diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,111 +1,4391 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 9.997955010224949,
-  "global_step": 610,
+  "epoch": 499.9979550102249,
+  "global_step": 30500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 39.944950103759766,
-      "eval_runtime": 1.5442,
-      "eval_samples_per_second": 79.651,
-      "eval_steps_per_second": 10.361,
+      "eval_loss": 39.79280090332031,
+      "eval_runtime": 1.532,
+      "eval_samples_per_second": 79.634,
+      "eval_steps_per_second": 10.444,
       "step": 61
     },
     {
       "epoch": 2.0,
-      "eval_loss": 39.573768615722656,
-      "eval_runtime": 1.5441,
-      "eval_samples_per_second": 79.656,
-      "eval_steps_per_second": 10.362,
+      "eval_loss": 39.81950378417969,
+      "eval_runtime": 1.5354,
+      "eval_samples_per_second": 79.461,
+      "eval_steps_per_second": 10.421,
       "step": 122
     },
     {
       "epoch": 3.0,
-      "eval_loss": 40.006507873535156,
-      "eval_runtime": 1.5451,
-      "eval_samples_per_second": 79.606,
-      "eval_steps_per_second": 10.355,
+      "eval_loss": 39.82277297973633,
+      "eval_runtime": 1.532,
+      "eval_samples_per_second": 79.635,
+      "eval_steps_per_second": 10.444,
       "step": 183
     },
     {
       "epoch": 4.0,
-      "eval_loss": 39.271480560302734,
-      "eval_runtime": 1.5464,
-      "eval_samples_per_second": 79.54,
-      "eval_steps_per_second": 10.347,
+      "eval_loss": 39.07925796508789,
+      "eval_runtime": 1.5323,
+      "eval_samples_per_second": 79.619,
+      "eval_steps_per_second": 10.442,
       "step": 244
     },
     {
       "epoch": 5.0,
-      "eval_loss": 38.749183654785156,
-      "eval_runtime": 1.5468,
-      "eval_samples_per_second": 79.518,
-      "eval_steps_per_second": 10.344,
+      "eval_loss": 38.66284942626953,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.671,
+      "eval_steps_per_second": 10.449,
       "step": 305
     },
     {
       "epoch": 6.0,
-      "eval_loss": 38.856666564941406,
-      "eval_runtime": 1.5467,
-      "eval_samples_per_second": 79.525,
-      "eval_steps_per_second": 10.345,
+      "eval_loss": 37.40141296386719,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.615,
+      "eval_steps_per_second": 10.441,
       "step": 366
     },
     {
       "epoch": 7.0,
-      "eval_loss": 38.75959396362305,
-      "eval_runtime": 1.5455,
-      "eval_samples_per_second": 79.588,
-      "eval_steps_per_second": 10.353,
+      "eval_loss": 36.820281982421875,
+      "eval_runtime": 1.5357,
+      "eval_samples_per_second": 79.443,
+      "eval_steps_per_second": 10.419,
       "step": 427
     },
     {
       "epoch": 8.0,
-      "eval_loss": 38.665225982666016,
-      "eval_runtime": 1.5457,
-      "eval_samples_per_second": 79.578,
-      "eval_steps_per_second": 10.352,
+      "eval_loss": 36.260650634765625,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.631,
+      "eval_steps_per_second": 10.443,
       "step": 488
     },
     {
       "epoch": 8.2,
-      "learning_rate": 9.01639344262295e-09,
-      "loss": 42.1342,
+      "learning_rate": 4.918032786885246e-08,
+      "loss": 41.0436,
       "step": 500
     },
     {
       "epoch": 9.0,
-      "eval_loss": 38.66032028198242,
-      "eval_runtime": 1.551,
-      "eval_samples_per_second": 79.304,
-      "eval_steps_per_second": 10.316,
+      "eval_loss": 35.931610107421875,
+      "eval_runtime": 1.5478,
+      "eval_samples_per_second": 78.82,
+      "eval_steps_per_second": 10.337,
       "step": 549
     },
     {
       "epoch": 10.0,
-      "eval_loss": 38.66507339477539,
-      "eval_runtime": 1.5562,
-      "eval_samples_per_second": 79.038,
-      "eval_steps_per_second": 10.281,
+      "eval_loss": 34.32823181152344,
+      "eval_runtime": 1.5396,
+      "eval_samples_per_second": 79.242,
+      "eval_steps_per_second": 10.392,
       "step": 610
     },
     {
-      "epoch": 10.0,
-      "step": 610,
-      "total_flos": 322666370343936.0,
-      "train_loss": 41.80477074795082,
-      "train_runtime": 454.8629,
-      "train_samples_per_second": 10.75,
-      "train_steps_per_second": 1.341
+      "epoch": 11.0,
+      "eval_loss": 33.36042404174805,
+      "eval_runtime": 1.5352,
+      "eval_samples_per_second": 79.469,
+      "eval_steps_per_second": 10.422,
+      "step": 671
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 32.44695281982422,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.623,
+      "eval_steps_per_second": 10.442,
+      "step": 732
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 31.54026985168457,
+      "eval_runtime": 1.5308,
+      "eval_samples_per_second": 79.698,
+      "eval_steps_per_second": 10.452,
+      "step": 793
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 30.966489791870117,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.689,
+      "eval_steps_per_second": 10.451,
+      "step": 854
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 29.417085647583008,
+      "eval_runtime": 1.5409,
+      "eval_samples_per_second": 79.175,
+      "eval_steps_per_second": 10.384,
+      "step": 915
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 28.458908081054688,
+      "eval_runtime": 1.5307,
+      "eval_samples_per_second": 79.7,
+      "eval_steps_per_second": 10.452,
+      "step": 976
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 4.8360655737704914e-08,
+      "loss": 32.5506,
+      "step": 1000
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 27.26991081237793,
+      "eval_runtime": 1.5386,
+      "eval_samples_per_second": 79.293,
+      "eval_steps_per_second": 10.399,
+      "step": 1037
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 26.061372756958008,
+      "eval_runtime": 1.5472,
+      "eval_samples_per_second": 78.852,
+      "eval_steps_per_second": 10.341,
+      "step": 1098
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 25.152942657470703,
+      "eval_runtime": 1.5408,
+      "eval_samples_per_second": 79.178,
+      "eval_steps_per_second": 10.384,
+      "step": 1159
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 24.308048248291016,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.616,
+      "eval_steps_per_second": 10.441,
+      "step": 1220
+    },
+    {
+      "epoch": 21.0,
+      "eval_loss": 23.15108871459961,
+      "eval_runtime": 1.5344,
+      "eval_samples_per_second": 79.509,
+      "eval_steps_per_second": 10.427,
+      "step": 1281
+    },
+    {
+      "epoch": 22.0,
+      "eval_loss": 22.333181381225586,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.691,
+      "eval_steps_per_second": 10.451,
+      "step": 1342
+    },
+    {
+      "epoch": 23.0,
+      "eval_loss": 21.71536636352539,
+      "eval_runtime": 1.5304,
+      "eval_samples_per_second": 79.718,
+      "eval_steps_per_second": 10.455,
+      "step": 1403
+    },
+    {
+      "epoch": 24.0,
+      "eval_loss": 20.736703872680664,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.67,
+      "eval_steps_per_second": 10.448,
+      "step": 1464
+    },
+    {
+      "epoch": 24.59,
+      "learning_rate": 4.7540983606557375e-08,
+      "loss": 24.3212,
+      "step": 1500
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 20.180402755737305,
+      "eval_runtime": 1.5364,
+      "eval_samples_per_second": 79.404,
+      "eval_steps_per_second": 10.414,
+      "step": 1525
+    },
+    {
+      "epoch": 26.0,
+      "eval_loss": 19.559112548828125,
+      "eval_runtime": 1.5443,
+      "eval_samples_per_second": 79.002,
+      "eval_steps_per_second": 10.361,
+      "step": 1586
+    },
+    {
+      "epoch": 27.0,
+      "eval_loss": 18.87604522705078,
+      "eval_runtime": 1.5495,
+      "eval_samples_per_second": 78.733,
+      "eval_steps_per_second": 10.326,
+      "step": 1647
+    },
+    {
+      "epoch": 28.0,
+      "eval_loss": 18.329124450683594,
+      "eval_runtime": 1.5379,
+      "eval_samples_per_second": 79.329,
+      "eval_steps_per_second": 10.404,
+      "step": 1708
+    },
+    {
+      "epoch": 29.0,
+      "eval_loss": 18.029600143432617,
+      "eval_runtime": 1.5333,
+      "eval_samples_per_second": 79.569,
+      "eval_steps_per_second": 10.435,
+      "step": 1769
+    },
+    {
+      "epoch": 30.0,
+      "eval_loss": 17.40913200378418,
+      "eval_runtime": 1.5337,
+      "eval_samples_per_second": 79.546,
+      "eval_steps_per_second": 10.432,
+      "step": 1830
+    },
+    {
+      "epoch": 31.0,
+      "eval_loss": 17.215015411376953,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.552,
+      "eval_steps_per_second": 10.433,
+      "step": 1891
+    },
+    {
+      "epoch": 32.0,
+      "eval_loss": 16.648334503173828,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.692,
+      "eval_steps_per_second": 10.451,
+      "step": 1952
+    },
+    {
+      "epoch": 32.79,
+      "learning_rate": 4.6721311475409836e-08,
+      "loss": 18.5662,
+      "step": 2000
+    },
+    {
+      "epoch": 33.0,
+      "eval_loss": 16.269676208496094,
+      "eval_runtime": 1.5361,
+      "eval_samples_per_second": 79.421,
+      "eval_steps_per_second": 10.416,
+      "step": 2013
+    },
+    {
+      "epoch": 34.0,
+      "eval_loss": 16.17144775390625,
+      "eval_runtime": 1.5342,
+      "eval_samples_per_second": 79.519,
+      "eval_steps_per_second": 10.429,
+      "step": 2074
+    },
+    {
+      "epoch": 35.0,
+      "eval_loss": 15.8008394241333,
+      "eval_runtime": 1.538,
+      "eval_samples_per_second": 79.323,
+      "eval_steps_per_second": 10.403,
+      "step": 2135
+    },
+    {
+      "epoch": 36.0,
+      "eval_loss": 15.264719009399414,
+      "eval_runtime": 1.5357,
+      "eval_samples_per_second": 79.444,
+      "eval_steps_per_second": 10.419,
+      "step": 2196
+    },
+    {
+      "epoch": 37.0,
+      "eval_loss": 15.41949462890625,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.387,
+      "eval_steps_per_second": 10.411,
+      "step": 2257
+    },
+    {
+      "epoch": 38.0,
+      "eval_loss": 14.989760398864746,
+      "eval_runtime": 1.5351,
+      "eval_samples_per_second": 79.476,
+      "eval_steps_per_second": 10.423,
+      "step": 2318
+    },
+    {
+      "epoch": 39.0,
+      "eval_loss": 14.859911918640137,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.676,
+      "eval_steps_per_second": 10.449,
+      "step": 2379
+    },
+    {
+      "epoch": 40.0,
+      "eval_loss": 14.611943244934082,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.677,
+      "eval_steps_per_second": 10.449,
+      "step": 2440
+    },
+    {
+      "epoch": 40.98,
+      "learning_rate": 4.590163934426229e-08,
+      "loss": 15.1141,
+      "step": 2500
+    },
+    {
+      "epoch": 41.0,
+      "eval_loss": 14.464351654052734,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.682,
+      "eval_steps_per_second": 10.45,
+      "step": 2501
+    },
+    {
+      "epoch": 42.0,
+      "eval_loss": 14.339770317077637,
+      "eval_runtime": 1.5431,
+      "eval_samples_per_second": 79.064,
+      "eval_steps_per_second": 10.369,
+      "step": 2562
+    },
+    {
+      "epoch": 43.0,
+      "eval_loss": 14.410659790039062,
+      "eval_runtime": 1.5584,
+      "eval_samples_per_second": 78.285,
+      "eval_steps_per_second": 10.267,
+      "step": 2623
+    },
+    {
+      "epoch": 44.0,
+      "eval_loss": 13.903946876525879,
+      "eval_runtime": 1.5462,
+      "eval_samples_per_second": 78.902,
+      "eval_steps_per_second": 10.348,
+      "step": 2684
+    },
+    {
+      "epoch": 45.0,
+      "eval_loss": 14.004327774047852,
+      "eval_runtime": 1.5337,
+      "eval_samples_per_second": 79.545,
+      "eval_steps_per_second": 10.432,
+      "step": 2745
+    },
+    {
+      "epoch": 46.0,
+      "eval_loss": 13.953380584716797,
+      "eval_runtime": 1.5354,
+      "eval_samples_per_second": 79.459,
+      "eval_steps_per_second": 10.421,
+      "step": 2806
+    },
+    {
+      "epoch": 47.0,
+      "eval_loss": 13.926652908325195,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.552,
+      "eval_steps_per_second": 10.433,
+      "step": 2867
+    },
+    {
+      "epoch": 48.0,
+      "eval_loss": 13.588834762573242,
+      "eval_runtime": 1.563,
+      "eval_samples_per_second": 78.054,
+      "eval_steps_per_second": 10.237,
+      "step": 2928
+    },
+    {
+      "epoch": 49.0,
+      "eval_loss": 13.63134479522705,
+      "eval_runtime": 1.5341,
+      "eval_samples_per_second": 79.524,
+      "eval_steps_per_second": 10.429,
+      "step": 2989
+    },
+    {
+      "epoch": 49.18,
+      "learning_rate": 4.508196721311475e-08,
+      "loss": 13.182,
+      "step": 3000
+    },
+    {
+      "epoch": 50.0,
+      "eval_loss": 13.540535926818848,
+      "eval_runtime": 1.5436,
+      "eval_samples_per_second": 79.038,
+      "eval_steps_per_second": 10.366,
+      "step": 3050
+    },
+    {
+      "epoch": 51.0,
+      "eval_loss": 13.337821960449219,
+      "eval_runtime": 1.5382,
+      "eval_samples_per_second": 79.316,
+      "eval_steps_per_second": 10.402,
+      "step": 3111
+    },
+    {
+      "epoch": 52.0,
+      "eval_loss": 13.316015243530273,
+      "eval_runtime": 1.5353,
+      "eval_samples_per_second": 79.465,
+      "eval_steps_per_second": 10.422,
+      "step": 3172
+    },
+    {
+      "epoch": 53.0,
+      "eval_loss": 13.135573387145996,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.692,
+      "eval_steps_per_second": 10.451,
+      "step": 3233
+    },
+    {
+      "epoch": 54.0,
+      "eval_loss": 13.248279571533203,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.688,
+      "eval_steps_per_second": 10.451,
+      "step": 3294
+    },
+    {
+      "epoch": 55.0,
+      "eval_loss": 13.125914573669434,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.693,
+      "eval_steps_per_second": 10.452,
+      "step": 3355
+    },
+    {
+      "epoch": 56.0,
+      "eval_loss": 13.177481651306152,
+      "eval_runtime": 1.5387,
+      "eval_samples_per_second": 79.286,
+      "eval_steps_per_second": 10.398,
+      "step": 3416
+    },
+    {
+      "epoch": 57.0,
+      "eval_loss": 13.111815452575684,
+      "eval_runtime": 1.5355,
+      "eval_samples_per_second": 79.455,
+      "eval_steps_per_second": 10.42,
+      "step": 3477
+    },
+    {
+      "epoch": 57.38,
+      "learning_rate": 4.426229508196721e-08,
+      "loss": 12.1712,
+      "step": 3500
+    },
+    {
+      "epoch": 58.0,
+      "eval_loss": 12.936331748962402,
+      "eval_runtime": 1.5524,
+      "eval_samples_per_second": 78.59,
+      "eval_steps_per_second": 10.307,
+      "step": 3538
+    },
+    {
+      "epoch": 59.0,
+      "eval_loss": 12.876540184020996,
+      "eval_runtime": 1.5376,
+      "eval_samples_per_second": 79.345,
+      "eval_steps_per_second": 10.406,
+      "step": 3599
+    },
+    {
+      "epoch": 60.0,
+      "eval_loss": 12.79231071472168,
+      "eval_runtime": 1.537,
+      "eval_samples_per_second": 79.375,
+      "eval_steps_per_second": 10.41,
+      "step": 3660
+    },
+    {
+      "epoch": 61.0,
+      "eval_loss": 12.973188400268555,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.532,
+      "eval_steps_per_second": 10.43,
+      "step": 3721
+    },
+    {
+      "epoch": 62.0,
+      "eval_loss": 12.860616683959961,
+      "eval_runtime": 1.5352,
+      "eval_samples_per_second": 79.468,
+      "eval_steps_per_second": 10.422,
+      "step": 3782
+    },
+    {
+      "epoch": 63.0,
+      "eval_loss": 12.789679527282715,
+      "eval_runtime": 1.5317,
+      "eval_samples_per_second": 79.652,
+      "eval_steps_per_second": 10.446,
+      "step": 3843
+    },
+    {
+      "epoch": 64.0,
+      "eval_loss": 12.651591300964355,
+      "eval_runtime": 1.5315,
+      "eval_samples_per_second": 79.662,
+      "eval_steps_per_second": 10.447,
+      "step": 3904
+    },
+    {
+      "epoch": 65.0,
+      "eval_loss": 12.624197959899902,
+      "eval_runtime": 1.5373,
+      "eval_samples_per_second": 79.36,
+      "eval_steps_per_second": 10.408,
+      "step": 3965
+    },
+    {
+      "epoch": 65.57,
+      "learning_rate": 4.344262295081967e-08,
+      "loss": 11.5853,
+      "step": 4000
+    },
+    {
+      "epoch": 66.0,
+      "eval_loss": 12.59536361694336,
+      "eval_runtime": 1.537,
+      "eval_samples_per_second": 79.375,
+      "eval_steps_per_second": 10.41,
+      "step": 4026
+    },
+    {
+      "epoch": 67.0,
+      "eval_loss": 12.450535774230957,
+      "eval_runtime": 1.5373,
+      "eval_samples_per_second": 79.357,
+      "eval_steps_per_second": 10.408,
+      "step": 4087
+    },
+    {
+      "epoch": 68.0,
+      "eval_loss": 12.559524536132812,
+      "eval_runtime": 1.5374,
+      "eval_samples_per_second": 79.356,
+      "eval_steps_per_second": 10.407,
+      "step": 4148
+    },
+    {
+      "epoch": 69.0,
+      "eval_loss": 12.428722381591797,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.665,
+      "eval_steps_per_second": 10.448,
+      "step": 4209
+    },
+    {
+      "epoch": 70.0,
+      "eval_loss": 12.476819038391113,
+      "eval_runtime": 1.5329,
+      "eval_samples_per_second": 79.59,
+      "eval_steps_per_second": 10.438,
+      "step": 4270
+    },
+    {
+      "epoch": 71.0,
+      "eval_loss": 12.400250434875488,
+      "eval_runtime": 1.5417,
+      "eval_samples_per_second": 79.132,
+      "eval_steps_per_second": 10.378,
+      "step": 4331
+    },
+    {
+      "epoch": 72.0,
+      "eval_loss": 12.432337760925293,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.387,
+      "eval_steps_per_second": 10.411,
+      "step": 4392
+    },
+    {
+      "epoch": 73.0,
+      "eval_loss": 12.41788101196289,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.531,
+      "eval_steps_per_second": 10.43,
+      "step": 4453
+    },
+    {
+      "epoch": 73.77,
+      "learning_rate": 4.262295081967213e-08,
+      "loss": 11.25,
+      "step": 4500
+    },
+    {
+      "epoch": 74.0,
+      "eval_loss": 12.376703262329102,
+      "eval_runtime": 1.5362,
+      "eval_samples_per_second": 79.418,
+      "eval_steps_per_second": 10.415,
+      "step": 4514
+    },
+    {
+      "epoch": 75.0,
+      "eval_loss": 12.434186935424805,
+      "eval_runtime": 1.543,
+      "eval_samples_per_second": 79.065,
+      "eval_steps_per_second": 10.369,
+      "step": 4575
+    },
+    {
+      "epoch": 76.0,
+      "eval_loss": 12.256501197814941,
+      "eval_runtime": 1.5414,
+      "eval_samples_per_second": 79.148,
+      "eval_steps_per_second": 10.38,
+      "step": 4636
+    },
+    {
+      "epoch": 77.0,
+      "eval_loss": 12.312841415405273,
+      "eval_runtime": 1.5542,
+      "eval_samples_per_second": 78.498,
+      "eval_steps_per_second": 10.295,
+      "step": 4697
+    },
+    {
+      "epoch": 78.0,
+      "eval_loss": 12.374871253967285,
+      "eval_runtime": 1.5323,
+      "eval_samples_per_second": 79.62,
+      "eval_steps_per_second": 10.442,
+      "step": 4758
+    },
+    {
+      "epoch": 79.0,
+      "eval_loss": 12.356457710266113,
+      "eval_runtime": 1.5352,
+      "eval_samples_per_second": 79.468,
+      "eval_steps_per_second": 10.422,
+      "step": 4819
+    },
+    {
+      "epoch": 80.0,
+      "eval_loss": 12.328751564025879,
+      "eval_runtime": 1.5365,
+      "eval_samples_per_second": 79.4,
+      "eval_steps_per_second": 10.413,
+      "step": 4880
+    },
+    {
+      "epoch": 81.0,
+      "eval_loss": 12.201430320739746,
+      "eval_runtime": 1.5343,
+      "eval_samples_per_second": 79.517,
+      "eval_steps_per_second": 10.428,
+      "step": 4941
+    },
+    {
+      "epoch": 81.97,
+      "learning_rate": 4.180327868852459e-08,
+      "loss": 11.0038,
+      "step": 5000
+    },
+    {
+      "epoch": 82.0,
+      "eval_loss": 12.177206039428711,
+      "eval_runtime": 1.5376,
+      "eval_samples_per_second": 79.345,
+      "eval_steps_per_second": 10.406,
+      "step": 5002
+    },
+    {
+      "epoch": 83.0,
+      "eval_loss": 12.2689790725708,
+      "eval_runtime": 1.5376,
+      "eval_samples_per_second": 79.346,
+      "eval_steps_per_second": 10.406,
+      "step": 5063
+    },
+    {
+      "epoch": 84.0,
+      "eval_loss": 12.198723793029785,
+      "eval_runtime": 1.5387,
+      "eval_samples_per_second": 79.286,
+      "eval_steps_per_second": 10.398,
+      "step": 5124
+    },
+    {
+      "epoch": 85.0,
+      "eval_loss": 12.0653715133667,
+      "eval_runtime": 1.5397,
+      "eval_samples_per_second": 79.234,
+      "eval_steps_per_second": 10.391,
+      "step": 5185
+    },
+    {
+      "epoch": 86.0,
+      "eval_loss": 12.12946891784668,
+      "eval_runtime": 1.5325,
+      "eval_samples_per_second": 79.61,
+      "eval_steps_per_second": 10.441,
+      "step": 5246
+    },
+    {
+      "epoch": 87.0,
+      "eval_loss": 12.156049728393555,
+      "eval_runtime": 1.5372,
+      "eval_samples_per_second": 79.367,
+      "eval_steps_per_second": 10.409,
+      "step": 5307
+    },
+    {
+      "epoch": 88.0,
+      "eval_loss": 11.96617317199707,
+      "eval_runtime": 1.5352,
+      "eval_samples_per_second": 79.471,
+      "eval_steps_per_second": 10.422,
+      "step": 5368
+    },
+    {
+      "epoch": 89.0,
+      "eval_loss": 12.113394737243652,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.693,
+      "eval_steps_per_second": 10.451,
+      "step": 5429
+    },
+    {
+      "epoch": 90.0,
+      "eval_loss": 12.029434204101562,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.626,
+      "eval_steps_per_second": 10.443,
+      "step": 5490
+    },
+    {
+      "epoch": 90.16,
+      "learning_rate": 4.0983606557377046e-08,
+      "loss": 10.8283,
+      "step": 5500
+    },
+    {
+      "epoch": 91.0,
+      "eval_loss": 12.111231803894043,
+      "eval_runtime": 1.5402,
+      "eval_samples_per_second": 79.212,
+      "eval_steps_per_second": 10.388,
+      "step": 5551
+    },
+    {
+      "epoch": 92.0,
+      "eval_loss": 12.091044425964355,
+      "eval_runtime": 1.5374,
+      "eval_samples_per_second": 79.357,
+      "eval_steps_per_second": 10.407,
+      "step": 5612
+    },
+    {
+      "epoch": 93.0,
+      "eval_loss": 12.008685111999512,
+      "eval_runtime": 1.5385,
+      "eval_samples_per_second": 79.298,
+      "eval_steps_per_second": 10.4,
+      "step": 5673
+    },
+    {
+      "epoch": 94.0,
+      "eval_loss": 11.980273246765137,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.688,
+      "eval_steps_per_second": 10.451,
+      "step": 5734
+    },
+    {
+      "epoch": 95.0,
+      "eval_loss": 11.909398078918457,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.689,
+      "eval_steps_per_second": 10.451,
+      "step": 5795
+    },
+    {
+      "epoch": 96.0,
+      "eval_loss": 12.159173965454102,
+      "eval_runtime": 1.5365,
+      "eval_samples_per_second": 79.401,
+      "eval_steps_per_second": 10.413,
+      "step": 5856
+    },
+    {
+      "epoch": 97.0,
+      "eval_loss": 11.936081886291504,
+      "eval_runtime": 1.5323,
+      "eval_samples_per_second": 79.617,
+      "eval_steps_per_second": 10.442,
+      "step": 5917
+    },
+    {
+      "epoch": 98.0,
+      "eval_loss": 11.959556579589844,
+      "eval_runtime": 1.5396,
+      "eval_samples_per_second": 79.243,
+      "eval_steps_per_second": 10.393,
+      "step": 5978
+    },
+    {
+      "epoch": 98.36,
+      "learning_rate": 4.016393442622951e-08,
+      "loss": 10.693,
+      "step": 6000
+    },
+    {
+      "epoch": 99.0,
+      "eval_loss": 11.902626037597656,
+      "eval_runtime": 1.5386,
+      "eval_samples_per_second": 79.293,
+      "eval_steps_per_second": 10.399,
+      "step": 6039
+    },
+    {
+      "epoch": 100.0,
+      "eval_loss": 12.003995895385742,
+      "eval_runtime": 1.54,
+      "eval_samples_per_second": 79.222,
+      "eval_steps_per_second": 10.39,
+      "step": 6100
+    },
+    {
+      "epoch": 101.0,
+      "eval_loss": 11.886563301086426,
+      "eval_runtime": 1.5404,
+      "eval_samples_per_second": 79.201,
+      "eval_steps_per_second": 10.387,
+      "step": 6161
+    },
+    {
+      "epoch": 102.0,
+      "eval_loss": 11.953593254089355,
+      "eval_runtime": 1.5318,
+      "eval_samples_per_second": 79.643,
+      "eval_steps_per_second": 10.445,
+      "step": 6222
+    },
+    {
+      "epoch": 103.0,
+      "eval_loss": 11.803376197814941,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.678,
+      "eval_steps_per_second": 10.45,
+      "step": 6283
+    },
+    {
+      "epoch": 104.0,
+      "eval_loss": 11.688481330871582,
+      "eval_runtime": 1.5386,
+      "eval_samples_per_second": 79.291,
+      "eval_steps_per_second": 10.399,
+      "step": 6344
+    },
+    {
+      "epoch": 105.0,
+      "eval_loss": 11.850458145141602,
+      "eval_runtime": 1.5318,
+      "eval_samples_per_second": 79.647,
+      "eval_steps_per_second": 10.446,
+      "step": 6405
+    },
+    {
+      "epoch": 106.0,
+      "eval_loss": 11.828042984008789,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.677,
+      "eval_steps_per_second": 10.449,
+      "step": 6466
+    },
+    {
+      "epoch": 106.56,
+      "learning_rate": 3.934426229508196e-08,
+      "loss": 10.5875,
+      "step": 6500
+    },
+    {
+      "epoch": 107.0,
+      "eval_loss": 11.787361145019531,
+      "eval_runtime": 1.5414,
+      "eval_samples_per_second": 79.149,
+      "eval_steps_per_second": 10.38,
+      "step": 6527
+    },
+    {
+      "epoch": 108.0,
+      "eval_loss": 11.734838485717773,
+      "eval_runtime": 1.5374,
+      "eval_samples_per_second": 79.353,
+      "eval_steps_per_second": 10.407,
+      "step": 6588
+    },
+    {
+      "epoch": 109.0,
+      "eval_loss": 11.776476860046387,
+      "eval_runtime": 1.5397,
+      "eval_samples_per_second": 79.238,
+      "eval_steps_per_second": 10.392,
+      "step": 6649
+    },
+    {
+      "epoch": 110.0,
+      "eval_loss": 11.752671241760254,
+      "eval_runtime": 1.538,
+      "eval_samples_per_second": 79.323,
+      "eval_steps_per_second": 10.403,
+      "step": 6710
+    },
+    {
+      "epoch": 111.0,
+      "eval_loss": 11.681644439697266,
+      "eval_runtime": 1.5307,
+      "eval_samples_per_second": 79.7,
+      "eval_steps_per_second": 10.452,
+      "step": 6771
+    },
+    {
+      "epoch": 112.0,
+      "eval_loss": 11.739622116088867,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.674,
+      "eval_steps_per_second": 10.449,
+      "step": 6832
+    },
+    {
+      "epoch": 113.0,
+      "eval_loss": 11.647455215454102,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.625,
+      "eval_steps_per_second": 10.443,
+      "step": 6893
+    },
+    {
+      "epoch": 114.0,
+      "eval_loss": 11.70103645324707,
+      "eval_runtime": 1.5307,
+      "eval_samples_per_second": 79.702,
+      "eval_steps_per_second": 10.453,
+      "step": 6954
+    },
+    {
+      "epoch": 114.75,
+      "learning_rate": 3.852459016393442e-08,
+      "loss": 10.5114,
+      "step": 7000
+    },
+    {
+      "epoch": 115.0,
+      "eval_loss": 11.704916000366211,
+      "eval_runtime": 1.543,
+      "eval_samples_per_second": 79.064,
+      "eval_steps_per_second": 10.369,
+      "step": 7015
+    },
+    {
+      "epoch": 116.0,
+      "eval_loss": 11.796719551086426,
+      "eval_runtime": 1.5451,
+      "eval_samples_per_second": 78.957,
+      "eval_steps_per_second": 10.355,
+      "step": 7076
+    },
+    {
+      "epoch": 117.0,
+      "eval_loss": 11.724781036376953,
+      "eval_runtime": 1.5406,
+      "eval_samples_per_second": 79.189,
+      "eval_steps_per_second": 10.385,
+      "step": 7137
+    },
+    {
+      "epoch": 118.0,
+      "eval_loss": 11.654914855957031,
+      "eval_runtime": 1.5376,
+      "eval_samples_per_second": 79.345,
+      "eval_steps_per_second": 10.406,
+      "step": 7198
+    },
+    {
+      "epoch": 119.0,
+      "eval_loss": 11.519423484802246,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.641,
+      "eval_steps_per_second": 10.445,
+      "step": 7259
+    },
+    {
+      "epoch": 120.0,
+      "eval_loss": 11.692423820495605,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.691,
+      "eval_steps_per_second": 10.451,
+      "step": 7320
+    },
+    {
+      "epoch": 121.0,
+      "eval_loss": 11.519417762756348,
+      "eval_runtime": 1.5318,
+      "eval_samples_per_second": 79.643,
+      "eval_steps_per_second": 10.445,
+      "step": 7381
+    },
+    {
+      "epoch": 122.0,
+      "eval_loss": 11.660749435424805,
+      "eval_runtime": 1.5564,
+      "eval_samples_per_second": 78.387,
+      "eval_steps_per_second": 10.28,
+      "step": 7442
+    },
+    {
+      "epoch": 122.95,
+      "learning_rate": 3.770491803278688e-08,
+      "loss": 10.4791,
+      "step": 7500
+    },
+    {
+      "epoch": 123.0,
+      "eval_loss": 11.518941879272461,
+      "eval_runtime": 1.5358,
+      "eval_samples_per_second": 79.437,
+      "eval_steps_per_second": 10.418,
+      "step": 7503
+    },
+    {
+      "epoch": 124.0,
+      "eval_loss": 11.552470207214355,
+      "eval_runtime": 1.5427,
+      "eval_samples_per_second": 79.083,
+      "eval_steps_per_second": 10.372,
+      "step": 7564
+    },
+    {
+      "epoch": 125.0,
+      "eval_loss": 11.522570610046387,
+      "eval_runtime": 1.5376,
+      "eval_samples_per_second": 79.342,
+      "eval_steps_per_second": 10.406,
+      "step": 7625
+    },
+    {
+      "epoch": 126.0,
+      "eval_loss": 11.495443344116211,
+      "eval_runtime": 1.5392,
+      "eval_samples_per_second": 79.264,
+      "eval_steps_per_second": 10.395,
+      "step": 7686
+    },
+    {
+      "epoch": 127.0,
+      "eval_loss": 11.59436321258545,
+      "eval_runtime": 1.5335,
+      "eval_samples_per_second": 79.555,
+      "eval_steps_per_second": 10.433,
+      "step": 7747
+    },
+    {
+      "epoch": 128.0,
+      "eval_loss": 11.705589294433594,
+      "eval_runtime": 1.5364,
+      "eval_samples_per_second": 79.408,
+      "eval_steps_per_second": 10.414,
+      "step": 7808
+    },
+    {
+      "epoch": 129.0,
+      "eval_loss": 11.746439933776855,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.665,
+      "eval_steps_per_second": 10.448,
+      "step": 7869
+    },
+    {
+      "epoch": 130.0,
+      "eval_loss": 11.48509407043457,
+      "eval_runtime": 1.5344,
+      "eval_samples_per_second": 79.51,
+      "eval_steps_per_second": 10.428,
+      "step": 7930
+    },
+    {
+      "epoch": 131.0,
+      "eval_loss": 11.310612678527832,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.688,
+      "eval_steps_per_second": 10.451,
+      "step": 7991
+    },
+    {
+      "epoch": 131.15,
+      "learning_rate": 3.6885245901639346e-08,
+      "loss": 10.4223,
+      "step": 8000
+    },
+    {
+      "epoch": 132.0,
+      "eval_loss": 11.661527633666992,
+      "eval_runtime": 1.543,
+      "eval_samples_per_second": 79.067,
+      "eval_steps_per_second": 10.369,
+      "step": 8052
+    },
+    {
+      "epoch": 133.0,
+      "eval_loss": 11.512996673583984,
+      "eval_runtime": 1.5421,
+      "eval_samples_per_second": 79.112,
+      "eval_steps_per_second": 10.375,
+      "step": 8113
+    },
+    {
+      "epoch": 134.0,
+      "eval_loss": 11.586601257324219,
+      "eval_runtime": 1.5394,
+      "eval_samples_per_second": 79.249,
+      "eval_steps_per_second": 10.393,
+      "step": 8174
+    },
+    {
+      "epoch": 135.0,
+      "eval_loss": 11.499874114990234,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.679,
+      "eval_steps_per_second": 10.45,
+      "step": 8235
+    },
+    {
+      "epoch": 136.0,
+      "eval_loss": 11.573627471923828,
+      "eval_runtime": 1.5328,
+      "eval_samples_per_second": 79.594,
+      "eval_steps_per_second": 10.439,
+      "step": 8296
+    },
+    {
+      "epoch": 137.0,
+      "eval_loss": 11.610065460205078,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.611,
+      "eval_steps_per_second": 10.441,
+      "step": 8357
+    },
+    {
+      "epoch": 138.0,
+      "eval_loss": 11.601760864257812,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.629,
+      "eval_steps_per_second": 10.443,
+      "step": 8418
+    },
+    {
+      "epoch": 139.0,
+      "eval_loss": 11.506176948547363,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.672,
+      "eval_steps_per_second": 10.449,
+      "step": 8479
+    },
+    {
+      "epoch": 139.34,
+      "learning_rate": 3.60655737704918e-08,
+      "loss": 10.3725,
+      "step": 8500
+    },
+    {
+      "epoch": 140.0,
+      "eval_loss": 11.469504356384277,
+      "eval_runtime": 1.5361,
+      "eval_samples_per_second": 79.423,
+      "eval_steps_per_second": 10.416,
+      "step": 8540
+    },
+    {
+      "epoch": 141.0,
+      "eval_loss": 11.574456214904785,
+      "eval_runtime": 1.5371,
+      "eval_samples_per_second": 79.368,
+      "eval_steps_per_second": 10.409,
+      "step": 8601
+    },
+    {
+      "epoch": 142.0,
+      "eval_loss": 11.213080406188965,
+      "eval_runtime": 1.5468,
+      "eval_samples_per_second": 78.874,
+      "eval_steps_per_second": 10.344,
+      "step": 8662
+    },
+    {
+      "epoch": 143.0,
+      "eval_loss": 11.486066818237305,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.628,
+      "eval_steps_per_second": 10.443,
+      "step": 8723
+    },
+    {
+      "epoch": 144.0,
+      "eval_loss": 11.352083206176758,
+      "eval_runtime": 1.5374,
+      "eval_samples_per_second": 79.354,
+      "eval_steps_per_second": 10.407,
+      "step": 8784
+    },
+    {
+      "epoch": 145.0,
+      "eval_loss": 11.430201530456543,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.692,
+      "eval_steps_per_second": 10.451,
+      "step": 8845
+    },
+    {
+      "epoch": 146.0,
+      "eval_loss": 11.389131546020508,
+      "eval_runtime": 1.532,
+      "eval_samples_per_second": 79.634,
+      "eval_steps_per_second": 10.444,
+      "step": 8906
+    },
+    {
+      "epoch": 147.0,
+      "eval_loss": 11.494805335998535,
+      "eval_runtime": 1.5431,
+      "eval_samples_per_second": 79.064,
+      "eval_steps_per_second": 10.369,
+      "step": 8967
+    },
+    {
+      "epoch": 147.54,
+      "learning_rate": 3.524590163934426e-08,
+      "loss": 10.3579,
+      "step": 9000
+    },
+    {
+      "epoch": 148.0,
+      "eval_loss": 11.413972854614258,
+      "eval_runtime": 1.5681,
+      "eval_samples_per_second": 77.803,
+      "eval_steps_per_second": 10.204,
+      "step": 9028
+    },
+    {
+      "epoch": 149.0,
+      "eval_loss": 11.28091812133789,
+      "eval_runtime": 1.5378,
+      "eval_samples_per_second": 79.334,
+      "eval_steps_per_second": 10.404,
+      "step": 9089
+    },
+    {
+      "epoch": 150.0,
+      "eval_loss": 11.508408546447754,
+      "eval_runtime": 1.5385,
+      "eval_samples_per_second": 79.299,
+      "eval_steps_per_second": 10.4,
+      "step": 9150
+    },
+    {
+      "epoch": 151.0,
+      "eval_loss": 11.494176864624023,
+      "eval_runtime": 1.5357,
+      "eval_samples_per_second": 79.441,
+      "eval_steps_per_second": 10.418,
+      "step": 9211
+    },
+    {
+      "epoch": 152.0,
+      "eval_loss": 11.431586265563965,
+      "eval_runtime": 1.535,
+      "eval_samples_per_second": 79.477,
+      "eval_steps_per_second": 10.423,
+      "step": 9272
+    },
+    {
+      "epoch": 153.0,
+      "eval_loss": 11.505435943603516,
+      "eval_runtime": 1.5329,
+      "eval_samples_per_second": 79.587,
+      "eval_steps_per_second": 10.438,
+      "step": 9333
+    },
+    {
+      "epoch": 154.0,
+      "eval_loss": 11.435380935668945,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.671,
+      "eval_steps_per_second": 10.449,
+      "step": 9394
+    },
+    {
+      "epoch": 155.0,
+      "eval_loss": 11.370787620544434,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.677,
+      "eval_steps_per_second": 10.449,
+      "step": 9455
+    },
+    {
+      "epoch": 155.74,
+      "learning_rate": 3.442622950819672e-08,
+      "loss": 10.3411,
+      "step": 9500
+    },
+    {
+      "epoch": 156.0,
+      "eval_loss": 11.60168170928955,
+      "eval_runtime": 1.541,
+      "eval_samples_per_second": 79.168,
+      "eval_steps_per_second": 10.383,
+      "step": 9516
+    },
+    {
+      "epoch": 157.0,
+      "eval_loss": 11.441522598266602,
+      "eval_runtime": 1.5413,
+      "eval_samples_per_second": 79.155,
+      "eval_steps_per_second": 10.381,
+      "step": 9577
+    },
+    {
+      "epoch": 158.0,
+      "eval_loss": 11.614983558654785,
+      "eval_runtime": 1.5394,
+      "eval_samples_per_second": 79.253,
+      "eval_steps_per_second": 10.394,
+      "step": 9638
+    },
+    {
+      "epoch": 159.0,
+      "eval_loss": 11.37887191772461,
+      "eval_runtime": 1.536,
+      "eval_samples_per_second": 79.425,
+      "eval_steps_per_second": 10.416,
+      "step": 9699
+    },
+    {
+      "epoch": 160.0,
+      "eval_loss": 11.334174156188965,
+      "eval_runtime": 1.532,
+      "eval_samples_per_second": 79.637,
+      "eval_steps_per_second": 10.444,
+      "step": 9760
+    },
+    {
+      "epoch": 161.0,
+      "eval_loss": 11.476101875305176,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.684,
+      "eval_steps_per_second": 10.45,
+      "step": 9821
+    },
+    {
+      "epoch": 162.0,
+      "eval_loss": 11.300040245056152,
+      "eval_runtime": 1.5341,
+      "eval_samples_per_second": 79.525,
+      "eval_steps_per_second": 10.43,
+      "step": 9882
+    },
+    {
+      "epoch": 163.0,
+      "eval_loss": 11.410900115966797,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.53,
+      "eval_steps_per_second": 10.43,
+      "step": 9943
+    },
+    {
+      "epoch": 163.93,
+      "learning_rate": 3.360655737704918e-08,
+      "loss": 10.3236,
+      "step": 10000
+    },
+    {
+      "epoch": 164.0,
+      "eval_loss": 11.425009727478027,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.612,
+      "eval_steps_per_second": 10.441,
+      "step": 10004
+    },
+    {
+      "epoch": 165.0,
+      "eval_loss": 11.32499885559082,
+      "eval_runtime": 1.5438,
+      "eval_samples_per_second": 79.024,
+      "eval_steps_per_second": 10.364,
+      "step": 10065
+    },
+    {
+      "epoch": 166.0,
+      "eval_loss": 11.42471981048584,
+      "eval_runtime": 1.5488,
+      "eval_samples_per_second": 78.773,
+      "eval_steps_per_second": 10.331,
+      "step": 10126
+    },
+    {
+      "epoch": 167.0,
+      "eval_loss": 11.16617488861084,
+      "eval_runtime": 1.5363,
+      "eval_samples_per_second": 79.414,
+      "eval_steps_per_second": 10.415,
+      "step": 10187
+    },
+    {
+      "epoch": 168.0,
+      "eval_loss": 11.44943904876709,
+      "eval_runtime": 1.5344,
+      "eval_samples_per_second": 79.507,
+      "eval_steps_per_second": 10.427,
+      "step": 10248
+    },
+    {
+      "epoch": 169.0,
+      "eval_loss": 11.387123107910156,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.551,
+      "eval_steps_per_second": 10.433,
+      "step": 10309
+    },
+    {
+      "epoch": 170.0,
+      "eval_loss": 11.296111106872559,
+      "eval_runtime": 1.5331,
+      "eval_samples_per_second": 79.578,
+      "eval_steps_per_second": 10.436,
+      "step": 10370
+    },
+    {
+      "epoch": 171.0,
+      "eval_loss": 11.357579231262207,
+      "eval_runtime": 1.5325,
+      "eval_samples_per_second": 79.607,
+      "eval_steps_per_second": 10.44,
+      "step": 10431
+    },
+    {
+      "epoch": 172.0,
+      "eval_loss": 11.435007095336914,
+      "eval_runtime": 1.5331,
+      "eval_samples_per_second": 79.577,
+      "eval_steps_per_second": 10.436,
+      "step": 10492
+    },
+    {
+      "epoch": 172.13,
+      "learning_rate": 3.278688524590163e-08,
+      "loss": 10.3059,
+      "step": 10500
+    },
+    {
+      "epoch": 173.0,
+      "eval_loss": 11.361226081848145,
+      "eval_runtime": 1.5571,
+      "eval_samples_per_second": 78.352,
+      "eval_steps_per_second": 10.276,
+      "step": 10553
+    },
+    {
+      "epoch": 174.0,
+      "eval_loss": 11.39792251586914,
+      "eval_runtime": 1.541,
+      "eval_samples_per_second": 79.167,
+      "eval_steps_per_second": 10.383,
+      "step": 10614
+    },
+    {
+      "epoch": 175.0,
+      "eval_loss": 11.371596336364746,
+      "eval_runtime": 1.5383,
+      "eval_samples_per_second": 79.306,
+      "eval_steps_per_second": 10.401,
+      "step": 10675
+    },
+    {
+      "epoch": 176.0,
+      "eval_loss": 11.40168571472168,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.669,
+      "eval_steps_per_second": 10.448,
+      "step": 10736
+    },
+    {
+      "epoch": 177.0,
+      "eval_loss": 11.534231185913086,
+      "eval_runtime": 1.533,
+      "eval_samples_per_second": 79.584,
+      "eval_steps_per_second": 10.437,
+      "step": 10797
+    },
+    {
+      "epoch": 178.0,
+      "eval_loss": 11.227432250976562,
+      "eval_runtime": 1.532,
+      "eval_samples_per_second": 79.634,
+      "eval_steps_per_second": 10.444,
+      "step": 10858
+    },
+    {
+      "epoch": 179.0,
+      "eval_loss": 11.432555198669434,
+      "eval_runtime": 1.5369,
+      "eval_samples_per_second": 79.379,
+      "eval_steps_per_second": 10.41,
+      "step": 10919
+    },
+    {
+      "epoch": 180.0,
+      "eval_loss": 11.477941513061523,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.671,
+      "eval_steps_per_second": 10.449,
+      "step": 10980
+    },
+    {
+      "epoch": 180.33,
+      "learning_rate": 3.19672131147541e-08,
+      "loss": 10.2637,
+      "step": 11000
+    },
+    {
+      "epoch": 181.0,
+      "eval_loss": 11.34237003326416,
+      "eval_runtime": 1.5403,
+      "eval_samples_per_second": 79.205,
+      "eval_steps_per_second": 10.388,
+      "step": 11041
+    },
+    {
+      "epoch": 182.0,
+      "eval_loss": 11.245933532714844,
+      "eval_runtime": 1.5372,
+      "eval_samples_per_second": 79.363,
+      "eval_steps_per_second": 10.408,
+      "step": 11102
+    },
+    {
+      "epoch": 183.0,
+      "eval_loss": 11.317768096923828,
+      "eval_runtime": 1.5396,
+      "eval_samples_per_second": 79.241,
+      "eval_steps_per_second": 10.392,
+      "step": 11163
+    },
+    {
+      "epoch": 184.0,
+      "eval_loss": 11.325362205505371,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.67,
+      "eval_steps_per_second": 10.449,
+      "step": 11224
+    },
+    {
+      "epoch": 185.0,
+      "eval_loss": 11.263531684875488,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.611,
+      "eval_steps_per_second": 10.441,
+      "step": 11285
+    },
+    {
+      "epoch": 186.0,
+      "eval_loss": 11.21447467803955,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.641,
+      "eval_steps_per_second": 10.445,
+      "step": 11346
+    },
+    {
+      "epoch": 187.0,
+      "eval_loss": 11.32798957824707,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.67,
+      "eval_steps_per_second": 10.449,
+      "step": 11407
+    },
+    {
+      "epoch": 188.0,
+      "eval_loss": 11.337279319763184,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.571,
+      "eval_steps_per_second": 10.436,
+      "step": 11468
+    },
+    {
+      "epoch": 188.52,
+      "learning_rate": 3.1147540983606555e-08,
+      "loss": 10.2837,
+      "step": 11500
+    },
+    {
+      "epoch": 189.0,
+      "eval_loss": 11.180814743041992,
+      "eval_runtime": 1.5382,
+      "eval_samples_per_second": 79.316,
+      "eval_steps_per_second": 10.402,
+      "step": 11529
+    },
+    {
+      "epoch": 190.0,
+      "eval_loss": 11.221953392028809,
+      "eval_runtime": 1.5474,
+      "eval_samples_per_second": 78.844,
+      "eval_steps_per_second": 10.34,
+      "step": 11590
+    },
+    {
+      "epoch": 191.0,
+      "eval_loss": 11.125133514404297,
+      "eval_runtime": 1.5562,
+      "eval_samples_per_second": 78.399,
+      "eval_steps_per_second": 10.282,
+      "step": 11651
+    },
+    {
+      "epoch": 192.0,
+      "eval_loss": 11.356901168823242,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.64,
+      "eval_steps_per_second": 10.445,
+      "step": 11712
+    },
+    {
+      "epoch": 193.0,
+      "eval_loss": 11.188824653625488,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.686,
+      "eval_steps_per_second": 10.451,
+      "step": 11773
+    },
+    {
+      "epoch": 194.0,
+      "eval_loss": 11.257243156433105,
+      "eval_runtime": 1.5371,
+      "eval_samples_per_second": 79.368,
+      "eval_steps_per_second": 10.409,
+      "step": 11834
+    },
+    {
+      "epoch": 195.0,
+      "eval_loss": 11.315752029418945,
+      "eval_runtime": 1.5316,
+      "eval_samples_per_second": 79.655,
+      "eval_steps_per_second": 10.447,
+      "step": 11895
+    },
+    {
+      "epoch": 196.0,
+      "eval_loss": 11.294753074645996,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.571,
+      "eval_steps_per_second": 10.436,
+      "step": 11956
+    },
+    {
+      "epoch": 196.72,
+      "learning_rate": 3.032786885245902e-08,
+      "loss": 10.2404,
+      "step": 12000
+    },
+    {
+      "epoch": 197.0,
+      "eval_loss": 11.21860408782959,
+      "eval_runtime": 1.5446,
+      "eval_samples_per_second": 78.984,
+      "eval_steps_per_second": 10.359,
+      "step": 12017
+    },
+    {
+      "epoch": 198.0,
+      "eval_loss": 11.28231143951416,
+      "eval_runtime": 1.5623,
+      "eval_samples_per_second": 78.089,
+      "eval_steps_per_second": 10.241,
+      "step": 12078
+    },
+    {
+      "epoch": 199.0,
+      "eval_loss": 11.250580787658691,
+      "eval_runtime": 1.5444,
+      "eval_samples_per_second": 78.996,
+      "eval_steps_per_second": 10.36,
+      "step": 12139
+    },
+    {
+      "epoch": 200.0,
+      "eval_loss": 11.55415153503418,
+      "eval_runtime": 1.5349,
+      "eval_samples_per_second": 79.484,
+      "eval_steps_per_second": 10.424,
+      "step": 12200
+    },
+    {
+      "epoch": 201.0,
+      "eval_loss": 11.313996315002441,
+      "eval_runtime": 1.5323,
+      "eval_samples_per_second": 79.619,
+      "eval_steps_per_second": 10.442,
+      "step": 12261
+    },
+    {
+      "epoch": 202.0,
+      "eval_loss": 11.200800895690918,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.664,
+      "eval_steps_per_second": 10.448,
+      "step": 12322
+    },
+    {
+      "epoch": 203.0,
+      "eval_loss": 11.157076835632324,
+      "eval_runtime": 1.5375,
+      "eval_samples_per_second": 79.352,
+      "eval_steps_per_second": 10.407,
+      "step": 12383
+    },
+    {
+      "epoch": 204.0,
+      "eval_loss": 11.233716011047363,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.671,
+      "eval_steps_per_second": 10.449,
+      "step": 12444
+    },
+    {
+      "epoch": 204.92,
+      "learning_rate": 2.9508196721311475e-08,
+      "loss": 10.2304,
+      "step": 12500
+    },
+    {
+      "epoch": 205.0,
+      "eval_loss": 11.270787239074707,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.615,
+      "eval_steps_per_second": 10.441,
+      "step": 12505
+    },
+    {
+      "epoch": 206.0,
+      "eval_loss": 11.30806827545166,
+      "eval_runtime": 1.5392,
+      "eval_samples_per_second": 79.261,
+      "eval_steps_per_second": 10.395,
+      "step": 12566
+    },
+    {
+      "epoch": 207.0,
+      "eval_loss": 11.11026382446289,
+      "eval_runtime": 1.5416,
+      "eval_samples_per_second": 79.137,
+      "eval_steps_per_second": 10.379,
+      "step": 12627
+    },
+    {
+      "epoch": 208.0,
+      "eval_loss": 11.123950958251953,
+      "eval_runtime": 1.538,
+      "eval_samples_per_second": 79.323,
+      "eval_steps_per_second": 10.403,
+      "step": 12688
+    },
+    {
+      "epoch": 209.0,
+      "eval_loss": 11.3450288772583,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.55,
+      "eval_steps_per_second": 10.433,
+      "step": 12749
+    },
+    {
+      "epoch": 210.0,
+      "eval_loss": 11.059000015258789,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.683,
+      "eval_steps_per_second": 10.45,
+      "step": 12810
+    },
+    {
+      "epoch": 211.0,
+      "eval_loss": 11.257308006286621,
+      "eval_runtime": 1.5316,
+      "eval_samples_per_second": 79.654,
+      "eval_steps_per_second": 10.446,
+      "step": 12871
+    },
+    {
+      "epoch": 212.0,
+      "eval_loss": 11.207562446594238,
+      "eval_runtime": 1.5416,
+      "eval_samples_per_second": 79.141,
+      "eval_steps_per_second": 10.379,
+      "step": 12932
+    },
+    {
+      "epoch": 213.0,
+      "eval_loss": 11.18131160736084,
+      "eval_runtime": 1.5315,
+      "eval_samples_per_second": 79.659,
+      "eval_steps_per_second": 10.447,
+      "step": 12993
+    },
+    {
+      "epoch": 213.11,
+      "learning_rate": 2.8688524590163933e-08,
+      "loss": 10.2405,
+      "step": 13000
+    },
+    {
+      "epoch": 214.0,
+      "eval_loss": 11.338172912597656,
+      "eval_runtime": 1.5378,
+      "eval_samples_per_second": 79.334,
+      "eval_steps_per_second": 10.404,
+      "step": 13054
+    },
+    {
+      "epoch": 215.0,
+      "eval_loss": 11.409613609313965,
+      "eval_runtime": 1.5389,
+      "eval_samples_per_second": 79.277,
+      "eval_steps_per_second": 10.397,
+      "step": 13115
+    },
+    {
+      "epoch": 216.0,
+      "eval_loss": 11.152140617370605,
+      "eval_runtime": 1.5382,
+      "eval_samples_per_second": 79.313,
+      "eval_steps_per_second": 10.402,
+      "step": 13176
+    },
+    {
+      "epoch": 217.0,
+      "eval_loss": 11.312527656555176,
+      "eval_runtime": 1.5316,
+      "eval_samples_per_second": 79.653,
+      "eval_steps_per_second": 10.446,
+      "step": 13237
+    },
+    {
+      "epoch": 218.0,
+      "eval_loss": 11.191665649414062,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.682,
+      "eval_steps_per_second": 10.45,
+      "step": 13298
+    },
+    {
+      "epoch": 219.0,
+      "eval_loss": 11.279202461242676,
+      "eval_runtime": 1.5384,
+      "eval_samples_per_second": 79.305,
+      "eval_steps_per_second": 10.401,
+      "step": 13359
+    },
+    {
+      "epoch": 220.0,
+      "eval_loss": 11.123604774475098,
+      "eval_runtime": 1.5361,
+      "eval_samples_per_second": 79.422,
+      "eval_steps_per_second": 10.416,
+      "step": 13420
+    },
+    {
+      "epoch": 221.0,
+      "eval_loss": 11.239697456359863,
+      "eval_runtime": 1.5358,
+      "eval_samples_per_second": 79.438,
+      "eval_steps_per_second": 10.418,
+      "step": 13481
+    },
+    {
+      "epoch": 221.31,
+      "learning_rate": 2.786885245901639e-08,
+      "loss": 10.2096,
+      "step": 13500
+    },
+    {
+      "epoch": 222.0,
+      "eval_loss": 11.187539100646973,
+      "eval_runtime": 1.5423,
+      "eval_samples_per_second": 79.104,
+      "eval_steps_per_second": 10.374,
+      "step": 13542
+    },
+    {
+      "epoch": 223.0,
+      "eval_loss": 11.311732292175293,
+      "eval_runtime": 1.5401,
+      "eval_samples_per_second": 79.213,
+      "eval_steps_per_second": 10.389,
+      "step": 13603
+    },
+    {
+      "epoch": 224.0,
+      "eval_loss": 11.156542778015137,
+      "eval_runtime": 1.5374,
+      "eval_samples_per_second": 79.353,
+      "eval_steps_per_second": 10.407,
+      "step": 13664
+    },
+    {
+      "epoch": 225.0,
+      "eval_loss": 11.416545867919922,
+      "eval_runtime": 1.533,
+      "eval_samples_per_second": 79.581,
+      "eval_steps_per_second": 10.437,
+      "step": 13725
+    },
+    {
+      "epoch": 226.0,
+      "eval_loss": 11.147370338439941,
+      "eval_runtime": 1.5328,
+      "eval_samples_per_second": 79.591,
+      "eval_steps_per_second": 10.438,
+      "step": 13786
+    },
+    {
+      "epoch": 227.0,
+      "eval_loss": 10.985366821289062,
+      "eval_runtime": 1.5331,
+      "eval_samples_per_second": 79.579,
+      "eval_steps_per_second": 10.437,
+      "step": 13847
+    },
+    {
+      "epoch": 228.0,
+      "eval_loss": 11.134580612182617,
+      "eval_runtime": 1.5339,
+      "eval_samples_per_second": 79.538,
+      "eval_steps_per_second": 10.431,
+      "step": 13908
+    },
+    {
+      "epoch": 229.0,
+      "eval_loss": 11.212328910827637,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.625,
+      "eval_steps_per_second": 10.443,
+      "step": 13969
+    },
+    {
+      "epoch": 229.51,
+      "learning_rate": 2.7049180327868852e-08,
+      "loss": 10.1998,
+      "step": 14000
+    },
+    {
+      "epoch": 230.0,
+      "eval_loss": 11.24535846710205,
+      "eval_runtime": 1.5423,
+      "eval_samples_per_second": 79.101,
+      "eval_steps_per_second": 10.374,
+      "step": 14030
+    },
+    {
+      "epoch": 231.0,
+      "eval_loss": 11.335315704345703,
+      "eval_runtime": 1.5438,
+      "eval_samples_per_second": 79.027,
+      "eval_steps_per_second": 10.364,
+      "step": 14091
+    },
+    {
+      "epoch": 232.0,
+      "eval_loss": 11.305168151855469,
+      "eval_runtime": 1.5455,
+      "eval_samples_per_second": 78.937,
+      "eval_steps_per_second": 10.352,
+      "step": 14152
+    },
+    {
+      "epoch": 233.0,
+      "eval_loss": 11.177278518676758,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.631,
+      "eval_steps_per_second": 10.443,
+      "step": 14213
+    },
+    {
+      "epoch": 234.0,
+      "eval_loss": 11.132712364196777,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.57,
+      "eval_steps_per_second": 10.435,
+      "step": 14274
+    },
+    {
+      "epoch": 235.0,
+      "eval_loss": 11.310881614685059,
+      "eval_runtime": 1.5363,
+      "eval_samples_per_second": 79.411,
+      "eval_steps_per_second": 10.415,
+      "step": 14335
+    },
+    {
+      "epoch": 236.0,
+      "eval_loss": 11.178828239440918,
+      "eval_runtime": 1.5317,
+      "eval_samples_per_second": 79.652,
+      "eval_steps_per_second": 10.446,
+      "step": 14396
+    },
+    {
+      "epoch": 237.0,
+      "eval_loss": 11.376667022705078,
+      "eval_runtime": 1.542,
+      "eval_samples_per_second": 79.118,
+      "eval_steps_per_second": 10.376,
+      "step": 14457
+    },
+    {
+      "epoch": 237.7,
+      "learning_rate": 2.622950819672131e-08,
+      "loss": 10.1947,
+      "step": 14500
+    },
+    {
+      "epoch": 238.0,
+      "eval_loss": 11.21568489074707,
+      "eval_runtime": 1.5435,
+      "eval_samples_per_second": 79.039,
+      "eval_steps_per_second": 10.366,
+      "step": 14518
+    },
+    {
+      "epoch": 239.0,
+      "eval_loss": 11.210213661193848,
+      "eval_runtime": 1.555,
+      "eval_samples_per_second": 78.455,
+      "eval_steps_per_second": 10.289,
+      "step": 14579
+    },
+    {
+      "epoch": 240.0,
+      "eval_loss": 11.184165954589844,
+      "eval_runtime": 1.5406,
+      "eval_samples_per_second": 79.191,
+      "eval_steps_per_second": 10.386,
+      "step": 14640
+    },
+    {
+      "epoch": 241.0,
+      "eval_loss": 11.139164924621582,
+      "eval_runtime": 1.5333,
+      "eval_samples_per_second": 79.568,
+      "eval_steps_per_second": 10.435,
+      "step": 14701
+    },
+    {
+      "epoch": 242.0,
+      "eval_loss": 11.139853477478027,
+      "eval_runtime": 1.5354,
+      "eval_samples_per_second": 79.456,
+      "eval_steps_per_second": 10.42,
+      "step": 14762
+    },
+    {
+      "epoch": 243.0,
+      "eval_loss": 11.163023948669434,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.55,
+      "eval_steps_per_second": 10.433,
+      "step": 14823
+    },
+    {
+      "epoch": 244.0,
+      "eval_loss": 11.19721794128418,
+      "eval_runtime": 1.5334,
+      "eval_samples_per_second": 79.56,
+      "eval_steps_per_second": 10.434,
+      "step": 14884
+    },
+    {
+      "epoch": 245.0,
+      "eval_loss": 11.054798126220703,
+      "eval_runtime": 1.5315,
+      "eval_samples_per_second": 79.661,
+      "eval_steps_per_second": 10.447,
+      "step": 14945
+    },
+    {
+      "epoch": 245.9,
+      "learning_rate": 2.5409836065573768e-08,
+      "loss": 10.1922,
+      "step": 15000
+    },
+    {
+      "epoch": 246.0,
+      "eval_loss": 11.127946853637695,
+      "eval_runtime": 1.5532,
+      "eval_samples_per_second": 78.545,
+      "eval_steps_per_second": 10.301,
+      "step": 15006
+    },
+    {
+      "epoch": 247.0,
+      "eval_loss": 11.096878051757812,
+      "eval_runtime": 1.5376,
+      "eval_samples_per_second": 79.343,
+      "eval_steps_per_second": 10.406,
+      "step": 15067
+    },
+    {
+      "epoch": 248.0,
+      "eval_loss": 11.234783172607422,
+      "eval_runtime": 1.5405,
+      "eval_samples_per_second": 79.193,
+      "eval_steps_per_second": 10.386,
+      "step": 15128
+    },
+    {
+      "epoch": 249.0,
+      "eval_loss": 11.115097045898438,
+      "eval_runtime": 1.5469,
+      "eval_samples_per_second": 78.865,
+      "eval_steps_per_second": 10.343,
+      "step": 15189
+    },
+    {
+      "epoch": 250.0,
+      "eval_loss": 11.53684139251709,
+      "eval_runtime": 1.5325,
+      "eval_samples_per_second": 79.611,
+      "eval_steps_per_second": 10.441,
+      "step": 15250
+    },
+    {
+      "epoch": 251.0,
+      "eval_loss": 11.224416732788086,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.627,
+      "eval_steps_per_second": 10.443,
+      "step": 15311
+    },
+    {
+      "epoch": 252.0,
+      "eval_loss": 11.210186958312988,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.387,
+      "eval_steps_per_second": 10.411,
+      "step": 15372
+    },
+    {
+      "epoch": 253.0,
+      "eval_loss": 11.273506164550781,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.622,
+      "eval_steps_per_second": 10.442,
+      "step": 15433
+    },
+    {
+      "epoch": 254.0,
+      "eval_loss": 11.3226900100708,
+      "eval_runtime": 1.5371,
+      "eval_samples_per_second": 79.371,
+      "eval_steps_per_second": 10.409,
+      "step": 15494
+    },
+    {
+      "epoch": 254.1,
+      "learning_rate": 2.459016393442623e-08,
+      "loss": 10.1994,
+      "step": 15500
+    },
+    {
+      "epoch": 255.0,
+      "eval_loss": 11.237663269042969,
+      "eval_runtime": 1.5421,
+      "eval_samples_per_second": 79.114,
+      "eval_steps_per_second": 10.376,
+      "step": 15555
+    },
+    {
+      "epoch": 256.0,
+      "eval_loss": 11.283459663391113,
+      "eval_runtime": 1.539,
+      "eval_samples_per_second": 79.272,
+      "eval_steps_per_second": 10.396,
+      "step": 15616
+    },
+    {
+      "epoch": 257.0,
+      "eval_loss": 11.347498893737793,
+      "eval_runtime": 1.5437,
+      "eval_samples_per_second": 79.032,
+      "eval_steps_per_second": 10.365,
+      "step": 15677
+    },
+    {
+      "epoch": 258.0,
+      "eval_loss": 11.209207534790039,
+      "eval_runtime": 1.5326,
+      "eval_samples_per_second": 79.605,
+      "eval_steps_per_second": 10.44,
+      "step": 15738
+    },
+    {
+      "epoch": 259.0,
+      "eval_loss": 11.19356918334961,
+      "eval_runtime": 1.5356,
+      "eval_samples_per_second": 79.448,
+      "eval_steps_per_second": 10.419,
+      "step": 15799
+    },
+    {
+      "epoch": 260.0,
+      "eval_loss": 11.0318603515625,
+      "eval_runtime": 1.5384,
+      "eval_samples_per_second": 79.303,
+      "eval_steps_per_second": 10.4,
+      "step": 15860
+    },
+    {
+      "epoch": 261.0,
+      "eval_loss": 11.191557884216309,
+      "eval_runtime": 1.5356,
+      "eval_samples_per_second": 79.447,
+      "eval_steps_per_second": 10.419,
+      "step": 15921
+    },
+    {
+      "epoch": 262.0,
+      "eval_loss": 11.135727882385254,
+      "eval_runtime": 1.5318,
+      "eval_samples_per_second": 79.643,
+      "eval_steps_per_second": 10.445,
+      "step": 15982
+    },
+    {
+      "epoch": 262.29,
+      "learning_rate": 2.3770491803278688e-08,
+      "loss": 10.1883,
+      "step": 16000
+    },
+    {
+      "epoch": 263.0,
+      "eval_loss": 10.97315788269043,
+      "eval_runtime": 1.5381,
+      "eval_samples_per_second": 79.317,
+      "eval_steps_per_second": 10.402,
+      "step": 16043
+    },
+    {
+      "epoch": 264.0,
+      "eval_loss": 11.183859825134277,
+      "eval_runtime": 1.5418,
+      "eval_samples_per_second": 79.131,
+      "eval_steps_per_second": 10.378,
+      "step": 16104
+    },
+    {
+      "epoch": 265.0,
+      "eval_loss": 11.070130348205566,
+      "eval_runtime": 1.5422,
+      "eval_samples_per_second": 79.109,
+      "eval_steps_per_second": 10.375,
+      "step": 16165
+    },
+    {
+      "epoch": 266.0,
+      "eval_loss": 11.161293029785156,
+      "eval_runtime": 1.5365,
+      "eval_samples_per_second": 79.403,
+      "eval_steps_per_second": 10.414,
+      "step": 16226
+    },
+    {
+      "epoch": 267.0,
+      "eval_loss": 11.130182266235352,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.572,
+      "eval_steps_per_second": 10.436,
+      "step": 16287
+    },
+    {
+      "epoch": 268.0,
+      "eval_loss": 11.095097541809082,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.642,
+      "eval_steps_per_second": 10.445,
+      "step": 16348
+    },
+    {
+      "epoch": 269.0,
+      "eval_loss": 11.05788803100586,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.552,
+      "eval_steps_per_second": 10.433,
+      "step": 16409
+    },
+    {
+      "epoch": 270.0,
+      "eval_loss": 11.145880699157715,
+      "eval_runtime": 1.5341,
+      "eval_samples_per_second": 79.523,
+      "eval_steps_per_second": 10.429,
+      "step": 16470
+    },
+    {
+      "epoch": 270.49,
+      "learning_rate": 2.2950819672131146e-08,
+      "loss": 10.1863,
+      "step": 16500
+    },
+    {
+      "epoch": 271.0,
+      "eval_loss": 11.196855545043945,
+      "eval_runtime": 1.5401,
+      "eval_samples_per_second": 79.218,
+      "eval_steps_per_second": 10.389,
+      "step": 16531
+    },
+    {
+      "epoch": 272.0,
+      "eval_loss": 11.127517700195312,
+      "eval_runtime": 1.5433,
+      "eval_samples_per_second": 79.049,
+      "eval_steps_per_second": 10.367,
+      "step": 16592
+    },
+    {
+      "epoch": 273.0,
+      "eval_loss": 11.111509323120117,
+      "eval_runtime": 1.5367,
+      "eval_samples_per_second": 79.393,
+      "eval_steps_per_second": 10.412,
+      "step": 16653
+    },
+    {
+      "epoch": 274.0,
+      "eval_loss": 11.128540992736816,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.616,
+      "eval_steps_per_second": 10.441,
+      "step": 16714
+    },
+    {
+      "epoch": 275.0,
+      "eval_loss": 11.105259895324707,
+      "eval_runtime": 1.5326,
+      "eval_samples_per_second": 79.601,
+      "eval_steps_per_second": 10.439,
+      "step": 16775
+    },
+    {
+      "epoch": 276.0,
+      "eval_loss": 11.010540008544922,
+      "eval_runtime": 1.5357,
+      "eval_samples_per_second": 79.442,
+      "eval_steps_per_second": 10.419,
+      "step": 16836
+    },
+    {
+      "epoch": 277.0,
+      "eval_loss": 11.137755393981934,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.53,
+      "eval_steps_per_second": 10.43,
+      "step": 16897
+    },
+    {
+      "epoch": 278.0,
+      "eval_loss": 11.077096939086914,
+      "eval_runtime": 1.533,
+      "eval_samples_per_second": 79.582,
+      "eval_steps_per_second": 10.437,
+      "step": 16958
+    },
+    {
+      "epoch": 278.69,
+      "learning_rate": 2.2131147540983604e-08,
+      "loss": 10.1614,
+      "step": 17000
+    },
+    {
+      "epoch": 279.0,
+      "eval_loss": 11.061995506286621,
+      "eval_runtime": 1.5452,
+      "eval_samples_per_second": 78.952,
+      "eval_steps_per_second": 10.354,
+      "step": 17019
+    },
+    {
+      "epoch": 280.0,
+      "eval_loss": 10.990643501281738,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.385,
+      "eval_steps_per_second": 10.411,
+      "step": 17080
+    },
+    {
+      "epoch": 281.0,
+      "eval_loss": 11.077146530151367,
+      "eval_runtime": 1.539,
+      "eval_samples_per_second": 79.273,
+      "eval_steps_per_second": 10.396,
+      "step": 17141
+    },
+    {
+      "epoch": 282.0,
+      "eval_loss": 11.035726547241211,
+      "eval_runtime": 1.5344,
+      "eval_samples_per_second": 79.51,
+      "eval_steps_per_second": 10.428,
+      "step": 17202
+    },
+    {
+      "epoch": 283.0,
+      "eval_loss": 11.0416841506958,
+      "eval_runtime": 1.5338,
+      "eval_samples_per_second": 79.542,
+      "eval_steps_per_second": 10.432,
+      "step": 17263
+    },
+    {
+      "epoch": 284.0,
+      "eval_loss": 11.028667449951172,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.612,
+      "eval_steps_per_second": 10.441,
+      "step": 17324
+    },
+    {
+      "epoch": 285.0,
+      "eval_loss": 11.117180824279785,
+      "eval_runtime": 1.5342,
+      "eval_samples_per_second": 79.518,
+      "eval_steps_per_second": 10.429,
+      "step": 17385
+    },
+    {
+      "epoch": 286.0,
+      "eval_loss": 10.925678253173828,
+      "eval_runtime": 1.5356,
+      "eval_samples_per_second": 79.449,
+      "eval_steps_per_second": 10.42,
+      "step": 17446
+    },
+    {
+      "epoch": 286.88,
+      "learning_rate": 2.1311475409836065e-08,
+      "loss": 10.1717,
+      "step": 17500
+    },
+    {
+      "epoch": 287.0,
+      "eval_loss": 11.231226921081543,
+      "eval_runtime": 1.542,
+      "eval_samples_per_second": 79.12,
+      "eval_steps_per_second": 10.376,
+      "step": 17507
+    },
+    {
+      "epoch": 288.0,
+      "eval_loss": 11.380009651184082,
+      "eval_runtime": 1.5602,
+      "eval_samples_per_second": 78.195,
+      "eval_steps_per_second": 10.255,
+      "step": 17568
+    },
+    {
+      "epoch": 289.0,
+      "eval_loss": 11.13857364654541,
+      "eval_runtime": 1.5491,
+      "eval_samples_per_second": 78.754,
+      "eval_steps_per_second": 10.328,
+      "step": 17629
+    },
+    {
+      "epoch": 290.0,
+      "eval_loss": 11.172418594360352,
+      "eval_runtime": 1.5472,
+      "eval_samples_per_second": 78.854,
+      "eval_steps_per_second": 10.341,
+      "step": 17690
+    },
+    {
+      "epoch": 291.0,
+      "eval_loss": 11.162772178649902,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.529,
+      "eval_steps_per_second": 10.43,
+      "step": 17751
+    },
+    {
+      "epoch": 292.0,
+      "eval_loss": 11.122556686401367,
+      "eval_runtime": 1.5424,
+      "eval_samples_per_second": 79.097,
+      "eval_steps_per_second": 10.373,
+      "step": 17812
+    },
+    {
+      "epoch": 293.0,
+      "eval_loss": 11.19551944732666,
+      "eval_runtime": 1.5328,
+      "eval_samples_per_second": 79.592,
+      "eval_steps_per_second": 10.438,
+      "step": 17873
+    },
+    {
+      "epoch": 294.0,
+      "eval_loss": 11.138795852661133,
+      "eval_runtime": 1.537,
+      "eval_samples_per_second": 79.376,
+      "eval_steps_per_second": 10.41,
+      "step": 17934
+    },
+    {
+      "epoch": 295.0,
+      "eval_loss": 11.087359428405762,
+      "eval_runtime": 1.5338,
+      "eval_samples_per_second": 79.543,
+      "eval_steps_per_second": 10.432,
+      "step": 17995
+    },
+    {
+      "epoch": 295.08,
+      "learning_rate": 2.0491803278688523e-08,
+      "loss": 10.1806,
+      "step": 18000
+    },
+    {
+      "epoch": 296.0,
+      "eval_loss": 11.081265449523926,
+      "eval_runtime": 1.5381,
+      "eval_samples_per_second": 79.319,
+      "eval_steps_per_second": 10.403,
+      "step": 18056
+    },
+    {
+      "epoch": 297.0,
+      "eval_loss": 11.147530555725098,
+      "eval_runtime": 1.5414,
+      "eval_samples_per_second": 79.151,
+      "eval_steps_per_second": 10.381,
+      "step": 18117
+    },
+    {
+      "epoch": 298.0,
+      "eval_loss": 11.167841911315918,
+      "eval_runtime": 1.5414,
+      "eval_samples_per_second": 79.146,
+      "eval_steps_per_second": 10.38,
+      "step": 18178
+    },
+    {
+      "epoch": 299.0,
+      "eval_loss": 11.244978904724121,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.623,
+      "eval_steps_per_second": 10.442,
+      "step": 18239
+    },
+    {
+      "epoch": 300.0,
+      "eval_loss": 11.193556785583496,
+      "eval_runtime": 1.5334,
+      "eval_samples_per_second": 79.563,
+      "eval_steps_per_second": 10.434,
+      "step": 18300
+    },
+    {
+      "epoch": 301.0,
+      "eval_loss": 11.102090835571289,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.642,
+      "eval_steps_per_second": 10.445,
+      "step": 18361
+    },
+    {
+      "epoch": 302.0,
+      "eval_loss": 11.187400817871094,
+      "eval_runtime": 1.5378,
+      "eval_samples_per_second": 79.332,
+      "eval_steps_per_second": 10.404,
+      "step": 18422
+    },
+    {
+      "epoch": 303.0,
+      "eval_loss": 11.171931266784668,
+      "eval_runtime": 1.5351,
+      "eval_samples_per_second": 79.473,
+      "eval_steps_per_second": 10.423,
+      "step": 18483
+    },
+    {
+      "epoch": 303.28,
+      "learning_rate": 1.967213114754098e-08,
+      "loss": 10.1683,
+      "step": 18500
+    },
+    {
+      "epoch": 304.0,
+      "eval_loss": 11.155352592468262,
+      "eval_runtime": 1.5403,
+      "eval_samples_per_second": 79.204,
+      "eval_steps_per_second": 10.387,
+      "step": 18544
+    },
+    {
+      "epoch": 305.0,
+      "eval_loss": 11.0771484375,
+      "eval_runtime": 1.5429,
+      "eval_samples_per_second": 79.074,
+      "eval_steps_per_second": 10.37,
+      "step": 18605
+    },
+    {
+      "epoch": 306.0,
+      "eval_loss": 11.067580223083496,
+      "eval_runtime": 1.5407,
+      "eval_samples_per_second": 79.183,
+      "eval_steps_per_second": 10.385,
+      "step": 18666
+    },
+    {
+      "epoch": 307.0,
+      "eval_loss": 11.128029823303223,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.63,
+      "eval_steps_per_second": 10.443,
+      "step": 18727
+    },
+    {
+      "epoch": 308.0,
+      "eval_loss": 11.023638725280762,
+      "eval_runtime": 1.535,
+      "eval_samples_per_second": 79.477,
+      "eval_steps_per_second": 10.423,
+      "step": 18788
+    },
+    {
+      "epoch": 309.0,
+      "eval_loss": 11.14404582977295,
+      "eval_runtime": 1.5352,
+      "eval_samples_per_second": 79.47,
+      "eval_steps_per_second": 10.422,
+      "step": 18849
+    },
+    {
+      "epoch": 310.0,
+      "eval_loss": 11.184294700622559,
+      "eval_runtime": 1.5344,
+      "eval_samples_per_second": 79.508,
+      "eval_steps_per_second": 10.427,
+      "step": 18910
+    },
+    {
+      "epoch": 311.0,
+      "eval_loss": 11.04742431640625,
+      "eval_runtime": 1.5383,
+      "eval_samples_per_second": 79.309,
+      "eval_steps_per_second": 10.401,
+      "step": 18971
+    },
+    {
+      "epoch": 311.47,
+      "learning_rate": 1.885245901639344e-08,
+      "loss": 10.1437,
+      "step": 19000
+    },
+    {
+      "epoch": 312.0,
+      "eval_loss": 11.039079666137695,
+      "eval_runtime": 1.5399,
+      "eval_samples_per_second": 79.224,
+      "eval_steps_per_second": 10.39,
+      "step": 19032
+    },
+    {
+      "epoch": 313.0,
+      "eval_loss": 10.914813995361328,
+      "eval_runtime": 1.5428,
+      "eval_samples_per_second": 79.078,
+      "eval_steps_per_second": 10.371,
+      "step": 19093
+    },
+    {
+      "epoch": 314.0,
+      "eval_loss": 11.057476997375488,
+      "eval_runtime": 1.5481,
+      "eval_samples_per_second": 78.806,
+      "eval_steps_per_second": 10.335,
+      "step": 19154
+    },
+    {
+      "epoch": 315.0,
+      "eval_loss": 11.195489883422852,
+      "eval_runtime": 1.536,
+      "eval_samples_per_second": 79.425,
+      "eval_steps_per_second": 10.416,
+      "step": 19215
+    },
+    {
+      "epoch": 316.0,
+      "eval_loss": 11.005314826965332,
+      "eval_runtime": 1.5327,
+      "eval_samples_per_second": 79.597,
+      "eval_steps_per_second": 10.439,
+      "step": 19276
+    },
+    {
+      "epoch": 317.0,
+      "eval_loss": 11.080973625183105,
+      "eval_runtime": 1.5413,
+      "eval_samples_per_second": 79.154,
+      "eval_steps_per_second": 10.381,
+      "step": 19337
+    },
+    {
+      "epoch": 318.0,
+      "eval_loss": 11.13598346710205,
+      "eval_runtime": 1.5363,
+      "eval_samples_per_second": 79.411,
+      "eval_steps_per_second": 10.415,
+      "step": 19398
+    },
+    {
+      "epoch": 319.0,
+      "eval_loss": 11.229124069213867,
+      "eval_runtime": 1.5344,
+      "eval_samples_per_second": 79.512,
+      "eval_steps_per_second": 10.428,
+      "step": 19459
+    },
+    {
+      "epoch": 319.67,
+      "learning_rate": 1.80327868852459e-08,
+      "loss": 10.1539,
+      "step": 19500
+    },
+    {
+      "epoch": 320.0,
+      "eval_loss": 11.023887634277344,
+      "eval_runtime": 1.5371,
+      "eval_samples_per_second": 79.369,
+      "eval_steps_per_second": 10.409,
+      "step": 19520
+    },
+    {
+      "epoch": 321.0,
+      "eval_loss": 11.121563911437988,
+      "eval_runtime": 1.5411,
+      "eval_samples_per_second": 79.162,
+      "eval_steps_per_second": 10.382,
+      "step": 19581
+    },
+    {
+      "epoch": 322.0,
+      "eval_loss": 11.251620292663574,
+      "eval_runtime": 1.545,
+      "eval_samples_per_second": 78.965,
+      "eval_steps_per_second": 10.356,
+      "step": 19642
+    },
+    {
+      "epoch": 323.0,
+      "eval_loss": 10.975918769836426,
+      "eval_runtime": 1.5345,
+      "eval_samples_per_second": 79.505,
+      "eval_steps_per_second": 10.427,
+      "step": 19703
+    },
+    {
+      "epoch": 324.0,
+      "eval_loss": 11.039804458618164,
+      "eval_runtime": 1.5356,
+      "eval_samples_per_second": 79.446,
+      "eval_steps_per_second": 10.419,
+      "step": 19764
+    },
+    {
+      "epoch": 325.0,
+      "eval_loss": 11.043091773986816,
+      "eval_runtime": 1.5325,
+      "eval_samples_per_second": 79.607,
+      "eval_steps_per_second": 10.44,
+      "step": 19825
+    },
+    {
+      "epoch": 326.0,
+      "eval_loss": 10.915124893188477,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.612,
+      "eval_steps_per_second": 10.441,
+      "step": 19886
+    },
+    {
+      "epoch": 327.0,
+      "eval_loss": 11.090548515319824,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.626,
+      "eval_steps_per_second": 10.443,
+      "step": 19947
+    },
+    {
+      "epoch": 327.87,
+      "learning_rate": 1.721311475409836e-08,
+      "loss": 10.1432,
+      "step": 20000
+    },
+    {
+      "epoch": 328.0,
+      "eval_loss": 11.009906768798828,
+      "eval_runtime": 1.539,
+      "eval_samples_per_second": 79.274,
+      "eval_steps_per_second": 10.397,
+      "step": 20008
+    },
+    {
+      "epoch": 329.0,
+      "eval_loss": 11.089301109313965,
+      "eval_runtime": 1.5522,
+      "eval_samples_per_second": 78.6,
+      "eval_steps_per_second": 10.308,
+      "step": 20069
+    },
+    {
+      "epoch": 330.0,
+      "eval_loss": 11.134428024291992,
+      "eval_runtime": 1.5358,
+      "eval_samples_per_second": 79.437,
+      "eval_steps_per_second": 10.418,
+      "step": 20130
+    },
+    {
+      "epoch": 331.0,
+      "eval_loss": 11.068151473999023,
+      "eval_runtime": 1.5446,
+      "eval_samples_per_second": 78.987,
+      "eval_steps_per_second": 10.359,
+      "step": 20191
+    },
+    {
+      "epoch": 332.0,
+      "eval_loss": 10.955802917480469,
+      "eval_runtime": 1.5328,
+      "eval_samples_per_second": 79.593,
+      "eval_steps_per_second": 10.438,
+      "step": 20252
+    },
+    {
+      "epoch": 333.0,
+      "eval_loss": 11.066947937011719,
+      "eval_runtime": 1.5339,
+      "eval_samples_per_second": 79.537,
+      "eval_steps_per_second": 10.431,
+      "step": 20313
+    },
+    {
+      "epoch": 334.0,
+      "eval_loss": 11.055612564086914,
+      "eval_runtime": 1.5381,
+      "eval_samples_per_second": 79.321,
+      "eval_steps_per_second": 10.403,
+      "step": 20374
+    },
+    {
+      "epoch": 335.0,
+      "eval_loss": 11.20965576171875,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.666,
+      "eval_steps_per_second": 10.448,
+      "step": 20435
+    },
+    {
+      "epoch": 336.0,
+      "eval_loss": 11.019977569580078,
+      "eval_runtime": 1.5335,
+      "eval_samples_per_second": 79.554,
+      "eval_steps_per_second": 10.433,
+      "step": 20496
+    },
+    {
+      "epoch": 336.07,
+      "learning_rate": 1.6393442622950816e-08,
+      "loss": 10.1343,
+      "step": 20500
+    },
+    {
+      "epoch": 337.0,
+      "eval_loss": 10.968345642089844,
+      "eval_runtime": 1.5381,
+      "eval_samples_per_second": 79.32,
+      "eval_steps_per_second": 10.403,
+      "step": 20557
+    },
+    {
+      "epoch": 338.0,
+      "eval_loss": 10.98238754272461,
+      "eval_runtime": 1.5536,
+      "eval_samples_per_second": 78.526,
+      "eval_steps_per_second": 10.299,
+      "step": 20618
+    },
+    {
+      "epoch": 339.0,
+      "eval_loss": 11.156255722045898,
+      "eval_runtime": 1.536,
+      "eval_samples_per_second": 79.425,
+      "eval_steps_per_second": 10.416,
+      "step": 20679
+    },
+    {
+      "epoch": 340.0,
+      "eval_loss": 11.148921012878418,
+      "eval_runtime": 1.5358,
+      "eval_samples_per_second": 79.438,
+      "eval_steps_per_second": 10.418,
+      "step": 20740
+    },
+    {
+      "epoch": 341.0,
+      "eval_loss": 11.138899803161621,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.686,
+      "eval_steps_per_second": 10.451,
+      "step": 20801
+    },
+    {
+      "epoch": 342.0,
+      "eval_loss": 11.112798690795898,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.638,
+      "eval_steps_per_second": 10.444,
+      "step": 20862
+    },
+    {
+      "epoch": 343.0,
+      "eval_loss": 11.043689727783203,
+      "eval_runtime": 1.5362,
+      "eval_samples_per_second": 79.418,
+      "eval_steps_per_second": 10.416,
+      "step": 20923
+    },
+    {
+      "epoch": 344.0,
+      "eval_loss": 11.100526809692383,
+      "eval_runtime": 1.5335,
+      "eval_samples_per_second": 79.556,
+      "eval_steps_per_second": 10.434,
+      "step": 20984
+    },
+    {
+      "epoch": 344.26,
+      "learning_rate": 1.5573770491803278e-08,
+      "loss": 10.143,
+      "step": 21000
+    },
+    {
+      "epoch": 345.0,
+      "eval_loss": 11.16964054107666,
+      "eval_runtime": 1.5442,
+      "eval_samples_per_second": 79.003,
+      "eval_steps_per_second": 10.361,
+      "step": 21045
+    },
+    {
+      "epoch": 346.0,
+      "eval_loss": 11.135641098022461,
+      "eval_runtime": 1.5566,
+      "eval_samples_per_second": 78.375,
+      "eval_steps_per_second": 10.279,
+      "step": 21106
+    },
+    {
+      "epoch": 347.0,
+      "eval_loss": 11.079752922058105,
+      "eval_runtime": 1.5382,
+      "eval_samples_per_second": 79.311,
+      "eval_steps_per_second": 10.401,
+      "step": 21167
+    },
+    {
+      "epoch": 348.0,
+      "eval_loss": 10.918296813964844,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.574,
+      "eval_steps_per_second": 10.436,
+      "step": 21228
+    },
+    {
+      "epoch": 349.0,
+      "eval_loss": 11.08788776397705,
+      "eval_runtime": 1.5381,
+      "eval_samples_per_second": 79.317,
+      "eval_steps_per_second": 10.402,
+      "step": 21289
+    },
+    {
+      "epoch": 350.0,
+      "eval_loss": 10.965096473693848,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.67,
+      "eval_steps_per_second": 10.449,
+      "step": 21350
+    },
+    {
+      "epoch": 351.0,
+      "eval_loss": 11.072400093078613,
+      "eval_runtime": 1.5343,
+      "eval_samples_per_second": 79.515,
+      "eval_steps_per_second": 10.428,
+      "step": 21411
+    },
+    {
+      "epoch": 352.0,
+      "eval_loss": 11.0264253616333,
+      "eval_runtime": 1.5325,
+      "eval_samples_per_second": 79.608,
+      "eval_steps_per_second": 10.44,
+      "step": 21472
+    },
+    {
+      "epoch": 352.46,
+      "learning_rate": 1.4754098360655737e-08,
+      "loss": 10.1456,
+      "step": 21500
+    },
+    {
+      "epoch": 353.0,
+      "eval_loss": 11.1398344039917,
+      "eval_runtime": 1.5366,
+      "eval_samples_per_second": 79.396,
+      "eval_steps_per_second": 10.413,
+      "step": 21533
+    },
+    {
+      "epoch": 354.0,
+      "eval_loss": 11.249741554260254,
+      "eval_runtime": 1.5389,
+      "eval_samples_per_second": 79.278,
+      "eval_steps_per_second": 10.397,
+      "step": 21594
+    },
+    {
+      "epoch": 355.0,
+      "eval_loss": 10.889848709106445,
+      "eval_runtime": 1.5359,
+      "eval_samples_per_second": 79.432,
+      "eval_steps_per_second": 10.417,
+      "step": 21655
+    },
+    {
+      "epoch": 356.0,
+      "eval_loss": 10.963099479675293,
+      "eval_runtime": 1.5354,
+      "eval_samples_per_second": 79.457,
+      "eval_steps_per_second": 10.421,
+      "step": 21716
+    },
+    {
+      "epoch": 357.0,
+      "eval_loss": 11.073355674743652,
+      "eval_runtime": 1.5372,
+      "eval_samples_per_second": 79.365,
+      "eval_steps_per_second": 10.408,
+      "step": 21777
+    },
+    {
+      "epoch": 358.0,
+      "eval_loss": 11.122593879699707,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.616,
+      "eval_steps_per_second": 10.441,
+      "step": 21838
+    },
+    {
+      "epoch": 359.0,
+      "eval_loss": 11.168622016906738,
+      "eval_runtime": 1.5326,
+      "eval_samples_per_second": 79.602,
+      "eval_steps_per_second": 10.44,
+      "step": 21899
+    },
+    {
+      "epoch": 360.0,
+      "eval_loss": 11.031410217285156,
+      "eval_runtime": 1.5309,
+      "eval_samples_per_second": 79.694,
+      "eval_steps_per_second": 10.452,
+      "step": 21960
+    },
+    {
+      "epoch": 360.65,
+      "learning_rate": 1.3934426229508195e-08,
+      "loss": 10.1345,
+      "step": 22000
+    },
+    {
+      "epoch": 361.0,
+      "eval_loss": 11.09403133392334,
+      "eval_runtime": 1.5394,
+      "eval_samples_per_second": 79.251,
+      "eval_steps_per_second": 10.394,
+      "step": 22021
+    },
+    {
+      "epoch": 362.0,
+      "eval_loss": 10.922209739685059,
+      "eval_runtime": 1.5403,
+      "eval_samples_per_second": 79.205,
+      "eval_steps_per_second": 10.388,
+      "step": 22082
+    },
+    {
+      "epoch": 363.0,
+      "eval_loss": 11.103551864624023,
+      "eval_runtime": 1.5393,
+      "eval_samples_per_second": 79.254,
+      "eval_steps_per_second": 10.394,
+      "step": 22143
+    },
+    {
+      "epoch": 364.0,
+      "eval_loss": 11.19053840637207,
+      "eval_runtime": 1.5388,
+      "eval_samples_per_second": 79.281,
+      "eval_steps_per_second": 10.398,
+      "step": 22204
+    },
+    {
+      "epoch": 365.0,
+      "eval_loss": 10.974089622497559,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.689,
+      "eval_steps_per_second": 10.451,
+      "step": 22265
+    },
+    {
+      "epoch": 366.0,
+      "eval_loss": 10.909160614013672,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.674,
+      "eval_steps_per_second": 10.449,
+      "step": 22326
+    },
+    {
+      "epoch": 367.0,
+      "eval_loss": 11.05643367767334,
+      "eval_runtime": 1.5337,
+      "eval_samples_per_second": 79.544,
+      "eval_steps_per_second": 10.432,
+      "step": 22387
+    },
+    {
+      "epoch": 368.0,
+      "eval_loss": 11.053413391113281,
+      "eval_runtime": 1.5382,
+      "eval_samples_per_second": 79.313,
+      "eval_steps_per_second": 10.402,
+      "step": 22448
+    },
+    {
+      "epoch": 368.85,
+      "learning_rate": 1.3114754098360655e-08,
+      "loss": 10.1354,
+      "step": 22500
+    },
+    {
+      "epoch": 369.0,
+      "eval_loss": 11.030744552612305,
+      "eval_runtime": 1.551,
+      "eval_samples_per_second": 78.66,
+      "eval_steps_per_second": 10.316,
+      "step": 22509
+    },
+    {
+      "epoch": 370.0,
+      "eval_loss": 11.14685344696045,
+      "eval_runtime": 1.5382,
+      "eval_samples_per_second": 79.313,
+      "eval_steps_per_second": 10.402,
+      "step": 22570
+    },
+    {
+      "epoch": 371.0,
+      "eval_loss": 11.056023597717285,
+      "eval_runtime": 1.5383,
+      "eval_samples_per_second": 79.307,
+      "eval_steps_per_second": 10.401,
+      "step": 22631
+    },
+    {
+      "epoch": 372.0,
+      "eval_loss": 11.023992538452148,
+      "eval_runtime": 1.5355,
+      "eval_samples_per_second": 79.454,
+      "eval_steps_per_second": 10.42,
+      "step": 22692
+    },
+    {
+      "epoch": 373.0,
+      "eval_loss": 10.986889839172363,
+      "eval_runtime": 1.5328,
+      "eval_samples_per_second": 79.593,
+      "eval_steps_per_second": 10.438,
+      "step": 22753
+    },
+    {
+      "epoch": 374.0,
+      "eval_loss": 11.000397682189941,
+      "eval_runtime": 1.5337,
+      "eval_samples_per_second": 79.547,
+      "eval_steps_per_second": 10.432,
+      "step": 22814
+    },
+    {
+      "epoch": 375.0,
+      "eval_loss": 11.137325286865234,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.665,
+      "eval_steps_per_second": 10.448,
+      "step": 22875
+    },
+    {
+      "epoch": 376.0,
+      "eval_loss": 11.095502853393555,
+      "eval_runtime": 1.5326,
+      "eval_samples_per_second": 79.601,
+      "eval_steps_per_second": 10.439,
+      "step": 22936
+    },
+    {
+      "epoch": 377.0,
+      "eval_loss": 11.05420207977295,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.667,
+      "eval_steps_per_second": 10.448,
+      "step": 22997
+    },
+    {
+      "epoch": 377.05,
+      "learning_rate": 1.2295081967213115e-08,
+      "loss": 10.1382,
+      "step": 23000
+    },
+    {
+      "epoch": 378.0,
+      "eval_loss": 10.981268882751465,
+      "eval_runtime": 1.542,
+      "eval_samples_per_second": 79.12,
+      "eval_steps_per_second": 10.376,
+      "step": 23058
+    },
+    {
+      "epoch": 379.0,
+      "eval_loss": 10.987357139587402,
+      "eval_runtime": 1.547,
+      "eval_samples_per_second": 78.861,
+      "eval_steps_per_second": 10.342,
+      "step": 23119
+    },
+    {
+      "epoch": 380.0,
+      "eval_loss": 10.973642349243164,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.386,
+      "eval_steps_per_second": 10.411,
+      "step": 23180
+    },
+    {
+      "epoch": 381.0,
+      "eval_loss": 11.129460334777832,
+      "eval_runtime": 1.5333,
+      "eval_samples_per_second": 79.565,
+      "eval_steps_per_second": 10.435,
+      "step": 23241
+    },
+    {
+      "epoch": 382.0,
+      "eval_loss": 10.87239933013916,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.672,
+      "eval_steps_per_second": 10.449,
+      "step": 23302
+    },
+    {
+      "epoch": 383.0,
+      "eval_loss": 10.936705589294434,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.573,
+      "eval_steps_per_second": 10.436,
+      "step": 23363
+    },
+    {
+      "epoch": 384.0,
+      "eval_loss": 11.051558494567871,
+      "eval_runtime": 1.5374,
+      "eval_samples_per_second": 79.353,
+      "eval_steps_per_second": 10.407,
+      "step": 23424
+    },
+    {
+      "epoch": 385.0,
+      "eval_loss": 11.027456283569336,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.575,
+      "eval_steps_per_second": 10.436,
+      "step": 23485
+    },
+    {
+      "epoch": 385.25,
+      "learning_rate": 1.1475409836065573e-08,
+      "loss": 10.1246,
+      "step": 23500
+    },
+    {
+      "epoch": 386.0,
+      "eval_loss": 11.018380165100098,
+      "eval_runtime": 1.5356,
+      "eval_samples_per_second": 79.45,
+      "eval_steps_per_second": 10.42,
+      "step": 23546
+    },
+    {
+      "epoch": 387.0,
+      "eval_loss": 11.057013511657715,
+      "eval_runtime": 1.5451,
+      "eval_samples_per_second": 78.957,
+      "eval_steps_per_second": 10.355,
+      "step": 23607
+    },
+    {
+      "epoch": 388.0,
+      "eval_loss": 11.024608612060547,
+      "eval_runtime": 1.5372,
+      "eval_samples_per_second": 79.364,
+      "eval_steps_per_second": 10.408,
+      "step": 23668
+    },
+    {
+      "epoch": 389.0,
+      "eval_loss": 11.013128280639648,
+      "eval_runtime": 1.5353,
+      "eval_samples_per_second": 79.466,
+      "eval_steps_per_second": 10.422,
+      "step": 23729
+    },
+    {
+      "epoch": 390.0,
+      "eval_loss": 11.016827583312988,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.53,
+      "eval_steps_per_second": 10.43,
+      "step": 23790
+    },
+    {
+      "epoch": 391.0,
+      "eval_loss": 11.08166217803955,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.551,
+      "eval_steps_per_second": 10.433,
+      "step": 23851
+    },
+    {
+      "epoch": 392.0,
+      "eval_loss": 10.894896507263184,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.681,
+      "eval_steps_per_second": 10.45,
+      "step": 23912
+    },
+    {
+      "epoch": 393.0,
+      "eval_loss": 10.769810676574707,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.625,
+      "eval_steps_per_second": 10.443,
+      "step": 23973
+    },
+    {
+      "epoch": 393.44,
+      "learning_rate": 1.0655737704918032e-08,
+      "loss": 10.1173,
+      "step": 24000
+    },
+    {
+      "epoch": 394.0,
+      "eval_loss": 11.004133224487305,
+      "eval_runtime": 1.545,
+      "eval_samples_per_second": 78.965,
+      "eval_steps_per_second": 10.356,
+      "step": 24034
+    },
+    {
+      "epoch": 395.0,
+      "eval_loss": 10.925704956054688,
+      "eval_runtime": 1.542,
+      "eval_samples_per_second": 79.12,
+      "eval_steps_per_second": 10.376,
+      "step": 24095
+    },
+    {
+      "epoch": 396.0,
+      "eval_loss": 10.9295015335083,
+      "eval_runtime": 1.5407,
+      "eval_samples_per_second": 79.187,
+      "eval_steps_per_second": 10.385,
+      "step": 24156
+    },
+    {
+      "epoch": 397.0,
+      "eval_loss": 10.947636604309082,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.552,
+      "eval_steps_per_second": 10.433,
+      "step": 24217
+    },
+    {
+      "epoch": 398.0,
+      "eval_loss": 11.058280944824219,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.682,
+      "eval_steps_per_second": 10.45,
+      "step": 24278
+    },
+    {
+      "epoch": 399.0,
+      "eval_loss": 11.00063419342041,
+      "eval_runtime": 1.5321,
+      "eval_samples_per_second": 79.627,
+      "eval_steps_per_second": 10.443,
+      "step": 24339
+    },
+    {
+      "epoch": 400.0,
+      "eval_loss": 10.971447944641113,
+      "eval_runtime": 1.5316,
+      "eval_samples_per_second": 79.656,
+      "eval_steps_per_second": 10.447,
+      "step": 24400
+    },
+    {
+      "epoch": 401.0,
+      "eval_loss": 11.0480318069458,
+      "eval_runtime": 1.5351,
+      "eval_samples_per_second": 79.474,
+      "eval_steps_per_second": 10.423,
+      "step": 24461
+    },
+    {
+      "epoch": 401.64,
+      "learning_rate": 9.83606557377049e-09,
+      "loss": 10.1253,
+      "step": 24500
+    },
+    {
+      "epoch": 402.0,
+      "eval_loss": 11.021344184875488,
+      "eval_runtime": 1.5405,
+      "eval_samples_per_second": 79.197,
+      "eval_steps_per_second": 10.386,
+      "step": 24522
+    },
+    {
+      "epoch": 403.0,
+      "eval_loss": 10.96359920501709,
+      "eval_runtime": 1.5391,
+      "eval_samples_per_second": 79.267,
+      "eval_steps_per_second": 10.396,
+      "step": 24583
+    },
+    {
+      "epoch": 404.0,
+      "eval_loss": 10.988608360290527,
+      "eval_runtime": 1.5469,
+      "eval_samples_per_second": 78.87,
+      "eval_steps_per_second": 10.344,
+      "step": 24644
+    },
+    {
+      "epoch": 405.0,
+      "eval_loss": 11.066388130187988,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.672,
+      "eval_steps_per_second": 10.449,
+      "step": 24705
+    },
+    {
+      "epoch": 406.0,
+      "eval_loss": 11.046152114868164,
+      "eval_runtime": 1.5351,
+      "eval_samples_per_second": 79.475,
+      "eval_steps_per_second": 10.423,
+      "step": 24766
+    },
+    {
+      "epoch": 407.0,
+      "eval_loss": 11.012222290039062,
+      "eval_runtime": 1.5324,
+      "eval_samples_per_second": 79.611,
+      "eval_steps_per_second": 10.441,
+      "step": 24827
+    },
+    {
+      "epoch": 408.0,
+      "eval_loss": 10.857176780700684,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.673,
+      "eval_steps_per_second": 10.449,
+      "step": 24888
+    },
+    {
+      "epoch": 409.0,
+      "eval_loss": 11.138158798217773,
+      "eval_runtime": 1.5365,
+      "eval_samples_per_second": 79.399,
+      "eval_steps_per_second": 10.413,
+      "step": 24949
+    },
+    {
+      "epoch": 409.83,
+      "learning_rate": 9.01639344262295e-09,
+      "loss": 10.1386,
+      "step": 25000
+    },
+    {
+      "epoch": 410.0,
+      "eval_loss": 11.070024490356445,
+      "eval_runtime": 1.5371,
+      "eval_samples_per_second": 79.369,
+      "eval_steps_per_second": 10.409,
+      "step": 25010
+    },
+    {
+      "epoch": 411.0,
+      "eval_loss": 10.96763801574707,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.384,
+      "eval_steps_per_second": 10.411,
+      "step": 25071
+    },
+    {
+      "epoch": 412.0,
+      "eval_loss": 11.1865234375,
+      "eval_runtime": 1.5408,
+      "eval_samples_per_second": 79.181,
+      "eval_steps_per_second": 10.384,
+      "step": 25132
+    },
+    {
+      "epoch": 413.0,
+      "eval_loss": 11.078502655029297,
+      "eval_runtime": 1.5486,
+      "eval_samples_per_second": 78.78,
+      "eval_steps_per_second": 10.332,
+      "step": 25193
+    },
+    {
+      "epoch": 414.0,
+      "eval_loss": 11.028984069824219,
+      "eval_runtime": 1.5323,
+      "eval_samples_per_second": 79.619,
+      "eval_steps_per_second": 10.442,
+      "step": 25254
+    },
+    {
+      "epoch": 415.0,
+      "eval_loss": 11.138273239135742,
+      "eval_runtime": 1.5322,
+      "eval_samples_per_second": 79.622,
+      "eval_steps_per_second": 10.442,
+      "step": 25315
+    },
+    {
+      "epoch": 416.0,
+      "eval_loss": 11.113880157470703,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.678,
+      "eval_steps_per_second": 10.45,
+      "step": 25376
+    },
+    {
+      "epoch": 417.0,
+      "eval_loss": 11.018509864807129,
+      "eval_runtime": 1.5361,
+      "eval_samples_per_second": 79.42,
+      "eval_steps_per_second": 10.416,
+      "step": 25437
+    },
+    {
+      "epoch": 418.0,
+      "eval_loss": 11.018741607666016,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.664,
+      "eval_steps_per_second": 10.448,
+      "step": 25498
+    },
+    {
+      "epoch": 418.03,
+      "learning_rate": 8.196721311475408e-09,
+      "loss": 10.1491,
+      "step": 25500
+    },
+    {
+      "epoch": 419.0,
+      "eval_loss": 11.089320182800293,
+      "eval_runtime": 1.5384,
+      "eval_samples_per_second": 79.303,
+      "eval_steps_per_second": 10.4,
+      "step": 25559
+    },
+    {
+      "epoch": 420.0,
+      "eval_loss": 11.034805297851562,
+      "eval_runtime": 1.5427,
+      "eval_samples_per_second": 79.083,
+      "eval_steps_per_second": 10.372,
+      "step": 25620
+    },
+    {
+      "epoch": 421.0,
+      "eval_loss": 10.993184089660645,
+      "eval_runtime": 1.538,
+      "eval_samples_per_second": 79.323,
+      "eval_steps_per_second": 10.403,
+      "step": 25681
+    },
+    {
+      "epoch": 422.0,
+      "eval_loss": 11.076547622680664,
+      "eval_runtime": 1.5334,
+      "eval_samples_per_second": 79.562,
+      "eval_steps_per_second": 10.434,
+      "step": 25742
+    },
+    {
+      "epoch": 423.0,
+      "eval_loss": 11.048792839050293,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.638,
+      "eval_steps_per_second": 10.444,
+      "step": 25803
+    },
+    {
+      "epoch": 424.0,
+      "eval_loss": 11.024137496948242,
+      "eval_runtime": 1.5337,
+      "eval_samples_per_second": 79.548,
+      "eval_steps_per_second": 10.433,
+      "step": 25864
+    },
+    {
+      "epoch": 425.0,
+      "eval_loss": 11.069534301757812,
+      "eval_runtime": 1.532,
+      "eval_samples_per_second": 79.632,
+      "eval_steps_per_second": 10.444,
+      "step": 25925
+    },
+    {
+      "epoch": 426.0,
+      "eval_loss": 10.898648262023926,
+      "eval_runtime": 1.5315,
+      "eval_samples_per_second": 79.662,
+      "eval_steps_per_second": 10.448,
+      "step": 25986
+    },
+    {
+      "epoch": 426.23,
+      "learning_rate": 7.377049180327869e-09,
+      "loss": 10.1184,
+      "step": 26000
+    },
+    {
+      "epoch": 427.0,
+      "eval_loss": 10.843330383300781,
+      "eval_runtime": 1.5388,
+      "eval_samples_per_second": 79.284,
+      "eval_steps_per_second": 10.398,
+      "step": 26047
+    },
+    {
+      "epoch": 428.0,
+      "eval_loss": 10.847617149353027,
+      "eval_runtime": 1.5395,
+      "eval_samples_per_second": 79.244,
+      "eval_steps_per_second": 10.393,
+      "step": 26108
+    },
+    {
+      "epoch": 429.0,
+      "eval_loss": 10.974745750427246,
+      "eval_runtime": 1.5392,
+      "eval_samples_per_second": 79.259,
+      "eval_steps_per_second": 10.395,
+      "step": 26169
+    },
+    {
+      "epoch": 430.0,
+      "eval_loss": 10.925885200500488,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.672,
+      "eval_steps_per_second": 10.449,
+      "step": 26230
+    },
+    {
+      "epoch": 431.0,
+      "eval_loss": 10.864691734313965,
+      "eval_runtime": 1.5326,
+      "eval_samples_per_second": 79.602,
+      "eval_steps_per_second": 10.44,
+      "step": 26291
+    },
+    {
+      "epoch": 432.0,
+      "eval_loss": 11.028008460998535,
+      "eval_runtime": 1.5318,
+      "eval_samples_per_second": 79.645,
+      "eval_steps_per_second": 10.445,
+      "step": 26352
+    },
+    {
+      "epoch": 433.0,
+      "eval_loss": 10.958173751831055,
+      "eval_runtime": 1.5349,
+      "eval_samples_per_second": 79.482,
+      "eval_steps_per_second": 10.424,
+      "step": 26413
+    },
+    {
+      "epoch": 434.0,
+      "eval_loss": 10.98095989227295,
+      "eval_runtime": 1.5339,
+      "eval_samples_per_second": 79.536,
+      "eval_steps_per_second": 10.431,
+      "step": 26474
+    },
+    {
+      "epoch": 434.43,
+      "learning_rate": 6.5573770491803275e-09,
+      "loss": 10.1396,
+      "step": 26500
+    },
+    {
+      "epoch": 435.0,
+      "eval_loss": 11.049097061157227,
+      "eval_runtime": 1.5478,
+      "eval_samples_per_second": 78.819,
+      "eval_steps_per_second": 10.337,
+      "step": 26535
+    },
+    {
+      "epoch": 436.0,
+      "eval_loss": 11.069966316223145,
+      "eval_runtime": 1.5391,
+      "eval_samples_per_second": 79.265,
+      "eval_steps_per_second": 10.395,
+      "step": 26596
+    },
+    {
+      "epoch": 437.0,
+      "eval_loss": 10.987845420837402,
+      "eval_runtime": 1.5481,
+      "eval_samples_per_second": 78.804,
+      "eval_steps_per_second": 10.335,
+      "step": 26657
+    },
+    {
+      "epoch": 438.0,
+      "eval_loss": 10.939976692199707,
+      "eval_runtime": 1.5375,
+      "eval_samples_per_second": 79.351,
+      "eval_steps_per_second": 10.407,
+      "step": 26718
+    },
+    {
+      "epoch": 439.0,
+      "eval_loss": 10.868154525756836,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.664,
+      "eval_steps_per_second": 10.448,
+      "step": 26779
+    },
+    {
+      "epoch": 440.0,
+      "eval_loss": 10.966716766357422,
+      "eval_runtime": 1.539,
+      "eval_samples_per_second": 79.27,
+      "eval_steps_per_second": 10.396,
+      "step": 26840
+    },
+    {
+      "epoch": 441.0,
+      "eval_loss": 11.011680603027344,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.674,
+      "eval_steps_per_second": 10.449,
+      "step": 26901
+    },
+    {
+      "epoch": 442.0,
+      "eval_loss": 11.037373542785645,
+      "eval_runtime": 1.5311,
+      "eval_samples_per_second": 79.68,
+      "eval_steps_per_second": 10.45,
+      "step": 26962
+    },
+    {
+      "epoch": 442.62,
+      "learning_rate": 5.737704918032786e-09,
+      "loss": 10.1337,
+      "step": 27000
+    },
+    {
+      "epoch": 443.0,
+      "eval_loss": 11.133712768554688,
+      "eval_runtime": 1.5458,
+      "eval_samples_per_second": 78.922,
+      "eval_steps_per_second": 10.35,
+      "step": 27023
+    },
+    {
+      "epoch": 444.0,
+      "eval_loss": 10.941494941711426,
+      "eval_runtime": 1.5386,
+      "eval_samples_per_second": 79.294,
+      "eval_steps_per_second": 10.399,
+      "step": 27084
+    },
+    {
+      "epoch": 445.0,
+      "eval_loss": 11.017427444458008,
+      "eval_runtime": 1.5391,
+      "eval_samples_per_second": 79.266,
+      "eval_steps_per_second": 10.396,
+      "step": 27145
+    },
+    {
+      "epoch": 446.0,
+      "eval_loss": 11.02389144897461,
+      "eval_runtime": 1.5319,
+      "eval_samples_per_second": 79.641,
+      "eval_steps_per_second": 10.445,
+      "step": 27206
+    },
+    {
+      "epoch": 447.0,
+      "eval_loss": 10.897884368896484,
+      "eval_runtime": 1.534,
+      "eval_samples_per_second": 79.53,
+      "eval_steps_per_second": 10.43,
+      "step": 27267
+    },
+    {
+      "epoch": 448.0,
+      "eval_loss": 10.921698570251465,
+      "eval_runtime": 1.5334,
+      "eval_samples_per_second": 79.564,
+      "eval_steps_per_second": 10.435,
+      "step": 27328
+    },
+    {
+      "epoch": 449.0,
+      "eval_loss": 10.892621994018555,
+      "eval_runtime": 1.5335,
+      "eval_samples_per_second": 79.557,
+      "eval_steps_per_second": 10.434,
+      "step": 27389
+    },
+    {
+      "epoch": 450.0,
+      "eval_loss": 11.121941566467285,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.668,
+      "eval_steps_per_second": 10.448,
+      "step": 27450
+    },
+    {
+      "epoch": 450.82,
+      "learning_rate": 4.918032786885245e-09,
+      "loss": 10.1168,
+      "step": 27500
+    },
+    {
+      "epoch": 451.0,
+      "eval_loss": 10.893115043640137,
+      "eval_runtime": 1.5424,
+      "eval_samples_per_second": 79.097,
+      "eval_steps_per_second": 10.373,
+      "step": 27511
+    },
+    {
+      "epoch": 452.0,
+      "eval_loss": 11.011245727539062,
+      "eval_runtime": 1.5529,
+      "eval_samples_per_second": 78.564,
+      "eval_steps_per_second": 10.303,
+      "step": 27572
+    },
+    {
+      "epoch": 453.0,
+      "eval_loss": 10.982295036315918,
+      "eval_runtime": 1.5411,
+      "eval_samples_per_second": 79.163,
+      "eval_steps_per_second": 10.382,
+      "step": 27633
+    },
+    {
+      "epoch": 454.0,
+      "eval_loss": 11.109077453613281,
+      "eval_runtime": 1.5438,
+      "eval_samples_per_second": 79.028,
+      "eval_steps_per_second": 10.364,
+      "step": 27694
+    },
+    {
+      "epoch": 455.0,
+      "eval_loss": 10.869407653808594,
+      "eval_runtime": 1.5332,
+      "eval_samples_per_second": 79.575,
+      "eval_steps_per_second": 10.436,
+      "step": 27755
+    },
+    {
+      "epoch": 456.0,
+      "eval_loss": 10.962514877319336,
+      "eval_runtime": 1.5316,
+      "eval_samples_per_second": 79.656,
+      "eval_steps_per_second": 10.447,
+      "step": 27816
+    },
+    {
+      "epoch": 457.0,
+      "eval_loss": 10.855345726013184,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.386,
+      "eval_steps_per_second": 10.411,
+      "step": 27877
+    },
+    {
+      "epoch": 458.0,
+      "eval_loss": 10.98884391784668,
+      "eval_runtime": 1.533,
+      "eval_samples_per_second": 79.582,
+      "eval_steps_per_second": 10.437,
+      "step": 27938
+    },
+    {
+      "epoch": 459.0,
+      "eval_loss": 10.929617881774902,
+      "eval_runtime": 1.5316,
+      "eval_samples_per_second": 79.658,
+      "eval_steps_per_second": 10.447,
+      "step": 27999
+    },
+    {
+      "epoch": 459.02,
+      "learning_rate": 4.098360655737704e-09,
+      "loss": 10.1229,
+      "step": 28000
+    },
+    {
+      "epoch": 460.0,
+      "eval_loss": 10.88948917388916,
+      "eval_runtime": 1.5368,
+      "eval_samples_per_second": 79.387,
+      "eval_steps_per_second": 10.411,
+      "step": 28060
+    },
+    {
+      "epoch": 461.0,
+      "eval_loss": 10.980300903320312,
+      "eval_runtime": 1.5401,
+      "eval_samples_per_second": 79.214,
+      "eval_steps_per_second": 10.389,
+      "step": 28121
+    },
+    {
+      "epoch": 462.0,
+      "eval_loss": 11.014444351196289,
+      "eval_runtime": 1.5678,
+      "eval_samples_per_second": 77.815,
+      "eval_steps_per_second": 10.205,
+      "step": 28182
+    },
+    {
+      "epoch": 463.0,
+      "eval_loss": 11.025659561157227,
+      "eval_runtime": 1.5329,
+      "eval_samples_per_second": 79.586,
+      "eval_steps_per_second": 10.438,
+      "step": 28243
+    },
+    {
+      "epoch": 464.0,
+      "eval_loss": 10.914053916931152,
+      "eval_runtime": 1.5336,
+      "eval_samples_per_second": 79.55,
+      "eval_steps_per_second": 10.433,
+      "step": 28304
+    },
+    {
+      "epoch": 465.0,
+      "eval_loss": 11.134551048278809,
+      "eval_runtime": 1.5313,
+      "eval_samples_per_second": 79.669,
+      "eval_steps_per_second": 10.448,
+      "step": 28365
+    },
+    {
+      "epoch": 466.0,
+      "eval_loss": 11.025362968444824,
+      "eval_runtime": 1.533,
+      "eval_samples_per_second": 79.58,
+      "eval_steps_per_second": 10.437,
+      "step": 28426
+    },
+    {
+      "epoch": 467.0,
+      "eval_loss": 11.038411140441895,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.664,
+      "eval_steps_per_second": 10.448,
+      "step": 28487
+    },
+    {
+      "epoch": 467.21,
+      "learning_rate": 3.2786885245901638e-09,
+      "loss": 10.1179,
+      "step": 28500
+    },
+    {
+      "epoch": 468.0,
+      "eval_loss": 10.849050521850586,
+      "eval_runtime": 1.5609,
+      "eval_samples_per_second": 78.159,
+      "eval_steps_per_second": 10.25,
+      "step": 28548
+    },
+    {
+      "epoch": 469.0,
+      "eval_loss": 11.046910285949707,
+      "eval_runtime": 1.5417,
+      "eval_samples_per_second": 79.132,
+      "eval_steps_per_second": 10.378,
+      "step": 28609
+    },
+    {
+      "epoch": 470.0,
+      "eval_loss": 10.967819213867188,
+      "eval_runtime": 1.539,
+      "eval_samples_per_second": 79.271,
+      "eval_steps_per_second": 10.396,
+      "step": 28670
+    },
+    {
+      "epoch": 471.0,
+      "eval_loss": 10.890240669250488,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.687,
+      "eval_steps_per_second": 10.451,
+      "step": 28731
+    },
+    {
+      "epoch": 472.0,
+      "eval_loss": 10.964876174926758,
+      "eval_runtime": 1.5349,
+      "eval_samples_per_second": 79.483,
+      "eval_steps_per_second": 10.424,
+      "step": 28792
+    },
+    {
+      "epoch": 473.0,
+      "eval_loss": 10.925169944763184,
+      "eval_runtime": 1.5315,
+      "eval_samples_per_second": 79.659,
+      "eval_steps_per_second": 10.447,
+      "step": 28853
+    },
+    {
+      "epoch": 474.0,
+      "eval_loss": 11.113059997558594,
+      "eval_runtime": 1.5314,
+      "eval_samples_per_second": 79.666,
+      "eval_steps_per_second": 10.448,
+      "step": 28914
+    },
+    {
+      "epoch": 475.0,
+      "eval_loss": 11.026731491088867,
+      "eval_runtime": 1.5312,
+      "eval_samples_per_second": 79.677,
+      "eval_steps_per_second": 10.449,
+      "step": 28975
+    },
+    {
+      "epoch": 475.41,
+      "learning_rate": 2.4590163934426226e-09,
+      "loss": 10.1189,
+      "step": 29000
+    },
+    {
+      "epoch": 476.0,
+      "eval_loss": 10.842777252197266,
+      "eval_runtime": 1.5457,
+      "eval_samples_per_second": 78.928,
+      "eval_steps_per_second": 10.351,
+      "step": 29036
+    },
+    {
+      "epoch": 477.0,
+      "eval_loss": 11.031394958496094,
+      "eval_runtime": 1.5433,
+      "eval_samples_per_second": 79.053,
+      "eval_steps_per_second": 10.368,
+      "step": 29097
+    },
+    {
+      "epoch": 478.0,
+      "eval_loss": 11.093620300292969,
+      "eval_runtime": 1.5431,
+      "eval_samples_per_second": 79.061,
+      "eval_steps_per_second": 10.369,
+      "step": 29158
+    },
+    {
+      "epoch": 479.0,
+      "eval_loss": 10.99679946899414,
+      "eval_runtime": 1.5405,
+      "eval_samples_per_second": 79.196,
+      "eval_steps_per_second": 10.386,
+      "step": 29219
+    },
+    {
+      "epoch": 480.0,
+      "eval_loss": 10.872136116027832,
+      "eval_runtime": 1.5317,
+      "eval_samples_per_second": 79.648,
+      "eval_steps_per_second": 10.446,
+      "step": 29280
+    },
+    {
+      "epoch": 481.0,
+      "eval_loss": 11.01526165008545,
+      "eval_runtime": 1.5338,
+      "eval_samples_per_second": 79.539,
+      "eval_steps_per_second": 10.431,
+      "step": 29341
+    },
+    {
+      "epoch": 482.0,
+      "eval_loss": 11.176105499267578,
+      "eval_runtime": 1.531,
+      "eval_samples_per_second": 79.688,
+      "eval_steps_per_second": 10.451,
+      "step": 29402
+    },
+    {
+      "epoch": 483.0,
+      "eval_loss": 10.983979225158691,
+      "eval_runtime": 1.5364,
+      "eval_samples_per_second": 79.408,
+      "eval_steps_per_second": 10.414,
+      "step": 29463
+    },
+    {
+      "epoch": 483.61,
+      "learning_rate": 1.6393442622950819e-09,
+      "loss": 10.1153,
+      "step": 29500
+    },
+    {
+      "epoch": 484.0,
+      "eval_loss": 10.964775085449219,
+      "eval_runtime": 1.5399,
+      "eval_samples_per_second": 79.225,
+      "eval_steps_per_second": 10.39,
+      "step": 29524
+    },
+    {
+      "epoch": 485.0,
+      "eval_loss": 11.114033699035645,
+      "eval_runtime": 1.5364,
+      "eval_samples_per_second": 79.409,
+      "eval_steps_per_second": 10.414,
+      "step": 29585
+    },
+    {
+      "epoch": 486.0,
+      "eval_loss": 11.021162986755371,
+      "eval_runtime": 1.5483,
+      "eval_samples_per_second": 78.798,
+      "eval_steps_per_second": 10.334,
+      "step": 29646
+    },
+    {
+      "epoch": 487.0,
+      "eval_loss": 10.919731140136719,
+      "eval_runtime": 1.5334,
+      "eval_samples_per_second": 79.563,
+      "eval_steps_per_second": 10.434,
+      "step": 29707
+    },
+    {
+      "epoch": 488.0,
+      "eval_loss": 10.979778289794922,
+      "eval_runtime": 1.5341,
+      "eval_samples_per_second": 79.523,
+      "eval_steps_per_second": 10.429,
+      "step": 29768
+    },
+    {
+      "epoch": 489.0,
+      "eval_loss": 10.904738426208496,
+      "eval_runtime": 1.5327,
+      "eval_samples_per_second": 79.596,
+      "eval_steps_per_second": 10.439,
+      "step": 29829
+    },
+    {
+      "epoch": 490.0,
+      "eval_loss": 11.014572143554688,
+      "eval_runtime": 1.5384,
+      "eval_samples_per_second": 79.303,
+      "eval_steps_per_second": 10.4,
+      "step": 29890
+    },
+    {
+      "epoch": 491.0,
+      "eval_loss": 11.057549476623535,
+      "eval_runtime": 1.5352,
+      "eval_samples_per_second": 79.469,
+      "eval_steps_per_second": 10.422,
+      "step": 29951
+    },
+    {
+      "epoch": 491.8,
+      "learning_rate": 8.196721311475409e-10,
+      "loss": 10.1141,
+      "step": 30000
+    },
+    {
+      "epoch": 492.0,
+      "eval_loss": 11.040027618408203,
+      "eval_runtime": 1.5409,
+      "eval_samples_per_second": 79.175,
+      "eval_steps_per_second": 10.384,
+      "step": 30012
+    },
+    {
+      "epoch": 493.0,
+      "eval_loss": 11.089771270751953,
+      "eval_runtime": 1.5412,
+      "eval_samples_per_second": 79.159,
+      "eval_steps_per_second": 10.382,
+      "step": 30073
+    },
+    {
+      "epoch": 494.0,
+      "eval_loss": 10.99104118347168,
+      "eval_runtime": 1.5377,
+      "eval_samples_per_second": 79.339,
+      "eval_steps_per_second": 10.405,
+      "step": 30134
+    },
+    {
+      "epoch": 495.0,
+      "eval_loss": 11.057866096496582,
+      "eval_runtime": 1.5447,
+      "eval_samples_per_second": 78.98,
+      "eval_steps_per_second": 10.358,
+      "step": 30195
+    },
+    {
+      "epoch": 496.0,
+      "eval_loss": 10.857988357543945,
+      "eval_runtime": 1.533,
+      "eval_samples_per_second": 79.581,
+      "eval_steps_per_second": 10.437,
+      "step": 30256
+    },
+    {
+      "epoch": 497.0,
+      "eval_loss": 10.944987297058105,
+      "eval_runtime": 1.5401,
+      "eval_samples_per_second": 79.216,
+      "eval_steps_per_second": 10.389,
+      "step": 30317
+    },
+    {
+      "epoch": 498.0,
+      "eval_loss": 11.052291870117188,
+      "eval_runtime": 1.5333,
+      "eval_samples_per_second": 79.568,
+      "eval_steps_per_second": 10.435,
+      "step": 30378
+    },
+    {
+      "epoch": 499.0,
+      "eval_loss": 11.122845649719238,
+      "eval_runtime": 1.5327,
+      "eval_samples_per_second": 79.598,
+      "eval_steps_per_second": 10.439,
+      "step": 30439
+    },
+    {
+      "epoch": 500.0,
+      "learning_rate": 0.0,
+      "loss": 10.1176,
+      "step": 30500
+    },
+    {
+      "epoch": 500.0,
+      "eval_loss": 11.049210548400879,
+      "eval_runtime": 1.5625,
+      "eval_samples_per_second": 78.08,
+      "eval_steps_per_second": 10.24,
+      "step": 30500
+    },
+    {
+      "epoch": 500.0,
+      "step": 30500,
+      "total_flos": 1.6136552440728576e+16,
+      "train_loss": 11.67972986039959,
+      "train_runtime": 24510.5451,
+      "train_samples_per_second": 9.975,
+      "train_steps_per_second": 1.244
     }
   ],
-  "max_steps": 610,
-  "num_train_epochs": 10,
-  "total_flos": 322666370343936.0,
+  "max_steps": 30500,
+  "num_train_epochs": 500,
+  "total_flos": 1.6136552440728576e+16,
   "trial_name": null,
   "trial_params": null
 }