703587803615088e-05,
-      "loss": 0.4477,
-      "step": 1292000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.6920501412791343e-05,
-      "loss": 0.4488,
-      "step": 1293000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.680519509358706e-05,
-      "loss": 0.4492,
-      "step": 1294000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.669007463003561e-05,
-      "loss": 0.4482,
-      "step": 1295000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4276537001132965,
-      "eval_runtime": 83.0062,
-      "eval_samples_per_second": 77.103,
-      "eval_steps_per_second": 0.602,
-      "step": 1295000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.6574909901474036e-05,
-      "loss": 0.4495,
-      "step": 1296000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.645981652979939e-05,
-      "loss": 0.4505,
-      "step": 1297000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.634490985133569e-05,
-      "loss": 0.4487,
-      "step": 1298000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.622996017276679e-05,
-      "loss": 0.45,
-      "step": 1299000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 5.6115082901861476e-05,
-      "loss": 0.4495,
-      "step": 1300000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4289723336696625,
-      "eval_runtime": 83.6952,
-      "eval_samples_per_second": 76.468,
-      "eval_steps_per_second": 0.597,
-      "step": 1300000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.60002783885559e-05,
-      "loss": 0.4497,
-      "step": 1301000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.5885776372183025e-05,
-      "loss": 0.4476,
-      "step": 1302000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.57711182757353e-05,
-      "loss": 0.4495,
-      "step": 1303000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.565664853197224e-05,
-      "loss": 0.449,
-      "step": 1304000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.554213832099068e-05,
-      "loss": 0.4497,
-      "step": 1305000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4274589419364929,
-      "eval_runtime": 77.8458,
-      "eval_samples_per_second": 82.214,
-      "eval_steps_per_second": 0.642,
-      "step": 1305000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.542781701127524e-05,
-      "loss": 0.4488,
-      "step": 1306000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.5313456079635704e-05,
-      "loss": 0.4485,
-      "step": 1307000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.519917034749017e-05,
-      "loss": 0.4495,
-      "step": 1308000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.5085074335305515e-05,
-      "loss": 0.45,
-      "step": 1309000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.4970939970251195e-05,
-      "loss": 0.4494,
-      "step": 1310000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.43162286281585693,
-      "eval_runtime": 78.176,
-      "eval_samples_per_second": 81.867,
-      "eval_steps_per_second": 0.64,
-      "step": 1310000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.485688184805438e-05,
-      "loss": 0.4483,
-      "step": 1311000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.474290031615589e-05,
-      "loss": 0.4486,
-      "step": 1312000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.462899572176321e-05,
-      "loss": 0.4492,
-      "step": 1313000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.4515395989108964e-05,
-      "loss": 0.447,
-      "step": 1314000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.440164615480369e-05,
-      "loss": 0.4483,
-      "step": 1315000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4310395121574402,
-      "eval_runtime": 78.2586,
-      "eval_samples_per_second": 81.78,
-      "eval_steps_per_second": 0.639,
-      "step": 1315000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.428797429752361e-05,
-      "loss": 0.4486,
-      "step": 1316000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.4174494317829243e-05,
-      "loss": 0.4494,
-      "step": 1317000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.4060979374311454e-05,
-      "loss": 0.4483,
-      "step": 1318000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.3947656841894844e-05,
-      "loss": 0.4478,
-      "step": 1319000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.3834300193900343e-05,
-      "loss": 0.4478,
-      "step": 1320000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4294278025627136,
-      "eval_runtime": 81.6018,
-      "eval_samples_per_second": 78.43,
-      "eval_steps_per_second": 0.613,
-      "step": 1320000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.372102325116521e-05,
-      "loss": 0.4487,
-      "step": 1321000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.3607826358750705e-05,
-      "loss": 0.4475,
-      "step": 1322000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.3494822937699356e-05,
-      "loss": 0.4471,
-      "step": 1323000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.3381787099221496e-05,
-      "loss": 0.4491,
-      "step": 1324000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.326883234443639e-05,
-      "loss": 0.4479,
-      "step": 1325000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4298754036426544,
-      "eval_runtime": 83.1912,
-      "eval_samples_per_second": 76.931,
-      "eval_steps_per_second": 0.601,
-      "step": 1325000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.315595901742376e-05,
-      "loss": 0.4486,
-      "step": 1326000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.304339296328977e-05,
-      "loss": 0.4472,
-      "step": 1327000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.293068335849521e-05,
-      "loss": 0.4466,
-      "step": 1328000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.281805621153364e-05,
-      "loss": 0.4469,
-      "step": 1329000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.270562436835969e-05,
-      "loss": 0.4483,
-      "step": 1330000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4302482604980469,
-      "eval_runtime": 78.9917,
-      "eval_samples_per_second": 81.021,
-      "eval_steps_per_second": 0.633,
-      "step": 1330000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.25931630827424e-05,
-      "loss": 0.4479,
-      "step": 1331000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.248078528310338e-05,
-      "loss": 0.4493,
-      "step": 1332000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.2368603563750103e-05,
-      "loss": 0.4483,
-      "step": 1333000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.225639367843791e-05,
-      "loss": 0.4482,
-      "step": 1334000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.2144380388008114e-05,
-      "loss": 0.4473,
-      "step": 1335000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4291970133781433,
-      "eval_runtime": 82.7148,
-      "eval_samples_per_second": 77.374,
-      "eval_steps_per_second": 0.604,
-      "step": 1335000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.203233978289791e-05,
-      "loss": 0.4475,
-      "step": 1336000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.192038437213101e-05,
-      "loss": 0.4462,
-      "step": 1337000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.1808514496743016e-05,
-      "loss": 0.4476,
-      "step": 1338000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.169684223849966e-05,
-      "loss": 0.4494,
-      "step": 1339000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.158514436954633e-05,
-      "loss": 0.4471,
-      "step": 1340000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4304357171058655,
-      "eval_runtime": 89.5128,
-      "eval_samples_per_second": 71.498,
-      "eval_steps_per_second": 0.559,
-      "step": 1340000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.147353305717107e-05,
-      "loss": 0.4485,
-      "step": 1341000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.1362120122259206e-05,
-      "loss": 0.4482,
-      "step": 1342000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.125079424891131e-05,
-      "loss": 0.4483,
-      "step": 1343000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.1139444469644865e-05,
-      "loss": 0.4466,
-      "step": 1344000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.10282938224747e-05,
-      "loss": 0.4483,
-      "step": 1345000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.4288882911205292,
-      "eval_runtime": 86.9362,
-      "eval_samples_per_second": 73.617,
-      "eval_steps_per_second": 0.575,
-      "step": 1345000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.091712012222336e-05,
-      "loss": 0.447,
-      "step": 1346000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.080603501346776e-05,
-      "loss": 0.4474,
-      "step": 1347000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.069503883459252e-05,
-      "loss": 0.4467,
-      "step": 1348000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.0584131923711175e-05,
-      "loss": 0.4483,
-      "step": 1349000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.047331461866543e-05,
-      "loss": 0.4482,
-      "step": 1350000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.4289691150188446,
-      "eval_runtime": 155.3438,
-      "eval_samples_per_second": 41.199,
-      "eval_steps_per_second": 0.322,
-      "step": 1350000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 5.036269793934667e-05,
-      "loss": 0.4474,
-      "step": 1351000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 5.02520607679552e-05,
-      "loss": 0.448,
-      "step": 1352000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 5.014162471512421e-05,
-      "loss": 0.4481,
-      "step": 1353000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 5.003116902411837e-05,
-      "loss": 0.4471,
-      "step": 1354000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 4.9920914942058365e-05,
-      "loss": 0.448,
-      "step": 1355000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.4262136220932007,
-      "eval_runtime": 84.0235,
-      "eval_samples_per_second": 76.169,
-      "eval_steps_per_second": 0.595,
-      "step": 1355000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 4.981075230294731e-05,
-      "loss": 0.4463,
-      "step": 1356000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 4.970057130673938e-05,
-      "loss": 0.4474,
-      "step": 1357000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 4.959048260784589e-05,
-      "loss": 0.4467,
-      "step": 1358000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 4.948048654161614e-05,
-      "loss": 0.4467,
-      "step": 1359000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.937069329966688e-05,
-      "loss": 0.447,
-      "step": 1360000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4244381785392761,
-      "eval_runtime": 78.4672,
-      "eval_samples_per_second": 81.563,
-      "eval_steps_per_second": 0.637,
-      "step": 1360000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.926088341021301e-05,
-      "loss": 0.4475,
-      "step": 1361000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.915116715743923e-05,
-      "loss": 0.4463,
-      "step": 1362000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.904154487556023e-05,
-      "loss": 0.4475,
-      "step": 1363000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.8932126379265246e-05,
-      "loss": 0.4473,
-      "step": 1364000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.8822692945868926e-05,
-      "loss": 0.4465,
-      "step": 1365000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.42896711826324463,
-      "eval_runtime": 78.3264,
-      "eval_samples_per_second": 81.709,
-      "eval_steps_per_second": 0.638,
-      "step": 1365000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.871346377486962e-05,
-      "loss": 0.4473,
-      "step": 1366000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.860422052203485e-05,
-      "loss": 0.4456,
-      "step": 1367000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.849507290618916e-05,
-      "loss": 0.4463,
-      "step": 1368000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.838613026341415e-05,
-      "loss": 0.4464,
-      "step": 1369000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.827717482223451e-05,
-      "loss": 0.4475,
-      "step": 1370000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4280179738998413,
-      "eval_runtime": 78.2271,
-      "eval_samples_per_second": 81.813,
-      "eval_steps_per_second": 0.639,
-      "step": 1370000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.8168316014281885e-05,
-      "loss": 0.4458,
-      "step": 1371000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.805966288445796e-05,
-      "loss": 0.4461,
-      "step": 1372000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.7950998240010834e-05,
-      "loss": 0.4456,
-      "step": 1373000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.784243122238121e-05,
-      "loss": 0.4463,
-      "step": 1374000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.773396216228298e-05,
-      "loss": 0.4462,
-      "step": 1375000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4259638488292694,
-      "eval_runtime": 79.4294,
-      "eval_samples_per_second": 80.575,
-      "eval_steps_per_second": 0.629,
-      "step": 1375000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.762569971169915e-05,
-      "loss": 0.4461,
-      "step": 1376000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.751764390469416e-05,
-      "loss": 0.4445,
-      "step": 1377000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.740947040114839e-05,
-      "loss": 0.4467,
-      "step": 1378000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.730139617400774e-05,
-      "loss": 0.4463,
-      "step": 1379000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.7193421552485106e-05,
-      "loss": 0.4463,
-      "step": 1380000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4267512559890747,
-      "eval_runtime": 79.7271,
-      "eval_samples_per_second": 80.274,
-      "eval_steps_per_second": 0.627,
-      "step": 1380000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.7085654690150224e-05,
-      "loss": 0.4468,
-      "step": 1381000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.697788016586024e-05,
-      "loss": 0.4458,
-      "step": 1382000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.687020623267417e-05,
-      "loss": 0.448,
-      "step": 1383000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.6762633218585406e-05,
-      "loss": 0.4468,
-      "step": 1384000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.6655161451280045e-05,
-      "loss": 0.4464,
-      "step": 1385000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.42754608392715454,
-      "eval_runtime": 79.2336,
-      "eval_samples_per_second": 80.774,
-      "eval_steps_per_second": 0.631,
-      "step": 1385000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.654779125813573e-05,
-      "loss": 0.4469,
-      "step": 1386000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.64406301835041e-05,
-      "loss": 0.446,
-      "step": 1387000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.633357113217997e-05,
-      "loss": 0.4465,
-      "step": 1388000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.622650741725104e-05,
-      "loss": 0.4453,
-      "step": 1389000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.6119546582237665e-05,
-      "loss": 0.4475,
-      "step": 1390000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.42643147706985474,
-      "eval_runtime": 78.7876,
-      "eval_samples_per_second": 81.231,
-      "eval_steps_per_second": 0.635,
-      "step": 1390000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.601279575893072e-05,
-      "loss": 0.4458,
-      "step": 1391000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.590604155720414e-05,
-      "loss": 0.4443,
-      "step": 1392000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.5799391211587674e-05,
-      "loss": 0.4465,
-      "step": 1393000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.569295154097499e-05,
-      "loss": 0.4457,
-      "step": 1394000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.5586509777220274e-05,
-      "loss": 0.4459,
-      "step": 1395000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.426248162984848,
-      "eval_runtime": 81.9909,
-      "eval_samples_per_second": 78.057,
-      "eval_steps_per_second": 0.61,
-      "step": 1395000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.54801728429249e-05,
-      "loss": 0.4461,
-      "step": 1396000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.537394106200966e-05,
-      "loss": 0.446,
-      "step": 1397000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.526792083158551e-05,
-      "loss": 0.4458,
-      "step": 1398000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 4.516190022194884e-05,
-      "loss": 0.4462,
-      "step": 1399000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.5056091596576796e-05,
-      "loss": 0.4467,
-      "step": 1400000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4303908944129944,
-      "eval_runtime": 78.2681,
-      "eval_samples_per_second": 81.77,
-      "eval_steps_per_second": 0.639,
-      "step": 1400000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.495028344875484e-05,
-      "loss": 0.4471,
-      "step": 1401000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.4844582068448006e-05,
-      "loss": 0.4459,
-      "step": 1402000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.4739093318333515e-05,
-      "loss": 0.447,
-      "step": 1403000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.463360633111313e-05,
-      "loss": 0.4464,
-      "step": 1404000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.452822707606128e-05,
-      "loss": 0.4466,
-      "step": 1405000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.42388811707496643,
-      "eval_runtime": 81.8794,
-      "eval_samples_per_second": 78.164,
-      "eval_steps_per_second": 0.611,
-      "step": 1405000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.4423061091303984e-05,
-      "loss": 0.4462,
-      "step": 1406000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.431789815473655e-05,
-      "loss": 0.4452,
-      "step": 1407000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.421284391203949e-05,
-      "loss": 0.4451,
-      "step": 1408000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.410789868322626e-05,
-      "loss": 0.4459,
-      "step": 1409000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.400316756915505e-05,
-      "loss": 0.4454,
-      "step": 1410000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4258207678794861,
-      "eval_runtime": 82.7222,
-      "eval_samples_per_second": 77.367,
-      "eval_steps_per_second": 0.604,
-      "step": 1410000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.3898441217008166e-05,
-      "loss": 0.446,
-      "step": 1411000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.3793929397813896e-05,
-      "loss": 0.446,
-      "step": 1412000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.368952764812006e-05,
-      "loss": 0.4444,
-      "step": 1413000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.358513194496827e-05,
-      "loss": 0.4461,
-      "step": 1414000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.348084716781979e-05,
-      "loss": 0.4454,
-      "step": 1415000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4262985587120056,
-      "eval_runtime": 82.7185,
-      "eval_samples_per_second": 77.371,
-      "eval_steps_per_second": 0.604,
-      "step": 1415000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.3376777752205805e-05,
-      "loss": 0.4457,
-      "step": 1416000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.3272819674268335e-05,
-      "loss": 0.4455,
-      "step": 1417000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.3168869355718386e-05,
-      "loss": 0.4452,
-      "step": 1418000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 4.306503123117983e-05,
-      "loss": 0.4456,
-      "step": 1419000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.296130561696156e-05,
-      "loss": 0.4447,
-      "step": 1420000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4243999421596527,
-      "eval_runtime": 78.935,
-      "eval_samples_per_second": 81.079,
-      "eval_steps_per_second": 0.633,
-      "step": 1420000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.2857692829029776e-05,
-      "loss": 0.4458,
-      "step": 1421000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.275419318300705e-05,
-      "loss": 0.4456,
-      "step": 1422000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.265091032358337e-05,
-      "loss": 0.4459,
-      "step": 1423000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.254774100853429e-05,
-      "loss": 0.4462,
-      "step": 1424000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.2444582450035355e-05,
-      "loss": 0.4449,
-      "step": 1425000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4202989935874939,
-      "eval_runtime": 81.6727,
-      "eval_samples_per_second": 78.362,
-      "eval_steps_per_second": 0.612,
-      "step": 1425000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.234153829185128e-05,
-      "loss": 0.4439,
-      "step": 1426000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.223871171991236e-05,
-      "loss": 0.4454,
-      "step": 1427000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.213589718849552e-05,
-      "loss": 0.446,
-      "step": 1428000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.203319799770243e-05,
-      "loss": 0.4448,
-      "step": 1429000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.193061446037263e-05,
-      "loss": 0.4449,
-      "step": 1430000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.423859179019928,
-      "eval_runtime": 81.673,
-      "eval_samples_per_second": 78.361,
-      "eval_steps_per_second": 0.612,
-      "step": 1430000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.182814688899332e-05,
-      "loss": 0.4453,
-      "step": 1431000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.172589788880715e-05,
-      "loss": 0.4448,
-      "step": 1432000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.1623663068631195e-05,
-      "loss": 0.4443,
-      "step": 1433000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.15216472088565e-05,
-      "loss": 0.4448,
-      "step": 1434000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.141964638433753e-05,
-      "loss": 0.4447,
-      "step": 1435000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4245039224624634,
-      "eval_runtime": 81.4224,
-      "eval_samples_per_second": 78.602,
-      "eval_steps_per_second": 0.614,
-      "step": 1435000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.131786490676274e-05,
-      "loss": 0.4453,
-      "step": 1436000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.1216099319506056e-05,
-      "loss": 0.4459,
-      "step": 1437000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.1114451874736935e-05,
-      "loss": 0.4457,
-      "step": 1438000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.101302435181392e-05,
-      "loss": 0.4445,
-      "step": 1439000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 4.091161400165054e-05,
-      "loss": 0.4451,
-      "step": 1440000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.42323625087738037,
-      "eval_runtime": 91.7597,
-      "eval_samples_per_second": 69.747,
-      "eval_steps_per_second": 0.545,
-      "step": 1440000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.0810322721489815e-05,
-      "loss": 0.4458,
-      "step": 1441000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.070915081988253e-05,
-      "loss": 0.4433,
-      "step": 1442000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.060819959734469e-05,
-      "loss": 0.4449,
-      "step": 1443000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.0507267256893e-05,
-      "loss": 0.4446,
-      "step": 1444000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.0406555970000004e-05,
-      "loss": 0.4442,
-      "step": 1445000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4247336685657501,
-      "eval_runtime": 79.9188,
-      "eval_samples_per_second": 80.081,
-      "eval_steps_per_second": 0.626,
-      "step": 1445000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.03059650505077e-05,
-      "loss": 0.4449,
-      "step": 1446000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.020539429396338e-05,
-      "loss": 0.4449,
-      "step": 1447000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.010494475868937e-05,
-      "loss": 0.4444,
-      "step": 1448000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 4.000461675067245e-05,
-      "loss": 0.4434,
-      "step": 1449000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.990451072074708e-05,
-      "loss": 0.4445,
-      "step": 1450000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.42264366149902344,
-      "eval_runtime": 81.9965,
-      "eval_samples_per_second": 78.052,
-      "eval_steps_per_second": 0.61,
-      "step": 1450000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.980442656143221e-05,
-      "loss": 0.4449,
-      "step": 1451000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.970456474526e-05,
-      "loss": 0.4452,
-      "step": 1452000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.960472565292187e-05,
-      "loss": 0.4445,
-      "step": 1453000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.9505009611594935e-05,
-      "loss": 0.4451,
-      "step": 1454000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.9405416925031554e-05,
-      "loss": 0.4441,
-      "step": 1455000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.42445144057273865,
-      "eval_runtime": 84.4516,
-      "eval_samples_per_second": 75.783,
-      "eval_steps_per_second": 0.592,
-      "step": 1455000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.930614671105286e-05,
-      "loss": 0.4443,
-      "step": 1456000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.9206801395545245e-05,
-      "loss": 0.4444,
-      "step": 1457000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.9107580343195134e-05,
-      "loss": 0.4442,
-      "step": 1458000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.9008582890413084e-05,
-      "loss": 0.4441,
-      "step": 1459000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.890961114571417e-05,
-      "loss": 0.445,
-      "step": 1460000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4252031743526459,
-      "eval_runtime": 78.9687,
-      "eval_samples_per_second": 81.045,
-      "eval_steps_per_second": 0.633,
-      "step": 1460000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.8810764569465806e-05,
-      "loss": 0.4443,
-      "step": 1461000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.87120434627718e-05,
-      "loss": 0.4439,
-      "step": 1462000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.86135466587679e-05,
-      "loss": 0.4449,
-      "step": 1463000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.85150772667438e-05,
-      "loss": 0.4434,
-      "step": 1464000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.841673424498882e-05,
-      "loss": 0.4442,
-      "step": 1465000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4248676598072052,
-      "eval_runtime": 80.6767,
-      "eval_samples_per_second": 79.329,
-      "eval_steps_per_second": 0.62,
-      "step": 1465000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.8318616046053615e-05,
-      "loss": 0.4442,
-      "step": 1466000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.822052653604248e-05,
-      "loss": 0.4435,
-      "step": 1467000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.812256429355308e-05,
-      "loss": 0.4448,
-      "step": 1468000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.802472961699541e-05,
-      "loss": 0.443,
-      "step": 1469000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.7927120447236364e-05,
-      "loss": 0.4449,
-      "step": 1470000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.42114773392677307,
-      "eval_runtime": 85.6392,
-      "eval_samples_per_second": 74.732,
-      "eval_steps_per_second": 0.584,
-      "step": 1470000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.782963918257054e-05,
-      "loss": 0.4439,
-      "step": 1471000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.7732188733163535e-05,
-      "loss": 0.4446,
-      "step": 1472000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.763486703883993e-05,
-      "loss": 0.4437,
-      "step": 1473000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.753777152414142e-05,
-      "loss": 0.4443,
-      "step": 1474000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.74407080994724e-05,
-      "loss": 0.4444,
-      "step": 1475000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4239863455295563,
-      "eval_runtime": 78.8432,
-      "eval_samples_per_second": 81.174,
-      "eval_steps_per_second": 0.634,
-      "step": 1475000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.734387118671403e-05,
-      "loss": 0.4443,
-      "step": 1476000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.724706721320442e-05,
-      "loss": 0.4444,
-      "step": 1477000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.715049008113378e-05,
-      "loss": 0.444,
-      "step": 1478000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 3.7053946737130206e-05,
-      "loss": 0.4445,
-      "step": 1479000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.6957534214252755e-05,
-      "loss": 0.4434,
-      "step": 1480000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4229904115200043,
-      "eval_runtime": 119.8175,
-      "eval_samples_per_second": 53.415,
-      "eval_steps_per_second": 0.417,
-      "step": 1480000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.6861349022009304e-05,
-      "loss": 0.4442,
-      "step": 1481000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.676519889049805e-05,
-      "loss": 0.4441,
-      "step": 1482000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.666918045968899e-05,
-      "loss": 0.4443,
-      "step": 1483000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.6573294022070874e-05,
-      "loss": 0.4442,
-      "step": 1484000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.6477635557709146e-05,
-      "loss": 0.4435,
-      "step": 1485000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.42340168356895447,
-      "eval_runtime": 87.4539,
-      "eval_samples_per_second": 73.181,
-      "eval_steps_per_second": 0.572,
-      "step": 1485000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.638210940499669e-05,
-      "loss": 0.4425,
-      "step": 1486000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.628662043183205e-05,
-      "loss": 0.4437,
-      "step": 1487000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.619126461720257e-05,
-      "loss": 0.4432,
-      "step": 1488000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.6096042251578594e-05,
-      "loss": 0.4433,
-      "step": 1489000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.600095362502394e-05,
-      "loss": 0.443,
-      "step": 1490000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.42291781306266785,
-      "eval_runtime": 80.5128,
-      "eval_samples_per_second": 79.49,
-      "eval_steps_per_second": 0.621,
-      "step": 1490000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.59060939147492e-05,
-      "loss": 0.4439,
-      "step": 1491000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.581127350043195e-05,
-      "loss": 0.4431,
-      "step": 1492000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.571658769263903e-05,
-      "loss": 0.4433,
-      "step": 1493000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.562213126323673e-05,
-      "loss": 0.4423,
-      "step": 1494000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.55278097462887e-05,
-      "loss": 0.4434,
-      "step": 1495000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.42214784026145935,
-      "eval_runtime": 79.6252,
-      "eval_samples_per_second": 80.377,
-      "eval_steps_per_second": 0.628,
-      "step": 1495000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.5433529215772676e-05,
-      "loss": 0.4433,
-      "step": 1496000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.533938444245558e-05,
-      "loss": 0.4442,
-      "step": 1497000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.5245375713118724e-05,
-      "loss": 0.4433,
-      "step": 1498000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.515159711833576e-05,
-      "loss": 0.4426,
-      "step": 1499000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 3.505786119888575e-05,
-      "loss": 0.4434,
-      "step": 1500000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4238176643848419,
-      "eval_runtime": 81.4375,
-      "eval_samples_per_second": 78.588,
-      "eval_steps_per_second": 0.614,
-      "step": 1500000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.496426218098463e-05,
-      "loss": 0.4434,
-      "step": 1501000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.4870893742962914e-05,
-      "loss": 0.4437,
-      "step": 1502000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.477756924548471e-05,
-      "loss": 0.4444,
-      "step": 1503000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.468447562121202e-05,
-      "loss": 0.4427,
-      "step": 1504000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.459142678014942e-05,
-      "loss": 0.4435,
-      "step": 1505000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.42447999119758606,
-      "eval_runtime": 81.5993,
-      "eval_samples_per_second": 78.432,
-      "eval_steps_per_second": 0.613,
-      "step": 1505000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.4498516261475824e-05,
-      "loss": 0.4427,
-      "step": 1506000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.440583705079855e-05,
-      "loss": 0.4426,
-      "step": 1507000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.431320388651632e-05,
-      "loss": 0.4418,
-      "step": 1508000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.422080231652349e-05,
-      "loss": 0.4432,
-      "step": 1509000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.412853991914283e-05,
-      "loss": 0.4435,
-      "step": 1510000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4248586893081665,
-      "eval_runtime": 79.0364,
-      "eval_samples_per_second": 80.975,
-      "eval_steps_per_second": 0.633,
-      "step": 1510000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.403632482966014e-05,
-      "loss": 0.4428,
-      "step": 1511000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.394424975350325e-05,
-      "loss": 0.444,
-      "step": 1512000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.385231497114882e-05,
-      "loss": 0.4432,
-      "step": 1513000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.376061248654489e-05,
-      "loss": 0.4421,
-      "step": 1514000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.366895899052203e-05,
-      "loss": 0.4443,
-      "step": 1515000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.42498764395713806,
-      "eval_runtime": 78.5642,
-      "eval_samples_per_second": 81.462,
-      "eval_steps_per_second": 0.636,
-      "step": 1515000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.357744662688502e-05,
-      "loss": 0.4423,
-      "step": 1516000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.3486166974621315e-05,
-      "loss": 0.4445,
-      "step": 1517000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.3395028728323046e-05,
-      "loss": 0.4426,
-      "step": 1518000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.330394114848308e-05,
-      "loss": 0.4414,
-      "step": 1519000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.321308668712888e-05,
-      "loss": 0.4425,
-      "step": 1520000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4216972291469574,
-      "eval_runtime": 78.7807,
-      "eval_samples_per_second": 81.238,
-      "eval_steps_per_second": 0.635,
-      "step": 1520000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.31222837302516e-05,
-      "loss": 0.4416,
-      "step": 1521000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.303162357102785e-05,
-      "loss": 0.4417,
-      "step": 1522000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.294119693115241e-05,
-      "loss": 0.4432,
-      "step": 1523000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.285082305181241e-05,
-      "loss": 0.4418,
-      "step": 1524000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.276059279704224e-05,
-      "loss": 0.4425,
-      "step": 1525000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4240274727344513,
-      "eval_runtime": 83.7734,
-      "eval_samples_per_second": 76.397,
-      "eval_steps_per_second": 0.597,
-      "step": 1525000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.267059645608504e-05,
-      "loss": 0.4433,
-      "step": 1526000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.258065413027638e-05,
-      "loss": 0.4423,
-      "step": 1527000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.249085625201892e-05,
-      "loss": 0.4416,
-      "step": 1528000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.240120309485257e-05,
-      "loss": 0.4424,
-      "step": 1529000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 3.2311694931876376e-05,
-      "loss": 0.4435,
-      "step": 1530000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4211861491203308,
-      "eval_runtime": 78.8009,
-      "eval_samples_per_second": 81.217,
-      "eval_steps_per_second": 0.635,
-      "step": 1530000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.222242132599248e-05,
-      "loss": 0.4433,
-      "step": 1531000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.21332038232514e-05,
-      "loss": 0.4409,
-      "step": 1532000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.2044221129842254e-05,
-      "loss": 0.4422,
-      "step": 1533000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.195529537333662e-05,
-      "loss": 0.442,
-      "step": 1534000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.186651596933338e-05,
-      "loss": 0.4432,
-      "step": 1535000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.42593932151794434,
-      "eval_runtime": 82.4195,
-      "eval_samples_per_second": 77.652,
-      "eval_steps_per_second": 0.607,
-      "step": 1535000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.177788318827007e-05,
-      "loss": 0.4427,
-      "step": 1536000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.168948571256281e-05,
-      "loss": 0.4431,
-      "step": 1537000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.1601146839607305e-05,
-      "loss": 0.4416,
-      "step": 1538000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.1512955397952055e-05,
-      "loss": 0.4426,
-      "step": 1539000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.142491165624354e-05,
-      "loss": 0.442,
-      "step": 1540000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.4219116270542145,
-      "eval_runtime": 79.7973,
-      "eval_samples_per_second": 80.203,
-      "eval_steps_per_second": 0.627,
-      "step": 1540000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.133719152637512e-05,
-      "loss": 0.4413,
-      "step": 1541000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.1249443691960104e-05,
-      "loss": 0.4426,
-      "step": 1542000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.1161844360194124e-05,
-      "loss": 0.4431,
-      "step": 1543000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.107448117408321e-05,
-      "loss": 0.4427,
-      "step": 1544000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.098717949852167e-05,
-      "loss": 0.4434,
-      "step": 1545000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.42146673798561096,
-      "eval_runtime": 79.3482,
-      "eval_samples_per_second": 80.657,
-      "eval_steps_per_second": 0.63,
-      "step": 1545000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.090002712451138e-05,
-      "loss": 0.4422,
-      "step": 1546000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.081311124554366e-05,
-      "loss": 0.4416,
-      "step": 1547000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.072625812065977e-05,
-      "loss": 0.4424,
-      "step": 1548000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3.063955509213895e-05,
-      "loss": 0.4428,
-      "step": 1549000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.055317537919465e-05,
-      "loss": 0.4437,
-      "step": 1550000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.42286813259124756,
-      "eval_runtime": 78.4047,
-      "eval_samples_per_second": 81.628,
-      "eval_steps_per_second": 0.638,
-      "step": 1550000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.04667730337682e-05,
-      "loss": 0.4412,
-      "step": 1551000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.0380521575141337e-05,
-      "loss": 0.4416,
-      "step": 1552000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.029442126605103e-05,
-      "loss": 0.4435,
-      "step": 1553000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.02084723687737e-05,
-      "loss": 0.4427,
-      "step": 1554000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.0122760866500258e-05,
-      "loss": 0.442,
-      "step": 1555000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4226168096065521,
-      "eval_runtime": 78.4147,
-      "eval_samples_per_second": 81.617,
-      "eval_steps_per_second": 0.638,
-      "step": 1555000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.003711542576724e-05,
-      "loss": 0.4425,
-      "step": 1556000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9951793015071306e-05,
-      "loss": 0.4418,
-      "step": 1557000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.986645192081631e-05,
-      "loss": 0.4432,
-      "step": 1558000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.978126354204314e-05,
-      "loss": 0.4428,
-      "step": 1559000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.969631309715692e-05,
-      "loss": 0.4426,
-      "step": 1560000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4223879873752594,
-      "eval_runtime": 80.6588,
-      "eval_samples_per_second": 79.347,
-      "eval_steps_per_second": 0.62,
-      "step": 1560000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9611430774013924e-05,
-      "loss": 0.4416,
-      "step": 1561000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9526701943191416e-05,
-      "loss": 0.4434,
-      "step": 1562000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9442126862788104e-05,
-      "loss": 0.4407,
-      "step": 1563000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9357705790434323e-05,
-      "loss": 0.4412,
-      "step": 1564000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.927352317295752e-05,
-      "loss": 0.4415,
-      "step": 1565000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4247082471847534,
-      "eval_runtime": 78.1913,
-      "eval_samples_per_second": 81.851,
-      "eval_steps_per_second": 0.639,
-      "step": 1565000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9189410733066673e-05,
-      "loss": 0.443,
-      "step": 1566000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.9105453071042536e-05,
-      "loss": 0.4409,
-      "step": 1567000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.902165044263471e-05,
-      "loss": 0.4409,
-      "step": 1568000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.893817024265179e-05,
-      "loss": 0.4414,
-      "step": 1569000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.885467813549419e-05,
-      "loss": 0.4429,
-      "step": 1570000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4239417612552643,
-      "eval_runtime": 77.6323,
-      "eval_samples_per_second": 82.44,
-      "eval_steps_per_second": 0.644,
-      "step": 1570000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.877134182585685e-05,
-      "loss": 0.4421,
-      "step": 1571000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.8688161567596623e-05,
-      "loss": 0.4423,
-      "step": 1572000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.8605220559890042e-05,
-      "loss": 0.4417,
-      "step": 1573000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.8522353007368517e-05,
-      "loss": 0.4417,
-      "step": 1574000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.843972489701952e-05,
-      "loss": 0.4405,
-      "step": 1575000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.42241203784942627,
-      "eval_runtime": 79.4533,
-      "eval_samples_per_second": 80.55,
-      "eval_steps_per_second": 0.629,
-      "step": 1575000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.8357171058942266e-05,
-      "loss": 0.4405,
-      "step": 1576000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.82747745338781e-05,
-      "loss": 0.4415,
-      "step": 1577000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.819269989332776e-05,
-      "loss": 0.4408,
-      "step": 1578000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.8110618430913386e-05,
-      "loss": 0.4413,
-      "step": 1579000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.802869503255417e-05,
-      "loss": 0.4422,
-      "step": 1580000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.42269888520240784,
-      "eval_runtime": 80.0177,
-      "eval_samples_per_second": 79.982,
-      "eval_steps_per_second": 0.625,
-      "step": 1580000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7946929947802964e-05,
-      "loss": 0.4411,
-      "step": 1581000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7865323425730444e-05,
-      "loss": 0.4409,
-      "step": 1582000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7783875714924202e-05,
-      "loss": 0.4419,
-      "step": 1583000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7702587063488073e-05,
-      "loss": 0.4409,
-      "step": 1584000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7621619818577606e-05,
-      "loss": 0.4429,
-      "step": 1585000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4247879087924957,
-      "eval_runtime": 79.4063,
-      "eval_samples_per_second": 80.598,
-      "eval_steps_per_second": 0.63,
-      "step": 1585000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7540649708899915e-05,
-      "loss": 0.4414,
-      "step": 1586000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7459839399500925e-05,
-      "loss": 0.4403,
-      "step": 1587000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7379189136542904e-05,
-      "loss": 0.4419,
-      "step": 1588000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7298699165700516e-05,
-      "loss": 0.4409,
-      "step": 1589000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.7218530230648304e-05,
-      "loss": 0.4421,
-      "step": 1590000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4219193756580353,
-      "eval_runtime": 77.0263,
-      "eval_samples_per_second": 83.089,
-      "eval_steps_per_second": 0.649,
-      "step": 1590000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.713836125729949e-05,
-      "loss": 0.4432,
-      "step": 1591000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.7058353309669814e-05,
-      "loss": 0.4422,
-      "step": 1592000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6978506631477342e-05,
-      "loss": 0.4413,
-      "step": 1593000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6898901070358037e-05,
-      "loss": 0.4409,
-      "step": 1594000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6819456941046587e-05,
-      "loss": 0.4412,
-      "step": 1595000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4237200617790222,
-      "eval_runtime": 79.3809,
-      "eval_samples_per_second": 80.624,
-      "eval_steps_per_second": 0.63,
-      "step": 1595000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6740095204321262e-05,
-      "loss": 0.4407,
-      "step": 1596000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6660895706502525e-05,
-      "loss": 0.4408,
-      "step": 1597000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.658185868884578e-05,
-      "loss": 0.4415,
-      "step": 1598000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.65030631850491e-05,
-      "loss": 0.4423,
-      "step": 1599000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.642435168642101e-05,
-      "loss": 0.4398,
-      "step": 1600000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.42158567905426025,
-      "eval_runtime": 77.7364,
-      "eval_samples_per_second": 82.329,
-      "eval_steps_per_second": 0.643,
-      "step": 1600000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6345803388508994e-05,
-      "loss": 0.4418,
-      "step": 1601000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6267418530584848e-05,
-      "loss": 0.4409,
-      "step": 1602000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.6189275490764684e-05,
-      "loss": 0.4408,
-      "step": 1603000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.611121806460332e-05,
-      "loss": 0.4402,
-      "step": 1604000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.603332479301739e-05,
-      "loss": 0.4404,
-      "step": 1605000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4226343035697937,
-      "eval_runtime": 78.8772,
-      "eval_samples_per_second": 81.139,
-      "eval_steps_per_second": 0.634,
-      "step": 1605000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.5955751206822222e-05,
-      "loss": 0.4412,
-      "step": 1606000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.5878186626222375e-05,
-      "loss": 0.4411,
-      "step": 1607000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.580078691005209e-05,
-      "loss": 0.4417,
-      "step": 1608000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2.5723629446154437e-05,
-      "loss": 0.4399,
-      "step": 1609000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.5646636986989314e-05,
-      "loss": 0.4403,
-      "step": 1610000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.42356759309768677,
-      "eval_runtime": 78.3383,
-      "eval_samples_per_second": 81.697,
-      "eval_steps_per_second": 0.638,
-      "step": 1610000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.556973294535873e-05,
-      "loss": 0.4407,
-      "step": 1611000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.5492994707761112e-05,
-      "loss": 0.4413,
-      "step": 1612000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.5416422507954494e-05,
-      "loss": 0.4402,
-      "step": 1613000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.5340016579191055e-05,
-      "loss": 0.4407,
-      "step": 1614000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.526385331039548e-05,
-      "loss": 0.4407,
-      "step": 1615000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4227603077888489,
-      "eval_runtime": 78.8019,
-      "eval_samples_per_second": 81.216,
-      "eval_steps_per_second": 0.635,
-      "step": 1615000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.518778045459646e-05,
-      "loss": 0.4418,
-      "step": 1616000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.5111874566323936e-05,
-      "loss": 0.4402,
-      "step": 1617000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.503621153189737e-05,
-      "loss": 0.4411,
-      "step": 1618000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4960640104291568e-05,
-      "loss": 0.4405,
-      "step": 1619000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4885236336121185e-05,
-      "loss": 0.4412,
-      "step": 1620000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4221682846546173,
-      "eval_runtime": 79.6082,
-      "eval_samples_per_second": 80.394,
-      "eval_steps_per_second": 0.628,
-      "step": 1620000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.481000045707917e-05,
-      "loss": 0.4392,
-      "step": 1621000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4735007680056548e-05,
-      "loss": 0.4411,
-      "step": 1622000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4660182913259782e-05,
-      "loss": 0.4397,
-      "step": 1623000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.458545173726531e-05,
-      "loss": 0.4421,
-      "step": 1624000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4510963841576386e-05,
-      "loss": 0.4412,
-      "step": 1625000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.41993048787117004,
-      "eval_runtime": 77.9039,
-      "eval_samples_per_second": 82.152,
-      "eval_steps_per_second": 0.642,
-      "step": 1625000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4436570328217597e-05,
-      "loss": 0.4407,
-      "step": 1626000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4362346070701672e-05,
-      "loss": 0.4406,
-      "step": 1627000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4288365265172765e-05,
-      "loss": 0.4405,
-      "step": 1628000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.421448002730611e-05,
-      "loss": 0.4401,
-      "step": 1629000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.4140838352157365e-05,
-      "loss": 0.4405,
-      "step": 1630000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4204855263233185,
-      "eval_runtime": 79.4366,
-      "eval_samples_per_second": 80.567,
-      "eval_steps_per_second": 0.629,
-      "step": 1630000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.4067293033308242e-05,
-      "loss": 0.4397,
-      "step": 1631000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3993918095184646e-05,
-      "loss": 0.4405,
-      "step": 1632000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.392078688034217e-05,
-      "loss": 0.4397,
-      "step": 1633000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.384782615102526e-05,
-      "loss": 0.4397,
-      "step": 1634000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3774963351754417e-05,
-      "loss": 0.4392,
-      "step": 1635000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4220183193683624,
-      "eval_runtime": 78.5206,
-      "eval_samples_per_second": 81.507,
-      "eval_steps_per_second": 0.637,
-      "step": 1635000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3702344429600395e-05,
-      "loss": 0.4404,
-      "step": 1636000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3629824222582374e-05,
-      "loss": 0.4405,
-      "step": 1637000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3557475728895742e-05,
-      "loss": 0.4406,
-      "step": 1638000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3485299168926543e-05,
-      "loss": 0.4399,
-      "step": 1639000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.341336668087976e-05,
-      "loss": 0.439,
-      "step": 1640000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.42630383372306824,
-      "eval_runtime": 79.5813,
-      "eval_samples_per_second": 80.421,
-      "eval_steps_per_second": 0.628,
-      "step": 1640000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3341534474925735e-05,
-      "loss": 0.4398,
-      "step": 1641000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3269874860483666e-05,
-      "loss": 0.4406,
-      "step": 1642000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.319838805584115e-05,
-      "loss": 0.441,
-      "step": 1643000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3127074278759437e-05,
-      "loss": 0.4402,
-      "step": 1644000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.3056004800396985e-05,
-      "loss": 0.4394,
-      "step": 1645000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.42043250799179077,
-      "eval_runtime": 77.5183,
-      "eval_samples_per_second": 82.561,
-      "eval_steps_per_second": 0.645,
-      "step": 1645000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.298510843657037e-05,
-      "loss": 0.4404,
-      "step": 1646000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.2914385402810497e-05,
-      "loss": 0.4403,
-      "step": 1647000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.2843765381030626e-05,
-      "loss": 0.4401,
-      "step": 1648000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.2773319467053702e-05,
-      "loss": 0.4388,
-      "step": 1649000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.2703047875470186e-05,
-      "loss": 0.4388,
-      "step": 1650000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4251132309436798,
-      "eval_runtime": 79.2673,
-      "eval_samples_per_second": 80.739,
-      "eval_steps_per_second": 0.631,
-      "step": 1650000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.263295082033955e-05,
-      "loss": 0.4402,
-      "step": 1651000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.256302851518958e-05,
-      "loss": 0.4408,
-      "step": 1652000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.2493281173015714e-05,
-      "loss": 0.4392,
-      "step": 1653000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.242370900628049e-05,
-      "loss": 0.4401,
-      "step": 1654000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.235438153601577e-05,
-      "loss": 0.4399,
-      "step": 1655000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.42048439383506775,
-      "eval_runtime": 77.8384,
-      "eval_samples_per_second": 82.222,
-      "eval_steps_per_second": 0.642,
-      "step": 1655000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.2285160179706007e-05,
-      "loss": 0.44,
-      "step": 1656000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.2216114632807524e-05,
-      "loss": 0.4404,
-      "step": 1657000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.214731388718044e-05,
-      "loss": 0.4406,
-      "step": 1658000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.2078620413208303e-05,
-      "loss": 0.4402,
-      "step": 1659000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.201010337780338e-05,
-      "loss": 0.4405,
-      "step": 1660000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.4228270649909973,
-      "eval_runtime": 78.2009,
-      "eval_samples_per_second": 81.841,
-      "eval_steps_per_second": 0.639,
-      "step": 1660000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.1941831241763897e-05,
-      "loss": 0.4402,
-      "step": 1661000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.1873667532140358e-05,
-      "loss": 0.439,
-      "step": 1662000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.1805748783540877e-05,
-      "loss": 0.44,
-      "step": 1663000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.1737939229421666e-05,
-      "loss": 0.4407,
-      "step": 1664000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.167037469500335e-05,
-      "loss": 0.4404,
-      "step": 1665000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.41908711194992065,
-      "eval_runtime": 78.8241,
-      "eval_samples_per_second": 81.193,
-      "eval_steps_per_second": 0.634,
-      "step": 1665000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.160292012180046e-05,
-      "loss": 0.4405,
-      "step": 1666000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.1535643436230335e-05,
-      "loss": 0.4401,
-      "step": 1667000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.146854484322948e-05,
-      "loss": 0.4403,
-      "step": 1668000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 2.140162454719184e-05,
-      "loss": 0.4418,
-      "step": 1669000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.1334882751968192e-05,
-      "loss": 0.4397,
-      "step": 1670000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.42122882604599,
-      "eval_runtime": 76.417,
-      "eval_samples_per_second": 83.751,
-      "eval_steps_per_second": 0.654,
-      "step": 1670000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.126838613462656e-05,
-      "loss": 0.4387,
-      "step": 1671000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.1202001771399895e-05,
-      "loss": 0.4387,
-      "step": 1672000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.1135796517072863e-05,
-      "loss": 0.4394,
-      "step": 1673000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.106977057331812e-05,
-      "loss": 0.4398,
-      "step": 1674000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.1003989897961326e-05,
-      "loss": 0.44,
-      "step": 1675000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.41976797580718994,
-      "eval_runtime": 77.5035,
-      "eval_samples_per_second": 82.577,
-      "eval_steps_per_second": 0.645,
-      "step": 1675000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0938388575438328e-05,
-      "loss": 0.4403,
-      "step": 1676000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0872901407947595e-05,
-      "loss": 0.4413,
-      "step": 1677000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.080759435185324e-05,
-      "loss": 0.4397,
-      "step": 1678000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0742467606091935e-05,
-      "loss": 0.4395,
-      "step": 1679000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0677586225058045e-05,
-      "loss": 0.4407,
-      "step": 1680000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.42079994082450867,
-      "eval_runtime": 79.6958,
-      "eval_samples_per_second": 80.305,
-      "eval_steps_per_second": 0.627,
-      "step": 1680000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0612885189152567e-05,
-      "loss": 0.4399,
-      "step": 1681000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0548300200510223e-05,
-      "loss": 0.4382,
-      "step": 1682000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.048389631205587e-05,
-      "loss": 0.4393,
-      "step": 1683000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.041967371997491e-05,
-      "loss": 0.4392,
-      "step": 1684000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0355760520841843e-05,
-      "loss": 0.4403,
-      "step": 1685000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.41910338401794434,
-      "eval_runtime": 79.2458,
-      "eval_samples_per_second": 80.761,
-      "eval_steps_per_second": 0.631,
-      "step": 1685000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0291900744285765e-05,
-      "loss": 0.4397,
-      "step": 1686000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.022822284895487e-05,
-      "loss": 0.4401,
-      "step": 1687000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.016472702882308e-05,
-      "loss": 0.4395,
-      "step": 1688000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 2.0101476699753774e-05,
-      "loss": 0.4394,
-      "step": 1689000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 2.003840846723428e-05,
-      "loss": 0.4408,
-      "step": 1690000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.41959914565086365,
-      "eval_runtime": 79.6028,
-      "eval_samples_per_second": 80.399,
-      "eval_steps_per_second": 0.628,
-      "step": 1690000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9975459665494844e-05,
-      "loss": 0.4406,
-      "step": 1691000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9912693708915007e-05,
-      "loss": 0.4403,
-      "step": 1692000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9850110788690757e-05,
-      "loss": 0.4391,
-      "step": 1693000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.978771109546051e-05,
-      "loss": 0.4388,
-      "step": 1694000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9725681193643978e-05,
-      "loss": 0.439,
-      "step": 1695000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4219348132610321,
-      "eval_runtime": 78.5115,
-      "eval_samples_per_second": 81.517,
-      "eval_steps_per_second": 0.637,
-      "step": 1695000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9663647972981225e-05,
-      "loss": 0.4389,
-      "step": 1696000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9601798547310563e-05,
-      "loss": 0.4396,
-      "step": 1697000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.954019467851605e-05,
-      "loss": 0.4405,
-      "step": 1698000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9478713223216454e-05,
-      "loss": 0.4403,
-      "step": 1699000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9417416126252245e-05,
-      "loss": 0.4394,
-      "step": 1700000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.42123520374298096,
-      "eval_runtime": 78.2147,
-      "eval_samples_per_second": 81.826,
-      "eval_steps_per_second": 0.639,
-      "step": 1700000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9356303574345033e-05,
-      "loss": 0.44,
-      "step": 1701000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9295375753654256e-05,
-      "loss": 0.4406,
-      "step": 1702000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9234693500252896e-05,
-      "loss": 0.4392,
-      "step": 1703000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9174195978495195e-05,
-      "loss": 0.4389,
-      "step": 1704000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9113823092023844e-05,
-      "loss": 0.4395,
-      "step": 1705000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.42103949189186096,
-      "eval_runtime": 79.7626,
-      "eval_samples_per_second": 80.238,
-      "eval_steps_per_second": 0.627,
-      "step": 1705000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.9053635675406527e-05,
-      "loss": 0.4387,
-      "step": 1706000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.899363391198454e-05,
-      "loss": 0.4405,
-      "step": 1707000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.893381798453365e-05,
-      "loss": 0.4395,
-      "step": 1708000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.887418807526355e-05,
-      "loss": 0.439,
-      "step": 1709000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.8814803716459616e-05,
-      "loss": 0.439,
-      "step": 1710000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.42225512862205505,
-      "eval_runtime": 77.1239,
-      "eval_samples_per_second": 82.983,
-      "eval_steps_per_second": 0.648,
-      "step": 1710000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.875560536579964e-05,
-      "loss": 0.439,
-      "step": 1711000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8696534225358292e-05,
-      "loss": 0.4392,
-      "step": 1712000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8637708616967782e-05,
-      "loss": 0.4397,
-      "step": 1713000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8579010950865316e-05,
-      "loss": 0.4385,
-      "step": 1714000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.852050038374741e-05,
-      "loss": 0.4391,
-      "step": 1715000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.420716792345047,
-      "eval_runtime": 78.5003,
-      "eval_samples_per_second": 81.528,
-      "eval_steps_per_second": 0.637,
-      "step": 1715000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8462235323533128e-05,
-      "loss": 0.4395,
-      "step": 1716000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8404099300970416e-05,
-      "loss": 0.4378,
-      "step": 1717000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8346208764813356e-05,
-      "loss": 0.4408,
-      "step": 1718000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8288447994466744e-05,
-      "loss": 0.4388,
-      "step": 1719000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8230932687039617e-05,
-      "loss": 0.439,
-      "step": 1720000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.41757285594940186,
-      "eval_runtime": 79.8473,
-      "eval_samples_per_second": 80.153,
-      "eval_steps_per_second": 0.626,
-      "step": 1720000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8173547872002242e-05,
-      "loss": 0.4384,
-      "step": 1721000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.811640849341029e-05,
-      "loss": 0.4401,
-      "step": 1722000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8059400332198968e-05,
-      "loss": 0.438,
-      "step": 1723000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.8002580852796262e-05,
-      "loss": 0.4401,
-      "step": 1724000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.7945950228284155e-05,
-      "loss": 0.4401,
-      "step": 1725000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.41903457045555115,
-      "eval_runtime": 77.4134,
-      "eval_samples_per_second": 82.673,
-      "eval_steps_per_second": 0.646,
-      "step": 1725000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.788950863116934e-05,
-      "loss": 0.4383,
-      "step": 1726000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.783331239121836e-05,
-      "loss": 0.4383,
-      "step": 1727000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.7777305143227536e-05,
-      "loss": 0.4401,
-      "step": 1728000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 1.772143127833117e-05,
-      "loss": 0.4391,
-      "step": 1729000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.766574712475575e-05,
-      "loss": 0.439,
-      "step": 1730000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4182310104370117,
-      "eval_runtime": 75.5121,
-      "eval_samples_per_second": 84.755,
-      "eval_steps_per_second": 0.662,
-      "step": 1730000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7610252852124898e-05,
-      "loss": 0.4387,
-      "step": 1731000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.755494862948377e-05,
-      "loss": 0.4382,
-      "step": 1732000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7499889644232756e-05,
-      "loss": 0.4385,
-      "step": 1733000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.744496583592041e-05,
-      "loss": 0.4408,
-      "step": 1734000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7390287219108524e-05,
-      "loss": 0.4401,
-      "step": 1735000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.4186602830886841,
-      "eval_runtime": 80.3977,
-      "eval_samples_per_second": 79.604,
-      "eval_steps_per_second": 0.622,
-      "step": 1735000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.733574449368513e-05,
-      "loss": 0.4391,
-      "step": 1736000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7281392654451555e-05,
-      "loss": 0.4401,
-      "step": 1737000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7227339997768454e-05,
-      "loss": 0.4405,
-      "step": 1738000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7173370044430122e-05,
-      "loss": 0.439,
-      "step": 1739000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7119591471902336e-05,
-      "loss": 0.4397,
-      "step": 1740000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.41898131370544434,
-      "eval_runtime": 77.8556,
-      "eval_samples_per_second": 82.203,
-      "eval_steps_per_second": 0.642,
-      "step": 1740000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7066004444003927e-05,
-      "loss": 0.4388,
-      "step": 1741000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.7012609123970294e-05,
-      "loss": 0.4388,
-      "step": 1742000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6959405674452816e-05,
-      "loss": 0.4393,
-      "step": 1743000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6906447172961322e-05,
-      "loss": 0.4386,
-      "step": 1744000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6853627757817506e-05,
-      "loss": 0.4379,
-      "step": 1745000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.41874217987060547,
-      "eval_runtime": 79.3439,
-      "eval_samples_per_second": 80.662,
-      "eval_steps_per_second": 0.63,
-      "step": 1745000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6801053228400387e-05,
-      "loss": 0.4407,
-      "step": 1746000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6748618490574697e-05,
-      "loss": 0.4398,
-      "step": 1747000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.669637642742642e-05,
-      "loss": 0.4385,
-      "step": 1748000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6644327198093962e-05,
-      "loss": 0.4376,
-      "step": 1749000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 1.6592522720912954e-05,
-      "loss": 0.4381,
-      "step": 1750000
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 0.42223626375198364,
-      "eval_runtime": 79.3599,
-      "eval_samples_per_second": 80.645,
-      "eval_steps_per_second": 0.63,
-      "step": 1750000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6540859441048118e-05,
-      "loss": 0.4394,
-      "step": 1751000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6489389468730806e-05,
-      "loss": 0.439,
-      "step": 1752000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6438215320582125e-05,
-      "loss": 0.4382,
-      "step": 1753000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.638713204573334e-05,
-      "loss": 0.4381,
-      "step": 1754000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6336293339368757e-05,
-      "loss": 0.4392,
-      "step": 1755000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.420003205537796,
-      "eval_runtime": 78.6855,
-      "eval_samples_per_second": 81.336,
-      "eval_steps_per_second": 0.635,
-      "step": 1755000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.628559757718579e-05,
-      "loss": 0.439,
-      "step": 1756000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6235095900122255e-05,
-      "loss": 0.4383,
-      "step": 1757000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.61848386723797e-05,
-      "loss": 0.4386,
-      "step": 1758000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6134725432005385e-05,
-      "loss": 0.4395,
-      "step": 1759000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6084856557803128e-05,
-      "loss": 0.438,
-      "step": 1760000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.4174318015575409,
-      "eval_runtime": 76.8661,
-      "eval_samples_per_second": 83.262,
-      "eval_steps_per_second": 0.65,
-      "step": 1760000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.6035132364121584e-05,
-      "loss": 0.4388,
-      "step": 1761000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5985603018519935e-05,
-      "loss": 0.4378,
-      "step": 1762000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5936317908767756e-05,
-      "loss": 0.4368,
-      "step": 1763000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5887178516132736e-05,
-      "loss": 0.4412,
-      "step": 1764000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5838283268763148e-05,
-      "loss": 0.4383,
-      "step": 1765000
-    },
-    {
-      "epoch": 0.13,
-      "eval_loss": 0.41864004731178284,
-      "eval_runtime": 86.8232,
-      "eval_samples_per_second": 73.713,
-      "eval_steps_per_second": 0.576,
-      "step": 1765000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5789583078410045e-05,
-      "loss": 0.4389,
-      "step": 1766000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.574102963743466e-05,
-      "loss": 0.439,
-      "step": 1767000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5692671940427092e-05,
-      "loss": 0.4385,
-      "step": 1768000
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 1.5644510134693248e-05,
-      "loss": 0.4384,
-      "step": 1769000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.559654436694238e-05,
-      "loss": 0.4392,
-      "step": 1770000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4172964096069336,
-      "eval_runtime": 79.9939,
-      "eval_samples_per_second": 80.006,
-      "eval_steps_per_second": 0.625,
-      "step": 1770000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5548822454827717e-05,
-      "loss": 0.4393,
-      "step": 1771000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5501249004379188e-05,
-      "loss": 0.4392,
-      "step": 1772000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.54539193071009e-05,
-      "loss": 0.4386,
-      "step": 1773000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5406738753042658e-05,
-      "loss": 0.4393,
-      "step": 1774000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5359754961260252e-05,
-      "loss": 0.4387,
-      "step": 1775000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4169865548610687,
-      "eval_runtime": 77.7982,
-      "eval_samples_per_second": 82.264,
-      "eval_steps_per_second": 0.643,
-      "step": 1775000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5312968074874446e-05,
-      "loss": 0.4381,
-      "step": 1776000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5266424727771944e-05,
-      "loss": 0.4396,
-      "step": 1777000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.522003188188146e-05,
-      "loss": 0.4395,
-      "step": 1778000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.517383636700831e-05,
-      "loss": 0.4382,
-      "step": 1779000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5127838323872036e-05,
-      "loss": 0.4364,
-      "step": 1780000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.41849958896636963,
-      "eval_runtime": 76.5418,
-      "eval_samples_per_second": 83.614,
-      "eval_steps_per_second": 0.653,
-      "step": 1780000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5082037892590664e-05,
-      "loss": 0.439,
-      "step": 1781000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.5036480716537045e-05,
-      "loss": 0.4393,
-      "step": 1782000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4991121035047137e-05,
-      "loss": 0.4383,
-      "step": 1783000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4945913877821996e-05,
-      "loss": 0.4383,
-      "step": 1784000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4900904886625165e-05,
-      "loss": 0.4377,
-      "step": 1785000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4204372465610504,
-      "eval_runtime": 76.9476,
-      "eval_samples_per_second": 83.173,
-      "eval_steps_per_second": 0.65,
-      "step": 1785000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4856138910151988e-05,
-      "loss": 0.4388,
-      "step": 1786000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4811526463215664e-05,
-      "loss": 0.4371,
-      "step": 1787000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.476715690631307e-05,
-      "loss": 0.438,
-      "step": 1788000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4722941546682392e-05,
-      "loss": 0.4381,
-      "step": 1789000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 1.4678968949438921e-05,
-      "loss": 0.4363,
-      "step": 1790000
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.4183988869190216,
-      "eval_runtime": 76.9826,
-      "eval_samples_per_second": 83.136,
-      "eval_steps_per_second": 0.649,
-      "step": 1790000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4635151215325466e-05,
-      "loss": 0.4366,
-      "step": 1791000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4591576112997706e-05,
-      "loss": 0.4391,
-      "step": 1792000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4548156537772989e-05,
-      "loss": 0.4391,
-      "step": 1793000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4504936340214418e-05,
-      "loss": 0.4385,
-      "step": 1794000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4461958572967858e-05,
-      "loss": 0.4378,
-      "step": 1795000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.4223540425300598,
-      "eval_runtime": 77.2417,
-      "eval_samples_per_second": 82.857,
-      "eval_steps_per_second": 0.647,
-      "step": 1795000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4419137325396865e-05,
-      "loss": 0.4389,
-      "step": 1796000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.437651584850691e-05,
-      "loss": 0.4386,
-      "step": 1797000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4334094272130413e-05,
-      "loss": 0.4367,
-      "step": 1798000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4291872725490842e-05,
-      "loss": 0.4384,
-      "step": 1799000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 1.4249893258568889e-05,
-      "loss": 0.4384,
-      "step": 1800000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.4171189069747925,
-      "eval_runtime": 76.9772,
-      "eval_samples_per_second": 83.142,
-      "eval_steps_per_second": 0.65,
-      "step": 1800000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4208113677502687e-05,
-      "loss": 0.4365,
-      "step": 1801000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4166492588365344e-05,
-      "loss": 0.4384,
-      "step": 1802000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4125072039508715e-05,
-      "loss": 0.4379,
-      "step": 1803000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4083852157106983e-05,
-      "loss": 0.4377,
-      "step": 1804000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4042833066723076e-05,
-      "loss": 0.4385,
-      "step": 1805000
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 0.42015281319618225,
-      "eval_runtime": 78.5938,
-      "eval_samples_per_second": 81.431,
-      "eval_steps_per_second": 0.636,
-      "step": 1805000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.4002055611082185e-05,
-      "loss": 0.4387,
-      "step": 1806000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.396143827787245e-05,
-      "loss": 0.4379,
-      "step": 1807000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.3921022109574423e-05,
-      "loss": 0.4373,
-      "step": 1808000
-    },
-    {
-      "epoch": 0.0,
-      "learning_rate": 1.3880847343598854e-05,
-      "loss": 0.4382,
-      "step": 1809000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.384087358540966e-05,
-      "loss": 0.438,
-      "step": 1810000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4181618392467499,
-      "eval_runtime": 77.0873,
-      "eval_samples_per_second": 83.023,
-      "eval_steps_per_second": 0.649,
-      "step": 1810000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3801061244895656e-05,
-      "loss": 0.4382,
-      "step": 1811000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3761450557829634e-05,
-      "loss": 0.4392,
-      "step": 1812000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.372204164487259e-05,
-      "loss": 0.4387,
-      "step": 1813000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.368283462607094e-05,
-      "loss": 0.4388,
-      "step": 1814000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3643868524915881e-05,
-      "loss": 0.4392,
-      "step": 1815000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.42043235898017883,
-      "eval_runtime": 79.2626,
-      "eval_samples_per_second": 80.744,
-      "eval_steps_per_second": 0.631,
-      "step": 1815000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3605065449912204e-05,
-      "loss": 0.4395,
-      "step": 1816000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3566464625393676e-05,
-      "loss": 0.4391,
-      "step": 1817000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.352810446627972e-05,
-      "loss": 0.4379,
-      "step": 1818000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3489908292326226e-05,
-      "loss": 0.4377,
-      "step": 1819000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3451952611981318e-05,
-      "loss": 0.4389,
-      "step": 1820000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.41748473048210144,
-      "eval_runtime": 77.7669,
-      "eval_samples_per_second": 82.297,
-      "eval_steps_per_second": 0.643,
-      "step": 1820000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3414161553535873e-05,
-      "loss": 0.4386,
-      "step": 1821000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3376573327101957e-05,
-      "loss": 0.4383,
-      "step": 1822000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.333918804717982e-05,
-      "loss": 0.4371,
-      "step": 1823000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3302079989360922e-05,
-      "loss": 0.4369,
-      "step": 1824000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3265100537030001e-05,
-      "loss": 0.4378,
-      "step": 1825000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4193136692047119,
-      "eval_runtime": 79.079,
-      "eval_samples_per_second": 80.932,
-      "eval_steps_per_second": 0.632,
-      "step": 1825000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3228324370776315e-05,
-      "loss": 0.4385,
-      "step": 1826000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.319175160262646e-05,
-      "loss": 0.4363,
-      "step": 1827000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3155418611556128e-05,
-      "loss": 0.438,
-      "step": 1828000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3119252769539538e-05,
-      "loss": 0.4378,
-      "step": 1829000
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.3083326518189592e-05,
-      "loss": 0.4377,
-      "step": 1830000
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 0.4179893732070923,
-      "eval_runtime": 78.6845,
-      "eval_samples_per_second": 81.337,
-      "eval_steps_per_second": 0.635,
-      "step": 1830000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.3047568042535075e-05,
-      "loss": 0.4388,
-      "step": 1831000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.3012013515599501e-05,
-      "loss": 0.439,
-      "step": 1832000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2976698294195656e-05,
-      "loss": 0.4392,
-      "step": 1833000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2941586829267356e-05,
-      "loss": 0.4378,
-      "step": 1834000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2906644387183456e-05,
-      "loss": 0.4372,
-      "step": 1835000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.4213528037071228,
-      "eval_runtime": 77.6423,
-      "eval_samples_per_second": 82.429,
-      "eval_steps_per_second": 0.644,
-      "step": 1835000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.287194095903841e-05,
-      "loss": 0.4367,
-      "step": 1836000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2837407174229876e-05,
-      "loss": 0.437,
-      "step": 1837000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2803077978326747e-05,
-      "loss": 0.4377,
-      "step": 1838000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2768953475901701e-05,
-      "loss": 0.4383,
-      "step": 1839000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2735101405857255e-05,
-      "loss": 0.4379,
-      "step": 1840000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.41641688346862793,
-      "eval_runtime": 79.1386,
-      "eval_samples_per_second": 80.871,
-      "eval_steps_per_second": 0.632,
-      "step": 1840000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2701386191707756e-05,
-      "loss": 0.4379,
-      "step": 1841000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2667875980807157e-05,
-      "loss": 0.4384,
-      "step": 1842000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2634570875233356e-05,
-      "loss": 0.4379,
-      "step": 1843000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2601470976439498e-05,
-      "loss": 0.4368,
-      "step": 1844000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2568576385253613e-05,
-      "loss": 0.4379,
-      "step": 1845000
-    },
-    {
-      "epoch": 0.02,
-      "eval_loss": 0.41581401228904724,
-      "eval_runtime": 81.6085,
-      "eval_samples_per_second": 78.423,
-      "eval_steps_per_second": 0.613,
-      "step": 1845000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2535919788427315e-05,
-      "loss": 0.4365,
-      "step": 1846000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2503435906882624e-05,
-      "loss": 0.4374,
-      "step": 1847000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.247115763157773e-05,
-      "loss": 0.4381,
-      "step": 1848000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.2439117030626584e-05,
-      "loss": 0.4368,
-      "step": 1849000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2407250056299487e-05,
-      "loss": 0.4383,
-      "step": 1850000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4171139597892761,
-      "eval_runtime": 78.6968,
-      "eval_samples_per_second": 81.325,
-      "eval_steps_per_second": 0.635,
-      "step": 1850000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2375588981210953e-05,
-      "loss": 0.4373,
-      "step": 1851000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2344133901806119e-05,
-      "loss": 0.4373,
-      "step": 1852000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2312916059916086e-05,
-      "loss": 0.4373,
-      "step": 1853000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2281873052469721e-05,
-      "loss": 0.4382,
-      "step": 1854000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.225106705983955e-05,
-      "loss": 0.4364,
-      "step": 1855000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4179363548755646,
-      "eval_runtime": 79.5482,
-      "eval_samples_per_second": 80.454,
-      "eval_steps_per_second": 0.629,
-      "step": 1855000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2220436502222735e-05,
-      "loss": 0.4369,
-      "step": 1856000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2190012412910979e-05,
-      "loss": 0.437,
-      "step": 1857000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2159794884581337e-05,
-      "loss": 0.4387,
-      "step": 1858000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.212981391690326e-05,
-      "loss": 0.437,
-      "step": 1859000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.21000392803005e-05,
-      "loss": 0.4385,
-      "step": 1860000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.41541191935539246,
-      "eval_runtime": 80.5191,
-      "eval_samples_per_second": 79.484,
-      "eval_steps_per_second": 0.621,
-      "step": 1860000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2070471065290428e-05,
-      "loss": 0.4373,
-      "step": 1861000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2041080074160751e-05,
-      "loss": 0.4387,
-      "step": 1862000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2011896097686919e-05,
-      "loss": 0.4385,
-      "step": 1863000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.1982948098163646e-05,
-      "loss": 0.4385,
-      "step": 1864000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.1954178209833276e-05,
-      "loss": 0.4375,
-      "step": 1865000
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 0.4207480549812317,
-      "eval_runtime": 78.9778,
-      "eval_samples_per_second": 81.035,
-      "eval_steps_per_second": 0.633,
-      "step": 1865000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.1925615600876905e-05,
-      "loss": 0.4376,
-      "step": 1866000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.1897260358301189e-05,
-      "loss": 0.4372,
-      "step": 1867000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.1869112568481112e-05,
-      "loss": 0.4371,
-      "step": 1868000
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.1841172317159714e-05,
-      "loss": 0.437,
-      "step": 1869000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1813494947438669e-05,
-      "loss": 0.4375,
-      "step": 1870000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4189639985561371,
-      "eval_runtime": 79.747,
-      "eval_samples_per_second": 80.254,
-      "eval_steps_per_second": 0.627,
-      "step": 1870000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1785969612314684e-05,
-      "loss": 0.4374,
-      "step": 1871000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.175865206895725e-05,
-      "loss": 0.4369,
-      "step": 1872000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1731542400580344e-05,
-      "loss": 0.4389,
-      "step": 1873000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.17046674875734e-05,
-      "loss": 0.4373,
-      "step": 1874000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1677973608186099e-05,
-      "loss": 0.437,
-      "step": 1875000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.41854363679885864,
-      "eval_runtime": 77.0323,
-      "eval_samples_per_second": 83.082,
-      "eval_steps_per_second": 0.649,
-      "step": 1875000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1651514231315332e-05,
-      "loss": 0.4382,
-      "step": 1876000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1625262639591158e-05,
-      "loss": 0.4386,
-      "step": 1877000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1599218912820604e-05,
-      "loss": 0.4363,
-      "step": 1878000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1573357372741149e-05,
-      "loss": 0.4384,
-      "step": 1879000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1547729821041103e-05,
-      "loss": 0.437,
-      "step": 1880000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.4172144830226898,
-      "eval_runtime": 79.3075,
-      "eval_samples_per_second": 80.699,
-      "eval_steps_per_second": 0.63,
-      "step": 1880000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1522285029389667e-05,
-      "loss": 0.4374,
-      "step": 1881000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.149704883273374e-05,
-      "loss": 0.4381,
-      "step": 1882000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1472021307947182e-05,
-      "loss": 0.4369,
-      "step": 1883000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1447227245749926e-05,
-      "loss": 0.4371,
-      "step": 1884000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1422617083919464e-05,
-      "loss": 0.4361,
-      "step": 1885000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.41780930757522583,
-      "eval_runtime": 80.7871,
-      "eval_samples_per_second": 79.221,
-      "eval_steps_per_second": 0.619,
-      "step": 1885000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1398215820690375e-05,
-      "loss": 0.4367,
-      "step": 1886000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1374023530393122e-05,
-      "loss": 0.4376,
-      "step": 1887000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1350040286721638e-05,
-      "loss": 0.4385,
-      "step": 1888000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1326289832377572e-05,
-      "loss": 0.437,
-      "step": 1889000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.1302724691263888e-05,
-      "loss": 0.437,
-      "step": 1890000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.41528433561325073,
-      "eval_runtime": 78.5162,
-      "eval_samples_per_second": 81.512,
-      "eval_steps_per_second": 0.637,
-      "step": 1890000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1279368813964581e-05,
-      "loss": 0.4374,
-      "step": 1891000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.125624531358172e-05,
-      "loss": 0.4375,
-      "step": 1892000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1233307967271214e-05,
-      "loss": 0.4375,
-      "step": 1893000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1210625342870107e-05,
-      "loss": 0.4372,
-      "step": 1894000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1188106597174481e-05,
-      "loss": 0.4367,
-      "step": 1895000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4200913608074188,
-      "eval_runtime": 79.1042,
-      "eval_samples_per_second": 80.906,
-      "eval_steps_per_second": 0.632,
-      "step": 1895000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1165797464439845e-05,
-      "loss": 0.4384,
-      "step": 1896000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1143720007317412e-05,
-      "loss": 0.4366,
-      "step": 1897000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1121830093956824e-05,
-      "loss": 0.4366,
-      "step": 1898000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1100149995446934e-05,
-      "loss": 0.4371,
-      "step": 1899000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1078679777829087e-05,
-      "loss": 0.4384,
-      "step": 1900000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.41656479239463806,
-      "eval_runtime": 80.5362,
-      "eval_samples_per_second": 79.467,
-      "eval_steps_per_second": 0.621,
-      "step": 1900000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.105744066188684e-05,
-      "loss": 0.4371,
-      "step": 1901000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1036390191576373e-05,
-      "loss": 0.4379,
-      "step": 1902000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.1015549796381372e-05,
-      "loss": 0.4373,
-      "step": 1903000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.0994960590538279e-05,
-      "loss": 0.4375,
-      "step": 1904000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.0974540114919287e-05,
-      "loss": 0.4363,
-      "step": 1905000
-    },
-    {
-      "epoch": 0.05,
-      "eval_loss": 0.4150693416595459,
-      "eval_runtime": 80.8571,
-      "eval_samples_per_second": 79.152,
-      "eval_steps_per_second": 0.618,
-      "step": 1905000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.0954329902821809e-05,
-      "loss": 0.4375,
-      "step": 1906000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.0934330015809674e-05,
-      "loss": 0.437,
-      "step": 1907000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.0914560199199067e-05,
-      "loss": 0.4379,
-      "step": 1908000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.0894980934009906e-05,
-      "loss": 0.4372,
-      "step": 1909000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0875612174693328e-05,
-      "loss": 0.437,
-      "step": 1910000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4164562225341797,
-      "eval_runtime": 79.4864,
-      "eval_samples_per_second": 80.517,
-      "eval_steps_per_second": 0.629,
-      "step": 1910000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0856473033247752e-05,
-      "loss": 0.4369,
-      "step": 1911000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0837525251384567e-05,
-      "loss": 0.4367,
-      "step": 1912000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0818806782260748e-05,
-      "loss": 0.4367,
-      "step": 1913000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0800280208492865e-05,
-      "loss": 0.4379,
-      "step": 1914000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0781982639541429e-05,
-      "loss": 0.4373,
-      "step": 1915000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4175663888454437,
-      "eval_runtime": 80.7249,
-      "eval_samples_per_second": 79.282,
-      "eval_steps_per_second": 0.619,
-      "step": 1915000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0763895499185767e-05,
-      "loss": 0.4362,
-      "step": 1916000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0746001053331784e-05,
-      "loss": 0.4367,
-      "step": 1917000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0728317567168942e-05,
-      "loss": 0.4373,
-      "step": 1918000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0710845094564199e-05,
-      "loss": 0.4377,
-      "step": 1919000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0693583688741745e-05,
-      "loss": 0.4364,
-      "step": 1920000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.4133068919181824,
-      "eval_runtime": 78.4611,
-      "eval_samples_per_second": 81.569,
-      "eval_steps_per_second": 0.637,
-      "step": 1920000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0676550347097805e-05,
-      "loss": 0.4376,
-      "step": 1921000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.06597110207435e-05,
-      "loss": 0.437,
-      "step": 1922000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0643082916934733e-05,
-      "loss": 0.4378,
-      "step": 1923000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0626682397606544e-05,
-      "loss": 0.4365,
-      "step": 1924000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0610492778999931e-05,
-      "loss": 0.4366,
-      "step": 1925000
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 0.41611722111701965,
-      "eval_runtime": 81.6547,
-      "eval_samples_per_second": 78.379,
-      "eval_steps_per_second": 0.612,
-      "step": 1925000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.059449822137189e-05,
-      "loss": 0.4372,
-      "step": 1926000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0578715084938887e-05,
-      "loss": 0.4374,
-      "step": 1927000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0563143417779096e-05,
-      "loss": 0.4366,
-      "step": 1928000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 1.0547798521808734e-05,
-      "loss": 0.437,
-      "step": 1929000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0532649723266384e-05,
-      "loss": 0.4365,
-      "step": 1930000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4162156581878662,
-      "eval_runtime": 80.1168,
-      "eval_samples_per_second": 79.883,
-      "eval_steps_per_second": 0.624,
-      "step": 1930000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0517727365795085e-05,
-      "loss": 0.4369,
-      "step": 1931000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0503001620268975e-05,
-      "loss": 0.4373,
-      "step": 1932000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0488487574652423e-05,
-      "loss": 0.4374,
-      "step": 1933000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0474199469678468e-05,
-      "loss": 0.437,
-      "step": 1934000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0460108744063674e-05,
-      "loss": 0.4369,
-      "step": 1935000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4142652451992035,
-      "eval_runtime": 77.8865,
-      "eval_samples_per_second": 82.171,
-      "eval_steps_per_second": 0.642,
-      "step": 1935000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0446243622089129e-05,
-      "loss": 0.4389,
-      "step": 1936000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0432576387995491e-05,
-      "loss": 0.4371,
-      "step": 1937000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0419121068338878e-05,
-      "loss": 0.4372,
-      "step": 1938000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0405877704106532e-05,
-      "loss": 0.4366,
-      "step": 1939000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0392859261103349e-05,
-      "loss": 0.4355,
-      "step": 1940000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4190742075443268,
-      "eval_runtime": 80.8959,
-      "eval_samples_per_second": 79.114,
-      "eval_steps_per_second": 0.618,
-      "step": 1940000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0380039716043426e-05,
-      "loss": 0.4357,
-      "step": 1941000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0367432245456347e-05,
-      "loss": 0.4362,
-      "step": 1942000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0355049177141353e-05,
-      "loss": 0.4362,
-      "step": 1943000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0342865757898152e-05,
-      "loss": 0.437,
-      "step": 1944000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0330906391597708e-05,
-      "loss": 0.4357,
-      "step": 1945000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.4155297577381134,
-      "eval_runtime": 77.8064,
-      "eval_samples_per_second": 82.255,
-      "eval_steps_per_second": 0.643,
-      "step": 1945000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0319147172001108e-05,
-      "loss": 0.4367,
-      "step": 1946000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0307600212366596e-05,
-      "loss": 0.4355,
-      "step": 1947000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.029627677647975e-05,
-      "loss": 0.4367,
-      "step": 1948000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0285154229298157e-05,
-      "loss": 0.4369,
-      "step": 1949000
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 1.0274244045627054e-05,
-      "loss": 0.4373,
-      "step": 1950000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.41703131794929504,
-      "eval_runtime": 77.2751,
-      "eval_samples_per_second": 82.821,
-      "eval_steps_per_second": 0.647,
-      "step": 1950000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.026354625870075e-05,
-      "loss": 0.4364,
-      "step": 1951000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0253060901106556e-05,
-      "loss": 0.4361,
-      "step": 1952000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0242798171546145e-05,
-      "loss": 0.4365,
-      "step": 1953000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0232747509747644e-05,
-      "loss": 0.4373,
-      "step": 1954000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0222899204125646e-05,
-      "loss": 0.4362,
-      "step": 1955000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.4164978265762329,
-      "eval_runtime": 80.0596,
-      "eval_samples_per_second": 79.94,
-      "eval_steps_per_second": 0.625,
-      "step": 1955000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0213263451653737e-05,
-      "loss": 0.4367,
-      "step": 1956000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0203849598659497e-05,
-      "loss": 0.4367,
-      "step": 1957000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0194638827271399e-05,
-      "loss": 0.4364,
-      "step": 1958000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0185640695119401e-05,
-      "loss": 0.4363,
-      "step": 1959000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.017685522961337e-05,
-      "loss": 0.4362,
-      "step": 1960000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.42052188515663147,
-      "eval_runtime": 77.8558,
-      "eval_samples_per_second": 82.203,
-      "eval_steps_per_second": 0.642,
-      "step": 1960000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0168282457515363e-05,
-      "loss": 0.4369,
-      "step": 1961000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0159930658730172e-05,
-      "loss": 0.4364,
-      "step": 1962000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0151791179631108e-05,
-      "loss": 0.4359,
-      "step": 1963000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0143856216286122e-05,
-      "loss": 0.4368,
-      "step": 1964000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0136134046869866e-05,
-      "loss": 0.4357,
-      "step": 1965000
-    },
-    {
-      "epoch": 0.08,
-      "eval_loss": 0.41740044951438904,
-      "eval_runtime": 78.1991,
-      "eval_samples_per_second": 81.842,
-      "eval_steps_per_second": 0.639,
-      "step": 1965000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0128632097947403e-05,
-      "loss": 0.4365,
-      "step": 1966000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.0121335373458022e-05,
-      "loss": 0.4362,
-      "step": 1967000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.011425151149977e-05,
-      "loss": 0.4361,
-      "step": 1968000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 1.010738729828653e-05,
-      "loss": 0.4375,
-      "step": 1969000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0100729012562797e-05,
-      "loss": 0.4372,
-      "step": 1970000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4145086705684662,
-      "eval_runtime": 79.8319,
-      "eval_samples_per_second": 80.168,
-      "eval_steps_per_second": 0.626,
-      "step": 1970000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0094289991138392e-05,
-      "loss": 0.4363,
-      "step": 1971000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0088057362697175e-05,
-      "loss": 0.4375,
-      "step": 1972000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0082049524936494e-05,
-      "loss": 0.4372,
-      "step": 1973000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0076242416653332e-05,
-      "loss": 0.4349,
-      "step": 1974000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0070648308262255e-05,
-      "loss": 0.436,
-      "step": 1975000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4151042103767395,
-      "eval_runtime": 79.0273,
-      "eval_samples_per_second": 80.985,
-      "eval_steps_per_second": 0.633,
-      "step": 1975000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.006526721680391e-05,
-      "loss": 0.4342,
-      "step": 1976000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0060099158670026e-05,
-      "loss": 0.4363,
-      "step": 1977000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0055148998189381e-05,
-      "loss": 0.437,
-      "step": 1978000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0050411475939925e-05,
-      "loss": 0.436,
-      "step": 1979000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0045882183469046e-05,
-      "loss": 0.4355,
-      "step": 1980000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4141569435596466,
-      "eval_runtime": 79.5726,
-      "eval_samples_per_second": 80.43,
-      "eval_steps_per_second": 0.628,
-      "step": 1980000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0041565983372807e-05,
-      "loss": 0.4359,
-      "step": 1981000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0037462888799093e-05,
-      "loss": 0.4362,
-      "step": 1982000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0033576695766748e-05,
-      "loss": 0.4376,
-      "step": 1983000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0029899635949539e-05,
-      "loss": 0.4373,
-      "step": 1984000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0026435717192568e-05,
-      "loss": 0.4367,
-      "step": 1985000
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 0.4171934127807617,
-      "eval_runtime": 77.9474,
-      "eval_samples_per_second": 82.107,
-      "eval_steps_per_second": 0.641,
-      "step": 1985000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0023184950047551e-05,
-      "loss": 0.4361,
-      "step": 1986000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.002015027554519e-05,
-      "loss": 0.4377,
-      "step": 1987000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0017325627506754e-05,
-      "loss": 0.4373,
-      "step": 1988000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 1.0014716663814055e-05,
-      "loss": 0.4368,
-      "step": 1989000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0012320461270247e-05,
-      "loss": 0.4358,
-      "step": 1990000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.41612717509269714,
-      "eval_runtime": 80.5577,
-      "eval_samples_per_second": 79.446,
-      "eval_steps_per_second": 0.621,
-      "step": 1990000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0010134948139825e-05,
-      "loss": 0.4366,
-      "step": 1991000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0008162636276321e-05,
-      "loss": 0.4369,
-      "step": 1992000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0006403531687724e-05,
-      "loss": 0.4372,
-      "step": 1993000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0004859079123212e-05,
-      "loss": 0.4361,
-      "step": 1994000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0003526191291106e-05,
-      "loss": 0.4369,
-      "step": 1995000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.4170204997062683,
-      "eval_runtime": 80.1918,
-      "eval_samples_per_second": 79.809,
-      "eval_steps_per_second": 0.624,
-      "step": 1995000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0002406524857334e-05,
-      "loss": 0.436,
-      "step": 1996000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0001500883167451e-05,
-      "loss": 0.4372,
-      "step": 1997000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0000807455884181e-05,
-      "loss": 0.4369,
-      "step": 1998000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.0000327631969819e-05,
-      "loss": 0.4362,
-      "step": 1999000
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 1.00000604522778e-05,
-      "loss": 0.4363,
-      "step": 2000000
-    },
-    {
-      "epoch": 0.1,
-      "eval_loss": 0.41442054510116577,
-      "eval_runtime": 79.9098,
-      "eval_samples_per_second": 80.09,
-      "eval_steps_per_second": 0.626,
-      "step": 2000000
-    }
-  ],
-  "max_steps": 2000000,
-  "num_train_epochs": 9223372036854775807,
-  "total_flos": 1.752506547830784e+22,
-  "trial_name": null,
-  "trial_params": null
-}