diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12816 @@
+{
+  "best_metric": 0.7345585823059082,
+  "best_model_checkpoint": "/home/khalid/Documents/github_rep/MyProjects/CodeBase/training/output/checkpoint-1000000",
+  "epoch": 1.0723469406083477,
+  "global_step": 1000000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 1.7484,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9969969969969972e-05,
+      "loss": 1.6173,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939939939939944e-05,
+      "loss": 1.5574,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9909909909909908e-05,
+      "loss": 1.5441,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987987987987988e-05,
+      "loss": 1.5047,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984984984984985e-05,
+      "loss": 1.4768,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9819819819819822e-05,
+      "loss": 1.4624,
+      "step": 3500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978978978978979e-05,
+      "loss": 1.4313,
+      "step": 4000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.975975975975976e-05,
+      "loss": 1.4138,
+      "step": 4500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972972972972973e-05,
+      "loss": 1.4181,
+      "step": 5000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96996996996997e-05,
+      "loss": 1.3907,
+      "step": 5500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9669669669669673e-05,
+      "loss": 1.377,
+      "step": 6000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9639699699699702e-05,
+      "loss": 1.383,
+      "step": 6500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960972972972973e-05,
+      "loss": 1.3738,
+      "step": 7000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.95796996996997e-05,
+      "loss": 1.3614,
+      "step": 7500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.954966966966967e-05,
+      "loss": 1.344,
+      "step": 8000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9519639639639642e-05,
+      "loss": 1.3477,
+      "step": 8500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.948960960960961e-05,
+      "loss": 1.3436,
+      "step": 9000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9459639639639642e-05,
+      "loss": 1.334,
+      "step": 9500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942966966966967e-05,
+      "loss": 1.3165,
+      "step": 10000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 1.2129223346710205,
+      "eval_runtime": 577.4417,
+      "eval_samples_per_second": 173.178,
+      "eval_steps_per_second": 43.294,
+      "step": 10000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9399639639639643e-05,
+      "loss": 1.3218,
+      "step": 10500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369609609609608e-05,
+      "loss": 1.3243,
+      "step": 11000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.933957957957958e-05,
+      "loss": 1.3016,
+      "step": 11500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.930954954954955e-05,
+      "loss": 1.3057,
+      "step": 12000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927951951951952e-05,
+      "loss": 1.2973,
+      "step": 12500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9249489489489493e-05,
+      "loss": 1.2911,
+      "step": 13000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9219459459459458e-05,
+      "loss": 1.294,
+      "step": 13500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.918942942942943e-05,
+      "loss": 1.291,
+      "step": 14000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91593993993994e-05,
+      "loss": 1.2678,
+      "step": 14500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9129369369369372e-05,
+      "loss": 1.27,
+      "step": 15000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.909933933933934e-05,
+      "loss": 1.2418,
+      "step": 15500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.906936936936937e-05,
+      "loss": 1.2373,
+      "step": 16000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903933933933934e-05,
+      "loss": 1.2567,
+      "step": 16500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900930930930931e-05,
+      "loss": 1.2449,
+      "step": 17000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897927927927928e-05,
+      "loss": 1.2554,
+      "step": 17500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.894924924924925e-05,
+      "loss": 1.2579,
+      "step": 18000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.891927927927928e-05,
+      "loss": 1.2351,
+      "step": 18500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888924924924925e-05,
+      "loss": 1.2227,
+      "step": 19000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885921921921922e-05,
+      "loss": 1.2436,
+      "step": 19500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882924924924925e-05,
+      "loss": 1.2256,
+      "step": 20000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 1.1297707557678223,
+      "eval_runtime": 613.186,
+      "eval_samples_per_second": 163.083,
+      "eval_steps_per_second": 40.771,
+      "step": 20000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879921921921922e-05,
+      "loss": 1.2092,
+      "step": 20500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8769189189189192e-05,
+      "loss": 1.2248,
+      "step": 21000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.873915915915916e-05,
+      "loss": 1.2378,
+      "step": 21500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8709129129129128e-05,
+      "loss": 1.2182,
+      "step": 22000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8679159159159157e-05,
+      "loss": 1.2111,
+      "step": 22500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.864912912912913e-05,
+      "loss": 1.2088,
+      "step": 23000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.86190990990991e-05,
+      "loss": 1.2104,
+      "step": 23500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.858906906906907e-05,
+      "loss": 1.1984,
+      "step": 24000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.8559039039039043e-05,
+      "loss": 1.206,
+      "step": 24500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.8529009009009007e-05,
+      "loss": 1.1827,
+      "step": 25000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.849897897897898e-05,
+      "loss": 1.1953,
+      "step": 25500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.8469009009009008e-05,
+      "loss": 1.1964,
+      "step": 26000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.843903903903904e-05,
+      "loss": 1.2048,
+      "step": 26500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.840900900900901e-05,
+      "loss": 1.1847,
+      "step": 27000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.837897897897898e-05,
+      "loss": 1.1808,
+      "step": 27500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.834894894894895e-05,
+      "loss": 1.1685,
+      "step": 28000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.831891891891892e-05,
+      "loss": 1.1838,
+      "step": 28500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.828888888888889e-05,
+      "loss": 1.1876,
+      "step": 29000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.825885885885886e-05,
+      "loss": 1.1851,
+      "step": 29500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.822882882882883e-05,
+      "loss": 1.1683,
+      "step": 30000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 1.090136170387268,
+      "eval_runtime": 615.6857,
+      "eval_samples_per_second": 162.421,
+      "eval_steps_per_second": 40.605,
+      "step": 30000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.8198798798798798e-05,
+      "loss": 1.1725,
+      "step": 30500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.816876876876877e-05,
+      "loss": 1.1879,
+      "step": 31000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.813873873873874e-05,
+      "loss": 1.1635,
+      "step": 31500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.810870870870871e-05,
+      "loss": 1.1892,
+      "step": 32000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.807873873873874e-05,
+      "loss": 1.1675,
+      "step": 32500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.804870870870871e-05,
+      "loss": 1.1766,
+      "step": 33000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.8018678678678677e-05,
+      "loss": 1.1699,
+      "step": 33500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.798864864864865e-05,
+      "loss": 1.1573,
+      "step": 34000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.7958678678678678e-05,
+      "loss": 1.1695,
+      "step": 34500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.792864864864865e-05,
+      "loss": 1.1419,
+      "step": 35000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.789861861861862e-05,
+      "loss": 1.1523,
+      "step": 35500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.7868588588588592e-05,
+      "loss": 1.1544,
+      "step": 36000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.783861861861862e-05,
+      "loss": 1.1501,
+      "step": 36500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.780864864864865e-05,
+      "loss": 1.1583,
+      "step": 37000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.777861861861862e-05,
+      "loss": 1.1455,
+      "step": 37500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.774858858858859e-05,
+      "loss": 1.1276,
+      "step": 38000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.7718558558558558e-05,
+      "loss": 1.139,
+      "step": 38500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.768852852852853e-05,
+      "loss": 1.1363,
+      "step": 39000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.7658498498498497e-05,
+      "loss": 1.1287,
+      "step": 39500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.762846846846847e-05,
+      "loss": 1.1429,
+      "step": 40000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 1.0513352155685425,
+      "eval_runtime": 613.186,
+      "eval_samples_per_second": 163.083,
+      "eval_steps_per_second": 40.771,
+      "step": 40000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.759843843843844e-05,
+      "loss": 1.1433,
+      "step": 40500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.756846846846847e-05,
+      "loss": 1.1378,
+      "step": 41000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.753843843843844e-05,
+      "loss": 1.1398,
+      "step": 41500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.750846846846847e-05,
+      "loss": 1.1185,
+      "step": 42000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.74784984984985e-05,
+      "loss": 1.1246,
+      "step": 42500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.744846846846847e-05,
+      "loss": 1.128,
+      "step": 43000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.741843843843844e-05,
+      "loss": 1.1205,
+      "step": 43500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.738840840840841e-05,
+      "loss": 1.1289,
+      "step": 44000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7358378378378378e-05,
+      "loss": 1.1206,
+      "step": 44500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.732834834834835e-05,
+      "loss": 1.1348,
+      "step": 45000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.729831831831832e-05,
+      "loss": 1.1181,
+      "step": 45500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.726828828828829e-05,
+      "loss": 1.1186,
+      "step": 46000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7238258258258257e-05,
+      "loss": 1.1241,
+      "step": 46500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7208228228228228e-05,
+      "loss": 1.1175,
+      "step": 47000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.717825825825826e-05,
+      "loss": 1.1179,
+      "step": 47500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.714822822822823e-05,
+      "loss": 1.1185,
+      "step": 48000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.71181981981982e-05,
+      "loss": 1.1051,
+      "step": 48500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7088168168168168e-05,
+      "loss": 1.1127,
+      "step": 49000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.705813813813814e-05,
+      "loss": 1.1079,
+      "step": 49500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7028108108108107e-05,
+      "loss": 1.09,
+      "step": 50000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 1.0235236883163452,
+      "eval_runtime": 630.7694,
+      "eval_samples_per_second": 158.537,
+      "eval_steps_per_second": 39.634,
+      "step": 50000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.699807807807808e-05,
+      "loss": 1.1017,
+      "step": 50500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.6968048048048047e-05,
+      "loss": 1.1153,
+      "step": 51000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.693807807807808e-05,
+      "loss": 1.1244,
+      "step": 51500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6908108108108112e-05,
+      "loss": 1.1129,
+      "step": 52000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6878078078078077e-05,
+      "loss": 1.1032,
+      "step": 52500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6848048048048048e-05,
+      "loss": 1.1024,
+      "step": 53000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.681801801801802e-05,
+      "loss": 1.1119,
+      "step": 53500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.678798798798799e-05,
+      "loss": 1.0847,
+      "step": 54000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.675795795795796e-05,
+      "loss": 1.1061,
+      "step": 54500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6727927927927927e-05,
+      "loss": 1.0983,
+      "step": 55000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6697897897897898e-05,
+      "loss": 1.0909,
+      "step": 55500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.666786786786787e-05,
+      "loss": 1.1001,
+      "step": 56000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.66378978978979e-05,
+      "loss": 1.0911,
+      "step": 56500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.660786786786787e-05,
+      "loss": 1.0993,
+      "step": 57000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6577837837837838e-05,
+      "loss": 1.09,
+      "step": 57500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.654780780780781e-05,
+      "loss": 1.0906,
+      "step": 58000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6517897897897897e-05,
+      "loss": 1.089,
+      "step": 58500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6487867867867868e-05,
+      "loss": 1.097,
+      "step": 59000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.645783783783784e-05,
+      "loss": 1.0785,
+      "step": 59500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.642780780780781e-05,
+      "loss": 1.0867,
+      "step": 60000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.0042845010757446,
+      "eval_runtime": 612.5588,
+      "eval_samples_per_second": 163.25,
+      "eval_steps_per_second": 40.812,
+      "step": 60000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.639777777777778e-05,
+      "loss": 1.0828,
+      "step": 60500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6367747747747747e-05,
+      "loss": 1.0736,
+      "step": 61000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6337717717717718e-05,
+      "loss": 1.0702,
+      "step": 61500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.630768768768769e-05,
+      "loss": 1.0765,
+      "step": 62000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.627771771771772e-05,
+      "loss": 1.0801,
+      "step": 62500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.624768768768769e-05,
+      "loss": 1.0709,
+      "step": 63000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.621771771771772e-05,
+      "loss": 1.0813,
+      "step": 63500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.618768768768769e-05,
+      "loss": 1.072,
+      "step": 64000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.615765765765766e-05,
+      "loss": 1.0789,
+      "step": 64500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6127627627627627e-05,
+      "loss": 1.0926,
+      "step": 65000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6097597597597598e-05,
+      "loss": 1.0765,
+      "step": 65500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.606756756756757e-05,
+      "loss": 1.0673,
+      "step": 66000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.60375975975976e-05,
+      "loss": 1.075,
+      "step": 66500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6007567567567567e-05,
+      "loss": 1.0678,
+      "step": 67000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.5977537537537538e-05,
+      "loss": 1.0645,
+      "step": 67500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.594750750750751e-05,
+      "loss": 1.0708,
+      "step": 68000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.591753753753754e-05,
+      "loss": 1.0746,
+      "step": 68500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.588750750750751e-05,
+      "loss": 1.0644,
+      "step": 69000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.5857477477477478e-05,
+      "loss": 1.0572,
+      "step": 69500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5827447447447446e-05,
+      "loss": 1.0777,
+      "step": 70000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.989311158657074,
+      "eval_runtime": 626.1693,
+      "eval_samples_per_second": 159.701,
+      "eval_steps_per_second": 39.925,
+      "step": 70000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5797477477477475e-05,
+      "loss": 1.0623,
+      "step": 70500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5767447447447447e-05,
+      "loss": 1.0469,
+      "step": 71000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5737417417417418e-05,
+      "loss": 1.0564,
+      "step": 71500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.570738738738739e-05,
+      "loss": 1.0434,
+      "step": 72000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.567735735735736e-05,
+      "loss": 1.0593,
+      "step": 72500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5647327327327325e-05,
+      "loss": 1.0587,
+      "step": 73000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5617297297297297e-05,
+      "loss": 1.0507,
+      "step": 73500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5587267267267268e-05,
+      "loss": 1.057,
+      "step": 74000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5557297297297297e-05,
+      "loss": 1.0607,
+      "step": 74500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.552726726726727e-05,
+      "loss": 1.0645,
+      "step": 75000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.549723723723724e-05,
+      "loss": 1.0603,
+      "step": 75500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5467207207207208e-05,
+      "loss": 1.0407,
+      "step": 76000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5437177177177176e-05,
+      "loss": 1.0573,
+      "step": 76500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5407147147147148e-05,
+      "loss": 1.0331,
+      "step": 77000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5377177177177177e-05,
+      "loss": 1.0456,
+      "step": 77500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5347147147147148e-05,
+      "loss": 1.0586,
+      "step": 78000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5317117117117116e-05,
+      "loss": 1.0591,
+      "step": 78500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.5287087087087088e-05,
+      "loss": 1.0519,
+      "step": 79000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.525705705705706e-05,
+      "loss": 1.0549,
+      "step": 79500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5227027027027027e-05,
+      "loss": 1.0588,
+      "step": 80000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.9711708426475525,
+      "eval_runtime": 621.1495,
+      "eval_samples_per_second": 160.992,
+      "eval_steps_per_second": 40.248,
+      "step": 80000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5196996996997e-05,
+      "loss": 1.0534,
+      "step": 80500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5166966966966966e-05,
+      "loss": 1.0529,
+      "step": 81000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5136996996996996e-05,
+      "loss": 1.0605,
+      "step": 81500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5106966966966967e-05,
+      "loss": 1.0639,
+      "step": 82000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.507693693693694e-05,
+      "loss": 1.0455,
+      "step": 82500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.504690690690691e-05,
+      "loss": 1.042,
+      "step": 83000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.501693693693694e-05,
+      "loss": 1.0267,
+      "step": 83500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.498690690690691e-05,
+      "loss": 1.0534,
+      "step": 84000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4956876876876875e-05,
+      "loss": 1.0345,
+      "step": 84500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4926906906906908e-05,
+      "loss": 1.0255,
+      "step": 85000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4896876876876876e-05,
+      "loss": 1.0277,
+      "step": 85500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4866846846846847e-05,
+      "loss": 1.0501,
+      "step": 86000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.483681681681682e-05,
+      "loss": 1.0417,
+      "step": 86500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4806846846846848e-05,
+      "loss": 1.0324,
+      "step": 87000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.477681681681682e-05,
+      "loss": 1.0552,
+      "step": 87500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4746786786786787e-05,
+      "loss": 1.038,
+      "step": 88000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.471675675675676e-05,
+      "loss": 1.0434,
+      "step": 88500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4686726726726726e-05,
+      "loss": 1.0349,
+      "step": 89000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4656696696696698e-05,
+      "loss": 1.0166,
+      "step": 89500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4626666666666666e-05,
+      "loss": 1.0389,
+      "step": 90000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.9617297649383545,
+      "eval_runtime": 593.47,
+      "eval_samples_per_second": 168.501,
+      "eval_steps_per_second": 42.125,
+      "step": 90000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4596636636636637e-05,
+      "loss": 1.0382,
+      "step": 90500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.456660660660661e-05,
+      "loss": 1.0389,
+      "step": 91000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4536576576576577e-05,
+      "loss": 1.0397,
+      "step": 91500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4506546546546548e-05,
+      "loss": 1.028,
+      "step": 92000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4476516516516516e-05,
+      "loss": 1.0317,
+      "step": 92500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4446546546546545e-05,
+      "loss": 1.0281,
+      "step": 93000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4416516516516517e-05,
+      "loss": 1.0385,
+      "step": 93500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4386486486486488e-05,
+      "loss": 1.0183,
+      "step": 94000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.435645645645646e-05,
+      "loss": 1.0204,
+      "step": 94500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.432648648648649e-05,
+      "loss": 1.044,
+      "step": 95000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4296516516516518e-05,
+      "loss": 1.0402,
+      "step": 95500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.426648648648649e-05,
+      "loss": 1.0346,
+      "step": 96000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4236456456456457e-05,
+      "loss": 1.0156,
+      "step": 96500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4206426426426425e-05,
+      "loss": 1.0054,
+      "step": 97000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4176396396396397e-05,
+      "loss": 1.0276,
+      "step": 97500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.4146366366366368e-05,
+      "loss": 1.022,
+      "step": 98000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.4116336336336336e-05,
+      "loss": 1.0351,
+      "step": 98500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.4086306306306307e-05,
+      "loss": 1.0306,
+      "step": 99000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.4056276276276275e-05,
+      "loss": 1.0183,
+      "step": 99500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.4026246246246247e-05,
+      "loss": 1.0245,
+      "step": 100000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.9445381760597229,
+      "eval_runtime": 618.5347,
+      "eval_samples_per_second": 161.672,
+      "eval_steps_per_second": 40.418,
+      "step": 100000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3996216216216218e-05,
+      "loss": 1.0274,
+      "step": 100500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3966186186186186e-05,
+      "loss": 1.007,
+      "step": 101000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3936276276276276e-05,
+      "loss": 1.021,
+      "step": 101500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3906246246246245e-05,
+      "loss": 1.0176,
+      "step": 102000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3876216216216216e-05,
+      "loss": 1.0279,
+      "step": 102500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3846186186186187e-05,
+      "loss": 1.0051,
+      "step": 103000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3816216216216216e-05,
+      "loss": 1.0184,
+      "step": 103500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3786186186186188e-05,
+      "loss": 1.0125,
+      "step": 104000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.375615615615616e-05,
+      "loss": 1.0202,
+      "step": 104500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3726126126126124e-05,
+      "loss": 1.0255,
+      "step": 105000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3696096096096095e-05,
+      "loss": 0.9965,
+      "step": 105500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3666066066066067e-05,
+      "loss": 1.0157,
+      "step": 106000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.3636036036036038e-05,
+      "loss": 1.0168,
+      "step": 106500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.360600600600601e-05,
+      "loss": 1.012,
+      "step": 107000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3575975975975974e-05,
+      "loss": 1.0102,
+      "step": 107500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3545945945945945e-05,
+      "loss": 1.0161,
+      "step": 108000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3515915915915917e-05,
+      "loss": 0.9957,
+      "step": 108500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3485945945945946e-05,
+      "loss": 1.0009,
+      "step": 109000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3455915915915917e-05,
+      "loss": 1.0096,
+      "step": 109500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3425885885885885e-05,
+      "loss": 1.0146,
+      "step": 110000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.934283435344696,
+      "eval_runtime": 601.7192,
+      "eval_samples_per_second": 166.19,
+      "eval_steps_per_second": 41.548,
+      "step": 110000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3395975975975976e-05,
+      "loss": 1.0183,
+      "step": 110500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3365945945945947e-05,
+      "loss": 1.0182,
+      "step": 111000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3335915915915915e-05,
+      "loss": 1.0091,
+      "step": 111500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3305885885885887e-05,
+      "loss": 1.0111,
+      "step": 112000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3275855855855858e-05,
+      "loss": 1.0015,
+      "step": 112500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3245825825825826e-05,
+      "loss": 1.0082,
+      "step": 113000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3215795795795794e-05,
+      "loss": 0.9918,
+      "step": 113500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3185765765765765e-05,
+      "loss": 1.0009,
+      "step": 114000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3155735735735737e-05,
+      "loss": 1.0195,
+      "step": 114500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3125705705705708e-05,
+      "loss": 1.0189,
+      "step": 115000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3095675675675676e-05,
+      "loss": 1.0038,
+      "step": 115500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3065645645645644e-05,
+      "loss": 0.9963,
+      "step": 116000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.3035615615615616e-05,
+      "loss": 1.0125,
+      "step": 116500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.3005585585585587e-05,
+      "loss": 0.988,
+      "step": 117000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.297555555555556e-05,
+      "loss": 0.9997,
+      "step": 117500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2945525525525523e-05,
+      "loss": 0.9983,
+      "step": 118000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.291555555555556e-05,
+      "loss": 1.0094,
+      "step": 118500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2885585585585588e-05,
+      "loss": 0.9988,
+      "step": 119000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2855555555555556e-05,
+      "loss": 0.9928,
+      "step": 119500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2825525525525524e-05,
+      "loss": 0.9951,
+      "step": 120000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.9260998964309692,
+      "eval_runtime": 605.8053,
+      "eval_samples_per_second": 165.07,
+      "eval_steps_per_second": 41.267,
+      "step": 120000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2795495495495496e-05,
+      "loss": 0.9923,
+      "step": 120500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2765465465465467e-05,
+      "loss": 1.0081,
+      "step": 121000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2735495495495496e-05,
+      "loss": 1.0009,
+      "step": 121500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2705465465465464e-05,
+      "loss": 0.9929,
+      "step": 122000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2675435435435436e-05,
+      "loss": 0.9956,
+      "step": 122500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2645405405405407e-05,
+      "loss": 1.0,
+      "step": 123000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2615435435435436e-05,
+      "loss": 0.9927,
+      "step": 123500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2585405405405408e-05,
+      "loss": 0.9882,
+      "step": 124000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2555375375375376e-05,
+      "loss": 0.9915,
+      "step": 124500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2525345345345344e-05,
+      "loss": 0.9897,
+      "step": 125000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2495315315315315e-05,
+      "loss": 1.0072,
+      "step": 125500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2465285285285286e-05,
+      "loss": 0.9896,
+      "step": 126000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2435315315315316e-05,
+      "loss": 0.9879,
+      "step": 126500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2405285285285287e-05,
+      "loss": 0.9952,
+      "step": 127000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.237525525525526e-05,
+      "loss": 0.9748,
+      "step": 127500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2345225225225223e-05,
+      "loss": 0.9967,
+      "step": 128000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2315195195195194e-05,
+      "loss": 1.0016,
+      "step": 128500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2285165165165166e-05,
+      "loss": 0.9801,
+      "step": 129000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2255135135135137e-05,
+      "loss": 0.9876,
+      "step": 129500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2225165165165166e-05,
+      "loss": 0.978,
+      "step": 130000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.9173310399055481,
+      "eval_runtime": 602.5766,
+      "eval_samples_per_second": 165.954,
+      "eval_steps_per_second": 41.489,
+      "step": 130000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2195135135135138e-05,
+      "loss": 0.9867,
+      "step": 130500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2165105105105106e-05,
+      "loss": 0.9824,
+      "step": 131000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2135075075075074e-05,
+      "loss": 0.9877,
+      "step": 131500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2105045045045045e-05,
+      "loss": 0.9828,
+      "step": 132000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2075015015015017e-05,
+      "loss": 0.9772,
+      "step": 132500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2044984984984985e-05,
+      "loss": 0.9906,
+      "step": 133000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.2015015015015014e-05,
+      "loss": 0.9847,
+      "step": 133500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.1984984984984985e-05,
+      "loss": 0.9826,
+      "step": 134000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.195507507507508e-05,
+      "loss": 0.9903,
+      "step": 134500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.1925045045045043e-05,
+      "loss": 0.9907,
+      "step": 135000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1895015015015015e-05,
+      "loss": 0.9909,
+      "step": 135500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1864984984984986e-05,
+      "loss": 0.9906,
+      "step": 136000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1834954954954958e-05,
+      "loss": 0.9933,
+      "step": 136500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1804924924924922e-05,
+      "loss": 0.9754,
+      "step": 137000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1774894894894894e-05,
+      "loss": 0.9774,
+      "step": 137500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1744864864864865e-05,
+      "loss": 0.9889,
+      "step": 138000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1714834834834836e-05,
+      "loss": 0.9972,
+      "step": 138500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1684804804804808e-05,
+      "loss": 0.9852,
+      "step": 139000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1654834834834837e-05,
+      "loss": 0.9615,
+      "step": 139500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1624804804804805e-05,
+      "loss": 0.9865,
+      "step": 140000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.909109354019165,
+      "eval_runtime": 589.6128,
+      "eval_samples_per_second": 169.603,
+      "eval_steps_per_second": 42.401,
+      "step": 140000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1594774774774773e-05,
+      "loss": 0.9732,
+      "step": 140500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1564744744744744e-05,
+      "loss": 0.9657,
+      "step": 141000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1534714714714716e-05,
+      "loss": 0.9651,
+      "step": 141500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1504684684684687e-05,
+      "loss": 0.9913,
+      "step": 142000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1474654654654655e-05,
+      "loss": 0.9685,
+      "step": 142500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1444624624624623e-05,
+      "loss": 0.9663,
+      "step": 143000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1414594594594595e-05,
+      "loss": 0.9743,
+      "step": 143500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1384564564564566e-05,
+      "loss": 0.9723,
+      "step": 144000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.1354534534534534e-05,
+      "loss": 0.9815,
+      "step": 144500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1324564564564563e-05,
+      "loss": 0.9714,
+      "step": 145000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1294534534534535e-05,
+      "loss": 0.9734,
+      "step": 145500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1264504504504506e-05,
+      "loss": 0.9679,
+      "step": 146000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1234474474474474e-05,
+      "loss": 0.9697,
+      "step": 146500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1204444444444445e-05,
+      "loss": 0.9941,
+      "step": 147000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1174474474474475e-05,
+      "loss": 0.9786,
+      "step": 147500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1144444444444443e-05,
+      "loss": 0.97,
+      "step": 148000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1114414414414414e-05,
+      "loss": 0.9672,
+      "step": 148500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1084384384384385e-05,
+      "loss": 0.9661,
+      "step": 149000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1054354354354357e-05,
+      "loss": 0.9867,
+      "step": 149500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.1024324324324325e-05,
+      "loss": 0.972,
+      "step": 150000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.8991417288780212,
+      "eval_runtime": 602.234,
+      "eval_samples_per_second": 166.048,
+      "eval_steps_per_second": 41.512,
+      "step": 150000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0994294294294293e-05,
+      "loss": 0.9681,
+      "step": 150500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0964264264264264e-05,
+      "loss": 0.9601,
+      "step": 151000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0934294294294293e-05,
+      "loss": 0.955,
+      "step": 151500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0904264264264265e-05,
+      "loss": 0.9703,
+      "step": 152000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0874234234234236e-05,
+      "loss": 0.9535,
+      "step": 152500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0844204204204204e-05,
+      "loss": 0.968,
+      "step": 153000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.0814174174174172e-05,
+      "loss": 0.9737,
+      "step": 153500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0784204204204205e-05,
+      "loss": 0.9714,
+      "step": 154000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0754174174174173e-05,
+      "loss": 0.9737,
+      "step": 154500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0724144144144144e-05,
+      "loss": 0.955,
+      "step": 155000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0694114114114116e-05,
+      "loss": 0.9677,
+      "step": 155500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0664084084084084e-05,
+      "loss": 0.964,
+      "step": 156000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0634054054054055e-05,
+      "loss": 0.9675,
+      "step": 156500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0604024024024023e-05,
+      "loss": 0.9705,
+      "step": 157000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0573993993993994e-05,
+      "loss": 0.9634,
+      "step": 157500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0544024024024024e-05,
+      "loss": 0.9583,
+      "step": 158000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0513993993993995e-05,
+      "loss": 0.9614,
+      "step": 158500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0483963963963963e-05,
+      "loss": 0.9707,
+      "step": 159000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0453933933933934e-05,
+      "loss": 0.95,
+      "step": 159500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0423903903903906e-05,
+      "loss": 0.9673,
+      "step": 160000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.8937882781028748,
+      "eval_runtime": 585.5551,
+      "eval_samples_per_second": 170.778,
+      "eval_steps_per_second": 42.695,
+      "step": 160000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0393873873873874e-05,
+      "loss": 0.9509,
+      "step": 160500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0363843843843842e-05,
+      "loss": 0.9537,
+      "step": 161000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0333813813813813e-05,
+      "loss": 0.9528,
+      "step": 161500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0303903903903907e-05,
+      "loss": 0.958,
+      "step": 162000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0273873873873875e-05,
+      "loss": 0.9773,
+      "step": 162500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.0243843843843843e-05,
+      "loss": 0.963,
+      "step": 163000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0213813813813814e-05,
+      "loss": 0.9617,
+      "step": 163500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0183843843843844e-05,
+      "loss": 0.9572,
+      "step": 164000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0153813813813815e-05,
+      "loss": 0.9535,
+      "step": 164500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0123783783783783e-05,
+      "loss": 0.9599,
+      "step": 165000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0093813813813816e-05,
+      "loss": 0.9612,
+      "step": 165500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0063783783783784e-05,
+      "loss": 0.9474,
+      "step": 166000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0033753753753755e-05,
+      "loss": 0.9534,
+      "step": 166500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.0003723723723726e-05,
+      "loss": 0.9509,
+      "step": 167000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9973693693693694e-05,
+      "loss": 0.955,
+      "step": 167500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9943663663663662e-05,
+      "loss": 0.9719,
+      "step": 168000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9913633633633634e-05,
+      "loss": 0.9597,
+      "step": 168500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9883603603603605e-05,
+      "loss": 0.9484,
+      "step": 169000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9853573573573577e-05,
+      "loss": 0.9539,
+      "step": 169500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9823543543543545e-05,
+      "loss": 0.9493,
+      "step": 170000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.8906692266464233,
+      "eval_runtime": 620.1154,
+      "eval_samples_per_second": 161.26,
+      "eval_steps_per_second": 40.315,
+      "step": 170000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9793513513513513e-05,
+      "loss": 0.9523,
+      "step": 170500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9763483483483484e-05,
+      "loss": 0.9462,
+      "step": 171000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9733453453453455e-05,
+      "loss": 0.9438,
+      "step": 171500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9703423423423423e-05,
+      "loss": 0.9407,
+      "step": 172000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.967339339339339e-05,
+      "loss": 0.9508,
+      "step": 172500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9643363363363363e-05,
+      "loss": 0.9544,
+      "step": 173000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9613393393393392e-05,
+      "loss": 0.9576,
+      "step": 173500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9583363363363363e-05,
+      "loss": 0.9528,
+      "step": 174000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9553333333333335e-05,
+      "loss": 0.9481,
+      "step": 174500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9523363363363364e-05,
+      "loss": 0.9393,
+      "step": 175000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9493333333333335e-05,
+      "loss": 0.9377,
+      "step": 175500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9463303303303303e-05,
+      "loss": 0.9342,
+      "step": 176000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9433273273273275e-05,
+      "loss": 0.9309,
+      "step": 176500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9403243243243243e-05,
+      "loss": 0.9607,
+      "step": 177000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9373213213213214e-05,
+      "loss": 0.9508,
+      "step": 177500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9343183183183186e-05,
+      "loss": 0.9361,
+      "step": 178000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9313153153153154e-05,
+      "loss": 0.9312,
+      "step": 178500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9283183183183183e-05,
+      "loss": 0.9503,
+      "step": 179000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9253153153153154e-05,
+      "loss": 0.9509,
+      "step": 179500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9223183183183183e-05,
+      "loss": 0.9311,
+      "step": 180000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.8821685910224915,
+      "eval_runtime": 603.0834,
+      "eval_samples_per_second": 165.815,
+      "eval_steps_per_second": 41.454,
+      "step": 180000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9193153153153155e-05,
+      "loss": 0.95,
+      "step": 180500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9163123123123126e-05,
+      "loss": 0.9439,
+      "step": 181000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9133093093093094e-05,
+      "loss": 0.9397,
+      "step": 181500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9103063063063062e-05,
+      "loss": 0.9353,
+      "step": 182000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9073033033033033e-05,
+      "loss": 0.9464,
+      "step": 182500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9043003003003005e-05,
+      "loss": 0.9346,
+      "step": 183000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9012972972972976e-05,
+      "loss": 0.9596,
+      "step": 183500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8983003003003005e-05,
+      "loss": 0.9653,
+      "step": 184000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8952972972972973e-05,
+      "loss": 0.9459,
+      "step": 184500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8923003003003006e-05,
+      "loss": 0.9249,
+      "step": 185000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8892972972972974e-05,
+      "loss": 0.9413,
+      "step": 185500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8862942942942942e-05,
+      "loss": 0.943,
+      "step": 186000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8832912912912913e-05,
+      "loss": 0.947,
+      "step": 186500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8802882882882885e-05,
+      "loss": 0.9567,
+      "step": 187000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8772912912912914e-05,
+      "loss": 0.9314,
+      "step": 187500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8742882882882882e-05,
+      "loss": 0.9512,
+      "step": 188000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8712852852852853e-05,
+      "loss": 0.9505,
+      "step": 188500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8682822822822825e-05,
+      "loss": 0.9338,
+      "step": 189000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8652792792792793e-05,
+      "loss": 0.9395,
+      "step": 189500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8622762762762764e-05,
+      "loss": 0.9462,
+      "step": 190000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.8739633560180664,
+      "eval_runtime": 600.0958,
+      "eval_samples_per_second": 166.64,
+      "eval_steps_per_second": 41.66,
+      "step": 190000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8592732732732732e-05,
+      "loss": 0.9298,
+      "step": 190500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8562702702702704e-05,
+      "loss": 0.9386,
+      "step": 191000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8532732732732733e-05,
+      "loss": 0.937,
+      "step": 191500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8502702702702704e-05,
+      "loss": 0.9436,
+      "step": 192000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8472672672672676e-05,
+      "loss": 0.9416,
+      "step": 192500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8442642642642644e-05,
+      "loss": 0.9432,
+      "step": 193000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.841261261261261e-05,
+      "loss": 0.9358,
+      "step": 193500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8382582582582583e-05,
+      "loss": 0.9322,
+      "step": 194000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8352552552552554e-05,
+      "loss": 0.9325,
+      "step": 194500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8322522522522526e-05,
+      "loss": 0.9309,
+      "step": 195000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8292552552552555e-05,
+      "loss": 0.9439,
+      "step": 195500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8262522522522523e-05,
+      "loss": 0.9186,
+      "step": 196000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.823249249249249e-05,
+      "loss": 0.9252,
+      "step": 196500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8202462462462462e-05,
+      "loss": 0.9458,
+      "step": 197000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.817249249249249e-05,
+      "loss": 0.9415,
+      "step": 197500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8142462462462463e-05,
+      "loss": 0.9387,
+      "step": 198000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8112432432432434e-05,
+      "loss": 0.922,
+      "step": 198500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8082462462462464e-05,
+      "loss": 0.9299,
+      "step": 199000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.805243243243243e-05,
+      "loss": 0.9296,
+      "step": 199500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8022402402402403e-05,
+      "loss": 0.9302,
+      "step": 200000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.868446946144104,
+      "eval_runtime": 599.5829,
+      "eval_samples_per_second": 166.783,
+      "eval_steps_per_second": 41.696,
+      "step": 200000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7992372372372374e-05,
+      "loss": 0.9236,
+      "step": 200500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7962342342342342e-05,
+      "loss": 0.9376,
+      "step": 201000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7932312312312314e-05,
+      "loss": 0.9373,
+      "step": 201500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7902342342342343e-05,
+      "loss": 0.9363,
+      "step": 202000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.787231231231231e-05,
+      "loss": 0.9386,
+      "step": 202500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7842282282282282e-05,
+      "loss": 0.9139,
+      "step": 203000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7812252252252254e-05,
+      "loss": 0.9403,
+      "step": 203500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7782282282282283e-05,
+      "loss": 0.9184,
+      "step": 204000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7752252252252254e-05,
+      "loss": 0.9281,
+      "step": 204500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7722222222222226e-05,
+      "loss": 0.9358,
+      "step": 205000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.769219219219219e-05,
+      "loss": 0.9337,
+      "step": 205500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7662162162162162e-05,
+      "loss": 0.9295,
+      "step": 206000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7632132132132133e-05,
+      "loss": 0.931,
+      "step": 206500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7602102102102105e-05,
+      "loss": 0.9348,
+      "step": 207000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7572072072072073e-05,
+      "loss": 0.9153,
+      "step": 207500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7542102102102102e-05,
+      "loss": 0.9248,
+      "step": 208000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7512072072072073e-05,
+      "loss": 0.9298,
+      "step": 208500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.748204204204204e-05,
+      "loss": 0.929,
+      "step": 209000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7452012012012013e-05,
+      "loss": 0.9327,
+      "step": 209500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7421981981981984e-05,
+      "loss": 0.921,
+      "step": 210000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.8622388243675232,
+      "eval_runtime": 575.3524,
+      "eval_samples_per_second": 173.807,
+      "eval_steps_per_second": 43.452,
+      "step": 210000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7392012012012013e-05,
+      "loss": 0.9248,
+      "step": 210500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.736198198198198e-05,
+      "loss": 0.9119,
+      "step": 211000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7331951951951952e-05,
+      "loss": 0.9253,
+      "step": 211500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7301921921921924e-05,
+      "loss": 0.9215,
+      "step": 212000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7271891891891892e-05,
+      "loss": 0.947,
+      "step": 212500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7241861861861863e-05,
+      "loss": 0.9279,
+      "step": 213000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.721183183183183e-05,
+      "loss": 0.9202,
+      "step": 213500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7181801801801803e-05,
+      "loss": 0.9283,
+      "step": 214000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7151831831831832e-05,
+      "loss": 0.917,
+      "step": 214500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.712186186186186e-05,
+      "loss": 0.9394,
+      "step": 215000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7091831831831832e-05,
+      "loss": 0.9431,
+      "step": 215500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7061801801801804e-05,
+      "loss": 0.9241,
+      "step": 216000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7031771771771775e-05,
+      "loss": 0.9305,
+      "step": 216500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.700174174174174e-05,
+      "loss": 0.9231,
+      "step": 217000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.6971771771771772e-05,
+      "loss": 0.9238,
+      "step": 217500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.694174174174174e-05,
+      "loss": 0.9168,
+      "step": 218000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.6911771771771773e-05,
+      "loss": 0.9201,
+      "step": 218500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.688174174174174e-05,
+      "loss": 0.9141,
+      "step": 219000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6851771771771774e-05,
+      "loss": 0.9076,
+      "step": 219500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.682174174174174e-05,
+      "loss": 0.9198,
+      "step": 220000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.85796719789505,
+      "eval_runtime": 609.8643,
+      "eval_samples_per_second": 163.971,
+      "eval_steps_per_second": 40.993,
+      "step": 220000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6791711711711713e-05,
+      "loss": 0.9256,
+      "step": 220500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.676168168168168e-05,
+      "loss": 0.9195,
+      "step": 221000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6731651651651652e-05,
+      "loss": 0.916,
+      "step": 221500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6701621621621624e-05,
+      "loss": 0.9283,
+      "step": 222000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6671591591591592e-05,
+      "loss": 0.912,
+      "step": 222500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.664156156156156e-05,
+      "loss": 0.9138,
+      "step": 223000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.661153153153153e-05,
+      "loss": 0.924,
+      "step": 223500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6581501501501503e-05,
+      "loss": 0.9147,
+      "step": 224000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6551471471471474e-05,
+      "loss": 0.9228,
+      "step": 224500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6521441441441442e-05,
+      "loss": 0.9203,
+      "step": 225000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.649141141141141e-05,
+      "loss": 0.9118,
+      "step": 225500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.646138138138138e-05,
+      "loss": 0.9151,
+      "step": 226000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6431351351351353e-05,
+      "loss": 0.9149,
+      "step": 226500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6401321321321324e-05,
+      "loss": 0.9162,
+      "step": 227000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.6371471471471472e-05,
+      "loss": 0.9191,
+      "step": 227500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.63415015015015e-05,
+      "loss": 0.9146,
+      "step": 228000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.6311471471471473e-05,
+      "loss": 0.9198,
+      "step": 228500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.628144144144144e-05,
+      "loss": 0.9175,
+      "step": 229000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.6251411411411412e-05,
+      "loss": 0.9145,
+      "step": 229500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.622138138138138e-05,
+      "loss": 0.9115,
+      "step": 230000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.8521190881729126,
+      "eval_runtime": 576.8956,
+      "eval_samples_per_second": 173.342,
+      "eval_steps_per_second": 43.335,
+      "step": 230000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.619141141141141e-05,
+      "loss": 0.9222,
+      "step": 230500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.616138138138138e-05,
+      "loss": 0.9218,
+      "step": 231000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.6131351351351352e-05,
+      "loss": 0.9173,
+      "step": 231500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.6101321321321324e-05,
+      "loss": 0.9215,
+      "step": 232000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.6071291291291292e-05,
+      "loss": 0.905,
+      "step": 232500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.604126126126126e-05,
+      "loss": 0.9172,
+      "step": 233000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.601123123123123e-05,
+      "loss": 0.9092,
+      "step": 233500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5981201201201203e-05,
+      "loss": 0.9137,
+      "step": 234000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5951171171171174e-05,
+      "loss": 0.9082,
+      "step": 234500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.592114114114114e-05,
+      "loss": 0.9221,
+      "step": 235000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.589111111111111e-05,
+      "loss": 0.9057,
+      "step": 235500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.586114114114114e-05,
+      "loss": 0.9145,
+      "step": 236000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.583111111111111e-05,
+      "loss": 0.9046,
+      "step": 236500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5801081081081082e-05,
+      "loss": 0.8992,
+      "step": 237000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5771051051051053e-05,
+      "loss": 0.9011,
+      "step": 237500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.574102102102102e-05,
+      "loss": 0.9066,
+      "step": 238000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.571099099099099e-05,
+      "loss": 0.9076,
+      "step": 238500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.568096096096096e-05,
+      "loss": 0.8947,
+      "step": 239000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5650930930930932e-05,
+      "loss": 0.8981,
+      "step": 239500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5620900900900904e-05,
+      "loss": 0.9177,
+      "step": 240000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.8470381498336792,
+      "eval_runtime": 609.4393,
+      "eval_samples_per_second": 164.085,
+      "eval_steps_per_second": 41.021,
+      "step": 240000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.559087087087087e-05,
+      "loss": 0.9091,
+      "step": 240500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.556084084084084e-05,
+      "loss": 0.9154,
+      "step": 241000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.553081081081081e-05,
+      "loss": 0.9091,
+      "step": 241500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5500780780780782e-05,
+      "loss": 0.9216,
+      "step": 242000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.547075075075075e-05,
+      "loss": 0.9147,
+      "step": 242500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5440720720720722e-05,
+      "loss": 0.9052,
+      "step": 243000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.541069069069069e-05,
+      "loss": 0.9056,
+      "step": 243500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5380720720720722e-05,
+      "loss": 0.9149,
+      "step": 244000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.535069069069069e-05,
+      "loss": 0.9082,
+      "step": 244500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5320660660660662e-05,
+      "loss": 0.9136,
+      "step": 245000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.529069069069069e-05,
+      "loss": 0.9214,
+      "step": 245500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.526066066066066e-05,
+      "loss": 0.9158,
+      "step": 246000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.523063063063063e-05,
+      "loss": 0.8907,
+      "step": 246500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.520066066066066e-05,
+      "loss": 0.8975,
+      "step": 247000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5170630630630631e-05,
+      "loss": 0.8965,
+      "step": 247500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5140600600600602e-05,
+      "loss": 0.8937,
+      "step": 248000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5110570570570572e-05,
+      "loss": 0.8949,
+      "step": 248500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.508054054054054e-05,
+      "loss": 0.9091,
+      "step": 249000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.505051051051051e-05,
+      "loss": 0.9114,
+      "step": 249500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5020480480480481e-05,
+      "loss": 0.8959,
+      "step": 250000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.8428720235824585,
+      "eval_runtime": 586.0492,
+      "eval_samples_per_second": 170.634,
+      "eval_steps_per_second": 42.659,
+      "step": 250000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4990510510510512e-05,
+      "loss": 0.904,
+      "step": 250500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.496048048048048e-05,
+      "loss": 0.9161,
+      "step": 251000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4930450450450451e-05,
+      "loss": 0.8939,
+      "step": 251500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4900420420420421e-05,
+      "loss": 0.9136,
+      "step": 252000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.487039039039039e-05,
+      "loss": 0.9042,
+      "step": 252500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.484036036036036e-05,
+      "loss": 0.9086,
+      "step": 253000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.481033033033033e-05,
+      "loss": 0.9049,
+      "step": 253500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4780300300300302e-05,
+      "loss": 0.9066,
+      "step": 254000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.475027027027027e-05,
+      "loss": 0.8934,
+      "step": 254500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4720240240240241e-05,
+      "loss": 0.909,
+      "step": 255000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.469021021021021e-05,
+      "loss": 0.8941,
+      "step": 255500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4660240240240242e-05,
+      "loss": 0.9059,
+      "step": 256000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.463021021021021e-05,
+      "loss": 0.9114,
+      "step": 256500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4600180180180181e-05,
+      "loss": 0.8954,
+      "step": 257000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.457015015015015e-05,
+      "loss": 0.9018,
+      "step": 257500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.454018018018018e-05,
+      "loss": 0.9021,
+      "step": 258000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4510150150150151e-05,
+      "loss": 0.8995,
+      "step": 258500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.448012012012012e-05,
+      "loss": 0.909,
+      "step": 259000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4450150150150152e-05,
+      "loss": 0.8923,
+      "step": 259500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.442012012012012e-05,
+      "loss": 0.9074,
+      "step": 260000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.8390381932258606,
+      "eval_runtime": 573.4535,
+      "eval_samples_per_second": 174.382,
+      "eval_steps_per_second": 43.596,
+      "step": 260000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4390090090090091e-05,
+      "loss": 0.9006,
+      "step": 260500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4360060060060061e-05,
+      "loss": 0.9005,
+      "step": 261000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.433003003003003e-05,
+      "loss": 0.9141,
+      "step": 261500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.43e-05,
+      "loss": 0.9145,
+      "step": 262000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.426996996996997e-05,
+      "loss": 0.899,
+      "step": 262500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.423993993993994e-05,
+      "loss": 0.9111,
+      "step": 263000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.420996996996997e-05,
+      "loss": 0.906,
+      "step": 263500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.417993993993994e-05,
+      "loss": 0.8911,
+      "step": 264000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4149909909909912e-05,
+      "loss": 0.8914,
+      "step": 264500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.411987987987988e-05,
+      "loss": 0.8978,
+      "step": 265000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4089849849849851e-05,
+      "loss": 0.8921,
+      "step": 265500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4059819819819819e-05,
+      "loss": 0.8938,
+      "step": 266000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.402978978978979e-05,
+      "loss": 0.9009,
+      "step": 266500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.3999759759759759e-05,
+      "loss": 0.8957,
+      "step": 267000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.396984984984985e-05,
+      "loss": 0.8944,
+      "step": 267500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.393981981981982e-05,
+      "loss": 0.8903,
+      "step": 268000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.390978978978979e-05,
+      "loss": 0.8932,
+      "step": 268500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.3879759759759761e-05,
+      "loss": 0.8853,
+      "step": 269000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.384972972972973e-05,
+      "loss": 0.8908,
+      "step": 269500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.38196996996997e-05,
+      "loss": 0.8888,
+      "step": 270000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.8343602418899536,
+      "eval_runtime": 589.1233,
+      "eval_samples_per_second": 169.744,
+      "eval_steps_per_second": 42.436,
+      "step": 270000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.3789669669669669e-05,
+      "loss": 0.8947,
+      "step": 270500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.375963963963964e-05,
+      "loss": 0.8939,
+      "step": 271000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.3729609609609612e-05,
+      "loss": 0.9018,
+      "step": 271500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.369957957957958e-05,
+      "loss": 0.9036,
+      "step": 272000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.3669549549549551e-05,
+      "loss": 0.8973,
+      "step": 272500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.3639519519519519e-05,
+      "loss": 0.885,
+      "step": 273000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.360954954954955e-05,
+      "loss": 0.8845,
+      "step": 273500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.357951951951952e-05,
+      "loss": 0.8789,
+      "step": 274000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.354948948948949e-05,
+      "loss": 0.8986,
+      "step": 274500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.351945945945946e-05,
+      "loss": 0.8909,
+      "step": 275000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.348942942942943e-05,
+      "loss": 0.9002,
+      "step": 275500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3459459459459461e-05,
+      "loss": 0.8946,
+      "step": 276000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.342942942942943e-05,
+      "loss": 0.8911,
+      "step": 276500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.33993993993994e-05,
+      "loss": 0.891,
+      "step": 277000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3369369369369369e-05,
+      "loss": 0.8833,
+      "step": 277500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.33393993993994e-05,
+      "loss": 0.8863,
+      "step": 278000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.330936936936937e-05,
+      "loss": 0.8765,
+      "step": 278500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.32793993993994e-05,
+      "loss": 0.8945,
+      "step": 279000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.324936936936937e-05,
+      "loss": 0.8936,
+      "step": 279500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.321933933933934e-05,
+      "loss": 0.9012,
+      "step": 280000
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.8308248519897461,
+      "eval_runtime": 569.4338,
+      "eval_samples_per_second": 175.613,
+      "eval_steps_per_second": 43.903,
+      "step": 280000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3189309309309311e-05,
+      "loss": 0.8835,
+      "step": 280500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3159279279279279e-05,
+      "loss": 0.8866,
+      "step": 281000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.312924924924925e-05,
+      "loss": 0.8876,
+      "step": 281500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.309927927927928e-05,
+      "loss": 0.8846,
+      "step": 282000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.306924924924925e-05,
+      "loss": 0.8833,
+      "step": 282500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3039219219219219e-05,
+      "loss": 0.8993,
+      "step": 283000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.300918918918919e-05,
+      "loss": 0.8802,
+      "step": 283500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.297915915915916e-05,
+      "loss": 0.8787,
+      "step": 284000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.294912912912913e-05,
+      "loss": 0.8809,
+      "step": 284500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.291915915915916e-05,
+      "loss": 0.8848,
+      "step": 285000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2889129129129129e-05,
+      "loss": 0.8847,
+      "step": 285500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.28590990990991e-05,
+      "loss": 0.891,
+      "step": 286000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2829069069069068e-05,
+      "loss": 0.8789,
+      "step": 286500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.279903903903904e-05,
+      "loss": 0.8962,
+      "step": 287000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.276900900900901e-05,
+      "loss": 0.8931,
+      "step": 287500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2738978978978979e-05,
+      "loss": 0.8789,
+      "step": 288000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.270900900900901e-05,
+      "loss": 0.8914,
+      "step": 288500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.267897897897898e-05,
+      "loss": 0.8902,
+      "step": 289000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2648948948948949e-05,
+      "loss": 0.8773,
+      "step": 289500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2618918918918919e-05,
+      "loss": 0.8996,
+      "step": 290000
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.8253816366195679,
+      "eval_runtime": 603.3986,
+      "eval_samples_per_second": 165.728,
+      "eval_steps_per_second": 41.432,
+      "step": 290000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2588888888888888e-05,
+      "loss": 0.8824,
+      "step": 290500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.255885885885886e-05,
+      "loss": 0.8916,
+      "step": 291000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.252882882882883e-05,
+      "loss": 0.887,
+      "step": 291500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.24987987987988e-05,
+      "loss": 0.8808,
+      "step": 292000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2468828828828828e-05,
+      "loss": 0.8912,
+      "step": 292500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.24387987987988e-05,
+      "loss": 0.8746,
+      "step": 293000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.2408768768768768e-05,
+      "loss": 0.8738,
+      "step": 293500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.237873873873874e-05,
+      "loss": 0.8837,
+      "step": 294000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.2348768768768768e-05,
+      "loss": 0.8826,
+      "step": 294500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.231873873873874e-05,
+      "loss": 0.875,
+      "step": 295000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.228870870870871e-05,
+      "loss": 0.8842,
+      "step": 295500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.225867867867868e-05,
+      "loss": 0.8778,
+      "step": 296000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.222870870870871e-05,
+      "loss": 0.8771,
+      "step": 296500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.2198678678678678e-05,
+      "loss": 0.8934,
+      "step": 297000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.216864864864865e-05,
+      "loss": 0.881,
+      "step": 297500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.2138618618618617e-05,
+      "loss": 0.8993,
+      "step": 298000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.2108648648648648e-05,
+      "loss": 0.8738,
+      "step": 298500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.2078618618618618e-05,
+      "loss": 0.8851,
+      "step": 299000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.204858858858859e-05,
+      "loss": 0.8802,
+      "step": 299500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.201855855855856e-05,
+      "loss": 0.8783,
+      "step": 300000
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.8227924108505249,
+      "eval_runtime": 588.0109,
+      "eval_samples_per_second": 170.065,
+      "eval_steps_per_second": 42.516,
+      "step": 300000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.198864864864865e-05,
+      "loss": 0.8749,
+      "step": 300500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.195861861861862e-05,
+      "loss": 0.8893,
+      "step": 301000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.1928588588588589e-05,
+      "loss": 0.8715,
+      "step": 301500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.1898558558558559e-05,
+      "loss": 0.8894,
+      "step": 302000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.1868588588588588e-05,
+      "loss": 0.875,
+      "step": 302500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.183855855855856e-05,
+      "loss": 0.8829,
+      "step": 303000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1808528528528529e-05,
+      "loss": 0.8872,
+      "step": 303500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1778498498498499e-05,
+      "loss": 0.8888,
+      "step": 304000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.174846846846847e-05,
+      "loss": 0.8716,
+      "step": 304500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1718438438438438e-05,
+      "loss": 0.8725,
+      "step": 305000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.168840840840841e-05,
+      "loss": 0.8749,
+      "step": 305500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1658378378378377e-05,
+      "loss": 0.8823,
+      "step": 306000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1628348348348349e-05,
+      "loss": 0.8788,
+      "step": 306500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1598378378378378e-05,
+      "loss": 0.8827,
+      "step": 307000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.156834834834835e-05,
+      "loss": 0.8728,
+      "step": 307500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1538318318318319e-05,
+      "loss": 0.8875,
+      "step": 308000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1508288288288289e-05,
+      "loss": 0.8812,
+      "step": 308500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1478378378378377e-05,
+      "loss": 0.8671,
+      "step": 309000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1448348348348349e-05,
+      "loss": 0.8723,
+      "step": 309500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1418318318318319e-05,
+      "loss": 0.8796,
+      "step": 310000
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.8185199499130249,
+      "eval_runtime": 589.4368,
+      "eval_samples_per_second": 169.653,
+      "eval_steps_per_second": 42.413,
+      "step": 310000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1388288288288288e-05,
+      "loss": 0.8775,
+      "step": 310500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1358318318318319e-05,
+      "loss": 0.875,
+      "step": 311000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1328288288288289e-05,
+      "loss": 0.8852,
+      "step": 311500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.1298258258258259e-05,
+      "loss": 0.877,
+      "step": 312000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1268228228228228e-05,
+      "loss": 0.8655,
+      "step": 312500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1238198198198198e-05,
+      "loss": 0.8779,
+      "step": 313000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1208228228228227e-05,
+      "loss": 0.8751,
+      "step": 313500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1178198198198199e-05,
+      "loss": 0.8678,
+      "step": 314000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.114816816816817e-05,
+      "loss": 0.8676,
+      "step": 314500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1118138138138138e-05,
+      "loss": 0.8695,
+      "step": 315000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.108810810810811e-05,
+      "loss": 0.8735,
+      "step": 315500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1058078078078077e-05,
+      "loss": 0.8809,
+      "step": 316000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1028048048048049e-05,
+      "loss": 0.8762,
+      "step": 316500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0998018018018018e-05,
+      "loss": 0.871,
+      "step": 317000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0968108108108109e-05,
+      "loss": 0.8726,
+      "step": 317500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0938078078078077e-05,
+      "loss": 0.8649,
+      "step": 318000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0908048048048048e-05,
+      "loss": 0.8849,
+      "step": 318500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.087801801801802e-05,
+      "loss": 0.8812,
+      "step": 319000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0848048048048049e-05,
+      "loss": 0.862,
+      "step": 319500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0818018018018018e-05,
+      "loss": 0.8739,
+      "step": 320000
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.8156814575195312,
+      "eval_runtime": 622.3771,
+      "eval_samples_per_second": 160.674,
+      "eval_steps_per_second": 40.169,
+      "step": 320000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0787987987987988e-05,
+      "loss": 0.8759,
+      "step": 320500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0757957957957958e-05,
+      "loss": 0.8817,
+      "step": 321000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.0727927927927928e-05,
+      "loss": 0.8834,
+      "step": 321500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0697897897897897e-05,
+      "loss": 0.8693,
+      "step": 322000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0667867867867869e-05,
+      "loss": 0.8618,
+      "step": 322500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0637837837837838e-05,
+      "loss": 0.8678,
+      "step": 323000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0607807807807808e-05,
+      "loss": 0.8805,
+      "step": 323500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0577777777777778e-05,
+      "loss": 0.8654,
+      "step": 324000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0547747747747747e-05,
+      "loss": 0.8551,
+      "step": 324500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0517717717717719e-05,
+      "loss": 0.8831,
+      "step": 325000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0487747747747748e-05,
+      "loss": 0.8771,
+      "step": 325500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0457717717717718e-05,
+      "loss": 0.8717,
+      "step": 326000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0427747747747749e-05,
+      "loss": 0.8588,
+      "step": 326500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0397717717717718e-05,
+      "loss": 0.882,
+      "step": 327000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0367687687687688e-05,
+      "loss": 0.8688,
+      "step": 327500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0337657657657658e-05,
+      "loss": 0.8675,
+      "step": 328000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0307627627627627e-05,
+      "loss": 0.8583,
+      "step": 328500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0277597597597597e-05,
+      "loss": 0.8745,
+      "step": 329000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0247567567567569e-05,
+      "loss": 0.8631,
+      "step": 329500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0217537537537537e-05,
+      "loss": 0.8637,
+      "step": 330000
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.8101323246955872,
+      "eval_runtime": 593.644,
+      "eval_samples_per_second": 168.451,
+      "eval_steps_per_second": 42.113,
+      "step": 330000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0187567567567569e-05,
+      "loss": 0.8626,
+      "step": 330500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.0157597597597598e-05,
+      "loss": 0.8651,
+      "step": 331000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.012762762762763e-05,
+      "loss": 0.8706,
+      "step": 331500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.0097597597597597e-05,
+      "loss": 0.8766,
+      "step": 332000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.0067567567567569e-05,
+      "loss": 0.8753,
+      "step": 332500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.0037537537537537e-05,
+      "loss": 0.8735,
+      "step": 333000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.0007567567567567e-05,
+      "loss": 0.8747,
+      "step": 333500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.977537537537537e-06,
+      "loss": 0.8596,
+      "step": 334000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.947507507507509e-06,
+      "loss": 0.868,
+      "step": 334500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.917477477477478e-06,
+      "loss": 0.8683,
+      "step": 335000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.887447447447448e-06,
+      "loss": 0.8763,
+      "step": 335500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.857417417417418e-06,
+      "loss": 0.8638,
+      "step": 336000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.827387387387387e-06,
+      "loss": 0.8677,
+      "step": 336500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.797357357357357e-06,
+      "loss": 0.8753,
+      "step": 337000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.767327327327328e-06,
+      "loss": 0.8708,
+      "step": 337500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.737297297297298e-06,
+      "loss": 0.8722,
+      "step": 338000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.707327327327329e-06,
+      "loss": 0.877,
+      "step": 338500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.677357357357358e-06,
+      "loss": 0.8641,
+      "step": 339000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.647327327327328e-06,
+      "loss": 0.8656,
+      "step": 339500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.617297297297298e-06,
+      "loss": 0.87,
+      "step": 340000
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.8081399202346802,
+      "eval_runtime": 594.2224,
+      "eval_samples_per_second": 168.287,
+      "eval_steps_per_second": 42.072,
+      "step": 340000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.587267267267267e-06,
+      "loss": 0.8514,
+      "step": 340500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.557237237237237e-06,
+      "loss": 0.8726,
+      "step": 341000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.527207207207207e-06,
+      "loss": 0.8674,
+      "step": 341500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.497177177177178e-06,
+      "loss": 0.8656,
+      "step": 342000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.467147147147148e-06,
+      "loss": 0.8688,
+      "step": 342500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.437177177177179e-06,
+      "loss": 0.8619,
+      "step": 343000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.407147147147147e-06,
+      "loss": 0.8576,
+      "step": 343500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.377117117117118e-06,
+      "loss": 0.861,
+      "step": 344000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.347087087087086e-06,
+      "loss": 0.8655,
+      "step": 344500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.317057057057058e-06,
+      "loss": 0.8534,
+      "step": 345000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.287087087087087e-06,
+      "loss": 0.8768,
+      "step": 345500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.257057057057058e-06,
+      "loss": 0.8584,
+      "step": 346000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.227027027027028e-06,
+      "loss": 0.8716,
+      "step": 346500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.196996996996997e-06,
+      "loss": 0.8568,
+      "step": 347000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.166966966966967e-06,
+      "loss": 0.8503,
+      "step": 347500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.136936936936937e-06,
+      "loss": 0.8573,
+      "step": 348000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.106906906906907e-06,
+      "loss": 0.8619,
+      "step": 348500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.076936936936936e-06,
+      "loss": 0.8588,
+      "step": 349000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.046906906906907e-06,
+      "loss": 0.8632,
+      "step": 349500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.016876876876879e-06,
+      "loss": 0.8517,
+      "step": 350000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.8035141229629517,
+      "eval_runtime": 567.0359,
+      "eval_samples_per_second": 176.356,
+      "eval_steps_per_second": 44.089,
+      "step": 350000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.986846846846847e-06,
+      "loss": 0.8759,
+      "step": 350500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.956816816816818e-06,
+      "loss": 0.8636,
+      "step": 351000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.926906906906907e-06,
+      "loss": 0.8523,
+      "step": 351500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.896876876876878e-06,
+      "loss": 0.8508,
+      "step": 352000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.866906906906907e-06,
+      "loss": 0.86,
+      "step": 352500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.836876876876877e-06,
+      "loss": 0.8693,
+      "step": 353000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.806846846846847e-06,
+      "loss": 0.8597,
+      "step": 353500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.776816816816818e-06,
+      "loss": 0.8509,
+      "step": 354000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.746786786786786e-06,
+      "loss": 0.8521,
+      "step": 354500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.716756756756757e-06,
+      "loss": 0.847,
+      "step": 355000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.686726726726727e-06,
+      "loss": 0.8739,
+      "step": 355500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.656696696696697e-06,
+      "loss": 0.8625,
+      "step": 356000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.626666666666667e-06,
+      "loss": 0.866,
+      "step": 356500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.596636636636636e-06,
+      "loss": 0.8592,
+      "step": 357000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.566606606606608e-06,
+      "loss": 0.8673,
+      "step": 357500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.536576576576577e-06,
+      "loss": 0.8658,
+      "step": 358000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.506546546546547e-06,
+      "loss": 0.8646,
+      "step": 358500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.476576576576578e-06,
+      "loss": 0.8572,
+      "step": 359000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.446546546546546e-06,
+      "loss": 0.8656,
+      "step": 359500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.416516516516517e-06,
+      "loss": 0.8523,
+      "step": 360000
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.802210807800293,
+      "eval_runtime": 630.0985,
+      "eval_samples_per_second": 158.705,
+      "eval_steps_per_second": 39.676,
+      "step": 360000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.386486486486485e-06,
+      "loss": 0.8687,
+      "step": 360500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.356456456456457e-06,
+      "loss": 0.8481,
+      "step": 361000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.326426426426428e-06,
+      "loss": 0.87,
+      "step": 361500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.296456456456457e-06,
+      "loss": 0.8499,
+      "step": 362000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.266486486486488e-06,
+      "loss": 0.8482,
+      "step": 362500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.236456456456456e-06,
+      "loss": 0.8452,
+      "step": 363000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.206426426426428e-06,
+      "loss": 0.8575,
+      "step": 363500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.176456456456457e-06,
+      "loss": 0.8567,
+      "step": 364000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.146426426426426e-06,
+      "loss": 0.8449,
+      "step": 364500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.116396396396396e-06,
+      "loss": 0.8634,
+      "step": 365000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.086366366366368e-06,
+      "loss": 0.8552,
+      "step": 365500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.056336336336337e-06,
+      "loss": 0.8593,
+      "step": 366000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.026306306306307e-06,
+      "loss": 0.846,
+      "step": 366500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.996276276276277e-06,
+      "loss": 0.8575,
+      "step": 367000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.966246246246246e-06,
+      "loss": 0.8659,
+      "step": 367500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.936216216216216e-06,
+      "loss": 0.8543,
+      "step": 368000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.906186186186186e-06,
+      "loss": 0.8501,
+      "step": 368500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.876156156156155e-06,
+      "loss": 0.8605,
+      "step": 369000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.846126126126127e-06,
+      "loss": 0.8541,
+      "step": 369500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.816096096096097e-06,
+      "loss": 0.85,
+      "step": 370000
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.798128604888916,
+      "eval_runtime": 622.5637,
+      "eval_samples_per_second": 160.626,
+      "eval_steps_per_second": 40.157,
+      "step": 370000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.786126126126127e-06,
+      "loss": 0.8495,
+      "step": 370500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.756096096096095e-06,
+      "loss": 0.8448,
+      "step": 371000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.726066066066067e-06,
+      "loss": 0.8508,
+      "step": 371500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.696036036036035e-06,
+      "loss": 0.8564,
+      "step": 372000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.666006006006006e-06,
+      "loss": 0.8434,
+      "step": 372500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.636036036036037e-06,
+      "loss": 0.853,
+      "step": 373000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.606006006006006e-06,
+      "loss": 0.8536,
+      "step": 373500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.5759759759759765e-06,
+      "loss": 0.8416,
+      "step": 374000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.545945945945945e-06,
+      "loss": 0.8509,
+      "step": 374500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.515915915915916e-06,
+      "loss": 0.852,
+      "step": 375000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.4858858858858865e-06,
+      "loss": 0.8481,
+      "step": 375500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.455855855855856e-06,
+      "loss": 0.8505,
+      "step": 376000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.425825825825826e-06,
+      "loss": 0.8425,
+      "step": 376500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.395855855855856e-06,
+      "loss": 0.8514,
+      "step": 377000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.365825825825826e-06,
+      "loss": 0.8478,
+      "step": 377500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.335795795795796e-06,
+      "loss": 0.8468,
+      "step": 378000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.305765765765766e-06,
+      "loss": 0.8621,
+      "step": 378500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.275795795795796e-06,
+      "loss": 0.8486,
+      "step": 379000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.245825825825827e-06,
+      "loss": 0.8491,
+      "step": 379500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.2157957957957965e-06,
+      "loss": 0.8329,
+      "step": 380000
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.795859694480896,
+      "eval_runtime": 607.5667,
+      "eval_samples_per_second": 164.591,
+      "eval_steps_per_second": 41.148,
+      "step": 380000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.185765765765766e-06,
+      "loss": 0.852,
+      "step": 380500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.155735735735736e-06,
+      "loss": 0.8493,
+      "step": 381000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.125765765765766e-06,
+      "loss": 0.8506,
+      "step": 381500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.095735735735736e-06,
+      "loss": 0.8537,
+      "step": 382000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.065765765765766e-06,
+      "loss": 0.8387,
+      "step": 382500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.035735735735736e-06,
+      "loss": 0.844,
+      "step": 383000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 7.005705705705706e-06,
+      "loss": 0.8467,
+      "step": 383500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.975675675675676e-06,
+      "loss": 0.8499,
+      "step": 384000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.945645645645646e-06,
+      "loss": 0.8463,
+      "step": 384500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.915615615615616e-06,
+      "loss": 0.8336,
+      "step": 385000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.8855855855855855e-06,
+      "loss": 0.8423,
+      "step": 385500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.855555555555555e-06,
+      "loss": 0.8323,
+      "step": 386000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.825525525525526e-06,
+      "loss": 0.8372,
+      "step": 386500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.795495495495496e-06,
+      "loss": 0.8564,
+      "step": 387000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.765465465465466e-06,
+      "loss": 0.8527,
+      "step": 387500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.735435435435436e-06,
+      "loss": 0.8376,
+      "step": 388000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.7054054054054054e-06,
+      "loss": 0.835,
+      "step": 388500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.675375375375375e-06,
+      "loss": 0.8449,
+      "step": 389000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.645345345345346e-06,
+      "loss": 0.8472,
+      "step": 389500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.615315315315315e-06,
+      "loss": 0.8473,
+      "step": 390000
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.7919498682022095,
+      "eval_runtime": 570.4158,
+      "eval_samples_per_second": 175.311,
+      "eval_steps_per_second": 43.828,
+      "step": 390000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.585285285285285e-06,
+      "loss": 0.8466,
+      "step": 390500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.555315315315316e-06,
+      "loss": 0.8593,
+      "step": 391000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.525285285285286e-06,
+      "loss": 0.8513,
+      "step": 391500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.495255255255255e-06,
+      "loss": 0.8398,
+      "step": 392000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.465225225225225e-06,
+      "loss": 0.8447,
+      "step": 392500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.4353153153153154e-06,
+      "loss": 0.8523,
+      "step": 393000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.405285285285285e-06,
+      "loss": 0.8435,
+      "step": 393500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.375255255255255e-06,
+      "loss": 0.8606,
+      "step": 394000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.345225225225225e-06,
+      "loss": 0.8538,
+      "step": 394500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.315195195195196e-06,
+      "loss": 0.8616,
+      "step": 395000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.285165165165166e-06,
+      "loss": 0.8418,
+      "step": 395500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.255135135135135e-06,
+      "loss": 0.8449,
+      "step": 396000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.225105105105105e-06,
+      "loss": 0.8382,
+      "step": 396500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.195135135135135e-06,
+      "loss": 0.8516,
+      "step": 397000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.165105105105105e-06,
+      "loss": 0.8472,
+      "step": 397500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.135135135135135e-06,
+      "loss": 0.8354,
+      "step": 398000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.1051051051051045e-06,
+      "loss": 0.856,
+      "step": 398500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.075075075075076e-06,
+      "loss": 0.8292,
+      "step": 399000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.045045045045046e-06,
+      "loss": 0.8403,
+      "step": 399500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.015015015015015e-06,
+      "loss": 0.8495,
+      "step": 400000
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.7893297672271729,
+      "eval_runtime": 586.3311,
+      "eval_samples_per_second": 170.552,
+      "eval_steps_per_second": 42.638,
+      "step": 400000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.985045045045045e-06,
+      "loss": 0.8557,
+      "step": 400500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.955015015015015e-06,
+      "loss": 0.8496,
+      "step": 401000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.925045045045045e-06,
+      "loss": 0.8406,
+      "step": 401500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.895015015015015e-06,
+      "loss": 0.8466,
+      "step": 402000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8649849849849845e-06,
+      "loss": 0.8401,
+      "step": 402500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.834954954954956e-06,
+      "loss": 0.8327,
+      "step": 403000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8049249249249256e-06,
+      "loss": 0.8292,
+      "step": 403500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.774894894894895e-06,
+      "loss": 0.8479,
+      "step": 404000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.744864864864865e-06,
+      "loss": 0.8425,
+      "step": 404500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.714834834834835e-06,
+      "loss": 0.8512,
+      "step": 405000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.684804804804804e-06,
+      "loss": 0.8414,
+      "step": 405500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.654774774774775e-06,
+      "loss": 0.8307,
+      "step": 406000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.624744744744745e-06,
+      "loss": 0.829,
+      "step": 406500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.5947747747747755e-06,
+      "loss": 0.8441,
+      "step": 407000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.564744744744745e-06,
+      "loss": 0.8467,
+      "step": 407500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.534714714714715e-06,
+      "loss": 0.8505,
+      "step": 408000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.504684684684685e-06,
+      "loss": 0.8398,
+      "step": 408500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.474714714714715e-06,
+      "loss": 0.8352,
+      "step": 409000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.444684684684684e-06,
+      "loss": 0.8452,
+      "step": 409500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.414654654654655e-06,
+      "loss": 0.8468,
+      "step": 410000
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.7882465124130249,
+      "eval_runtime": 564.1275,
+      "eval_samples_per_second": 177.265,
+      "eval_steps_per_second": 44.316,
+      "step": 410000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.384684684684685e-06,
+      "loss": 0.8428,
+      "step": 410500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3546546546546555e-06,
+      "loss": 0.8524,
+      "step": 411000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.324624624624625e-06,
+      "loss": 0.8399,
+      "step": 411500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.294594594594595e-06,
+      "loss": 0.8295,
+      "step": 412000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.2645645645645646e-06,
+      "loss": 0.8332,
+      "step": 412500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.234534534534534e-06,
+      "loss": 0.8444,
+      "step": 413000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.204504504504505e-06,
+      "loss": 0.8372,
+      "step": 413500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.1744744744744745e-06,
+      "loss": 0.8402,
+      "step": 414000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.144444444444445e-06,
+      "loss": 0.8521,
+      "step": 414500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.114414414414415e-06,
+      "loss": 0.8418,
+      "step": 415000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.0843843843843845e-06,
+      "loss": 0.8454,
+      "step": 415500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.0544144144144145e-06,
+      "loss": 0.8176,
+      "step": 416000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.024384384384384e-06,
+      "loss": 0.8556,
+      "step": 416500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.994354354354355e-06,
+      "loss": 0.8245,
+      "step": 417000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.9643243243243245e-06,
+      "loss": 0.8418,
+      "step": 417500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.934294294294294e-06,
+      "loss": 0.8351,
+      "step": 418000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.904264264264265e-06,
+      "loss": 0.841,
+      "step": 418500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.874234234234234e-06,
+      "loss": 0.8335,
+      "step": 419000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.844204204204204e-06,
+      "loss": 0.8299,
+      "step": 419500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.814234234234234e-06,
+      "loss": 0.8355,
+      "step": 420000
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.7847426533699036,
+      "eval_runtime": 565.6865,
+      "eval_samples_per_second": 176.776,
+      "eval_steps_per_second": 44.194,
+      "step": 420000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.784204204204205e-06,
+      "loss": 0.8216,
+      "step": 420500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.754174174174174e-06,
+      "loss": 0.8486,
+      "step": 421000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.724144144144144e-06,
+      "loss": 0.8411,
+      "step": 421500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.694114114114114e-06,
+      "loss": 0.8447,
+      "step": 422000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.664144144144145e-06,
+      "loss": 0.8363,
+      "step": 422500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.634114114114114e-06,
+      "loss": 0.83,
+      "step": 423000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.604084084084084e-06,
+      "loss": 0.8398,
+      "step": 423500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.574114114114114e-06,
+      "loss": 0.8441,
+      "step": 424000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.544084084084084e-06,
+      "loss": 0.8373,
+      "step": 424500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.514054054054054e-06,
+      "loss": 0.8403,
+      "step": 425000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.484024024024024e-06,
+      "loss": 0.8391,
+      "step": 425500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.453993993993994e-06,
+      "loss": 0.832,
+      "step": 426000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.423963963963964e-06,
+      "loss": 0.8524,
+      "step": 426500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.393933933933934e-06,
+      "loss": 0.8396,
+      "step": 427000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.3639039039039046e-06,
+      "loss": 0.8363,
+      "step": 427500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.333993993993994e-06,
+      "loss": 0.834,
+      "step": 428000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.303963963963964e-06,
+      "loss": 0.8476,
+      "step": 428500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2739339339339335e-06,
+      "loss": 0.8509,
+      "step": 429000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.243903903903904e-06,
+      "loss": 0.8368,
+      "step": 429500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.213873873873874e-06,
+      "loss": 0.8366,
+      "step": 430000
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.7822093963623047,
+      "eval_runtime": 571.4633,
+      "eval_samples_per_second": 174.989,
+      "eval_steps_per_second": 43.747,
+      "step": 430000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.183843843843844e-06,
+      "loss": 0.8383,
+      "step": 430500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.153813813813814e-06,
+      "loss": 0.8404,
+      "step": 431000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.123783783783784e-06,
+      "loss": 0.8348,
+      "step": 431500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.093813813813814e-06,
+      "loss": 0.8332,
+      "step": 432000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.063783783783783e-06,
+      "loss": 0.8323,
+      "step": 432500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.033753753753754e-06,
+      "loss": 0.8299,
+      "step": 433000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.003783783783784e-06,
+      "loss": 0.831,
+      "step": 433500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.973813813813814e-06,
+      "loss": 0.837,
+      "step": 434000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.9437837837837846e-06,
+      "loss": 0.8455,
+      "step": 434500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.913753753753754e-06,
+      "loss": 0.8317,
+      "step": 435000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.883723723723724e-06,
+      "loss": 0.817,
+      "step": 435500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.853693693693694e-06,
+      "loss": 0.8346,
+      "step": 436000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.823663663663663e-06,
+      "loss": 0.8277,
+      "step": 436500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.7936336336336343e-06,
+      "loss": 0.8357,
+      "step": 437000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.763603603603604e-06,
+      "loss": 0.8382,
+      "step": 437500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.7335735735735737e-06,
+      "loss": 0.8501,
+      "step": 438000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.7035435435435434e-06,
+      "loss": 0.8233,
+      "step": 438500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.673513513513514e-06,
+      "loss": 0.8342,
+      "step": 439000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.6434834834834837e-06,
+      "loss": 0.8502,
+      "step": 439500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.6135135135135137e-06,
+      "loss": 0.8418,
+      "step": 440000
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.7793955206871033,
+      "eval_runtime": 566.9399,
+      "eval_samples_per_second": 176.386,
+      "eval_steps_per_second": 44.096,
+      "step": 440000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.5834834834834834e-06,
+      "loss": 0.8153,
+      "step": 440500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.5534534534534536e-06,
+      "loss": 0.8435,
+      "step": 441000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.5234234234234237e-06,
+      "loss": 0.8288,
+      "step": 441500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.4933933933933934e-06,
+      "loss": 0.8185,
+      "step": 442000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 3.4633633633633635e-06,
+      "loss": 0.819,
+      "step": 442500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.4333333333333336e-06,
+      "loss": 0.8296,
+      "step": 443000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.4033033033033033e-06,
+      "loss": 0.815,
+      "step": 443500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.3733933933933933e-06,
+      "loss": 0.8266,
+      "step": 444000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.3433633633633634e-06,
+      "loss": 0.8252,
+      "step": 444500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.3133333333333335e-06,
+      "loss": 0.8336,
+      "step": 445000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.2833033033033036e-06,
+      "loss": 0.8363,
+      "step": 445500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.2532732732732733e-06,
+      "loss": 0.8297,
+      "step": 446000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.2233033033033034e-06,
+      "loss": 0.8247,
+      "step": 446500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.1932732732732735e-06,
+      "loss": 0.8205,
+      "step": 447000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.163243243243243e-06,
+      "loss": 0.8311,
+      "step": 447500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.1332132132132133e-06,
+      "loss": 0.8398,
+      "step": 448000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.103183183183183e-06,
+      "loss": 0.8344,
+      "step": 448500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.0731531531531536e-06,
+      "loss": 0.8124,
+      "step": 449000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.043183183183183e-06,
+      "loss": 0.8206,
+      "step": 449500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.0131531531531533e-06,
+      "loss": 0.8271,
+      "step": 450000
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.7778518199920654,
+      "eval_runtime": 574.7064,
+      "eval_samples_per_second": 174.002,
+      "eval_steps_per_second": 43.5,
+      "step": 450000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.983123123123123e-06,
+      "loss": 0.8326,
+      "step": 450500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.953093093093093e-06,
+      "loss": 0.8222,
+      "step": 451000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9230630630630633e-06,
+      "loss": 0.8365,
+      "step": 451500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.893033033033033e-06,
+      "loss": 0.8341,
+      "step": 452000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.863003003003003e-06,
+      "loss": 0.8402,
+      "step": 452500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8329729729729732e-06,
+      "loss": 0.8316,
+      "step": 453000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8030030030030032e-06,
+      "loss": 0.8329,
+      "step": 453500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.772972972972973e-06,
+      "loss": 0.8108,
+      "step": 454000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.7429429429429426e-06,
+      "loss": 0.815,
+      "step": 454500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.712912912912913e-06,
+      "loss": 0.834,
+      "step": 455000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.683003003003003e-06,
+      "loss": 0.8178,
+      "step": 455500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.652972972972973e-06,
+      "loss": 0.8206,
+      "step": 456000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.622942942942943e-06,
+      "loss": 0.8338,
+      "step": 456500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.592912912912913e-06,
+      "loss": 0.825,
+      "step": 457000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5628828828828828e-06,
+      "loss": 0.818,
+      "step": 457500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.532852852852853e-06,
+      "loss": 0.8246,
+      "step": 458000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.502822822822823e-06,
+      "loss": 0.8359,
+      "step": 458500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.472852852852853e-06,
+      "loss": 0.8253,
+      "step": 459000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.4428228228228228e-06,
+      "loss": 0.8227,
+      "step": 459500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.412792792792793e-06,
+      "loss": 0.8394,
+      "step": 460000
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.7761884331703186,
+      "eval_runtime": 598.2224,
+      "eval_samples_per_second": 167.162,
+      "eval_steps_per_second": 41.79,
+      "step": 460000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.382762762762763e-06,
+      "loss": 0.8264,
+      "step": 460500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.3527327327327327e-06,
+      "loss": 0.8397,
+      "step": 461000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.3227627627627627e-06,
+      "loss": 0.8178,
+      "step": 461500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.292732732732733e-06,
+      "loss": 0.8224,
+      "step": 462000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.262702702702703e-06,
+      "loss": 0.8338,
+      "step": 462500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.2326726726726727e-06,
+      "loss": 0.8179,
+      "step": 463000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.2027027027027027e-06,
+      "loss": 0.8332,
+      "step": 463500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.1726726726726724e-06,
+      "loss": 0.8318,
+      "step": 464000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.142642642642643e-06,
+      "loss": 0.8298,
+      "step": 464500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.1126126126126127e-06,
+      "loss": 0.8298,
+      "step": 465000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.082582582582583e-06,
+      "loss": 0.818,
+      "step": 465500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.0525525525525525e-06,
+      "loss": 0.8202,
+      "step": 466000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.0225225225225226e-06,
+      "loss": 0.8174,
+      "step": 466500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9924924924924928e-06,
+      "loss": 0.8165,
+      "step": 467000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9625225225225224e-06,
+      "loss": 0.8184,
+      "step": 467500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9324924924924925e-06,
+      "loss": 0.823,
+      "step": 468000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9024624624624624e-06,
+      "loss": 0.8224,
+      "step": 468500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8724324324324325e-06,
+      "loss": 0.8266,
+      "step": 469000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8424024024024024e-06,
+      "loss": 0.8199,
+      "step": 469500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8123723723723726e-06,
+      "loss": 0.8319,
+      "step": 470000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.7738833427429199,
+      "eval_runtime": 565.2199,
+      "eval_samples_per_second": 176.922,
+      "eval_steps_per_second": 44.231,
+      "step": 470000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.7823423423423423e-06,
+      "loss": 0.834,
+      "step": 470500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7523123123123124e-06,
+      "loss": 0.8208,
+      "step": 471000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7223423423423424e-06,
+      "loss": 0.8136,
+      "step": 471500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6923123123123123e-06,
+      "loss": 0.8286,
+      "step": 472000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6622822822822823e-06,
+      "loss": 0.8284,
+      "step": 472500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6323123123123125e-06,
+      "loss": 0.8276,
+      "step": 473000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6022822822822822e-06,
+      "loss": 0.8387,
+      "step": 473500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5722522522522523e-06,
+      "loss": 0.825,
+      "step": 474000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5422222222222222e-06,
+      "loss": 0.8339,
+      "step": 474500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5121921921921924e-06,
+      "loss": 0.8279,
+      "step": 475000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.482162162162162e-06,
+      "loss": 0.8162,
+      "step": 475500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.4521321321321322e-06,
+      "loss": 0.8312,
+      "step": 476000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.422102102102102e-06,
+      "loss": 0.826,
+      "step": 476500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.3921321321321321e-06,
+      "loss": 0.8415,
+      "step": 477000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.362102102102102e-06,
+      "loss": 0.8175,
+      "step": 477500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.3320720720720722e-06,
+      "loss": 0.8104,
+      "step": 478000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.302042042042042e-06,
+      "loss": 0.8368,
+      "step": 478500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.272012012012012e-06,
+      "loss": 0.8226,
+      "step": 479000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.242042042042042e-06,
+      "loss": 0.819,
+      "step": 479500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.2120120120120121e-06,
+      "loss": 0.8125,
+      "step": 480000
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.7732232809066772,
+      "eval_runtime": 573.449,
+      "eval_samples_per_second": 174.383,
+      "eval_steps_per_second": 43.596,
+      "step": 480000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.1819819819819819e-06,
+      "loss": 0.8312,
+      "step": 480500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.151951951951952e-06,
+      "loss": 0.8299,
+      "step": 481000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.121981981981982e-06,
+      "loss": 0.8081,
+      "step": 481500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.091951951951952e-06,
+      "loss": 0.8267,
+      "step": 482000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.061981981981982e-06,
+      "loss": 0.8213,
+      "step": 482500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.031951951951952e-06,
+      "loss": 0.8024,
+      "step": 483000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.0019219219219218e-06,
+      "loss": 0.8223,
+      "step": 483500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.71891891891892e-07,
+      "loss": 0.8254,
+      "step": 484000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.418618618618619e-07,
+      "loss": 0.8258,
+      "step": 484500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.118318318318318e-07,
+      "loss": 0.8303,
+      "step": 485000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 8.818018018018019e-07,
+      "loss": 0.8197,
+      "step": 485500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 8.517717717717718e-07,
+      "loss": 0.824,
+      "step": 486000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 8.218018018018018e-07,
+      "loss": 0.8255,
+      "step": 486500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 7.917717717717718e-07,
+      "loss": 0.8235,
+      "step": 487000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 7.618018018018018e-07,
+      "loss": 0.8036,
+      "step": 487500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 7.317717717717718e-07,
+      "loss": 0.818,
+      "step": 488000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 7.017417417417418e-07,
+      "loss": 0.8402,
+      "step": 488500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 6.717117117117117e-07,
+      "loss": 0.8439,
+      "step": 489000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 6.416816816816817e-07,
+      "loss": 0.8121,
+      "step": 489500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 6.116516516516516e-07,
+      "loss": 0.8065,
+      "step": 490000
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.7721747756004333,
+      "eval_runtime": 582.9907,
+      "eval_samples_per_second": 171.529,
+      "eval_steps_per_second": 42.882,
+      "step": 490000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.816216216216216e-07,
+      "loss": 0.827,
+      "step": 490500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.515915915915916e-07,
+      "loss": 0.8138,
+      "step": 491000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.216216216216216e-07,
+      "loss": 0.8212,
+      "step": 491500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.915915915915916e-07,
+      "loss": 0.8166,
+      "step": 492000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6156156156156157e-07,
+      "loss": 0.8246,
+      "step": 492500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.3153153153153154e-07,
+      "loss": 0.821,
+      "step": 493000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.015015015015015e-07,
+      "loss": 0.8265,
+      "step": 493500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.7153153153153153e-07,
+      "loss": 0.8297,
+      "step": 494000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.415015015015015e-07,
+      "loss": 0.8053,
+      "step": 494500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 3.1147147147147147e-07,
+      "loss": 0.8171,
+      "step": 495000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.8144144144144143e-07,
+      "loss": 0.8261,
+      "step": 495500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.5147147147147146e-07,
+      "loss": 0.8216,
+      "step": 496000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.2144144144144145e-07,
+      "loss": 0.8195,
+      "step": 496500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9141141141141142e-07,
+      "loss": 0.8303,
+      "step": 497000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.613813813813814e-07,
+      "loss": 0.836,
+      "step": 497500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.3135135135135136e-07,
+      "loss": 0.833,
+      "step": 498000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.0132132132132131e-07,
+      "loss": 0.8264,
+      "step": 498500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 7.129129129129129e-08,
+      "loss": 0.8173,
+      "step": 499000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1261261261261266e-08,
+      "loss": 0.8153,
+      "step": 499500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.1291291291291292e-08,
+      "loss": 0.8383,
+      "step": 500000
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.7712512016296387,
+      "eval_runtime": 603.0742,
+      "eval_samples_per_second": 165.817,
+      "eval_steps_per_second": 41.454,
+      "step": 500000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4998139069534767e-05,
+      "loss": 0.8465,
+      "step": 500500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4983131565782893e-05,
+      "loss": 0.8468,
+      "step": 501000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4968124062031015e-05,
+      "loss": 0.8479,
+      "step": 501500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.495311655827914e-05,
+      "loss": 0.8614,
+      "step": 502000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4938139069534766e-05,
+      "loss": 0.8439,
+      "step": 502500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4923131565782892e-05,
+      "loss": 0.8538,
+      "step": 503000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4908124062031016e-05,
+      "loss": 0.8375,
+      "step": 503500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4893116558279141e-05,
+      "loss": 0.8551,
+      "step": 504000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4878109054527264e-05,
+      "loss": 0.8507,
+      "step": 504500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4863101550775387e-05,
+      "loss": 0.8539,
+      "step": 505000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4848094047023513e-05,
+      "loss": 0.8515,
+      "step": 505500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.483311655827914e-05,
+      "loss": 0.8389,
+      "step": 506000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4818109054527264e-05,
+      "loss": 0.8546,
+      "step": 506500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4803101550775388e-05,
+      "loss": 0.844,
+      "step": 507000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4788094047023512e-05,
+      "loss": 0.8389,
+      "step": 507500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.4773086543271636e-05,
+      "loss": 0.8542,
+      "step": 508000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.475807903951976e-05,
+      "loss": 0.8595,
+      "step": 508500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4743071535767885e-05,
+      "loss": 0.8535,
+      "step": 509000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4728064032016008e-05,
+      "loss": 0.8618,
+      "step": 509500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4713086543271636e-05,
+      "loss": 0.8531,
+      "step": 510000
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.7975139021873474,
+      "eval_runtime": 604.9535,
+      "eval_samples_per_second": 165.302,
+      "eval_steps_per_second": 41.325,
+      "step": 510000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4698079039519759e-05,
+      "loss": 0.844,
+      "step": 510500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4683071535767884e-05,
+      "loss": 0.8603,
+      "step": 511000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4668064032016008e-05,
+      "loss": 0.8658,
+      "step": 511500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4653056528264134e-05,
+      "loss": 0.8574,
+      "step": 512000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.463807903951976e-05,
+      "loss": 0.8587,
+      "step": 512500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4623071535767885e-05,
+      "loss": 0.8496,
+      "step": 513000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4608064032016009e-05,
+      "loss": 0.8558,
+      "step": 513500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4593056528264133e-05,
+      "loss": 0.8604,
+      "step": 514000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4578079039519762e-05,
+      "loss": 0.8619,
+      "step": 514500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4563101550775387e-05,
+      "loss": 0.8637,
+      "step": 515000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4548094047023513e-05,
+      "loss": 0.8512,
+      "step": 515500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4533086543271637e-05,
+      "loss": 0.8522,
+      "step": 516000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.451807903951976e-05,
+      "loss": 0.8531,
+      "step": 516500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4503071535767884e-05,
+      "loss": 0.859,
+      "step": 517000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4488064032016008e-05,
+      "loss": 0.8399,
+      "step": 517500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4473056528264132e-05,
+      "loss": 0.8597,
+      "step": 518000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4458049024512256e-05,
+      "loss": 0.8739,
+      "step": 518500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4443071535767885e-05,
+      "loss": 0.8528,
+      "step": 519000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4428064032016007e-05,
+      "loss": 0.8613,
+      "step": 519500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4413056528264133e-05,
+      "loss": 0.8647,
+      "step": 520000
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.8006455302238464,
+      "eval_runtime": 614.931,
+      "eval_samples_per_second": 162.62,
+      "eval_steps_per_second": 40.655,
+      "step": 520000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4398049024512257e-05,
+      "loss": 0.8663,
+      "step": 520500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4383071535767884e-05,
+      "loss": 0.8476,
+      "step": 521000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4368064032016008e-05,
+      "loss": 0.8491,
+      "step": 521500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4353056528264133e-05,
+      "loss": 0.8642,
+      "step": 522000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4338049024512256e-05,
+      "loss": 0.8615,
+      "step": 522500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4323071535767884e-05,
+      "loss": 0.8509,
+      "step": 523000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4308064032016008e-05,
+      "loss": 0.8413,
+      "step": 523500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4293056528264132e-05,
+      "loss": 0.8668,
+      "step": 524000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4278049024512256e-05,
+      "loss": 0.8579,
+      "step": 524500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.426304152076038e-05,
+      "loss": 0.8592,
+      "step": 525000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4248034017008506e-05,
+      "loss": 0.8657,
+      "step": 525500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4233026513256628e-05,
+      "loss": 0.8439,
+      "step": 526000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4218019009504754e-05,
+      "loss": 0.8613,
+      "step": 526500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4203041520760379e-05,
+      "loss": 0.8532,
+      "step": 527000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4188034017008505e-05,
+      "loss": 0.8533,
+      "step": 527500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4173026513256629e-05,
+      "loss": 0.8421,
+      "step": 528000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4158019009504754e-05,
+      "loss": 0.8463,
+      "step": 528500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4143041520760381e-05,
+      "loss": 0.8644,
+      "step": 529000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4128034017008505e-05,
+      "loss": 0.8666,
+      "step": 529500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4113056528264132e-05,
+      "loss": 0.8528,
+      "step": 530000
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.7983748316764832,
+      "eval_runtime": 645.3764,
+      "eval_samples_per_second": 154.948,
+      "eval_steps_per_second": 38.737,
+      "step": 530000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4098049024512256e-05,
+      "loss": 0.8359,
+      "step": 530500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4083041520760382e-05,
+      "loss": 0.8603,
+      "step": 531000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4068034017008504e-05,
+      "loss": 0.8479,
+      "step": 531500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4053026513256628e-05,
+      "loss": 0.8581,
+      "step": 532000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4038019009504754e-05,
+      "loss": 0.8553,
+      "step": 532500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.402304152076038e-05,
+      "loss": 0.8564,
+      "step": 533000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4008034017008505e-05,
+      "loss": 0.855,
+      "step": 533500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.3993026513256629e-05,
+      "loss": 0.8573,
+      "step": 534000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.3978019009504753e-05,
+      "loss": 0.8689,
+      "step": 534500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.3963041520760381e-05,
+      "loss": 0.8597,
+      "step": 535000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.3948034017008505e-05,
+      "loss": 0.8429,
+      "step": 535500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.3933026513256627e-05,
+      "loss": 0.8561,
+      "step": 536000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3918019009504753e-05,
+      "loss": 0.8651,
+      "step": 536500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3903011505752877e-05,
+      "loss": 0.8675,
+      "step": 537000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3888004002001001e-05,
+      "loss": 0.8466,
+      "step": 537500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3872996498249125e-05,
+      "loss": 0.8629,
+      "step": 538000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3857988994497249e-05,
+      "loss": 0.8458,
+      "step": 538500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3843011505752876e-05,
+      "loss": 0.8502,
+      "step": 539000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3828004002001e-05,
+      "loss": 0.8528,
+      "step": 539500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3812996498249125e-05,
+      "loss": 0.8557,
+      "step": 540000
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.7993029952049255,
+      "eval_runtime": 619.1901,
+      "eval_samples_per_second": 161.501,
+      "eval_steps_per_second": 40.375,
+      "step": 540000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.379798899449725e-05,
+      "loss": 0.8428,
+      "step": 540500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3783011505752876e-05,
+      "loss": 0.8375,
+      "step": 541000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3768004002001e-05,
+      "loss": 0.8569,
+      "step": 541500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3752996498249126e-05,
+      "loss": 0.8465,
+      "step": 542000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3738019009504753e-05,
+      "loss": 0.8545,
+      "step": 542500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3723011505752877e-05,
+      "loss": 0.8584,
+      "step": 543000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3708004002001001e-05,
+      "loss": 0.855,
+      "step": 543500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3692996498249125e-05,
+      "loss": 0.8386,
+      "step": 544000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3677988994497249e-05,
+      "loss": 0.8451,
+      "step": 544500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3663011505752876e-05,
+      "loss": 0.8503,
+      "step": 545000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.3648004002001002e-05,
+      "loss": 0.8647,
+      "step": 545500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3632996498249125e-05,
+      "loss": 0.8711,
+      "step": 546000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3617988994497248e-05,
+      "loss": 0.8475,
+      "step": 546500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3602981490745373e-05,
+      "loss": 0.8646,
+      "step": 547000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3587973986993497e-05,
+      "loss": 0.8595,
+      "step": 547500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3572966483241621e-05,
+      "loss": 0.8466,
+      "step": 548000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3557958979489745e-05,
+      "loss": 0.8538,
+      "step": 548500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3542981490745374e-05,
+      "loss": 0.849,
+      "step": 549000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3527973986993498e-05,
+      "loss": 0.8513,
+      "step": 549500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.351296648324162e-05,
+      "loss": 0.8447,
+      "step": 550000
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.7979006171226501,
+      "eval_runtime": 598.1411,
+      "eval_samples_per_second": 167.185,
+      "eval_steps_per_second": 41.796,
+      "step": 550000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3497958979489746e-05,
+      "loss": 0.839,
+      "step": 550500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3482981490745373e-05,
+      "loss": 0.8704,
+      "step": 551000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3467973986993497e-05,
+      "loss": 0.8562,
+      "step": 551500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.345296648324162e-05,
+      "loss": 0.8529,
+      "step": 552000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3437958979489746e-05,
+      "loss": 0.8441,
+      "step": 552500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3422981490745373e-05,
+      "loss": 0.856,
+      "step": 553000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3407973986993497e-05,
+      "loss": 0.849,
+      "step": 553500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3392966483241621e-05,
+      "loss": 0.8478,
+      "step": 554000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.3377958979489745e-05,
+      "loss": 0.8524,
+      "step": 554500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3362951475737869e-05,
+      "loss": 0.8477,
+      "step": 555000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3347943971985995e-05,
+      "loss": 0.8488,
+      "step": 555500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3332996498249125e-05,
+      "loss": 0.8426,
+      "step": 556000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3317988994497249e-05,
+      "loss": 0.8702,
+      "step": 556500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3302981490745373e-05,
+      "loss": 0.854,
+      "step": 557000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3287973986993497e-05,
+      "loss": 0.8539,
+      "step": 557500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.327296648324162e-05,
+      "loss": 0.8529,
+      "step": 558000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3257958979489745e-05,
+      "loss": 0.8548,
+      "step": 558500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3242951475737869e-05,
+      "loss": 0.8587,
+      "step": 559000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3227943971985994e-05,
+      "loss": 0.8469,
+      "step": 559500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3212936468234118e-05,
+      "loss": 0.8529,
+      "step": 560000
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.7950595021247864,
+      "eval_runtime": 602.8951,
+      "eval_samples_per_second": 165.866,
+      "eval_steps_per_second": 41.467,
+      "step": 560000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.319792896448224e-05,
+      "loss": 0.8347,
+      "step": 560500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3182921460730366e-05,
+      "loss": 0.8425,
+      "step": 561000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.316791395697849e-05,
+      "loss": 0.8658,
+      "step": 561500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3152936468234117e-05,
+      "loss": 0.8605,
+      "step": 562000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3137928964482241e-05,
+      "loss": 0.8643,
+      "step": 562500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3122921460730367e-05,
+      "loss": 0.8543,
+      "step": 563000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3107913956978489e-05,
+      "loss": 0.8436,
+      "step": 563500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.3092906453226613e-05,
+      "loss": 0.8369,
+      "step": 564000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.3077928964482241e-05,
+      "loss": 0.8563,
+      "step": 564500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.3062921460730365e-05,
+      "loss": 0.8431,
+      "step": 565000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.304791395697849e-05,
+      "loss": 0.8434,
+      "step": 565500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.3032906453226613e-05,
+      "loss": 0.8527,
+      "step": 566000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.3017898949474737e-05,
+      "loss": 0.8626,
+      "step": 566500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.3002891445722861e-05,
+      "loss": 0.8644,
+      "step": 567000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2987883941970987e-05,
+      "loss": 0.8478,
+      "step": 567500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.297287643821911e-05,
+      "loss": 0.8429,
+      "step": 568000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2957898949474738e-05,
+      "loss": 0.8492,
+      "step": 568500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2942891445722862e-05,
+      "loss": 0.8498,
+      "step": 569000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2927883941970987e-05,
+      "loss": 0.8469,
+      "step": 569500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2912906453226614e-05,
+      "loss": 0.8568,
+      "step": 570000
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.7931195497512817,
+      "eval_runtime": 601.1433,
+      "eval_samples_per_second": 166.35,
+      "eval_steps_per_second": 41.587,
+      "step": 570000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2897898949474738e-05,
+      "loss": 0.8439,
+      "step": 570500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.288289144572286e-05,
+      "loss": 0.8329,
+      "step": 571000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2867883941970986e-05,
+      "loss": 0.864,
+      "step": 571500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.285287643821911e-05,
+      "loss": 0.8517,
+      "step": 572000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2837898949474737e-05,
+      "loss": 0.8559,
+      "step": 572500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.2822891445722861e-05,
+      "loss": 0.8606,
+      "step": 573000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.280791395697849e-05,
+      "loss": 0.8449,
+      "step": 573500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2792906453226614e-05,
+      "loss": 0.8377,
+      "step": 574000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2777898949474738e-05,
+      "loss": 0.8525,
+      "step": 574500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2762891445722862e-05,
+      "loss": 0.8573,
+      "step": 575000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2747883941970986e-05,
+      "loss": 0.851,
+      "step": 575500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.273287643821911e-05,
+      "loss": 0.8596,
+      "step": 576000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2717898949474738e-05,
+      "loss": 0.8489,
+      "step": 576500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.270289144572286e-05,
+      "loss": 0.8385,
+      "step": 577000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2687883941970986e-05,
+      "loss": 0.839,
+      "step": 577500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.267287643821911e-05,
+      "loss": 0.8542,
+      "step": 578000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2657868934467232e-05,
+      "loss": 0.8386,
+      "step": 578500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2642861430715358e-05,
+      "loss": 0.8556,
+      "step": 579000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2627853926963482e-05,
+      "loss": 0.8598,
+      "step": 579500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2612846423211606e-05,
+      "loss": 0.858,
+      "step": 580000
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.7923389077186584,
+      "eval_runtime": 603.6948,
+      "eval_samples_per_second": 165.647,
+      "eval_steps_per_second": 41.412,
+      "step": 580000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.259783891945973e-05,
+      "loss": 0.8411,
+      "step": 580500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2582831415707854e-05,
+      "loss": 0.8443,
+      "step": 581000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2567853926963483e-05,
+      "loss": 0.8564,
+      "step": 581500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.2552846423211606e-05,
+      "loss": 0.8485,
+      "step": 582000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.253783891945973e-05,
+      "loss": 0.8542,
+      "step": 582500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2522831415707854e-05,
+      "loss": 0.846,
+      "step": 583000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2507823911955978e-05,
+      "loss": 0.8492,
+      "step": 583500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2492816408204102e-05,
+      "loss": 0.8476,
+      "step": 584000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2477808904452226e-05,
+      "loss": 0.85,
+      "step": 584500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2462801400700352e-05,
+      "loss": 0.8449,
+      "step": 585000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2447853926963482e-05,
+      "loss": 0.8547,
+      "step": 585500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2432846423211608e-05,
+      "loss": 0.835,
+      "step": 586000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.241783891945973e-05,
+      "loss": 0.8502,
+      "step": 586500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2402831415707854e-05,
+      "loss": 0.8596,
+      "step": 587000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.238782391195598e-05,
+      "loss": 0.8586,
+      "step": 587500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2372816408204102e-05,
+      "loss": 0.8478,
+      "step": 588000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2357808904452226e-05,
+      "loss": 0.8472,
+      "step": 588500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2342831415707854e-05,
+      "loss": 0.8523,
+      "step": 589000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2327823911955978e-05,
+      "loss": 0.8395,
+      "step": 589500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2312816408204102e-05,
+      "loss": 0.8573,
+      "step": 590000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.7895151376724243,
+      "eval_runtime": 593.5374,
+      "eval_samples_per_second": 168.481,
+      "eval_steps_per_second": 42.12,
+      "step": 590000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2297808904452226e-05,
+      "loss": 0.8431,
+      "step": 590500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.228280140070035e-05,
+      "loss": 0.8416,
+      "step": 591000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2267793896948474e-05,
+      "loss": 0.8563,
+      "step": 591500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.22527863931966e-05,
+      "loss": 0.8352,
+      "step": 592000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2237778889444724e-05,
+      "loss": 0.8399,
+      "step": 592500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.222280140070035e-05,
+      "loss": 0.841,
+      "step": 593000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.220782391195598e-05,
+      "loss": 0.8455,
+      "step": 593500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2192816408204102e-05,
+      "loss": 0.8493,
+      "step": 594000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2177808904452226e-05,
+      "loss": 0.8465,
+      "step": 594500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2162801400700351e-05,
+      "loss": 0.8478,
+      "step": 595000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2147793896948473e-05,
+      "loss": 0.8447,
+      "step": 595500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2132786393196599e-05,
+      "loss": 0.8333,
+      "step": 596000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2117778889444723e-05,
+      "loss": 0.8496,
+      "step": 596500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2102771385692847e-05,
+      "loss": 0.8357,
+      "step": 597000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2087793896948474e-05,
+      "loss": 0.8534,
+      "step": 597500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.20727863931966e-05,
+      "loss": 0.8443,
+      "step": 598000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2057778889444722e-05,
+      "loss": 0.8482,
+      "step": 598500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2042771385692846e-05,
+      "loss": 0.8462,
+      "step": 599000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2027823911955978e-05,
+      "loss": 0.8487,
+      "step": 599500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.2012816408204102e-05,
+      "loss": 0.8527,
+      "step": 600000
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.7874695062637329,
+      "eval_runtime": 597.2742,
+      "eval_samples_per_second": 167.427,
+      "eval_steps_per_second": 41.857,
+      "step": 600000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.1997808904452227e-05,
+      "loss": 0.8327,
+      "step": 600500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.1982801400700351e-05,
+      "loss": 0.8275,
+      "step": 601000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1967793896948473e-05,
+      "loss": 0.8332,
+      "step": 601500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1952786393196599e-05,
+      "loss": 0.8389,
+      "step": 602000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1937778889444723e-05,
+      "loss": 0.8557,
+      "step": 602500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1922771385692845e-05,
+      "loss": 0.8465,
+      "step": 603000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1907793896948474e-05,
+      "loss": 0.8416,
+      "step": 603500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.18927863931966e-05,
+      "loss": 0.8341,
+      "step": 604000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1877778889444722e-05,
+      "loss": 0.8451,
+      "step": 604500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1862771385692846e-05,
+      "loss": 0.8517,
+      "step": 605000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1847763881940971e-05,
+      "loss": 0.8417,
+      "step": 605500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1832756378189095e-05,
+      "loss": 0.8435,
+      "step": 606000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.181774887443722e-05,
+      "loss": 0.855,
+      "step": 606500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1802741370685343e-05,
+      "loss": 0.8568,
+      "step": 607000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1787793896948474e-05,
+      "loss": 0.8465,
+      "step": 607500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1772786393196599e-05,
+      "loss": 0.8427,
+      "step": 608000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1757778889444723e-05,
+      "loss": 0.8507,
+      "step": 608500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1742771385692845e-05,
+      "loss": 0.8348,
+      "step": 609000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1727763881940971e-05,
+      "loss": 0.8327,
+      "step": 609500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1712756378189095e-05,
+      "loss": 0.8321,
+      "step": 610000
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.7858200073242188,
+      "eval_runtime": 639.6935,
+      "eval_samples_per_second": 156.325,
+      "eval_steps_per_second": 39.081,
+      "step": 610000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.1697748874437219e-05,
+      "loss": 0.8321,
+      "step": 610500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1682741370685343e-05,
+      "loss": 0.8484,
+      "step": 611000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1667733866933467e-05,
+      "loss": 0.8498,
+      "step": 611500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1652726363181592e-05,
+      "loss": 0.8476,
+      "step": 612000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1637718859429715e-05,
+      "loss": 0.8507,
+      "step": 612500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1622711355677838e-05,
+      "loss": 0.8308,
+      "step": 613000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1607733866933467e-05,
+      "loss": 0.8443,
+      "step": 613500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1592726363181591e-05,
+      "loss": 0.8424,
+      "step": 614000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1577718859429715e-05,
+      "loss": 0.8445,
+      "step": 614500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1562711355677839e-05,
+      "loss": 0.8503,
+      "step": 615000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1547703851925963e-05,
+      "loss": 0.8312,
+      "step": 615500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1532696348174087e-05,
+      "loss": 0.8474,
+      "step": 616000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.151768884442221e-05,
+      "loss": 0.8477,
+      "step": 616500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1502711355677838e-05,
+      "loss": 0.8313,
+      "step": 617000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1487703851925964e-05,
+      "loss": 0.8423,
+      "step": 617500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1472696348174087e-05,
+      "loss": 0.8389,
+      "step": 618000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1457688844422211e-05,
+      "loss": 0.8361,
+      "step": 618500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1442711355677838e-05,
+      "loss": 0.847,
+      "step": 619000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1427703851925964e-05,
+      "loss": 0.8392,
+      "step": 619500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.1412696348174086e-05,
+      "loss": 0.828,
+      "step": 620000
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.7836601734161377,
+      "eval_runtime": 599.8827,
+      "eval_samples_per_second": 166.699,
+      "eval_steps_per_second": 41.675,
+      "step": 620000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1397688844422212e-05,
+      "loss": 0.8334,
+      "step": 620500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1382681340670336e-05,
+      "loss": 0.8389,
+      "step": 621000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.136767383691846e-05,
+      "loss": 0.8306,
+      "step": 621500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1352666333166584e-05,
+      "loss": 0.8488,
+      "step": 622000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1337658829414708e-05,
+      "loss": 0.8244,
+      "step": 622500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1322681340670335e-05,
+      "loss": 0.8449,
+      "step": 623000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1307673836918459e-05,
+      "loss": 0.8417,
+      "step": 623500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1292666333166584e-05,
+      "loss": 0.8319,
+      "step": 624000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1277658829414708e-05,
+      "loss": 0.8309,
+      "step": 624500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.126265132566283e-05,
+      "loss": 0.8326,
+      "step": 625000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.124767383691846e-05,
+      "loss": 0.8243,
+      "step": 625500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1232696348174086e-05,
+      "loss": 0.834,
+      "step": 626000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1217688844422212e-05,
+      "loss": 0.8355,
+      "step": 626500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1202711355677839e-05,
+      "loss": 0.8326,
+      "step": 627000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1187703851925963e-05,
+      "loss": 0.8316,
+      "step": 627500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1172696348174087e-05,
+      "loss": 0.8426,
+      "step": 628000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1157688844422213e-05,
+      "loss": 0.8407,
+      "step": 628500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.1142681340670335e-05,
+      "loss": 0.8292,
+      "step": 629000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1127673836918459e-05,
+      "loss": 0.8412,
+      "step": 629500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1112666333166584e-05,
+      "loss": 0.8378,
+      "step": 630000
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.7810372114181519,
+      "eval_runtime": 608.8449,
+      "eval_samples_per_second": 164.245,
+      "eval_steps_per_second": 41.061,
+      "step": 630000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1097658829414707e-05,
+      "loss": 0.832,
+      "step": 630500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.108265132566283e-05,
+      "loss": 0.8393,
+      "step": 631000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.106767383691846e-05,
+      "loss": 0.8387,
+      "step": 631500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1052666333166583e-05,
+      "loss": 0.8366,
+      "step": 632000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1037688844422212e-05,
+      "loss": 0.8145,
+      "step": 632500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1022681340670336e-05,
+      "loss": 0.8275,
+      "step": 633000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.1007673836918458e-05,
+      "loss": 0.8372,
+      "step": 633500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0992666333166584e-05,
+      "loss": 0.8224,
+      "step": 634000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0977658829414708e-05,
+      "loss": 0.8247,
+      "step": 634500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0962651325662832e-05,
+      "loss": 0.8397,
+      "step": 635000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0947643821910956e-05,
+      "loss": 0.8302,
+      "step": 635500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0932666333166584e-05,
+      "loss": 0.8334,
+      "step": 636000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0917658829414707e-05,
+      "loss": 0.8377,
+      "step": 636500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0902651325662832e-05,
+      "loss": 0.816,
+      "step": 637000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0887643821910956e-05,
+      "loss": 0.8293,
+      "step": 637500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.087263631815908e-05,
+      "loss": 0.8414,
+      "step": 638000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.0857658829414707e-05,
+      "loss": 0.8348,
+      "step": 638500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0842651325662833e-05,
+      "loss": 0.8368,
+      "step": 639000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0827643821910955e-05,
+      "loss": 0.8547,
+      "step": 639500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0812636318159079e-05,
+      "loss": 0.842,
+      "step": 640000
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.7787752151489258,
+      "eval_runtime": 597.0129,
+      "eval_samples_per_second": 167.501,
+      "eval_steps_per_second": 41.875,
+      "step": 640000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0797628814407205e-05,
+      "loss": 0.8327,
+      "step": 640500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0782621310655329e-05,
+      "loss": 0.8286,
+      "step": 641000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.076761380690345e-05,
+      "loss": 0.8293,
+      "step": 641500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0752606303151576e-05,
+      "loss": 0.8377,
+      "step": 642000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.07375987993997e-05,
+      "loss": 0.8163,
+      "step": 642500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0722591295647823e-05,
+      "loss": 0.8428,
+      "step": 643000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0707583791895948e-05,
+      "loss": 0.8245,
+      "step": 643500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0692576288144072e-05,
+      "loss": 0.8317,
+      "step": 644000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.06775987993997e-05,
+      "loss": 0.8223,
+      "step": 644500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0662591295647825e-05,
+      "loss": 0.8298,
+      "step": 645000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0647583791895949e-05,
+      "loss": 0.8304,
+      "step": 645500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0632576288144073e-05,
+      "loss": 0.8417,
+      "step": 646000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.06175987993997e-05,
+      "loss": 0.8283,
+      "step": 646500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0602591295647825e-05,
+      "loss": 0.8296,
+      "step": 647000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0587583791895948e-05,
+      "loss": 0.825,
+      "step": 647500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.0572576288144072e-05,
+      "loss": 0.8312,
+      "step": 648000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0557568784392197e-05,
+      "loss": 0.8393,
+      "step": 648500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0542561280640321e-05,
+      "loss": 0.838,
+      "step": 649000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0527553776888443e-05,
+      "loss": 0.8271,
+      "step": 649500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0512576288144072e-05,
+      "loss": 0.8348,
+      "step": 650000
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.7790135145187378,
+      "eval_runtime": 606.4656,
+      "eval_samples_per_second": 164.89,
+      "eval_steps_per_second": 41.222,
+      "step": 650000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0497568784392196e-05,
+      "loss": 0.8234,
+      "step": 650500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.048256128064032e-05,
+      "loss": 0.8434,
+      "step": 651000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0467553776888444e-05,
+      "loss": 0.8387,
+      "step": 651500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.045254627313657e-05,
+      "loss": 0.8255,
+      "step": 652000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0437538769384692e-05,
+      "loss": 0.8211,
+      "step": 652500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0422531265632816e-05,
+      "loss": 0.8452,
+      "step": 653000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0407553776888445e-05,
+      "loss": 0.8278,
+      "step": 653500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0392546273136568e-05,
+      "loss": 0.8275,
+      "step": 654000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0377538769384692e-05,
+      "loss": 0.8328,
+      "step": 654500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0362531265632818e-05,
+      "loss": 0.8127,
+      "step": 655000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.034752376188094e-05,
+      "loss": 0.8171,
+      "step": 655500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0332516258129064e-05,
+      "loss": 0.8297,
+      "step": 656000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.031750875437719e-05,
+      "loss": 0.8299,
+      "step": 656500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.0302501250625314e-05,
+      "loss": 0.8343,
+      "step": 657000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.028752376188094e-05,
+      "loss": 0.8344,
+      "step": 657500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.027254627313657e-05,
+      "loss": 0.8232,
+      "step": 658000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0257538769384692e-05,
+      "loss": 0.83,
+      "step": 658500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0242531265632817e-05,
+      "loss": 0.8321,
+      "step": 659000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0227523761880941e-05,
+      "loss": 0.8327,
+      "step": 659500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0212516258129064e-05,
+      "loss": 0.8217,
+      "step": 660000
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.7772210836410522,
+      "eval_runtime": 623.462,
+      "eval_samples_per_second": 160.395,
+      "eval_steps_per_second": 40.099,
+      "step": 660000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.019750875437719e-05,
+      "loss": 0.8316,
+      "step": 660500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0182501250625313e-05,
+      "loss": 0.8346,
+      "step": 661000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0167493746873437e-05,
+      "loss": 0.8353,
+      "step": 661500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0152516258129064e-05,
+      "loss": 0.8295,
+      "step": 662000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.013750875437719e-05,
+      "loss": 0.8361,
+      "step": 662500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0122501250625312e-05,
+      "loss": 0.8363,
+      "step": 663000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0107493746873436e-05,
+      "loss": 0.8393,
+      "step": 663500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0092516258129065e-05,
+      "loss": 0.837,
+      "step": 664000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0077538769384692e-05,
+      "loss": 0.8273,
+      "step": 664500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0062531265632817e-05,
+      "loss": 0.8308,
+      "step": 665000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0047523761880941e-05,
+      "loss": 0.8322,
+      "step": 665500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0032516258129064e-05,
+      "loss": 0.8109,
+      "step": 666000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.0017538769384692e-05,
+      "loss": 0.8194,
+      "step": 666500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.0002531265632816e-05,
+      "loss": 0.8341,
+      "step": 667000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.98752376188094e-06,
+      "loss": 0.8371,
+      "step": 667500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.972516258129064e-06,
+      "loss": 0.8252,
+      "step": 668000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.95750875437719e-06,
+      "loss": 0.8276,
+      "step": 668500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.942501250625312e-06,
+      "loss": 0.8333,
+      "step": 669000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.927493746873436e-06,
+      "loss": 0.8287,
+      "step": 669500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.912516258129065e-06,
+      "loss": 0.8309,
+      "step": 670000
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.7751027941703796,
+      "eval_runtime": 657.5511,
+      "eval_samples_per_second": 152.079,
+      "eval_steps_per_second": 38.02,
+      "step": 670000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.897508754377189e-06,
+      "loss": 0.8265,
+      "step": 670500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.882501250625313e-06,
+      "loss": 0.8296,
+      "step": 671000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.867493746873438e-06,
+      "loss": 0.8207,
+      "step": 671500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.85248624312156e-06,
+      "loss": 0.8193,
+      "step": 672000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.837478739369684e-06,
+      "loss": 0.8323,
+      "step": 672500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.822501250625313e-06,
+      "loss": 0.8389,
+      "step": 673000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.807493746873437e-06,
+      "loss": 0.8307,
+      "step": 673500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.792486243121561e-06,
+      "loss": 0.8229,
+      "step": 674000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.777478739369685e-06,
+      "loss": 0.8343,
+      "step": 674500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.762471235617809e-06,
+      "loss": 0.8317,
+      "step": 675000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.747493746873438e-06,
+      "loss": 0.8269,
+      "step": 675500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.732486243121562e-06,
+      "loss": 0.8238,
+      "step": 676000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.717478739369684e-06,
+      "loss": 0.8096,
+      "step": 676500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.70247123561781e-06,
+      "loss": 0.8251,
+      "step": 677000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.687463731865933e-06,
+      "loss": 0.8358,
+      "step": 677500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.672456228114057e-06,
+      "loss": 0.8155,
+      "step": 678000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.657448724362181e-06,
+      "loss": 0.8232,
+      "step": 678500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.642441220610305e-06,
+      "loss": 0.8185,
+      "step": 679000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.627463731865932e-06,
+      "loss": 0.8348,
+      "step": 679500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.612456228114056e-06,
+      "loss": 0.8094,
+      "step": 680000
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.7731601595878601,
+      "eval_runtime": 620.6555,
+      "eval_samples_per_second": 161.12,
+      "eval_steps_per_second": 40.28,
+      "step": 680000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.597448724362182e-06,
+      "loss": 0.8311,
+      "step": 680500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.582441220610306e-06,
+      "loss": 0.818,
+      "step": 681000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.56743371685843e-06,
+      "loss": 0.8162,
+      "step": 681500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.552426213106554e-06,
+      "loss": 0.8283,
+      "step": 682000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.537418709354678e-06,
+      "loss": 0.8117,
+      "step": 682500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.522441220610305e-06,
+      "loss": 0.8371,
+      "step": 683000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.50743371685843e-06,
+      "loss": 0.8156,
+      "step": 683500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.492426213106554e-06,
+      "loss": 0.828,
+      "step": 684000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.477418709354677e-06,
+      "loss": 0.8315,
+      "step": 684500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.462411205602802e-06,
+      "loss": 0.8242,
+      "step": 685000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.447403701850926e-06,
+      "loss": 0.8334,
+      "step": 685500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.43239619809905e-06,
+      "loss": 0.8207,
+      "step": 686000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.417388694347174e-06,
+      "loss": 0.8318,
+      "step": 686500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.402411205602803e-06,
+      "loss": 0.8288,
+      "step": 687000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.38743371685843e-06,
+      "loss": 0.8319,
+      "step": 687500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.372426213106554e-06,
+      "loss": 0.8174,
+      "step": 688000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.357418709354678e-06,
+      "loss": 0.8295,
+      "step": 688500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.342411205602802e-06,
+      "loss": 0.8265,
+      "step": 689000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.327403701850926e-06,
+      "loss": 0.8263,
+      "step": 689500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.31239619809905e-06,
+      "loss": 0.826,
+      "step": 690000
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.7714752554893494,
+      "eval_runtime": 616.7754,
+      "eval_samples_per_second": 162.134,
+      "eval_steps_per_second": 40.533,
+      "step": 690000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.297418709354677e-06,
+      "loss": 0.8202,
+      "step": 690500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.282411205602802e-06,
+      "loss": 0.8123,
+      "step": 691000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.267403701850926e-06,
+      "loss": 0.8179,
+      "step": 691500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.252396198099048e-06,
+      "loss": 0.8245,
+      "step": 692000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.237388694347174e-06,
+      "loss": 0.8257,
+      "step": 692500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.222381190595298e-06,
+      "loss": 0.8111,
+      "step": 693000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.207403701850925e-06,
+      "loss": 0.821,
+      "step": 693500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.192396198099049e-06,
+      "loss": 0.8179,
+      "step": 694000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.177388694347175e-06,
+      "loss": 0.8151,
+      "step": 694500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.162381190595297e-06,
+      "loss": 0.8411,
+      "step": 695000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.147403701850926e-06,
+      "loss": 0.8441,
+      "step": 695500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.132396198099051e-06,
+      "loss": 0.8214,
+      "step": 696000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.117388694347173e-06,
+      "loss": 0.8247,
+      "step": 696500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.102381190595297e-06,
+      "loss": 0.8196,
+      "step": 697000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.087373686843423e-06,
+      "loss": 0.8204,
+      "step": 697500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.072366183091547e-06,
+      "loss": 0.8126,
+      "step": 698000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.057388694347174e-06,
+      "loss": 0.8344,
+      "step": 698500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.042381190595298e-06,
+      "loss": 0.825,
+      "step": 699000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.027373686843422e-06,
+      "loss": 0.8234,
+      "step": 699500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.01239619809905e-06,
+      "loss": 0.8289,
+      "step": 700000
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.7685481905937195,
+      "eval_runtime": 617.0203,
+      "eval_samples_per_second": 162.069,
+      "eval_steps_per_second": 40.517,
+      "step": 700000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.997388694347175e-06,
+      "loss": 0.8284,
+      "step": 700500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.982381190595297e-06,
+      "loss": 0.8129,
+      "step": 701000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.967373686843422e-06,
+      "loss": 0.8313,
+      "step": 701500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.952366183091546e-06,
+      "loss": 0.813,
+      "step": 702000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.93735867933967e-06,
+      "loss": 0.8221,
+      "step": 702500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.922351175587794e-06,
+      "loss": 0.8273,
+      "step": 703000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.907343671835918e-06,
+      "loss": 0.8329,
+      "step": 703500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.892366183091545e-06,
+      "loss": 0.8303,
+      "step": 704000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.87735867933967e-06,
+      "loss": 0.8078,
+      "step": 704500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.862381190595298e-06,
+      "loss": 0.8184,
+      "step": 705000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.847373686843422e-06,
+      "loss": 0.8168,
+      "step": 705500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.832366183091546e-06,
+      "loss": 0.8302,
+      "step": 706000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.817388694347175e-06,
+      "loss": 0.8239,
+      "step": 706500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.802381190595297e-06,
+      "loss": 0.8246,
+      "step": 707000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.787373686843422e-06,
+      "loss": 0.8196,
+      "step": 707500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.772366183091546e-06,
+      "loss": 0.8266,
+      "step": 708000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.757358679339669e-06,
+      "loss": 0.8321,
+      "step": 708500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.742351175587794e-06,
+      "loss": 0.8268,
+      "step": 709000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.727343671835918e-06,
+      "loss": 0.8253,
+      "step": 709500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.712366183091545e-06,
+      "loss": 0.8222,
+      "step": 710000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.7680653929710388,
+      "eval_runtime": 586.5219,
+      "eval_samples_per_second": 170.497,
+      "eval_steps_per_second": 42.624,
+      "step": 710000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.697358679339671e-06,
+      "loss": 0.8269,
+      "step": 710500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.682351175587795e-06,
+      "loss": 0.8123,
+      "step": 711000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.667343671835917e-06,
+      "loss": 0.8246,
+      "step": 711500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.652336168084043e-06,
+      "loss": 0.808,
+      "step": 712000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.637328664332167e-06,
+      "loss": 0.8297,
+      "step": 712500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.62232116058029e-06,
+      "loss": 0.8164,
+      "step": 713000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.607313656828414e-06,
+      "loss": 0.8028,
+      "step": 713500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.592306153076538e-06,
+      "loss": 0.8048,
+      "step": 714000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.577298649324662e-06,
+      "loss": 0.8233,
+      "step": 714500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.562291145572786e-06,
+      "loss": 0.8238,
+      "step": 715000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.54728364182091e-06,
+      "loss": 0.8322,
+      "step": 715500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.532276138069034e-06,
+      "loss": 0.8237,
+      "step": 716000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.51726863431716e-06,
+      "loss": 0.8338,
+      "step": 716500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.502261130565282e-06,
+      "loss": 0.8221,
+      "step": 717000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.487253626813408e-06,
+      "loss": 0.809,
+      "step": 717500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.472276138069036e-06,
+      "loss": 0.8317,
+      "step": 718000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.457268634317159e-06,
+      "loss": 0.8149,
+      "step": 718500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.442261130565283e-06,
+      "loss": 0.8168,
+      "step": 719000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.427253626813408e-06,
+      "loss": 0.8143,
+      "step": 719500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.41224612306153e-06,
+      "loss": 0.8208,
+      "step": 720000
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.7673591375350952,
+      "eval_runtime": 614.272,
+      "eval_samples_per_second": 162.794,
+      "eval_steps_per_second": 40.699,
+      "step": 720000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.397238619309654e-06,
+      "loss": 0.825,
+      "step": 720500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.38223111555778e-06,
+      "loss": 0.8217,
+      "step": 721000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.367223611805904e-06,
+      "loss": 0.8191,
+      "step": 721500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.352246123061531e-06,
+      "loss": 0.8185,
+      "step": 722000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.337238619309655e-06,
+      "loss": 0.8198,
+      "step": 722500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.322231115557779e-06,
+      "loss": 0.8174,
+      "step": 723000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.307223611805903e-06,
+      "loss": 0.8028,
+      "step": 723500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.292246123061532e-06,
+      "loss": 0.8118,
+      "step": 724000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.277238619309654e-06,
+      "loss": 0.8277,
+      "step": 724500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.262261130565283e-06,
+      "loss": 0.8261,
+      "step": 725000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.247253626813408e-06,
+      "loss": 0.8166,
+      "step": 725500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.23224612306153e-06,
+      "loss": 0.8226,
+      "step": 726000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.217238619309654e-06,
+      "loss": 0.8205,
+      "step": 726500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.20223111555778e-06,
+      "loss": 0.8266,
+      "step": 727000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.187223611805902e-06,
+      "loss": 0.8173,
+      "step": 727500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.172246123061531e-06,
+      "loss": 0.8102,
+      "step": 728000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.157238619309655e-06,
+      "loss": 0.8065,
+      "step": 728500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.142231115557779e-06,
+      "loss": 0.8224,
+      "step": 729000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.127223611805903e-06,
+      "loss": 0.8121,
+      "step": 729500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.112216108054028e-06,
+      "loss": 0.8068,
+      "step": 730000
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.7651957869529724,
+      "eval_runtime": 605.1423,
+      "eval_samples_per_second": 165.25,
+      "eval_steps_per_second": 41.313,
+      "step": 730000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.09720860430215e-06,
+      "loss": 0.8107,
+      "step": 730500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.082201100550275e-06,
+      "loss": 0.82,
+      "step": 731000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.0671935967984e-06,
+      "loss": 0.8167,
+      "step": 731500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.052216108054027e-06,
+      "loss": 0.8155,
+      "step": 732000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.037208604302151e-06,
+      "loss": 0.8169,
+      "step": 732500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.022201100550275e-06,
+      "loss": 0.8167,
+      "step": 733000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.007223611805902e-06,
+      "loss": 0.8225,
+      "step": 733500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.992216108054028e-06,
+      "loss": 0.8243,
+      "step": 734000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.977208604302152e-06,
+      "loss": 0.8146,
+      "step": 734500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.962201100550274e-06,
+      "loss": 0.8102,
+      "step": 735000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.9471935967984e-06,
+      "loss": 0.814,
+      "step": 735500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.932186093046524e-06,
+      "loss": 0.8221,
+      "step": 736000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.917178589294648e-06,
+      "loss": 0.8207,
+      "step": 736500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.902171085542771e-06,
+      "loss": 0.817,
+      "step": 737000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.8871935967984e-06,
+      "loss": 0.8043,
+      "step": 737500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.872186093046524e-06,
+      "loss": 0.8082,
+      "step": 738000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.857178589294646e-06,
+      "loss": 0.8119,
+      "step": 738500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.842171085542772e-06,
+      "loss": 0.8165,
+      "step": 739000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.827193596798399e-06,
+      "loss": 0.8235,
+      "step": 739500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.812186093046523e-06,
+      "loss": 0.8118,
+      "step": 740000
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.7625706195831299,
+      "eval_runtime": 571.8371,
+      "eval_samples_per_second": 174.875,
+      "eval_steps_per_second": 43.719,
+      "step": 740000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.797178589294647e-06,
+      "loss": 0.8077,
+      "step": 740500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.782171085542773e-06,
+      "loss": 0.8171,
+      "step": 741000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.7671935967984e-06,
+      "loss": 0.807,
+      "step": 741500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.752186093046524e-06,
+      "loss": 0.8073,
+      "step": 742000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.737178589294648e-06,
+      "loss": 0.824,
+      "step": 742500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.722171085542771e-06,
+      "loss": 0.8159,
+      "step": 743000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.7071935967984e-06,
+      "loss": 0.8057,
+      "step": 743500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.692216108054027e-06,
+      "loss": 0.8145,
+      "step": 744000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.677208604302151e-06,
+      "loss": 0.8241,
+      "step": 744500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.662201100550275e-06,
+      "loss": 0.8223,
+      "step": 745000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.6471935967984e-06,
+      "loss": 0.8127,
+      "step": 745500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.632186093046523e-06,
+      "loss": 0.8108,
+      "step": 746000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.617178589294648e-06,
+      "loss": 0.8077,
+      "step": 746500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.602201100550274e-06,
+      "loss": 0.8255,
+      "step": 747000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.5871935967984e-06,
+      "loss": 0.8239,
+      "step": 747500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.572216108054028e-06,
+      "loss": 0.815,
+      "step": 748000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.557208604302151e-06,
+      "loss": 0.8117,
+      "step": 748500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.542201100550276e-06,
+      "loss": 0.8061,
+      "step": 749000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.5271935967983995e-06,
+      "loss": 0.8132,
+      "step": 749500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.512186093046523e-06,
+      "loss": 0.8134,
+      "step": 750000
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.761015772819519,
+      "eval_runtime": 586.4258,
+      "eval_samples_per_second": 170.525,
+      "eval_steps_per_second": 42.631,
+      "step": 750000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.497178589294647e-06,
+      "loss": 0.8265,
+      "step": 750500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.482171085542771e-06,
+      "loss": 0.8161,
+      "step": 751000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.467163581790896e-06,
+      "loss": 0.8179,
+      "step": 751500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.452156078039019e-06,
+      "loss": 0.8219,
+      "step": 752000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.437178589294647e-06,
+      "loss": 0.8043,
+      "step": 752500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.422171085542772e-06,
+      "loss": 0.8199,
+      "step": 753000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.407163581790896e-06,
+      "loss": 0.809,
+      "step": 753500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.39215607803902e-06,
+      "loss": 0.8071,
+      "step": 754000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.377148574287144e-06,
+      "loss": 0.8114,
+      "step": 754500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.3621710855427716e-06,
+      "loss": 0.8021,
+      "step": 755000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.3471635817908955e-06,
+      "loss": 0.8117,
+      "step": 755500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.33215607803902e-06,
+      "loss": 0.8195,
+      "step": 756000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.317148574287143e-06,
+      "loss": 0.807,
+      "step": 756500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.302141070535268e-06,
+      "loss": 0.8047,
+      "step": 757000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.287163581790896e-06,
+      "loss": 0.8079,
+      "step": 757500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.27215607803902e-06,
+      "loss": 0.821,
+      "step": 758000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.257148574287144e-06,
+      "loss": 0.7959,
+      "step": 758500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.242141070535268e-06,
+      "loss": 0.8177,
+      "step": 759000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.227163581790896e-06,
+      "loss": 0.8155,
+      "step": 759500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.21215607803902e-06,
+      "loss": 0.8152,
+      "step": 760000
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.7589353919029236,
+      "eval_runtime": 594.6054,
+      "eval_samples_per_second": 168.179,
+      "eval_steps_per_second": 42.045,
+      "step": 760000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.197148574287144e-06,
+      "loss": 0.8072,
+      "step": 760500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.1821410705352676e-06,
+      "loss": 0.8052,
+      "step": 761000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.167133566783392e-06,
+      "loss": 0.8271,
+      "step": 761500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.152126063031516e-06,
+      "loss": 0.8098,
+      "step": 762000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.13711855927964e-06,
+      "loss": 0.8081,
+      "step": 762500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.122111055527764e-06,
+      "loss": 0.8172,
+      "step": 763000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.107103551775888e-06,
+      "loss": 0.8152,
+      "step": 763500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.092096048024012e-06,
+      "loss": 0.8143,
+      "step": 764000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.077088544272136e-06,
+      "loss": 0.8038,
+      "step": 764500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.062081040520261e-06,
+      "loss": 0.8107,
+      "step": 765000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.047103551775889e-06,
+      "loss": 0.7959,
+      "step": 765500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.032096048024012e-06,
+      "loss": 0.8229,
+      "step": 766000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.0170885442721365e-06,
+      "loss": 0.8147,
+      "step": 766500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.0020810405202604e-06,
+      "loss": 0.8026,
+      "step": 767000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 6.9871335667833915e-06,
+      "loss": 0.8015,
+      "step": 767500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 6.972126063031516e-06,
+      "loss": 0.8041,
+      "step": 768000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 6.957118559279639e-06,
+      "loss": 0.8258,
+      "step": 768500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 6.942111055527764e-06,
+      "loss": 0.8184,
+      "step": 769000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.927103551775888e-06,
+      "loss": 0.8076,
+      "step": 769500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.912096048024013e-06,
+      "loss": 0.8106,
+      "step": 770000
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.7582979202270508,
+      "eval_runtime": 581.7955,
+      "eval_samples_per_second": 171.882,
+      "eval_steps_per_second": 42.97,
+      "step": 770000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.897088544272136e-06,
+      "loss": 0.8126,
+      "step": 770500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.882111055527764e-06,
+      "loss": 0.8049,
+      "step": 771000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.867103551775888e-06,
+      "loss": 0.8076,
+      "step": 771500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.852096048024013e-06,
+      "loss": 0.8063,
+      "step": 772000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.837088544272136e-06,
+      "loss": 0.8151,
+      "step": 772500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.8220810405202605e-06,
+      "loss": 0.8293,
+      "step": 773000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.807073536768384e-06,
+      "loss": 0.8075,
+      "step": 773500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.792096048024012e-06,
+      "loss": 0.8077,
+      "step": 774000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.777088544272136e-06,
+      "loss": 0.8028,
+      "step": 774500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.76208104052026e-06,
+      "loss": 0.8217,
+      "step": 775000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.747073536768384e-06,
+      "loss": 0.8213,
+      "step": 775500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.732066033016509e-06,
+      "loss": 0.8068,
+      "step": 776000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.717058529264632e-06,
+      "loss": 0.802,
+      "step": 776500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.702051025512757e-06,
+      "loss": 0.8199,
+      "step": 777000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.687043521760881e-06,
+      "loss": 0.8121,
+      "step": 777500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.672036018009005e-06,
+      "loss": 0.8154,
+      "step": 778000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 6.6570585292646326e-06,
+      "loss": 0.804,
+      "step": 778500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.64208104052026e-06,
+      "loss": 0.8103,
+      "step": 779000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.6270735367683844e-06,
+      "loss": 0.7978,
+      "step": 779500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.612096048024012e-06,
+      "loss": 0.812,
+      "step": 780000
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.7558347582817078,
+      "eval_runtime": 600.211,
+      "eval_samples_per_second": 166.608,
+      "eval_steps_per_second": 41.652,
+      "step": 780000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.597088544272136e-06,
+      "loss": 0.7998,
+      "step": 780500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.58208104052026e-06,
+      "loss": 0.8003,
+      "step": 781000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.567073536768384e-06,
+      "loss": 0.8138,
+      "step": 781500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.552096048024012e-06,
+      "loss": 0.7981,
+      "step": 782000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.537088544272137e-06,
+      "loss": 0.8065,
+      "step": 782500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.52208104052026e-06,
+      "loss": 0.7939,
+      "step": 783000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.507073536768384e-06,
+      "loss": 0.7971,
+      "step": 783500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.492066033016509e-06,
+      "loss": 0.8023,
+      "step": 784000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.477058529264633e-06,
+      "loss": 0.8124,
+      "step": 784500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.4620510255127565e-06,
+      "loss": 0.8006,
+      "step": 785000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.4470435217608805e-06,
+      "loss": 0.8008,
+      "step": 785500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.432036018009005e-06,
+      "loss": 0.8218,
+      "step": 786000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.417028514257128e-06,
+      "loss": 0.8166,
+      "step": 786500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.402021010505252e-06,
+      "loss": 0.7998,
+      "step": 787000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.387013506753377e-06,
+      "loss": 0.814,
+      "step": 787500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.372006003001501e-06,
+      "loss": 0.8006,
+      "step": 788000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.356998499249625e-06,
+      "loss": 0.8054,
+      "step": 788500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.341990995497749e-06,
+      "loss": 0.8008,
+      "step": 789000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.326983491745874e-06,
+      "loss": 0.8155,
+      "step": 789500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.3120060030015015e-06,
+      "loss": 0.8077,
+      "step": 790000
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.7543774843215942,
+      "eval_runtime": 577.6914,
+      "eval_samples_per_second": 173.103,
+      "eval_steps_per_second": 43.276,
+      "step": 790000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.296998499249625e-06,
+      "loss": 0.7926,
+      "step": 790500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.2819909954977485e-06,
+      "loss": 0.8047,
+      "step": 791000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.2670135067533765e-06,
+      "loss": 0.808,
+      "step": 791500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.252006003001501e-06,
+      "loss": 0.7919,
+      "step": 792000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.237028514257129e-06,
+      "loss": 0.7999,
+      "step": 792500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.222021010505252e-06,
+      "loss": 0.806,
+      "step": 793000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.207013506753377e-06,
+      "loss": 0.7944,
+      "step": 793500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.192006003001501e-06,
+      "loss": 0.8031,
+      "step": 794000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.177028514257129e-06,
+      "loss": 0.8094,
+      "step": 794500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.162021010505252e-06,
+      "loss": 0.8119,
+      "step": 795000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.147013506753377e-06,
+      "loss": 0.8092,
+      "step": 795500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.132006003001501e-06,
+      "loss": 0.7965,
+      "step": 796000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.1169984992496255e-06,
+      "loss": 0.8127,
+      "step": 796500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.1019909954977486e-06,
+      "loss": 0.8076,
+      "step": 797000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.086983491745873e-06,
+      "loss": 0.8006,
+      "step": 797500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.071975987993997e-06,
+      "loss": 0.8039,
+      "step": 798000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.056968484242122e-06,
+      "loss": 0.8076,
+      "step": 798500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.041960980490245e-06,
+      "loss": 0.7912,
+      "step": 799000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.026953476738369e-06,
+      "loss": 0.8113,
+      "step": 799500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.011945972986494e-06,
+      "loss": 0.8112,
+      "step": 800000
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.7531115412712097,
+      "eval_runtime": 659.7719,
+      "eval_samples_per_second": 151.568,
+      "eval_steps_per_second": 37.892,
+      "step": 800000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.996968484242122e-06,
+      "loss": 0.7955,
+      "step": 800500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.981960980490245e-06,
+      "loss": 0.8159,
+      "step": 801000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.96695347673837e-06,
+      "loss": 0.8022,
+      "step": 801500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.9519459729864936e-06,
+      "loss": 0.806,
+      "step": 802000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.9369684842421215e-06,
+      "loss": 0.806,
+      "step": 802500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.921960980490245e-06,
+      "loss": 0.8175,
+      "step": 803000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.906953476738369e-06,
+      "loss": 0.8015,
+      "step": 803500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.891975987993997e-06,
+      "loss": 0.8171,
+      "step": 804000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.876968484242121e-06,
+      "loss": 0.8022,
+      "step": 804500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.861960980490245e-06,
+      "loss": 0.81,
+      "step": 805000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.846953476738369e-06,
+      "loss": 0.7946,
+      "step": 805500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.831945972986494e-06,
+      "loss": 0.8044,
+      "step": 806000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.816968484242122e-06,
+      "loss": 0.7969,
+      "step": 806500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.801960980490245e-06,
+      "loss": 0.8136,
+      "step": 807000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.786953476738369e-06,
+      "loss": 0.814,
+      "step": 807500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.771945972986494e-06,
+      "loss": 0.7976,
+      "step": 808000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.7569384692346175e-06,
+      "loss": 0.793,
+      "step": 808500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.7419309654827415e-06,
+      "loss": 0.8028,
+      "step": 809000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.726923461730865e-06,
+      "loss": 0.8023,
+      "step": 809500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.71191595797899e-06,
+      "loss": 0.8162,
+      "step": 810000
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.7523924708366394,
+      "eval_runtime": 648.5551,
+      "eval_samples_per_second": 154.189,
+      "eval_steps_per_second": 38.547,
+      "step": 810000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.696938469234618e-06,
+      "loss": 0.809,
+      "step": 810500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.681930965482741e-06,
+      "loss": 0.8036,
+      "step": 811000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.666923461730865e-06,
+      "loss": 0.8188,
+      "step": 811500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.65191595797899e-06,
+      "loss": 0.8053,
+      "step": 812000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.636938469234618e-06,
+      "loss": 0.802,
+      "step": 812500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.621960980490246e-06,
+      "loss": 0.8064,
+      "step": 813000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.606953476738369e-06,
+      "loss": 0.8014,
+      "step": 813500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.591945972986494e-06,
+      "loss": 0.8038,
+      "step": 814000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5769384692346176e-06,
+      "loss": 0.8161,
+      "step": 814500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5619309654827415e-06,
+      "loss": 0.7968,
+      "step": 815000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.546953476738369e-06,
+      "loss": 0.7984,
+      "step": 815500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.531945972986493e-06,
+      "loss": 0.8034,
+      "step": 816000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.516938469234617e-06,
+      "loss": 0.8058,
+      "step": 816500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.501930965482742e-06,
+      "loss": 0.8205,
+      "step": 817000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.486923461730865e-06,
+      "loss": 0.794,
+      "step": 817500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.47191595797899e-06,
+      "loss": 0.8056,
+      "step": 818000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.456938469234618e-06,
+      "loss": 0.7963,
+      "step": 818500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.441930965482742e-06,
+      "loss": 0.8053,
+      "step": 819000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.426923461730865e-06,
+      "loss": 0.8077,
+      "step": 819500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.41191595797899e-06,
+      "loss": 0.7988,
+      "step": 820000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.7511362433433533,
+      "eval_runtime": 622.036,
+      "eval_samples_per_second": 160.762,
+      "eval_steps_per_second": 40.191,
+      "step": 820000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.396908454227114e-06,
+      "loss": 0.8059,
+      "step": 820500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3819309654827415e-06,
+      "loss": 0.8038,
+      "step": 821000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3669234617308654e-06,
+      "loss": 0.7953,
+      "step": 821500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.351915957978989e-06,
+      "loss": 0.7935,
+      "step": 822000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.336908454227114e-06,
+      "loss": 0.8052,
+      "step": 822500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.321900950475238e-06,
+      "loss": 0.8097,
+      "step": 823000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.306893446723361e-06,
+      "loss": 0.808,
+      "step": 823500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.291915957978989e-06,
+      "loss": 0.7991,
+      "step": 824000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.276908454227114e-06,
+      "loss": 0.8031,
+      "step": 824500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.261900950475238e-06,
+      "loss": 0.8172,
+      "step": 825000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.246893446723362e-06,
+      "loss": 0.7948,
+      "step": 825500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.231885942971486e-06,
+      "loss": 0.8089,
+      "step": 826000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.2168784392196105e-06,
+      "loss": 0.7969,
+      "step": 826500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.201870935467734e-06,
+      "loss": 0.8,
+      "step": 827000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.1868634317158575e-06,
+      "loss": 0.7848,
+      "step": 827500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.171885942971485e-06,
+      "loss": 0.8025,
+      "step": 828000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.15687843921961e-06,
+      "loss": 0.7814,
+      "step": 828500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.141900950475238e-06,
+      "loss": 0.7938,
+      "step": 829000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.126893446723362e-06,
+      "loss": 0.8011,
+      "step": 829500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.111885942971486e-06,
+      "loss": 0.8038,
+      "step": 830000
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.7489431500434875,
+      "eval_runtime": 627.3417,
+      "eval_samples_per_second": 159.403,
+      "eval_steps_per_second": 39.851,
+      "step": 830000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.09687843921961e-06,
+      "loss": 0.8,
+      "step": 830500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.081870935467734e-06,
+      "loss": 0.7896,
+      "step": 831000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.066863431715858e-06,
+      "loss": 0.81,
+      "step": 831500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.051855927963982e-06,
+      "loss": 0.8004,
+      "step": 832000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.0368484242121065e-06,
+      "loss": 0.8019,
+      "step": 832500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.021870935467734e-06,
+      "loss": 0.794,
+      "step": 833000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.0068634317158575e-06,
+      "loss": 0.8003,
+      "step": 833500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.9918859429714854e-06,
+      "loss": 0.808,
+      "step": 834000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.97687843921961e-06,
+      "loss": 0.7961,
+      "step": 834500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.961870935467734e-06,
+      "loss": 0.7936,
+      "step": 835000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.946863431715858e-06,
+      "loss": 0.8034,
+      "step": 835500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.931855927963982e-06,
+      "loss": 0.7837,
+      "step": 836000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.916848424212106e-06,
+      "loss": 0.791,
+      "step": 836500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.901840920460231e-06,
+      "loss": 0.8138,
+      "step": 837000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.886833416708354e-06,
+      "loss": 0.8031,
+      "step": 837500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.871855927963982e-06,
+      "loss": 0.7994,
+      "step": 838000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.8568484242121065e-06,
+      "loss": 0.7981,
+      "step": 838500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.8418409204602304e-06,
+      "loss": 0.7969,
+      "step": 839000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.826833416708354e-06,
+      "loss": 0.7991,
+      "step": 839500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.8118559279639815e-06,
+      "loss": 0.7978,
+      "step": 840000
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.7486736178398132,
+      "eval_runtime": 614.7602,
+      "eval_samples_per_second": 162.665,
+      "eval_steps_per_second": 40.666,
+      "step": 840000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.796848424212106e-06,
+      "loss": 0.7944,
+      "step": 840500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.78184092046023e-06,
+      "loss": 0.7894,
+      "step": 841000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.766833416708355e-06,
+      "loss": 0.8049,
+      "step": 841500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.751855927963982e-06,
+      "loss": 0.7895,
+      "step": 842000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.73687843921961e-06,
+      "loss": 0.7994,
+      "step": 842500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.721870935467734e-06,
+      "loss": 0.7957,
+      "step": 843000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.706863431715858e-06,
+      "loss": 0.8035,
+      "step": 843500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.691855927963983e-06,
+      "loss": 0.8004,
+      "step": 844000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.676848424212106e-06,
+      "loss": 0.8055,
+      "step": 844500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.661870935467734e-06,
+      "loss": 0.7879,
+      "step": 845000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.646863431715858e-06,
+      "loss": 0.7818,
+      "step": 845500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.631855927963982e-06,
+      "loss": 0.7879,
+      "step": 846000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.616848424212106e-06,
+      "loss": 0.7961,
+      "step": 846500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.60184092046023e-06,
+      "loss": 0.7985,
+      "step": 847000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.586833416708354e-06,
+      "loss": 0.7991,
+      "step": 847500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.571825912956478e-06,
+      "loss": 0.7947,
+      "step": 848000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.556818409204602e-06,
+      "loss": 0.7986,
+      "step": 848500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.541810905452727e-06,
+      "loss": 0.7947,
+      "step": 849000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.526863431715858e-06,
+      "loss": 0.7932,
+      "step": 849500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.511855927963983e-06,
+      "loss": 0.7925,
+      "step": 850000
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.74729984998703,
+      "eval_runtime": 633.8245,
+      "eval_samples_per_second": 157.772,
+      "eval_steps_per_second": 39.443,
+      "step": 850000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.496848424212106e-06,
+      "loss": 0.7857,
+      "step": 850500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.481870935467734e-06,
+      "loss": 0.7798,
+      "step": 851000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.466863431715858e-06,
+      "loss": 0.8151,
+      "step": 851500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.451855927963982e-06,
+      "loss": 0.805,
+      "step": 852000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.436848424212106e-06,
+      "loss": 0.8009,
+      "step": 852500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.42184092046023e-06,
+      "loss": 0.7944,
+      "step": 853000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.406833416708354e-06,
+      "loss": 0.8048,
+      "step": 853500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.391825912956479e-06,
+      "loss": 0.7969,
+      "step": 854000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.376818409204602e-06,
+      "loss": 0.8027,
+      "step": 854500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.361810905452727e-06,
+      "loss": 0.7892,
+      "step": 855000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.346803401700851e-06,
+      "loss": 0.7969,
+      "step": 855500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.331795897948975e-06,
+      "loss": 0.8099,
+      "step": 856000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.316788394197099e-06,
+      "loss": 0.8151,
+      "step": 856500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.3018109054527265e-06,
+      "loss": 0.7951,
+      "step": 857000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.2868334167083545e-06,
+      "loss": 0.7952,
+      "step": 857500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.271825912956478e-06,
+      "loss": 0.7942,
+      "step": 858000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.256818409204603e-06,
+      "loss": 0.7968,
+      "step": 858500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.241810905452726e-06,
+      "loss": 0.7929,
+      "step": 859000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.22680340170085e-06,
+      "loss": 0.8059,
+      "step": 859500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.211825912956478e-06,
+      "loss": 0.7919,
+      "step": 860000
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.7454802393913269,
+      "eval_runtime": 606.3303,
+      "eval_samples_per_second": 164.927,
+      "eval_steps_per_second": 41.232,
+      "step": 860000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.196818409204603e-06,
+      "loss": 0.7904,
+      "step": 860500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.181810905452726e-06,
+      "loss": 0.8019,
+      "step": 861000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.166803401700851e-06,
+      "loss": 0.7888,
+      "step": 861500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.151795897948975e-06,
+      "loss": 0.7882,
+      "step": 862000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.136788394197099e-06,
+      "loss": 0.802,
+      "step": 862500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.1217808904452225e-06,
+      "loss": 0.798,
+      "step": 863000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.1068034017008505e-06,
+      "loss": 0.7931,
+      "step": 863500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.091825912956478e-06,
+      "loss": 0.8102,
+      "step": 864000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.076818409204602e-06,
+      "loss": 0.7989,
+      "step": 864500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.061810905452726e-06,
+      "loss": 0.7932,
+      "step": 865000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.04680340170085e-06,
+      "loss": 0.7873,
+      "step": 865500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.031795897948975e-06,
+      "loss": 0.7947,
+      "step": 866000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.016788394197099e-06,
+      "loss": 0.7919,
+      "step": 866500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.001780890445223e-06,
+      "loss": 0.7963,
+      "step": 867000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.986773386693347e-06,
+      "loss": 0.7956,
+      "step": 867500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.971765882941471e-06,
+      "loss": 0.7991,
+      "step": 868000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.956758379189595e-06,
+      "loss": 0.8067,
+      "step": 868500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.9417508754377186e-06,
+      "loss": 0.7864,
+      "step": 869000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.926743371685843e-06,
+      "loss": 0.7947,
+      "step": 869500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.911765882941471e-06,
+      "loss": 0.7883,
+      "step": 870000
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.743596076965332,
+      "eval_runtime": 589.6574,
+      "eval_samples_per_second": 169.59,
+      "eval_steps_per_second": 42.398,
+      "step": 870000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.896758379189595e-06,
+      "loss": 0.7853,
+      "step": 870500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.881780890445222e-06,
+      "loss": 0.7879,
+      "step": 871000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.866773386693346e-06,
+      "loss": 0.797,
+      "step": 871500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.851765882941471e-06,
+      "loss": 0.8057,
+      "step": 872000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.836758379189595e-06,
+      "loss": 0.7998,
+      "step": 872500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.821750875437719e-06,
+      "loss": 0.8032,
+      "step": 873000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.806743371685843e-06,
+      "loss": 0.7969,
+      "step": 873500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.791735867933967e-06,
+      "loss": 0.8039,
+      "step": 874000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.7767283641820915e-06,
+      "loss": 0.7789,
+      "step": 874500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.7617508754377186e-06,
+      "loss": 0.7955,
+      "step": 875000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.746743371685843e-06,
+      "loss": 0.7855,
+      "step": 875500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.7317358679339673e-06,
+      "loss": 0.7946,
+      "step": 876000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.7167283641820912e-06,
+      "loss": 0.797,
+      "step": 876500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.7017208604302156e-06,
+      "loss": 0.7964,
+      "step": 877000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.6867433716858427e-06,
+      "loss": 0.7946,
+      "step": 877500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.671735867933967e-06,
+      "loss": 0.7962,
+      "step": 878000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.656728364182091e-06,
+      "loss": 0.7968,
+      "step": 878500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.6417208604302153e-06,
+      "loss": 0.7812,
+      "step": 879000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.6267133566783392e-06,
+      "loss": 0.797,
+      "step": 879500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.6117058529264636e-06,
+      "loss": 0.7819,
+      "step": 880000
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.7429205179214478,
+      "eval_runtime": 604.5124,
+      "eval_samples_per_second": 165.423,
+      "eval_steps_per_second": 41.356,
+      "step": 880000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5966983491745875e-06,
+      "loss": 0.7879,
+      "step": 880500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5816908454227115e-06,
+      "loss": 0.7936,
+      "step": 881000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.566713356678339e-06,
+      "loss": 0.8068,
+      "step": 881500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5517058529264633e-06,
+      "loss": 0.7924,
+      "step": 882000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5366983491745873e-06,
+      "loss": 0.7907,
+      "step": 882500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5216908454227116e-06,
+      "loss": 0.7876,
+      "step": 883000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.506713356678339e-06,
+      "loss": 0.7992,
+      "step": 883500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4917058529264635e-06,
+      "loss": 0.7884,
+      "step": 884000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4766983491745874e-06,
+      "loss": 0.7833,
+      "step": 884500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4616908454227118e-06,
+      "loss": 0.7935,
+      "step": 885000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4467133566783393e-06,
+      "loss": 0.7985,
+      "step": 885500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.431705852926463e-06,
+      "loss": 0.7935,
+      "step": 886000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.416698349174587e-06,
+      "loss": 0.7787,
+      "step": 886500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4016908454227115e-06,
+      "loss": 0.7971,
+      "step": 887000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.3867133566783394e-06,
+      "loss": 0.7929,
+      "step": 887500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.3717058529264634e-06,
+      "loss": 0.7943,
+      "step": 888000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.3566983491745877e-06,
+      "loss": 0.7842,
+      "step": 888500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.3416908454227116e-06,
+      "loss": 0.7941,
+      "step": 889000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.326713356678339e-06,
+      "loss": 0.7939,
+      "step": 889500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.311735867933967e-06,
+      "loss": 0.7854,
+      "step": 890000
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.742018461227417,
+      "eval_runtime": 599.0341,
+      "eval_samples_per_second": 166.935,
+      "eval_steps_per_second": 41.734,
+      "step": 890000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.296728364182091e-06,
+      "loss": 0.7893,
+      "step": 890500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.281720860430215e-06,
+      "loss": 0.803,
+      "step": 891000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2667133566783393e-06,
+      "loss": 0.802,
+      "step": 891500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2517058529264632e-06,
+      "loss": 0.7993,
+      "step": 892000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2366983491745876e-06,
+      "loss": 0.8021,
+      "step": 892500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2216908454227115e-06,
+      "loss": 0.7878,
+      "step": 893000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2066833416708354e-06,
+      "loss": 0.781,
+      "step": 893500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1916758379189594e-06,
+      "loss": 0.8049,
+      "step": 894000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1766683341670837e-06,
+      "loss": 0.8089,
+      "step": 894500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1616608304152077e-06,
+      "loss": 0.7842,
+      "step": 895000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1466533266633316e-06,
+      "loss": 0.7958,
+      "step": 895500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.131645822911456e-06,
+      "loss": 0.7933,
+      "step": 896000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.116668334167084e-06,
+      "loss": 0.7816,
+      "step": 896500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.101660830415208e-06,
+      "loss": 0.798,
+      "step": 897000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.0866833416708353e-06,
+      "loss": 0.7837,
+      "step": 897500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.0716758379189593e-06,
+      "loss": 0.7781,
+      "step": 898000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.0566683341670836e-06,
+      "loss": 0.7966,
+      "step": 898500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.0416608304152075e-06,
+      "loss": 0.7896,
+      "step": 899000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.026653326663332e-06,
+      "loss": 0.776,
+      "step": 899500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.011645822911456e-06,
+      "loss": 0.7784,
+      "step": 900000
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.7412576675415039,
+      "eval_runtime": 624.3312,
+      "eval_samples_per_second": 160.171,
+      "eval_steps_per_second": 40.043,
+      "step": 900000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.99663831915958e-06,
+      "loss": 0.7892,
+      "step": 900500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.9816608304152077e-06,
+      "loss": 0.7944,
+      "step": 901000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.9666533266633316e-06,
+      "loss": 0.7818,
+      "step": 901500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.9516458229114555e-06,
+      "loss": 0.7921,
+      "step": 902000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.93663831915958e-06,
+      "loss": 0.7927,
+      "step": 902500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.921630815407704e-06,
+      "loss": 0.7877,
+      "step": 903000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.906623311655828e-06,
+      "loss": 0.7997,
+      "step": 903500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.8916458229114557e-06,
+      "loss": 0.7847,
+      "step": 904000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.87663831915958e-06,
+      "loss": 0.7988,
+      "step": 904500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.861630815407704e-06,
+      "loss": 0.7929,
+      "step": 905000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.846623311655828e-06,
+      "loss": 0.7907,
+      "step": 905500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.831615807903952e-06,
+      "loss": 0.7932,
+      "step": 906000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.816608304152076e-06,
+      "loss": 0.7834,
+      "step": 906500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.8016008004002e-06,
+      "loss": 0.7859,
+      "step": 907000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7865932966483245e-06,
+      "loss": 0.7873,
+      "step": 907500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.771615807903952e-06,
+      "loss": 0.786,
+      "step": 908000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7566083041520763e-06,
+      "loss": 0.7941,
+      "step": 908500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.7416308154077043e-06,
+      "loss": 0.7962,
+      "step": 909000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.726623311655828e-06,
+      "loss": 0.7934,
+      "step": 909500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7116158079039517e-06,
+      "loss": 0.7874,
+      "step": 910000
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.7403942942619324,
+      "eval_runtime": 608.9146,
+      "eval_samples_per_second": 164.227,
+      "eval_steps_per_second": 41.057,
+      "step": 910000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6966383191595797e-06,
+      "loss": 0.7744,
+      "step": 910500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.681630815407704e-06,
+      "loss": 0.8014,
+      "step": 911000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.666623311655828e-06,
+      "loss": 0.786,
+      "step": 911500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6516158079039523e-06,
+      "loss": 0.7952,
+      "step": 912000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6366083041520762e-06,
+      "loss": 0.7962,
+      "step": 912500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.621630815407704e-06,
+      "loss": 0.793,
+      "step": 913000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6066233116558277e-06,
+      "loss": 0.7885,
+      "step": 913500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.591615807903952e-06,
+      "loss": 0.7883,
+      "step": 914000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.576608304152076e-06,
+      "loss": 0.7811,
+      "step": 914500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5616008004002003e-06,
+      "loss": 0.7862,
+      "step": 915000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5465932966483242e-06,
+      "loss": 0.7899,
+      "step": 915500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.531585792896448e-06,
+      "loss": 0.771,
+      "step": 916000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5165782891445725e-06,
+      "loss": 0.7835,
+      "step": 916500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5015707853926965e-06,
+      "loss": 0.779,
+      "step": 917000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.486593296648324e-06,
+      "loss": 0.7986,
+      "step": 917500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.4715857928964483e-06,
+      "loss": 0.7895,
+      "step": 918000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.4565782891445723e-06,
+      "loss": 0.7882,
+      "step": 918500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.441570785392696e-06,
+      "loss": 0.786,
+      "step": 919000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.4265632816408205e-06,
+      "loss": 0.7866,
+      "step": 919500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.4115557778889445e-06,
+      "loss": 0.7727,
+      "step": 920000
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.7387272715568542,
+      "eval_runtime": 611.0966,
+      "eval_samples_per_second": 163.64,
+      "eval_steps_per_second": 40.91,
+      "step": 920000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3965782891445724e-06,
+      "loss": 0.79,
+      "step": 920500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3815707853926968e-06,
+      "loss": 0.7841,
+      "step": 921000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3665632816408203e-06,
+      "loss": 0.797,
+      "step": 921500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.351555777888944e-06,
+      "loss": 0.7758,
+      "step": 922000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3365482741370685e-06,
+      "loss": 0.7854,
+      "step": 922500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3215707853926965e-06,
+      "loss": 0.7877,
+      "step": 923000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.3065632816408204e-06,
+      "loss": 0.787,
+      "step": 923500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2915557778889448e-06,
+      "loss": 0.7823,
+      "step": 924000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2765482741370687e-06,
+      "loss": 0.7831,
+      "step": 924500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.261540770385193e-06,
+      "loss": 0.7846,
+      "step": 925000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2465332666333166e-06,
+      "loss": 0.7889,
+      "step": 925500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2315257628814405e-06,
+      "loss": 0.7796,
+      "step": 926000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.216518259129565e-06,
+      "loss": 0.785,
+      "step": 926500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2015407703851928e-06,
+      "loss": 0.8018,
+      "step": 927000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.1865332666333167e-06,
+      "loss": 0.787,
+      "step": 927500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.171525762881441e-06,
+      "loss": 0.7962,
+      "step": 928000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.156518259129565e-06,
+      "loss": 0.783,
+      "step": 928500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.141540770385193e-06,
+      "loss": 0.8057,
+      "step": 929000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1265332666333164e-06,
+      "loss": 0.7904,
+      "step": 929500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.111525762881441e-06,
+      "loss": 0.7756,
+      "step": 930000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.7374897003173828,
+      "eval_runtime": 578.7073,
+      "eval_samples_per_second": 172.799,
+      "eval_steps_per_second": 43.2,
+      "step": 930000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0965182591295647e-06,
+      "loss": 0.7893,
+      "step": 930500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.081510755377689e-06,
+      "loss": 0.7763,
+      "step": 931000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.066503251625813e-06,
+      "loss": 0.7817,
+      "step": 931500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.051525762881441e-06,
+      "loss": 0.7807,
+      "step": 932000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.036518259129565e-06,
+      "loss": 0.7849,
+      "step": 932500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0215107553776892e-06,
+      "loss": 0.7735,
+      "step": 933000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.0065032516258127e-06,
+      "loss": 0.7658,
+      "step": 933500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.991495747873937e-06,
+      "loss": 0.7667,
+      "step": 934000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.976488244122061e-06,
+      "loss": 0.7655,
+      "step": 934500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.961480740370185e-06,
+      "loss": 0.7673,
+      "step": 935000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9464732366183093e-06,
+      "loss": 0.7603,
+      "step": 935500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9314957478739372e-06,
+      "loss": 0.7651,
+      "step": 936000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.916488244122061e-06,
+      "loss": 0.76,
+      "step": 936500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9014807403701853e-06,
+      "loss": 0.762,
+      "step": 937000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8864732366183092e-06,
+      "loss": 0.7732,
+      "step": 937500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8714657328664334e-06,
+      "loss": 0.764,
+      "step": 938000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8564582291145573e-06,
+      "loss": 0.7633,
+      "step": 938500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.841480740370185e-06,
+      "loss": 0.7701,
+      "step": 939000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8264732366183092e-06,
+      "loss": 0.7654,
+      "step": 939500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8114657328664333e-06,
+      "loss": 0.7588,
+      "step": 940000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.737705409526825,
+      "eval_runtime": 579.0479,
+      "eval_samples_per_second": 172.697,
+      "eval_steps_per_second": 43.174,
+      "step": 940000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7964582291145573e-06,
+      "loss": 0.7784,
+      "step": 940500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7814507253626814e-06,
+      "loss": 0.7521,
+      "step": 941000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7664432216108055e-06,
+      "loss": 0.7575,
+      "step": 941500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7514357178589297e-06,
+      "loss": 0.7786,
+      "step": 942000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7364582291145572e-06,
+      "loss": 0.7631,
+      "step": 942500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7214507253626813e-06,
+      "loss": 0.756,
+      "step": 943000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7064432216108055e-06,
+      "loss": 0.7576,
+      "step": 943500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6914357178589296e-06,
+      "loss": 0.7712,
+      "step": 944000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6764282141070535e-06,
+      "loss": 0.7553,
+      "step": 944500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6614507253626813e-06,
+      "loss": 0.7592,
+      "step": 945000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6464432216108054e-06,
+      "loss": 0.7677,
+      "step": 945500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6314357178589296e-06,
+      "loss": 0.7628,
+      "step": 946000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6164282141070535e-06,
+      "loss": 0.7724,
+      "step": 946500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.6014207103551776e-06,
+      "loss": 0.7541,
+      "step": 947000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5864432216108056e-06,
+      "loss": 0.7656,
+      "step": 947500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5714357178589295e-06,
+      "loss": 0.7726,
+      "step": 948000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5564282141070534e-06,
+      "loss": 0.755,
+      "step": 948500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5414207103551776e-06,
+      "loss": 0.7555,
+      "step": 949000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5264132066033017e-06,
+      "loss": 0.765,
+      "step": 949500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5114057028514259e-06,
+      "loss": 0.7734,
+      "step": 950000
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 0.736987292766571,
+      "eval_runtime": 575.6732,
+      "eval_samples_per_second": 173.71,
+      "eval_steps_per_second": 43.427,
+      "step": 950000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4963981990995498e-06,
+      "loss": 0.7605,
+      "step": 950500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4814207103551775e-06,
+      "loss": 0.7677,
+      "step": 951000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4664132066033016e-06,
+      "loss": 0.764,
+      "step": 951500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4514057028514258e-06,
+      "loss": 0.7535,
+      "step": 952000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4363981990995497e-06,
+      "loss": 0.7746,
+      "step": 952500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4213906953476739e-06,
+      "loss": 0.7642,
+      "step": 953000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.406383191595798e-06,
+      "loss": 0.7574,
+      "step": 953500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.391375687843922e-06,
+      "loss": 0.7556,
+      "step": 954000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3763981990995497e-06,
+      "loss": 0.7684,
+      "step": 954500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3613906953476738e-06,
+      "loss": 0.7678,
+      "step": 955000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.346383191595798e-06,
+      "loss": 0.7582,
+      "step": 955500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.331375687843922e-06,
+      "loss": 0.7699,
+      "step": 956000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.316368184092046e-06,
+      "loss": 0.7704,
+      "step": 956500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.3013606803401702e-06,
+      "loss": 0.7631,
+      "step": 957000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2863531765882943e-06,
+      "loss": 0.7781,
+      "step": 957500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2713456728364182e-06,
+      "loss": 0.7707,
+      "step": 958000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.256368184092046e-06,
+      "loss": 0.7599,
+      "step": 958500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.24136068034017e-06,
+      "loss": 0.7639,
+      "step": 959000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2263531765882942e-06,
+      "loss": 0.7784,
+      "step": 959500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2113456728364182e-06,
+      "loss": 0.7563,
+      "step": 960000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 0.7364427447319031,
+      "eval_runtime": 576.5734,
+      "eval_samples_per_second": 173.438,
+      "eval_steps_per_second": 43.36,
+      "step": 960000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.196368184092046e-06,
+      "loss": 0.7694,
+      "step": 960500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.18136068034017e-06,
+      "loss": 0.7588,
+      "step": 961000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.1663531765882942e-06,
+      "loss": 0.7441,
+      "step": 961500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.1513756878439221e-06,
+      "loss": 0.7665,
+      "step": 962000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.136368184092046e-06,
+      "loss": 0.7515,
+      "step": 962500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.1213606803401702e-06,
+      "loss": 0.7683,
+      "step": 963000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.1063531765882941e-06,
+      "loss": 0.7637,
+      "step": 963500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0913456728364183e-06,
+      "loss": 0.7537,
+      "step": 964000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.076368184092046e-06,
+      "loss": 0.7667,
+      "step": 964500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0613606803401701e-06,
+      "loss": 0.7663,
+      "step": 965000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.0463531765882943e-06,
+      "loss": 0.7604,
+      "step": 965500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.0313456728364184e-06,
+      "loss": 0.7534,
+      "step": 966000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.0163381690845423e-06,
+      "loss": 0.771,
+      "step": 966500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.0013306653326663e-06,
+      "loss": 0.7567,
+      "step": 967000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.863531765882942e-07,
+      "loss": 0.7632,
+      "step": 967500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.713456728364183e-07,
+      "loss": 0.7728,
+      "step": 968000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.563381690845423e-07,
+      "loss": 0.754,
+      "step": 968500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.413306653326664e-07,
+      "loss": 0.7629,
+      "step": 969000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.263231615807905e-07,
+      "loss": 0.7696,
+      "step": 969500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.113156578289145e-07,
+      "loss": 0.7753,
+      "step": 970000
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 0.7353302240371704,
+      "eval_runtime": 645.1885,
+      "eval_samples_per_second": 154.993,
+      "eval_steps_per_second": 38.748,
+      "step": 970000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.963081540770385e-07,
+      "loss": 0.7666,
+      "step": 970500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.813006503251626e-07,
+      "loss": 0.7675,
+      "step": 971000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.663231615807905e-07,
+      "loss": 0.777,
+      "step": 971500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.513156578289144e-07,
+      "loss": 0.7657,
+      "step": 972000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.363081540770386e-07,
+      "loss": 0.7698,
+      "step": 972500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.213006503251626e-07,
+      "loss": 0.7571,
+      "step": 973000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 8.063231615807904e-07,
+      "loss": 0.7628,
+      "step": 973500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.913156578289145e-07,
+      "loss": 0.7672,
+      "step": 974000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.763081540770386e-07,
+      "loss": 0.7677,
+      "step": 974500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.613306653326663e-07,
+      "loss": 0.7624,
+      "step": 975000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.463231615807905e-07,
+      "loss": 0.7617,
+      "step": 975500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.313156578289144e-07,
+      "loss": 0.7627,
+      "step": 976000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.163081540770386e-07,
+      "loss": 0.7499,
+      "step": 976500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.013006503251626e-07,
+      "loss": 0.7753,
+      "step": 977000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.862931465732867e-07,
+      "loss": 0.7663,
+      "step": 977500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.713156578289145e-07,
+      "loss": 0.7773,
+      "step": 978000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.563381690845423e-07,
+      "loss": 0.7573,
+      "step": 978500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.413306653326663e-07,
+      "loss": 0.7679,
+      "step": 979000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.263231615807905e-07,
+      "loss": 0.7665,
+      "step": 979500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.113156578289144e-07,
+      "loss": 0.7651,
+      "step": 980000
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 0.7347660064697266,
+      "eval_runtime": 596.5908,
+      "eval_samples_per_second": 167.619,
+      "eval_steps_per_second": 41.905,
+      "step": 980000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.963081540770385e-07,
+      "loss": 0.7661,
+      "step": 980500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.813006503251626e-07,
+      "loss": 0.7551,
+      "step": 981000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.662931465732867e-07,
+      "loss": 0.7647,
+      "step": 981500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.512856428214107e-07,
+      "loss": 0.7543,
+      "step": 982000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.362781390695348e-07,
+      "loss": 0.7666,
+      "step": 982500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.212706353176588e-07,
+      "loss": 0.7721,
+      "step": 983000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.06263131565783e-07,
+      "loss": 0.7605,
+      "step": 983500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.912856428214107e-07,
+      "loss": 0.7602,
+      "step": 984000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.762781390695348e-07,
+      "loss": 0.7553,
+      "step": 984500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.612706353176588e-07,
+      "loss": 0.7499,
+      "step": 985000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.462631315657829e-07,
+      "loss": 0.7592,
+      "step": 985500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.3125562781390695e-07,
+      "loss": 0.7653,
+      "step": 986000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.1624812406203104e-07,
+      "loss": 0.7705,
+      "step": 986500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.012406203101551e-07,
+      "loss": 0.7631,
+      "step": 987000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.862631315657829e-07,
+      "loss": 0.7501,
+      "step": 987500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.7125562781390694e-07,
+      "loss": 0.7564,
+      "step": 988000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.5624812406203103e-07,
+      "loss": 0.7495,
+      "step": 988500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.4124062031015507e-07,
+      "loss": 0.7605,
+      "step": 989000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.2623311655827916e-07,
+      "loss": 0.7698,
+      "step": 989500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.1125562781390693e-07,
+      "loss": 0.7526,
+      "step": 990000
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 0.7347068190574646,
+      "eval_runtime": 582.4895,
+      "eval_samples_per_second": 171.677,
+      "eval_steps_per_second": 42.919,
+      "step": 990000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.96248124062031e-07,
+      "loss": 0.7615,
+      "step": 990500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8124062031015506e-07,
+      "loss": 0.7677,
+      "step": 991000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6623311655827915e-07,
+      "loss": 0.754,
+      "step": 991500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.512256128064032e-07,
+      "loss": 0.7632,
+      "step": 992000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.3621810905452727e-07,
+      "loss": 0.7677,
+      "step": 992500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.2121060530265134e-07,
+      "loss": 0.7546,
+      "step": 993000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.0623311655827914e-07,
+      "loss": 0.7602,
+      "step": 993500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9125562781390697e-07,
+      "loss": 0.7498,
+      "step": 994000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.7624812406203103e-07,
+      "loss": 0.7587,
+      "step": 994500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.612406203101551e-07,
+      "loss": 0.7581,
+      "step": 995000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4623311655827916e-07,
+      "loss": 0.7569,
+      "step": 995500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.3122561280640322e-07,
+      "loss": 0.7696,
+      "step": 996000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.1621810905452727e-07,
+      "loss": 0.7511,
+      "step": 996500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.0121060530265133e-07,
+      "loss": 0.7672,
+      "step": 997000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 8.620310155077538e-08,
+      "loss": 0.7461,
+      "step": 997500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 7.119559779889944e-08,
+      "loss": 0.7562,
+      "step": 998000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.6188094047023514e-08,
+      "loss": 0.7609,
+      "step": 998500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.1210605302651325e-08,
+      "loss": 0.7588,
+      "step": 999000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6203101550775387e-08,
+      "loss": 0.7513,
+      "step": 999500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.119559779889945e-08,
+      "loss": 0.755,
+      "step": 1000000
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 0.7345585823059082,
+      "eval_runtime": 579.2567,
+      "eval_samples_per_second": 172.635,
+      "eval_steps_per_second": 43.159,
+      "step": 1000000
+    }
+  ],
+  "max_steps": 1000000,
+  "num_train_epochs": 2,
+  "total_flos": 2.242924955382055e+18,
+  "trial_name": null,
+  "trial_params": null
+}