diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,4673 +1,13946 @@
 {
-  "best_metric": 0.011437707580626011,
-  "best_model_checkpoint": "/home/paperspace/Data/models/asianpaints/llm3br256/checkpoint-400",
-  "epoch": 4.9655172413793105,
-  "eval_steps": 5,
-  "global_step": 540,
+  "best_metric": 0.015713628381490707,
+  "best_model_checkpoint": "/home/paperspace/Data/models/asianpaints/llm3br256-v1.5/checkpoint-1700",
+  "epoch": 9.178743961352657,
+  "eval_steps": 25,
+  "global_step": 1900,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.009195402298850575,
-      "grad_norm": 0.13789793848991394,
-      "learning_rate": 1.8518518518518519e-06,
-      "loss": 0.0591,
+      "epoch": 0.004830917874396135,
+      "grad_norm": 0.5997226238250732,
+      "learning_rate": 1.9305019305019306e-07,
+      "loss": 0.1739,
       "step": 1
     },
     {
-      "epoch": 0.01839080459770115,
-      "grad_norm": 0.13098689913749695,
-      "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.0572,
+      "epoch": 0.00966183574879227,
+      "grad_norm": 0.586310863494873,
+      "learning_rate": 3.8610038610038613e-07,
+      "loss": 0.1804,
       "step": 2
     },
     {
-      "epoch": 0.027586206896551724,
-      "grad_norm": 0.12751829624176025,
-      "learning_rate": 5.555555555555556e-06,
-      "loss": 0.0586,
+      "epoch": 0.014492753623188406,
+      "grad_norm": 0.6426325440406799,
+      "learning_rate": 5.791505791505791e-07,
+      "loss": 0.1612,
       "step": 3
     },
     {
-      "epoch": 0.0367816091954023,
-      "grad_norm": 0.12321432679891586,
-      "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.0521,
+      "epoch": 0.01932367149758454,
+      "grad_norm": 0.5951049327850342,
+      "learning_rate": 7.722007722007723e-07,
+      "loss": 0.1851,
       "step": 4
     },
     {
-      "epoch": 0.04597701149425287,
-      "grad_norm": 0.12428242713212967,
-      "learning_rate": 9.259259259259259e-06,
-      "loss": 0.0567,
-      "step": 5
-    },
-    {
-      "epoch": 0.04597701149425287,
-      "eval_loss": 0.0583593025803566,
-      "eval_runtime": 10.2736,
-      "eval_samples_per_second": 4.867,
-      "eval_steps_per_second": 1.265,
+      "epoch": 0.024154589371980676,
+      "grad_norm": 0.574536144733429,
+      "learning_rate": 9.652509652509653e-07,
+      "loss": 0.1837,
       "step": 5
     },
     {
-      "epoch": 0.05517241379310345,
-      "grad_norm": 0.10534916073083878,
-      "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.0485,
+      "epoch": 0.028985507246376812,
+      "grad_norm": 0.5631237030029297,
+      "learning_rate": 1.1583011583011583e-06,
+      "loss": 0.1905,
       "step": 6
     },
     {
-      "epoch": 0.06436781609195402,
-      "grad_norm": 0.08470993489027023,
-      "learning_rate": 1.2962962962962962e-05,
-      "loss": 0.0439,
+      "epoch": 0.033816425120772944,
+      "grad_norm": 0.5657174587249756,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.1719,
       "step": 7
     },
     {
-      "epoch": 0.0735632183908046,
-      "grad_norm": 0.07256518304347992,
-      "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.0402,
+      "epoch": 0.03864734299516908,
+      "grad_norm": 0.5333519577980042,
+      "learning_rate": 1.5444015444015445e-06,
+      "loss": 0.1662,
       "step": 8
     },
     {
-      "epoch": 0.08275862068965517,
-      "grad_norm": 0.060682956129312515,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.0374,
+      "epoch": 0.043478260869565216,
+      "grad_norm": 0.5621788501739502,
+      "learning_rate": 1.7374517374517375e-06,
+      "loss": 0.1862,
       "step": 9
     },
     {
-      "epoch": 0.09195402298850575,
-      "grad_norm": 0.05973079800605774,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.0378,
-      "step": 10
-    },
-    {
-      "epoch": 0.09195402298850575,
-      "eval_loss": 0.0383976586163044,
-      "eval_runtime": 7.5937,
-      "eval_samples_per_second": 6.584,
-      "eval_steps_per_second": 1.712,
+      "epoch": 0.04830917874396135,
+      "grad_norm": 0.5728170275688171,
+      "learning_rate": 1.9305019305019305e-06,
+      "loss": 0.1817,
       "step": 10
     },
     {
-      "epoch": 0.10114942528735632,
-      "grad_norm": 0.06681109964847565,
-      "learning_rate": 2.037037037037037e-05,
-      "loss": 0.0335,
+      "epoch": 0.05314009661835749,
+      "grad_norm": 0.5463429689407349,
+      "learning_rate": 2.1235521235521236e-06,
+      "loss": 0.1787,
       "step": 11
     },
     {
-      "epoch": 0.1103448275862069,
-      "grad_norm": 0.07075705379247665,
-      "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.0314,
+      "epoch": 0.057971014492753624,
+      "grad_norm": 0.5315811038017273,
+      "learning_rate": 2.3166023166023166e-06,
+      "loss": 0.1844,
       "step": 12
     },
     {
-      "epoch": 0.11954022988505747,
-      "grad_norm": 0.05252094194293022,
-      "learning_rate": 2.4074074074074074e-05,
-      "loss": 0.0347,
+      "epoch": 0.06280193236714976,
+      "grad_norm": 0.4727146625518799,
+      "learning_rate": 2.5096525096525096e-06,
+      "loss": 0.1704,
       "step": 13
     },
     {
-      "epoch": 0.12873563218390804,
-      "grad_norm": 0.04202576354146004,
-      "learning_rate": 2.5925925925925925e-05,
-      "loss": 0.03,
+      "epoch": 0.06763285024154589,
+      "grad_norm": 0.4733094573020935,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.1517,
       "step": 14
     },
     {
-      "epoch": 0.13793103448275862,
-      "grad_norm": 0.03977281600236893,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 0.0301,
-      "step": 15
-    },
-    {
-      "epoch": 0.13793103448275862,
-      "eval_loss": 0.0318371020257473,
-      "eval_runtime": 7.612,
-      "eval_samples_per_second": 6.569,
-      "eval_steps_per_second": 1.708,
+      "epoch": 0.07246376811594203,
+      "grad_norm": 0.42338845133781433,
+      "learning_rate": 2.895752895752896e-06,
+      "loss": 0.175,
       "step": 15
     },
     {
-      "epoch": 0.1471264367816092,
-      "grad_norm": 0.04173818975687027,
-      "learning_rate": 2.962962962962963e-05,
-      "loss": 0.0286,
+      "epoch": 0.07729468599033816,
+      "grad_norm": 0.3983526825904846,
+      "learning_rate": 3.088803088803089e-06,
+      "loss": 0.1692,
       "step": 16
     },
     {
-      "epoch": 0.15632183908045977,
-      "grad_norm": 0.039005085825920105,
-      "learning_rate": 3.148148148148148e-05,
-      "loss": 0.0281,
+      "epoch": 0.0821256038647343,
+      "grad_norm": 0.3729923963546753,
+      "learning_rate": 3.2818532818532816e-06,
+      "loss": 0.1597,
       "step": 17
     },
     {
-      "epoch": 0.16551724137931034,
-      "grad_norm": 0.03674433380365372,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.0248,
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.3619980216026306,
+      "learning_rate": 3.474903474903475e-06,
+      "loss": 0.1753,
       "step": 18
     },
     {
-      "epoch": 0.17471264367816092,
-      "grad_norm": 0.035750504583120346,
-      "learning_rate": 3.518518518518519e-05,
-      "loss": 0.0256,
+      "epoch": 0.09178743961352658,
+      "grad_norm": 0.3523077070713043,
+      "learning_rate": 3.6679536679536685e-06,
+      "loss": 0.1581,
       "step": 19
     },
     {
-      "epoch": 0.1839080459770115,
-      "grad_norm": 0.027469012886285782,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.0248,
-      "step": 20
-    },
-    {
-      "epoch": 0.1839080459770115,
-      "eval_loss": 0.028074176982045174,
-      "eval_runtime": 7.6024,
-      "eval_samples_per_second": 6.577,
-      "eval_steps_per_second": 1.71,
+      "epoch": 0.0966183574879227,
+      "grad_norm": 0.33768779039382935,
+      "learning_rate": 3.861003861003861e-06,
+      "loss": 0.1625,
       "step": 20
     },
     {
-      "epoch": 0.19310344827586207,
-      "grad_norm": 0.026447350159287453,
-      "learning_rate": 3.888888888888889e-05,
-      "loss": 0.0231,
+      "epoch": 0.10144927536231885,
+      "grad_norm": 0.31936874985694885,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.1505,
       "step": 21
     },
     {
-      "epoch": 0.20229885057471264,
-      "grad_norm": 0.031306445598602295,
-      "learning_rate": 4.074074074074074e-05,
-      "loss": 0.0236,
+      "epoch": 0.10628019323671498,
+      "grad_norm": 0.324440598487854,
+      "learning_rate": 4.247104247104247e-06,
+      "loss": 0.1468,
       "step": 22
     },
     {
-      "epoch": 0.21149425287356322,
-      "grad_norm": 0.029817912727594376,
-      "learning_rate": 4.259259259259259e-05,
-      "loss": 0.0227,
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.3510577380657196,
+      "learning_rate": 4.4401544401544405e-06,
+      "loss": 0.1703,
       "step": 23
     },
     {
-      "epoch": 0.2206896551724138,
-      "grad_norm": 0.03197651728987694,
-      "learning_rate": 4.4444444444444447e-05,
-      "loss": 0.0277,
+      "epoch": 0.11594202898550725,
+      "grad_norm": 0.295952707529068,
+      "learning_rate": 4.633204633204633e-06,
+      "loss": 0.1359,
       "step": 24
     },
     {
-      "epoch": 0.22988505747126436,
-      "grad_norm": 0.03044748678803444,
-      "learning_rate": 4.62962962962963e-05,
-      "loss": 0.0241,
+      "epoch": 0.12077294685990338,
+      "grad_norm": 0.284309059381485,
+      "learning_rate": 4.8262548262548266e-06,
+      "loss": 0.1302,
       "step": 25
     },
     {
-      "epoch": 0.22988505747126436,
-      "eval_loss": 0.025574278086423874,
-      "eval_runtime": 7.6017,
-      "eval_samples_per_second": 6.577,
-      "eval_steps_per_second": 1.71,
+      "epoch": 0.12077294685990338,
+      "eval_loss": 0.14291973412036896,
+      "eval_runtime": 23.4942,
+      "eval_samples_per_second": 4.256,
+      "eval_steps_per_second": 0.128,
       "step": 25
     },
     {
-      "epoch": 0.23908045977011494,
-      "grad_norm": 0.02689642831683159,
-      "learning_rate": 4.814814814814815e-05,
-      "loss": 0.0253,
+      "epoch": 0.12560386473429952,
+      "grad_norm": 0.2970901131629944,
+      "learning_rate": 5.019305019305019e-06,
+      "loss": 0.131,
       "step": 26
     },
     {
-      "epoch": 0.2482758620689655,
-      "grad_norm": 0.0260293111205101,
-      "learning_rate": 5e-05,
-      "loss": 0.0231,
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.2529473304748535,
+      "learning_rate": 5.212355212355213e-06,
+      "loss": 0.1251,
       "step": 27
     },
     {
-      "epoch": 0.2574712643678161,
-      "grad_norm": 0.025041967630386353,
-      "learning_rate": 5.185185185185185e-05,
-      "loss": 0.0244,
+      "epoch": 0.13526570048309178,
+      "grad_norm": 0.2643989026546478,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.1462,
       "step": 28
     },
     {
-      "epoch": 0.26666666666666666,
-      "grad_norm": 0.025733083486557007,
-      "learning_rate": 5.370370370370371e-05,
-      "loss": 0.0219,
+      "epoch": 0.14009661835748793,
+      "grad_norm": 0.2184019535779953,
+      "learning_rate": 5.598455598455599e-06,
+      "loss": 0.1332,
       "step": 29
     },
     {
-      "epoch": 0.27586206896551724,
-      "grad_norm": 0.023718759417533875,
-      "learning_rate": 5.555555555555556e-05,
-      "loss": 0.021,
-      "step": 30
-    },
-    {
-      "epoch": 0.27586206896551724,
-      "eval_loss": 0.0234465803951025,
-      "eval_runtime": 7.6024,
-      "eval_samples_per_second": 6.577,
-      "eval_steps_per_second": 1.71,
+      "epoch": 0.14492753623188406,
+      "grad_norm": 0.19679449498653412,
+      "learning_rate": 5.791505791505792e-06,
+      "loss": 0.1343,
       "step": 30
     },
     {
-      "epoch": 0.2850574712643678,
-      "grad_norm": 0.02254614047706127,
-      "learning_rate": 5.740740740740741e-05,
-      "loss": 0.0213,
+      "epoch": 0.1497584541062802,
+      "grad_norm": 0.19991901516914368,
+      "learning_rate": 5.984555984555985e-06,
+      "loss": 0.1373,
       "step": 31
     },
     {
-      "epoch": 0.2942528735632184,
-      "grad_norm": 0.024035446345806122,
-      "learning_rate": 5.925925925925926e-05,
-      "loss": 0.0215,
+      "epoch": 0.15458937198067632,
+      "grad_norm": 0.23242047429084778,
+      "learning_rate": 6.177606177606178e-06,
+      "loss": 0.1496,
       "step": 32
     },
     {
-      "epoch": 0.30344827586206896,
-      "grad_norm": 0.02866807021200657,
-      "learning_rate": 6.111111111111112e-05,
-      "loss": 0.0211,
+      "epoch": 0.15942028985507245,
+      "grad_norm": 0.21723514795303345,
+      "learning_rate": 6.370656370656371e-06,
+      "loss": 0.1167,
       "step": 33
     },
     {
-      "epoch": 0.31264367816091954,
-      "grad_norm": 0.02149030566215515,
-      "learning_rate": 6.296296296296296e-05,
-      "loss": 0.0196,
+      "epoch": 0.1642512077294686,
+      "grad_norm": 0.1968252956867218,
+      "learning_rate": 6.563706563706563e-06,
+      "loss": 0.125,
       "step": 34
     },
     {
-      "epoch": 0.3218390804597701,
-      "grad_norm": 0.021923156455159187,
-      "learning_rate": 6.481481481481482e-05,
-      "loss": 0.0213,
-      "step": 35
-    },
-    {
-      "epoch": 0.3218390804597701,
-      "eval_loss": 0.022549211978912354,
-      "eval_runtime": 7.607,
-      "eval_samples_per_second": 6.573,
-      "eval_steps_per_second": 1.709,
+      "epoch": 0.16908212560386474,
+      "grad_norm": 0.1917056292295456,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.1311,
       "step": 35
     },
     {
-      "epoch": 0.3310344827586207,
-      "grad_norm": 0.022337742149829865,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 0.0191,
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.18152444064617157,
+      "learning_rate": 6.94980694980695e-06,
+      "loss": 0.1158,
       "step": 36
     },
     {
-      "epoch": 0.34022988505747126,
-      "grad_norm": 0.025172578170895576,
-      "learning_rate": 6.851851851851852e-05,
-      "loss": 0.0194,
+      "epoch": 0.178743961352657,
+      "grad_norm": 0.20013165473937988,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.1153,
       "step": 37
     },
     {
-      "epoch": 0.34942528735632183,
-      "grad_norm": 0.01893443800508976,
-      "learning_rate": 7.037037037037038e-05,
-      "loss": 0.0209,
+      "epoch": 0.18357487922705315,
+      "grad_norm": 0.190288707613945,
+      "learning_rate": 7.335907335907337e-06,
+      "loss": 0.1039,
       "step": 38
     },
     {
-      "epoch": 0.3586206896551724,
-      "grad_norm": 0.022978560999035835,
-      "learning_rate": 7.222222222222222e-05,
-      "loss": 0.0195,
+      "epoch": 0.18840579710144928,
+      "grad_norm": 0.19227007031440735,
+      "learning_rate": 7.52895752895753e-06,
+      "loss": 0.1261,
       "step": 39
     },
     {
-      "epoch": 0.367816091954023,
-      "grad_norm": 0.021215129643678665,
-      "learning_rate": 7.407407407407407e-05,
-      "loss": 0.0211,
-      "step": 40
-    },
-    {
-      "epoch": 0.367816091954023,
-      "eval_loss": 0.0213603638112545,
-      "eval_runtime": 7.5795,
-      "eval_samples_per_second": 6.597,
-      "eval_steps_per_second": 1.715,
+      "epoch": 0.1932367149758454,
+      "grad_norm": 0.19719913601875305,
+      "learning_rate": 7.722007722007722e-06,
+      "loss": 0.1233,
       "step": 40
     },
     {
-      "epoch": 0.37701149425287356,
-      "grad_norm": 0.022082313895225525,
-      "learning_rate": 7.592592592592593e-05,
-      "loss": 0.0199,
+      "epoch": 0.19806763285024154,
+      "grad_norm": 0.1750050187110901,
+      "learning_rate": 7.915057915057915e-06,
+      "loss": 0.1132,
       "step": 41
     },
     {
-      "epoch": 0.38620689655172413,
-      "grad_norm": 0.020405782386660576,
-      "learning_rate": 7.777777777777778e-05,
-      "loss": 0.0188,
+      "epoch": 0.2028985507246377,
+      "grad_norm": 0.16355034708976746,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.119,
       "step": 42
     },
     {
-      "epoch": 0.3954022988505747,
-      "grad_norm": 0.020502161234617233,
-      "learning_rate": 7.962962962962964e-05,
-      "loss": 0.0203,
+      "epoch": 0.20772946859903382,
+      "grad_norm": 0.16448749601840973,
+      "learning_rate": 8.301158301158302e-06,
+      "loss": 0.1124,
       "step": 43
     },
     {
-      "epoch": 0.4045977011494253,
-      "grad_norm": 0.020312348380684853,
-      "learning_rate": 8.148148148148148e-05,
-      "loss": 0.0184,
+      "epoch": 0.21256038647342995,
+      "grad_norm": 0.16512104868888855,
+      "learning_rate": 8.494208494208494e-06,
+      "loss": 0.0985,
       "step": 44
     },
     {
-      "epoch": 0.41379310344827586,
-      "grad_norm": 0.01888488046824932,
-      "learning_rate": 8.333333333333334e-05,
-      "loss": 0.0185,
-      "step": 45
-    },
-    {
-      "epoch": 0.41379310344827586,
-      "eval_loss": 0.020027954131364822,
-      "eval_runtime": 7.5799,
-      "eval_samples_per_second": 6.596,
-      "eval_steps_per_second": 1.715,
+      "epoch": 0.21739130434782608,
+      "grad_norm": 0.18671417236328125,
+      "learning_rate": 8.687258687258689e-06,
+      "loss": 0.1016,
       "step": 45
     },
     {
-      "epoch": 0.42298850574712643,
-      "grad_norm": 0.016626283526420593,
-      "learning_rate": 8.518518518518518e-05,
-      "loss": 0.0175,
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.18079911172389984,
+      "learning_rate": 8.880308880308881e-06,
+      "loss": 0.1053,
       "step": 46
     },
     {
-      "epoch": 0.432183908045977,
-      "grad_norm": 0.01808527484536171,
-      "learning_rate": 8.703703703703704e-05,
-      "loss": 0.017,
+      "epoch": 0.22705314009661837,
+      "grad_norm": 0.16191694140434265,
+      "learning_rate": 9.073359073359074e-06,
+      "loss": 0.1205,
       "step": 47
     },
     {
-      "epoch": 0.4413793103448276,
-      "grad_norm": 0.02142912708222866,
-      "learning_rate": 8.888888888888889e-05,
-      "loss": 0.018,
+      "epoch": 0.2318840579710145,
+      "grad_norm": 0.15231011807918549,
+      "learning_rate": 9.266409266409266e-06,
+      "loss": 0.098,
       "step": 48
     },
     {
-      "epoch": 0.45057471264367815,
-      "grad_norm": 0.017742585390806198,
-      "learning_rate": 9.074074074074075e-05,
-      "loss": 0.0165,
+      "epoch": 0.23671497584541062,
+      "grad_norm": 0.14875712990760803,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.089,
       "step": 49
     },
     {
-      "epoch": 0.45977011494252873,
-      "grad_norm": 0.022063937038183212,
-      "learning_rate": 9.25925925925926e-05,
-      "loss": 0.0162,
+      "epoch": 0.24154589371980675,
+      "grad_norm": 0.13682980835437775,
+      "learning_rate": 9.652509652509653e-06,
+      "loss": 0.0909,
       "step": 50
     },
     {
-      "epoch": 0.45977011494252873,
-      "eval_loss": 0.019592924043536186,
-      "eval_runtime": 7.5839,
-      "eval_samples_per_second": 6.593,
-      "eval_steps_per_second": 1.714,
+      "epoch": 0.24154589371980675,
+      "eval_loss": 0.09681461751461029,
+      "eval_runtime": 20.6089,
+      "eval_samples_per_second": 4.852,
+      "eval_steps_per_second": 0.146,
       "step": 50
     },
     {
-      "epoch": 0.4689655172413793,
-      "grad_norm": 0.018968017771840096,
-      "learning_rate": 9.444444444444444e-05,
-      "loss": 0.0183,
+      "epoch": 0.2463768115942029,
+      "grad_norm": 0.14475570619106293,
+      "learning_rate": 9.845559845559846e-06,
+      "loss": 0.0881,
       "step": 51
     },
     {
-      "epoch": 0.4781609195402299,
-      "grad_norm": 0.024733267724514008,
-      "learning_rate": 9.62962962962963e-05,
-      "loss": 0.0189,
+      "epoch": 0.25120772946859904,
+      "grad_norm": 0.13333068788051605,
+      "learning_rate": 1.0038610038610038e-05,
+      "loss": 0.0858,
       "step": 52
     },
     {
-      "epoch": 0.48735632183908045,
-      "grad_norm": 0.021883539855480194,
-      "learning_rate": 9.814814814814815e-05,
-      "loss": 0.0188,
+      "epoch": 0.2560386473429952,
+      "grad_norm": 0.15850719809532166,
+      "learning_rate": 1.0231660231660233e-05,
+      "loss": 0.1189,
       "step": 53
     },
     {
-      "epoch": 0.496551724137931,
-      "grad_norm": 0.017187397927045822,
-      "learning_rate": 0.0001,
-      "loss": 0.0167,
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.1379750370979309,
+      "learning_rate": 1.0424710424710425e-05,
+      "loss": 0.0839,
       "step": 54
     },
     {
-      "epoch": 0.5057471264367817,
-      "grad_norm": 0.018074609339237213,
-      "learning_rate": 9.99989553622803e-05,
-      "loss": 0.0177,
-      "step": 55
-    },
-    {
-      "epoch": 0.5057471264367817,
-      "eval_loss": 0.0189231988042593,
-      "eval_runtime": 7.6176,
-      "eval_samples_per_second": 6.564,
-      "eval_steps_per_second": 1.707,
+      "epoch": 0.26570048309178745,
+      "grad_norm": 0.1474492847919464,
+      "learning_rate": 1.0617760617760618e-05,
+      "loss": 0.0921,
       "step": 55
     },
     {
-      "epoch": 0.5149425287356322,
-      "grad_norm": 0.017995629459619522,
-      "learning_rate": 9.999582149277187e-05,
-      "loss": 0.0162,
+      "epoch": 0.27053140096618356,
+      "grad_norm": 0.12119033187627792,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 0.0814,
       "step": 56
     },
     {
-      "epoch": 0.5241379310344828,
-      "grad_norm": 0.018664268776774406,
-      "learning_rate": 9.999059852242507e-05,
-      "loss": 0.0153,
+      "epoch": 0.2753623188405797,
+      "grad_norm": 0.1305706650018692,
+      "learning_rate": 1.1003861003861005e-05,
+      "loss": 0.0896,
       "step": 57
     },
     {
-      "epoch": 0.5333333333333333,
-      "grad_norm": 0.017265547066926956,
-      "learning_rate": 9.998328666948438e-05,
-      "loss": 0.0189,
+      "epoch": 0.28019323671497587,
+      "grad_norm": 0.127845898270607,
+      "learning_rate": 1.1196911196911197e-05,
+      "loss": 0.0744,
       "step": 58
     },
     {
-      "epoch": 0.542528735632184,
-      "grad_norm": 0.01721198670566082,
-      "learning_rate": 9.997388623947928e-05,
-      "loss": 0.0158,
+      "epoch": 0.28502415458937197,
+      "grad_norm": 0.1521938443183899,
+      "learning_rate": 1.138996138996139e-05,
+      "loss": 0.087,
       "step": 59
     },
     {
-      "epoch": 0.5517241379310345,
-      "grad_norm": 0.0194217711687088,
-      "learning_rate": 9.996239762521151e-05,
-      "loss": 0.0168,
-      "step": 60
-    },
-    {
-      "epoch": 0.5517241379310345,
-      "eval_loss": 0.018419023603200912,
-      "eval_runtime": 7.5901,
-      "eval_samples_per_second": 6.588,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.2898550724637681,
+      "grad_norm": 0.12970903515815735,
+      "learning_rate": 1.1583011583011584e-05,
+      "loss": 0.0721,
       "step": 60
     },
     {
-      "epoch": 0.5609195402298851,
-      "grad_norm": 0.019492702558636665,
-      "learning_rate": 9.994882130673868e-05,
-      "loss": 0.0186,
+      "epoch": 0.2946859903381642,
+      "grad_norm": 0.13914500176906586,
+      "learning_rate": 1.1776061776061777e-05,
+      "loss": 0.0765,
       "step": 61
     },
     {
-      "epoch": 0.5701149425287356,
-      "grad_norm": 0.01645253598690033,
-      "learning_rate": 9.993315785135416e-05,
-      "loss": 0.0175,
+      "epoch": 0.2995169082125604,
+      "grad_norm": 0.1305667757987976,
+      "learning_rate": 1.196911196911197e-05,
+      "loss": 0.0822,
       "step": 62
     },
     {
-      "epoch": 0.5793103448275863,
-      "grad_norm": 0.018438637256622314,
-      "learning_rate": 9.991540791356342e-05,
-      "loss": 0.0148,
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.15017420053482056,
+      "learning_rate": 1.2162162162162164e-05,
+      "loss": 0.1075,
       "step": 63
     },
     {
-      "epoch": 0.5885057471264368,
-      "grad_norm": 0.016131950542330742,
-      "learning_rate": 9.989557223505661e-05,
-      "loss": 0.0163,
+      "epoch": 0.30917874396135264,
+      "grad_norm": 0.15811654925346375,
+      "learning_rate": 1.2355212355212356e-05,
+      "loss": 0.0992,
       "step": 64
     },
     {
-      "epoch": 0.5977011494252874,
-      "grad_norm": 0.018596597015857697,
-      "learning_rate": 9.987365164467767e-05,
-      "loss": 0.017,
-      "step": 65
-    },
-    {
-      "epoch": 0.5977011494252874,
-      "eval_loss": 0.01821504347026348,
-      "eval_runtime": 7.5805,
-      "eval_samples_per_second": 6.596,
-      "eval_steps_per_second": 1.715,
+      "epoch": 0.3140096618357488,
+      "grad_norm": 0.17078033089637756,
+      "learning_rate": 1.2548262548262549e-05,
+      "loss": 0.108,
       "step": 65
     },
     {
-      "epoch": 0.6068965517241379,
-      "grad_norm": 0.017436878755688667,
-      "learning_rate": 9.98496470583896e-05,
-      "loss": 0.016,
+      "epoch": 0.3188405797101449,
+      "grad_norm": 0.14823614060878754,
+      "learning_rate": 1.2741312741312741e-05,
+      "loss": 0.0707,
       "step": 66
     },
     {
-      "epoch": 0.6160919540229886,
-      "grad_norm": 0.017340335994958878,
-      "learning_rate": 9.982355947923629e-05,
-      "loss": 0.017,
+      "epoch": 0.32367149758454106,
+      "grad_norm": 0.11846248060464859,
+      "learning_rate": 1.2934362934362934e-05,
+      "loss": 0.0836,
       "step": 67
     },
     {
-      "epoch": 0.6252873563218391,
-      "grad_norm": 0.017135461792349815,
-      "learning_rate": 9.979538999730047e-05,
-      "loss": 0.0149,
+      "epoch": 0.3285024154589372,
+      "grad_norm": 0.1331644058227539,
+      "learning_rate": 1.3127413127413127e-05,
+      "loss": 0.0847,
       "step": 68
     },
     {
-      "epoch": 0.6344827586206897,
-      "grad_norm": 0.018239280208945274,
-      "learning_rate": 9.976513978965829e-05,
-      "loss": 0.0165,
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.12533916532993317,
+      "learning_rate": 1.3320463320463322e-05,
+      "loss": 0.0616,
       "step": 69
     },
     {
-      "epoch": 0.6436781609195402,
-      "grad_norm": 0.016211749985814095,
-      "learning_rate": 9.973281012033007e-05,
-      "loss": 0.0143,
-      "step": 70
-    },
-    {
-      "epoch": 0.6436781609195402,
-      "eval_loss": 0.017700908705592155,
-      "eval_runtime": 7.5856,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 0.33816425120772947,
+      "grad_norm": 0.13879841566085815,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 0.0886,
       "step": 70
     },
     {
-      "epoch": 0.6528735632183909,
-      "grad_norm": 0.016638092696666718,
-      "learning_rate": 9.969840234022749e-05,
-      "loss": 0.0154,
+      "epoch": 0.34299516908212563,
+      "grad_norm": 0.13229133188724518,
+      "learning_rate": 1.3706563706563708e-05,
+      "loss": 0.081,
       "step": 71
     },
     {
-      "epoch": 0.6620689655172414,
-      "grad_norm": 0.020088642835617065,
-      "learning_rate": 9.966191788709716e-05,
-      "loss": 0.0163,
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.12827053666114807,
+      "learning_rate": 1.38996138996139e-05,
+      "loss": 0.0778,
       "step": 72
     },
     {
-      "epoch": 0.671264367816092,
-      "grad_norm": 0.01989225298166275,
-      "learning_rate": 9.962335828546048e-05,
-      "loss": 0.0155,
+      "epoch": 0.3526570048309179,
+      "grad_norm": 0.11902689933776855,
+      "learning_rate": 1.4092664092664093e-05,
+      "loss": 0.0695,
       "step": 73
     },
     {
-      "epoch": 0.6804597701149425,
-      "grad_norm": 0.01787324622273445,
-      "learning_rate": 9.958272514655006e-05,
-      "loss": 0.0145,
+      "epoch": 0.357487922705314,
+      "grad_norm": 0.14065556228160858,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.0702,
       "step": 74
     },
     {
-      "epoch": 0.6896551724137931,
-      "grad_norm": 0.018312491476535797,
-      "learning_rate": 9.954002016824227e-05,
-      "loss": 0.0143,
+      "epoch": 0.36231884057971014,
+      "grad_norm": 0.11380023509263992,
+      "learning_rate": 1.4478764478764478e-05,
+      "loss": 0.062,
       "step": 75
     },
     {
-      "epoch": 0.6896551724137931,
-      "eval_loss": 0.017578113824129105,
-      "eval_runtime": 7.6044,
-      "eval_samples_per_second": 6.575,
-      "eval_steps_per_second": 1.71,
+      "epoch": 0.36231884057971014,
+      "eval_loss": 0.07478907704353333,
+      "eval_runtime": 20.612,
+      "eval_samples_per_second": 4.852,
+      "eval_steps_per_second": 0.146,
       "step": 75
     },
     {
-      "epoch": 0.6988505747126437,
-      "grad_norm": 0.019170604646205902,
-      "learning_rate": 9.949524513498636e-05,
-      "loss": 0.0168,
+      "epoch": 0.3671497584541063,
+      "grad_norm": 0.13626235723495483,
+      "learning_rate": 1.4671814671814674e-05,
+      "loss": 0.0897,
       "step": 76
     },
     {
-      "epoch": 0.7080459770114943,
-      "grad_norm": 0.022306382656097412,
-      "learning_rate": 9.944840191772987e-05,
-      "loss": 0.0159,
+      "epoch": 0.3719806763285024,
+      "grad_norm": 0.11516878753900528,
+      "learning_rate": 1.4864864864864867e-05,
+      "loss": 0.0613,
       "step": 77
     },
     {
-      "epoch": 0.7172413793103448,
-      "grad_norm": 0.01719992235302925,
-      "learning_rate": 9.939949247384046e-05,
-      "loss": 0.0167,
+      "epoch": 0.37681159420289856,
+      "grad_norm": 0.11802902072668076,
+      "learning_rate": 1.505791505791506e-05,
+      "loss": 0.0661,
       "step": 78
     },
     {
-      "epoch": 0.7264367816091954,
-      "grad_norm": 0.020334061235189438,
-      "learning_rate": 9.934851884702414e-05,
-      "loss": 0.0159,
+      "epoch": 0.38164251207729466,
+      "grad_norm": 0.14430475234985352,
+      "learning_rate": 1.5250965250965252e-05,
+      "loss": 0.0765,
       "step": 79
     },
     {
-      "epoch": 0.735632183908046,
-      "grad_norm": 0.016988877207040787,
-      "learning_rate": 9.929548316723982e-05,
-      "loss": 0.0155,
-      "step": 80
-    },
-    {
-      "epoch": 0.735632183908046,
-      "eval_loss": 0.01759597845375538,
-      "eval_runtime": 7.5797,
-      "eval_samples_per_second": 6.597,
-      "eval_steps_per_second": 1.715,
+      "epoch": 0.3864734299516908,
+      "grad_norm": 0.11286959052085876,
+      "learning_rate": 1.5444015444015444e-05,
+      "loss": 0.0699,
       "step": 80
     },
     {
-      "epoch": 0.7448275862068966,
-      "grad_norm": 0.01999749429523945,
-      "learning_rate": 9.924038765061042e-05,
-      "loss": 0.0156,
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.12507657706737518,
+      "learning_rate": 1.5637065637065637e-05,
+      "loss": 0.0734,
       "step": 81
     },
     {
-      "epoch": 0.7540229885057471,
-      "grad_norm": 0.021031200885772705,
-      "learning_rate": 9.918323459933005e-05,
-      "loss": 0.017,
+      "epoch": 0.3961352657004831,
+      "grad_norm": 0.12890633940696716,
+      "learning_rate": 1.583011583011583e-05,
+      "loss": 0.0715,
       "step": 82
     },
     {
-      "epoch": 0.7632183908045977,
-      "grad_norm": 0.015833275392651558,
-      "learning_rate": 9.912402640156811e-05,
-      "loss": 0.0151,
+      "epoch": 0.40096618357487923,
+      "grad_norm": 0.12751871347427368,
+      "learning_rate": 1.6023166023166026e-05,
+      "loss": 0.0699,
       "step": 83
     },
     {
-      "epoch": 0.7724137931034483,
-      "grad_norm": 0.02128308266401291,
-      "learning_rate": 9.906276553136923e-05,
-      "loss": 0.0174,
+      "epoch": 0.4057971014492754,
+      "grad_norm": 0.1366698145866394,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 0.0765,
       "step": 84
     },
     {
-      "epoch": 0.7816091954022989,
-      "grad_norm": 0.01704588159918785,
-      "learning_rate": 9.899945454855006e-05,
-      "loss": 0.0162,
-      "step": 85
-    },
-    {
-      "epoch": 0.7816091954022989,
-      "eval_loss": 0.016866466030478477,
-      "eval_runtime": 7.5833,
-      "eval_samples_per_second": 6.593,
-      "eval_steps_per_second": 1.714,
+      "epoch": 0.4106280193236715,
+      "grad_norm": 0.10429048538208008,
+      "learning_rate": 1.640926640926641e-05,
+      "loss": 0.057,
       "step": 85
     },
     {
-      "epoch": 0.7908045977011494,
-      "grad_norm": 0.01643301546573639,
-      "learning_rate": 9.893409609859222e-05,
-      "loss": 0.0152,
+      "epoch": 0.41545893719806765,
+      "grad_norm": 0.10773543268442154,
+      "learning_rate": 1.6602316602316603e-05,
+      "loss": 0.0666,
       "step": 86
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.017949821427464485,
-      "learning_rate": 9.88666929125318e-05,
-      "loss": 0.0156,
+      "epoch": 0.42028985507246375,
+      "grad_norm": 0.110495425760746,
+      "learning_rate": 1.6795366795366796e-05,
+      "loss": 0.0635,
       "step": 87
     },
     {
-      "epoch": 0.8091954022988506,
-      "grad_norm": 0.018931446596980095,
-      "learning_rate": 9.879724780684519e-05,
-      "loss": 0.0162,
+      "epoch": 0.4251207729468599,
+      "grad_norm": 0.139565110206604,
+      "learning_rate": 1.698841698841699e-05,
+      "loss": 0.0874,
       "step": 88
     },
     {
-      "epoch": 0.8183908045977012,
-      "grad_norm": 0.012615452520549297,
-      "learning_rate": 9.872576368333151e-05,
-      "loss": 0.0138,
+      "epoch": 0.42995169082125606,
+      "grad_norm": 0.1371033936738968,
+      "learning_rate": 1.718146718146718e-05,
+      "loss": 0.0669,
       "step": 89
     },
     {
-      "epoch": 0.8275862068965517,
-      "grad_norm": 0.016196150332689285,
-      "learning_rate": 9.865224352899119e-05,
-      "loss": 0.0164,
-      "step": 90
-    },
-    {
-      "epoch": 0.8275862068965517,
-      "eval_loss": 0.01638174243271351,
-      "eval_runtime": 7.5993,
-      "eval_samples_per_second": 6.58,
-      "eval_steps_per_second": 1.711,
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.14201398193836212,
+      "learning_rate": 1.7374517374517377e-05,
+      "loss": 0.0638,
       "step": 90
     },
     {
-      "epoch": 0.8367816091954023,
-      "grad_norm": 0.016620157286524773,
-      "learning_rate": 9.857669041590134e-05,
-      "loss": 0.0158,
+      "epoch": 0.4396135265700483,
+      "grad_norm": 0.13561639189720154,
+      "learning_rate": 1.756756756756757e-05,
+      "loss": 0.0654,
       "step": 91
     },
     {
-      "epoch": 0.8459770114942529,
-      "grad_norm": 0.014475880190730095,
-      "learning_rate": 9.849910750108717e-05,
-      "loss": 0.0149,
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.11701712012290955,
+      "learning_rate": 1.7760617760617762e-05,
+      "loss": 0.0589,
       "step": 92
     },
     {
-      "epoch": 0.8551724137931035,
-      "grad_norm": 0.016380243003368378,
-      "learning_rate": 9.84194980263903e-05,
-      "loss": 0.0147,
+      "epoch": 0.4492753623188406,
+      "grad_norm": 0.11068889498710632,
+      "learning_rate": 1.7953667953667955e-05,
+      "loss": 0.0707,
       "step": 93
     },
     {
-      "epoch": 0.864367816091954,
-      "grad_norm": 0.015107197687029839,
-      "learning_rate": 9.83378653183331e-05,
-      "loss": 0.0148,
+      "epoch": 0.45410628019323673,
+      "grad_norm": 0.11085647344589233,
+      "learning_rate": 1.8146718146718147e-05,
+      "loss": 0.0604,
       "step": 94
     },
     {
-      "epoch": 0.8735632183908046,
-      "grad_norm": 0.01710473746061325,
-      "learning_rate": 9.825421278797983e-05,
-      "loss": 0.0154,
-      "step": 95
-    },
-    {
-      "epoch": 0.8735632183908046,
-      "eval_loss": 0.0161591824144125,
-      "eval_runtime": 7.6104,
-      "eval_samples_per_second": 6.57,
-      "eval_steps_per_second": 1.708,
+      "epoch": 0.45893719806763283,
+      "grad_norm": 0.11306110769510269,
+      "learning_rate": 1.833976833976834e-05,
+      "loss": 0.0526,
       "step": 95
     },
     {
-      "epoch": 0.8827586206896552,
-      "grad_norm": 0.015920141711831093,
-      "learning_rate": 9.816854393079403e-05,
-      "loss": 0.0159,
+      "epoch": 0.463768115942029,
+      "grad_norm": 0.12925806641578674,
+      "learning_rate": 1.8532818532818533e-05,
+      "loss": 0.0541,
       "step": 96
     },
     {
-      "epoch": 0.8919540229885058,
-      "grad_norm": 0.015056249685585499,
-      "learning_rate": 9.808086232649246e-05,
-      "loss": 0.0154,
+      "epoch": 0.46859903381642515,
+      "grad_norm": 0.11406856775283813,
+      "learning_rate": 1.8725868725868725e-05,
+      "loss": 0.0575,
       "step": 97
     },
     {
-      "epoch": 0.9011494252873563,
-      "grad_norm": 0.016236839815974236,
-      "learning_rate": 9.799117163889559e-05,
-      "loss": 0.0152,
+      "epoch": 0.47342995169082125,
+      "grad_norm": 0.11861402541399002,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 0.0593,
       "step": 98
     },
     {
-      "epoch": 0.9103448275862069,
-      "grad_norm": 0.014755873009562492,
-      "learning_rate": 9.789947561577445e-05,
-      "loss": 0.0135,
+      "epoch": 0.4782608695652174,
+      "grad_norm": 0.15513893961906433,
+      "learning_rate": 1.9111969111969114e-05,
+      "loss": 0.0565,
       "step": 99
     },
     {
-      "epoch": 0.9195402298850575,
-      "grad_norm": 0.015870464965701103,
-      "learning_rate": 9.780577808869398e-05,
-      "loss": 0.0164,
+      "epoch": 0.4830917874396135,
+      "grad_norm": 0.1187349483370781,
+      "learning_rate": 1.9305019305019306e-05,
+      "loss": 0.0602,
       "step": 100
     },
     {
-      "epoch": 0.9195402298850575,
-      "eval_loss": 0.015854647383093834,
-      "eval_runtime": 7.5871,
-      "eval_samples_per_second": 6.59,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.4830917874396135,
+      "eval_loss": 0.06083356589078903,
+      "eval_runtime": 20.6118,
+      "eval_samples_per_second": 4.852,
+      "eval_steps_per_second": 0.146,
       "step": 100
     },
     {
-      "epoch": 0.9287356321839081,
-      "grad_norm": 0.015317806974053383,
-      "learning_rate": 9.771008297285307e-05,
-      "loss": 0.0144,
+      "epoch": 0.48792270531400966,
+      "grad_norm": 0.11506832391023636,
+      "learning_rate": 1.94980694980695e-05,
+      "loss": 0.0572,
       "step": 101
     },
     {
-      "epoch": 0.9379310344827586,
-      "grad_norm": 0.015459987334907055,
-      "learning_rate": 9.761239426692077e-05,
-      "loss": 0.0136,
+      "epoch": 0.4927536231884058,
+      "grad_norm": 0.12325363606214523,
+      "learning_rate": 1.969111969111969e-05,
+      "loss": 0.0615,
       "step": 102
     },
     {
-      "epoch": 0.9471264367816092,
-      "grad_norm": 0.01605418138206005,
-      "learning_rate": 9.751271605286941e-05,
-      "loss": 0.0133,
+      "epoch": 0.4975845410628019,
+      "grad_norm": 0.13748331367969513,
+      "learning_rate": 1.9884169884169884e-05,
+      "loss": 0.0644,
       "step": 103
     },
     {
-      "epoch": 0.9563218390804598,
-      "grad_norm": 0.01653791405260563,
-      "learning_rate": 9.741105249580383e-05,
-      "loss": 0.0142,
+      "epoch": 0.5024154589371981,
+      "grad_norm": 0.12451450526714325,
+      "learning_rate": 2.0077220077220077e-05,
+      "loss": 0.0532,
       "step": 104
     },
     {
-      "epoch": 0.9655172413793104,
-      "grad_norm": 0.017847595736384392,
-      "learning_rate": 9.730740784378753e-05,
-      "loss": 0.0156,
-      "step": 105
-    },
-    {
-      "epoch": 0.9655172413793104,
-      "eval_loss": 0.01596776954829693,
-      "eval_runtime": 7.5993,
-      "eval_samples_per_second": 6.58,
-      "eval_steps_per_second": 1.711,
+      "epoch": 0.5072463768115942,
+      "grad_norm": 0.13660487532615662,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 0.0505,
       "step": 105
     },
     {
-      "epoch": 0.9747126436781609,
-      "grad_norm": 0.01591021567583084,
-      "learning_rate": 9.7201786427665e-05,
-      "loss": 0.0142,
+      "epoch": 0.5120772946859904,
+      "grad_norm": 0.11977998912334442,
+      "learning_rate": 2.0463320463320465e-05,
+      "loss": 0.0587,
       "step": 106
     },
     {
-      "epoch": 0.9839080459770115,
-      "grad_norm": 0.019531626254320145,
-      "learning_rate": 9.709419266088086e-05,
-      "loss": 0.0149,
+      "epoch": 0.5169082125603864,
+      "grad_norm": 0.11822707206010818,
+      "learning_rate": 2.0656370656370658e-05,
+      "loss": 0.0465,
       "step": 107
     },
     {
-      "epoch": 0.993103448275862,
-      "grad_norm": 0.017293043434619904,
-      "learning_rate": 9.698463103929542e-05,
-      "loss": 0.0145,
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.13716371357440948,
+      "learning_rate": 2.084942084942085e-05,
+      "loss": 0.0452,
       "step": 108
     },
     {
-      "epoch": 1.0022988505747126,
-      "grad_norm": 0.01688973978161812,
-      "learning_rate": 9.687310614099675e-05,
-      "loss": 0.0165,
+      "epoch": 0.5265700483091788,
+      "grad_norm": 0.1545065939426422,
+      "learning_rate": 2.1042471042471043e-05,
+      "loss": 0.0589,
       "step": 109
     },
     {
-      "epoch": 1.0114942528735633,
-      "grad_norm": 0.021913839504122734,
-      "learning_rate": 9.67596226261095e-05,
-      "loss": 0.0145,
-      "step": 110
-    },
-    {
-      "epoch": 1.0114942528735633,
-      "eval_loss": 0.01590101048350334,
-      "eval_runtime": 7.5856,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 0.5314009661835749,
+      "grad_norm": 0.12922868132591248,
+      "learning_rate": 2.1235521235521236e-05,
+      "loss": 0.0758,
       "step": 110
     },
     {
-      "epoch": 1.0206896551724138,
-      "grad_norm": 0.018519939854741096,
-      "learning_rate": 9.664418523660004e-05,
-      "loss": 0.0131,
+      "epoch": 0.5362318840579711,
+      "grad_norm": 0.17041181027889252,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.0653,
       "step": 111
     },
     {
-      "epoch": 1.0298850574712644,
-      "grad_norm": 0.020266039296984673,
-      "learning_rate": 9.652679879607843e-05,
-      "loss": 0.0129,
+      "epoch": 0.5410628019323671,
+      "grad_norm": 0.14169234037399292,
+      "learning_rate": 2.1621621621621624e-05,
+      "loss": 0.0567,
       "step": 112
     },
     {
-      "epoch": 1.0390804597701149,
-      "grad_norm": 0.021503202617168427,
-      "learning_rate": 9.640746820959684e-05,
-      "loss": 0.0147,
+      "epoch": 0.5458937198067633,
+      "grad_norm": 0.158858060836792,
+      "learning_rate": 2.1814671814671817e-05,
+      "loss": 0.0492,
       "step": 113
     },
     {
-      "epoch": 1.0482758620689656,
-      "grad_norm": 0.01816105656325817,
-      "learning_rate": 9.628619846344454e-05,
-      "loss": 0.0131,
+      "epoch": 0.5507246376811594,
+      "grad_norm": 0.17282356321811676,
+      "learning_rate": 2.200772200772201e-05,
+      "loss": 0.0561,
       "step": 114
     },
     {
-      "epoch": 1.0574712643678161,
-      "grad_norm": 0.02214297465980053,
-      "learning_rate": 9.616299462493952e-05,
-      "loss": 0.0133,
-      "step": 115
-    },
-    {
-      "epoch": 1.0574712643678161,
-      "eval_loss": 0.015581479296088219,
-      "eval_runtime": 7.5911,
-      "eval_samples_per_second": 6.587,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.12390092015266418,
+      "learning_rate": 2.2200772200772202e-05,
+      "loss": 0.0671,
       "step": 115
     },
     {
-      "epoch": 1.0666666666666667,
-      "grad_norm": 0.022218503057956696,
-      "learning_rate": 9.603786184221693e-05,
-      "loss": 0.0154,
+      "epoch": 0.5603864734299517,
+      "grad_norm": 0.1567428857088089,
+      "learning_rate": 2.2393822393822394e-05,
+      "loss": 0.0556,
       "step": 116
     },
     {
-      "epoch": 1.0758620689655172,
-      "grad_norm": 0.019172191619873047,
-      "learning_rate": 9.591080534401371e-05,
-      "loss": 0.0136,
+      "epoch": 0.5652173913043478,
+      "grad_norm": 0.16888275742530823,
+      "learning_rate": 2.2586872586872587e-05,
+      "loss": 0.0521,
       "step": 117
     },
     {
-      "epoch": 1.085057471264368,
-      "grad_norm": 0.021756643429398537,
-      "learning_rate": 9.57818304394503e-05,
-      "loss": 0.0142,
+      "epoch": 0.5700483091787439,
+      "grad_norm": 0.1285606473684311,
+      "learning_rate": 2.277992277992278e-05,
+      "loss": 0.0558,
       "step": 118
     },
     {
-      "epoch": 1.0942528735632184,
-      "grad_norm": 0.018046511337161064,
-      "learning_rate": 9.565094251780871e-05,
-      "loss": 0.0144,
+      "epoch": 0.5748792270531401,
+      "grad_norm": 0.12908753752708435,
+      "learning_rate": 2.2972972972972976e-05,
+      "loss": 0.0872,
       "step": 119
     },
     {
-      "epoch": 1.103448275862069,
-      "grad_norm": 0.02018323354423046,
-      "learning_rate": 9.551814704830734e-05,
-      "loss": 0.0126,
-      "step": 120
-    },
-    {
-      "epoch": 1.103448275862069,
-      "eval_loss": 0.01549120806157589,
-      "eval_runtime": 7.5874,
-      "eval_samples_per_second": 6.59,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.5797101449275363,
+      "grad_norm": 0.13818618655204773,
+      "learning_rate": 2.3166023166023168e-05,
+      "loss": 0.0486,
       "step": 120
     },
     {
-      "epoch": 1.1126436781609195,
-      "grad_norm": 0.02136109583079815,
-      "learning_rate": 9.538344957987244e-05,
-      "loss": 0.0144,
+      "epoch": 0.5845410628019324,
+      "grad_norm": 0.14711114764213562,
+      "learning_rate": 2.335907335907336e-05,
+      "loss": 0.0608,
       "step": 121
     },
     {
-      "epoch": 1.1218390804597702,
-      "grad_norm": 0.0185832716524601,
-      "learning_rate": 9.524685574090627e-05,
-      "loss": 0.0149,
+      "epoch": 0.5893719806763285,
+      "grad_norm": 0.1265874058008194,
+      "learning_rate": 2.3552123552123553e-05,
+      "loss": 0.0816,
       "step": 122
     },
     {
-      "epoch": 1.1310344827586207,
-      "grad_norm": 0.01844456046819687,
-      "learning_rate": 9.51083712390519e-05,
-      "loss": 0.0137,
+      "epoch": 0.5942028985507246,
+      "grad_norm": 0.1758117824792862,
+      "learning_rate": 2.3745173745173746e-05,
+      "loss": 0.0599,
       "step": 123
     },
     {
-      "epoch": 1.1402298850574712,
-      "grad_norm": 0.018193133175373077,
-      "learning_rate": 9.496800186095466e-05,
-      "loss": 0.0131,
+      "epoch": 0.5990338164251208,
+      "grad_norm": 0.17134052515029907,
+      "learning_rate": 2.393822393822394e-05,
+      "loss": 0.0512,
       "step": 124
     },
     {
-      "epoch": 1.1494252873563218,
-      "grad_norm": 0.01977015659213066,
-      "learning_rate": 9.482575347202047e-05,
-      "loss": 0.0145,
+      "epoch": 0.6038647342995169,
+      "grad_norm": 0.12125252187252045,
+      "learning_rate": 2.413127413127413e-05,
+      "loss": 0.057,
       "step": 125
     },
     {
-      "epoch": 1.1494252873563218,
-      "eval_loss": 0.015383351594209671,
-      "eval_runtime": 7.6046,
-      "eval_samples_per_second": 6.575,
-      "eval_steps_per_second": 1.709,
+      "epoch": 0.6038647342995169,
+      "eval_loss": 0.05518662929534912,
+      "eval_runtime": 20.6274,
+      "eval_samples_per_second": 4.848,
+      "eval_steps_per_second": 0.145,
       "step": 125
     },
     {
-      "epoch": 1.1586206896551725,
-      "grad_norm": 0.016328683122992516,
-      "learning_rate": 9.468163201617062e-05,
-      "loss": 0.0133,
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.16893912851810455,
+      "learning_rate": 2.4324324324324327e-05,
+      "loss": 0.0477,
       "step": 126
     },
     {
-      "epoch": 1.167816091954023,
-      "grad_norm": 0.0158162210136652,
-      "learning_rate": 9.453564351559348e-05,
-      "loss": 0.0145,
+      "epoch": 0.6135265700483091,
+      "grad_norm": 0.16074183583259583,
+      "learning_rate": 2.451737451737452e-05,
+      "loss": 0.0544,
       "step": 127
     },
     {
-      "epoch": 1.1770114942528735,
-      "grad_norm": 0.01766177825629711,
-      "learning_rate": 9.438779407049281e-05,
-      "loss": 0.0133,
+      "epoch": 0.6183574879227053,
+      "grad_norm": 0.14464056491851807,
+      "learning_rate": 2.4710424710424712e-05,
+      "loss": 0.0578,
       "step": 128
     },
     {
-      "epoch": 1.186206896551724,
-      "grad_norm": 0.015718543902039528,
-      "learning_rate": 9.423808985883289e-05,
-      "loss": 0.013,
+      "epoch": 0.6231884057971014,
+      "grad_norm": 0.14202292263507843,
+      "learning_rate": 2.4903474903474905e-05,
+      "loss": 0.0398,
       "step": 129
     },
     {
-      "epoch": 1.1954022988505748,
-      "grad_norm": 0.01557285524904728,
-      "learning_rate": 9.40865371360804e-05,
-      "loss": 0.0125,
-      "step": 130
-    },
-    {
-      "epoch": 1.1954022988505748,
-      "eval_loss": 0.014952810481190681,
-      "eval_runtime": 7.5808,
-      "eval_samples_per_second": 6.596,
-      "eval_steps_per_second": 1.715,
+      "epoch": 0.6280193236714976,
+      "grad_norm": 0.16005530953407288,
+      "learning_rate": 2.5096525096525097e-05,
+      "loss": 0.0527,
       "step": 130
     },
     {
-      "epoch": 1.2045977011494253,
-      "grad_norm": 0.014815889298915863,
-      "learning_rate": 9.393314223494296e-05,
-      "loss": 0.0111,
+      "epoch": 0.6328502415458938,
+      "grad_norm": 0.10633774846792221,
+      "learning_rate": 2.528957528957529e-05,
+      "loss": 0.0532,
       "step": 131
     },
     {
-      "epoch": 1.2137931034482758,
-      "grad_norm": 0.01469665952026844,
-      "learning_rate": 9.377791156510455e-05,
-      "loss": 0.0132,
+      "epoch": 0.6376811594202898,
+      "grad_norm": 0.13469672203063965,
+      "learning_rate": 2.5482625482625483e-05,
+      "loss": 0.0477,
       "step": 132
     },
     {
-      "epoch": 1.2229885057471264,
-      "grad_norm": 0.014182367362082005,
-      "learning_rate": 9.362085161295769e-05,
-      "loss": 0.0119,
+      "epoch": 0.642512077294686,
+      "grad_norm": 0.13366642594337463,
+      "learning_rate": 2.5675675675675675e-05,
+      "loss": 0.0473,
       "step": 133
     },
     {
-      "epoch": 1.232183908045977,
-      "grad_norm": 0.016701558604836464,
-      "learning_rate": 9.346196894133239e-05,
-      "loss": 0.0119,
+      "epoch": 0.6473429951690821,
+      "grad_norm": 0.13137054443359375,
+      "learning_rate": 2.5868725868725868e-05,
+      "loss": 0.0829,
       "step": 134
     },
     {
-      "epoch": 1.2413793103448276,
-      "grad_norm": 0.017275458201766014,
-      "learning_rate": 9.330127018922194e-05,
-      "loss": 0.0122,
-      "step": 135
-    },
-    {
-      "epoch": 1.2413793103448276,
-      "eval_loss": 0.014751402661204338,
-      "eval_runtime": 7.5882,
-      "eval_samples_per_second": 6.589,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.6521739130434783,
+      "grad_norm": 0.14475256204605103,
+      "learning_rate": 2.606177606177606e-05,
+      "loss": 0.0612,
       "step": 135
     },
     {
-      "epoch": 1.2505747126436781,
-      "grad_norm": 0.014996697194874287,
-      "learning_rate": 9.313876207150543e-05,
-      "loss": 0.0128,
+      "epoch": 0.6570048309178744,
+      "grad_norm": 0.16254907846450806,
+      "learning_rate": 2.6254826254826253e-05,
+      "loss": 0.0557,
       "step": 136
     },
     {
-      "epoch": 1.2597701149425287,
-      "grad_norm": 0.01727742701768875,
-      "learning_rate": 9.297445137866727e-05,
-      "loss": 0.0119,
+      "epoch": 0.6618357487922706,
+      "grad_norm": 0.11103875190019608,
+      "learning_rate": 2.6447876447876452e-05,
+      "loss": 0.0495,
       "step": 137
     },
     {
-      "epoch": 1.2689655172413792,
-      "grad_norm": 0.017734119668602943,
-      "learning_rate": 9.280834497651334e-05,
-      "loss": 0.0133,
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.15056666731834412,
+      "learning_rate": 2.6640926640926645e-05,
+      "loss": 0.0443,
       "step": 138
     },
     {
-      "epoch": 1.27816091954023,
-      "grad_norm": 0.017755163833498955,
-      "learning_rate": 9.264044980588416e-05,
-      "loss": 0.0129,
+      "epoch": 0.6714975845410628,
+      "grad_norm": 0.1170201525092125,
+      "learning_rate": 2.6833976833976838e-05,
+      "loss": 0.0515,
       "step": 139
     },
     {
-      "epoch": 1.2873563218390804,
-      "grad_norm": 0.015634773299098015,
-      "learning_rate": 9.247077288236488e-05,
-      "loss": 0.0127,
-      "step": 140
-    },
-    {
-      "epoch": 1.2873563218390804,
-      "eval_loss": 0.014655789360404015,
-      "eval_runtime": 7.5803,
-      "eval_samples_per_second": 6.596,
-      "eval_steps_per_second": 1.715,
+      "epoch": 0.6763285024154589,
+      "grad_norm": 0.13160322606563568,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 0.0513,
       "step": 140
     },
     {
-      "epoch": 1.296551724137931,
-      "grad_norm": 0.014848074875772,
-      "learning_rate": 9.229932129599205e-05,
-      "loss": 0.0126,
+      "epoch": 0.6811594202898551,
+      "grad_norm": 0.11938180029392242,
+      "learning_rate": 2.7220077220077223e-05,
+      "loss": 0.0473,
       "step": 141
     },
     {
-      "epoch": 1.3057471264367817,
-      "grad_norm": 0.015292002819478512,
-      "learning_rate": 9.212610221095748e-05,
-      "loss": 0.0126,
+      "epoch": 0.6859903381642513,
+      "grad_norm": 0.11169520020484924,
+      "learning_rate": 2.7413127413127415e-05,
+      "loss": 0.0668,
       "step": 142
     },
     {
-      "epoch": 1.3149425287356322,
-      "grad_norm": 0.01488974317908287,
-      "learning_rate": 9.195112286530873e-05,
-      "loss": 0.0113,
+      "epoch": 0.6908212560386473,
+      "grad_norm": 0.12496872991323471,
+      "learning_rate": 2.7606177606177608e-05,
+      "loss": 0.0609,
       "step": 143
     },
     {
-      "epoch": 1.3241379310344827,
-      "grad_norm": 0.015857763588428497,
-      "learning_rate": 9.177439057064683e-05,
-      "loss": 0.0118,
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.150443434715271,
+      "learning_rate": 2.77992277992278e-05,
+      "loss": 0.0782,
       "step": 144
     },
     {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 0.01526150293648243,
-      "learning_rate": 9.159591271182058e-05,
-      "loss": 0.0139,
-      "step": 145
-    },
-    {
-      "epoch": 1.3333333333333333,
-      "eval_loss": 0.014428853057324886,
-      "eval_runtime": 7.5934,
-      "eval_samples_per_second": 6.585,
-      "eval_steps_per_second": 1.712,
+      "epoch": 0.7004830917874396,
+      "grad_norm": 0.11996805667877197,
+      "learning_rate": 2.7992277992277993e-05,
+      "loss": 0.0475,
       "step": 145
     },
     {
-      "epoch": 1.3425287356321838,
-      "grad_norm": 0.015166752971708775,
-      "learning_rate": 9.141569674661817e-05,
-      "loss": 0.0122,
+      "epoch": 0.7053140096618358,
+      "grad_norm": 0.11406496167182922,
+      "learning_rate": 2.8185328185328186e-05,
+      "loss": 0.0663,
       "step": 146
     },
     {
-      "epoch": 1.3517241379310345,
-      "grad_norm": 0.014061689376831055,
-      "learning_rate": 9.123375020545535e-05,
-      "loss": 0.0123,
+      "epoch": 0.7101449275362319,
+      "grad_norm": 0.10466250032186508,
+      "learning_rate": 2.8378378378378378e-05,
+      "loss": 0.0483,
       "step": 147
     },
     {
-      "epoch": 1.360919540229885,
-      "grad_norm": 0.015425407327711582,
-      "learning_rate": 9.105008069106093e-05,
-      "loss": 0.0117,
+      "epoch": 0.714975845410628,
+      "grad_norm": 0.11609887331724167,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.0452,
       "step": 148
     },
     {
-      "epoch": 1.3701149425287356,
-      "grad_norm": 0.018539879471063614,
-      "learning_rate": 9.086469587815904e-05,
-      "loss": 0.0123,
+      "epoch": 0.7198067632850241,
+      "grad_norm": 0.12703728675842285,
+      "learning_rate": 2.8764478764478763e-05,
+      "loss": 0.0504,
       "step": 149
     },
     {
-      "epoch": 1.3793103448275863,
-      "grad_norm": 0.014866286888718605,
-      "learning_rate": 9.067760351314838e-05,
-      "loss": 0.0122,
+      "epoch": 0.7246376811594203,
+      "grad_norm": 0.11724966764450073,
+      "learning_rate": 2.8957528957528956e-05,
+      "loss": 0.0456,
       "step": 150
     },
     {
-      "epoch": 1.3793103448275863,
-      "eval_loss": 0.014377254992723465,
-      "eval_runtime": 7.6093,
-      "eval_samples_per_second": 6.571,
-      "eval_steps_per_second": 1.708,
+      "epoch": 0.7246376811594203,
+      "eval_loss": 0.05008581280708313,
+      "eval_runtime": 20.7274,
+      "eval_samples_per_second": 4.825,
+      "eval_steps_per_second": 0.145,
       "step": 150
     },
     {
-      "epoch": 1.3885057471264368,
-      "grad_norm": 0.014935806393623352,
-      "learning_rate": 9.048881141377863e-05,
-      "loss": 0.0126,
+      "epoch": 0.7294685990338164,
+      "grad_norm": 0.11703478544950485,
+      "learning_rate": 2.915057915057915e-05,
+      "loss": 0.0445,
       "step": 151
     },
     {
-      "epoch": 1.3977011494252873,
-      "grad_norm": 0.017417676746845245,
-      "learning_rate": 9.029832746882371e-05,
-      "loss": 0.013,
+      "epoch": 0.7342995169082126,
+      "grad_norm": 0.1328444480895996,
+      "learning_rate": 2.9343629343629348e-05,
+      "loss": 0.0534,
       "step": 152
     },
     {
-      "epoch": 1.4068965517241379,
-      "grad_norm": 0.015856370329856873,
-      "learning_rate": 9.01061596377522e-05,
-      "loss": 0.0114,
+      "epoch": 0.7391304347826086,
+      "grad_norm": 0.1250855177640915,
+      "learning_rate": 2.953667953667954e-05,
+      "loss": 0.0655,
       "step": 153
     },
     {
-      "epoch": 1.4160919540229884,
-      "grad_norm": 0.015301518142223358,
-      "learning_rate": 8.991231595039465e-05,
-      "loss": 0.0136,
+      "epoch": 0.7439613526570048,
+      "grad_norm": 0.0979803130030632,
+      "learning_rate": 2.9729729729729733e-05,
+      "loss": 0.0484,
       "step": 154
     },
     {
-      "epoch": 1.4252873563218391,
-      "grad_norm": 0.017993515357375145,
-      "learning_rate": 8.97168045066082e-05,
-      "loss": 0.0138,
-      "step": 155
-    },
-    {
-      "epoch": 1.4252873563218391,
-      "eval_loss": 0.013892178423702717,
-      "eval_runtime": 7.5856,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 0.748792270531401,
+      "grad_norm": 0.10639014095067978,
+      "learning_rate": 2.9922779922779926e-05,
+      "loss": 0.0458,
       "step": 155
     },
     {
-      "epoch": 1.4344827586206896,
-      "grad_norm": 0.014389860443770885,
-      "learning_rate": 8.951963347593797e-05,
-      "loss": 0.0118,
+      "epoch": 0.7536231884057971,
+      "grad_norm": 0.10299219191074371,
+      "learning_rate": 3.011583011583012e-05,
+      "loss": 0.0343,
       "step": 156
     },
     {
-      "epoch": 1.4436781609195402,
-      "grad_norm": 0.013852115720510483,
-      "learning_rate": 8.932081109727582e-05,
-      "loss": 0.011,
+      "epoch": 0.7584541062801933,
+      "grad_norm": 0.12237931042909622,
+      "learning_rate": 3.030888030888031e-05,
+      "loss": 0.0493,
       "step": 157
     },
     {
-      "epoch": 1.452873563218391,
-      "grad_norm": 0.01491858996450901,
-      "learning_rate": 8.912034567851599e-05,
-      "loss": 0.0115,
+      "epoch": 0.7632850241545893,
+      "grad_norm": 0.14812174439430237,
+      "learning_rate": 3.0501930501930504e-05,
+      "loss": 0.0526,
       "step": 158
     },
     {
-      "epoch": 1.4620689655172414,
-      "grad_norm": 0.014816755428910255,
-      "learning_rate": 8.891824559620801e-05,
-      "loss": 0.0115,
+      "epoch": 0.7681159420289855,
+      "grad_norm": 0.1112482100725174,
+      "learning_rate": 3.0694980694980696e-05,
+      "loss": 0.0427,
       "step": 159
     },
     {
-      "epoch": 1.471264367816092,
-      "grad_norm": 0.017679313197731972,
-      "learning_rate": 8.871451929520663e-05,
-      "loss": 0.0143,
-      "step": 160
-    },
-    {
-      "epoch": 1.471264367816092,
-      "eval_loss": 0.013947720639407635,
-      "eval_runtime": 7.588,
-      "eval_samples_per_second": 6.589,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.7729468599033816,
+      "grad_norm": 0.13047181069850922,
+      "learning_rate": 3.088803088803089e-05,
+      "loss": 0.0584,
       "step": 160
     },
     {
-      "epoch": 1.4804597701149425,
-      "grad_norm": 0.015570229850709438,
-      "learning_rate": 8.850917528831899e-05,
-      "loss": 0.0131,
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.1432395875453949,
+      "learning_rate": 3.108108108108108e-05,
+      "loss": 0.0499,
       "step": 161
     },
     {
-      "epoch": 1.489655172413793,
-      "grad_norm": 0.014714114367961884,
-      "learning_rate": 8.83022221559489e-05,
-      "loss": 0.0109,
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.1461172252893448,
+      "learning_rate": 3.1274131274131274e-05,
+      "loss": 0.0636,
       "step": 162
     },
     {
-      "epoch": 1.4988505747126437,
-      "grad_norm": 0.015797005966305733,
-      "learning_rate": 8.809366854573831e-05,
-      "loss": 0.0117,
+      "epoch": 0.7874396135265701,
+      "grad_norm": 0.11835489422082901,
+      "learning_rate": 3.1467181467181466e-05,
+      "loss": 0.0412,
       "step": 163
     },
     {
-      "epoch": 1.5080459770114942,
-      "grad_norm": 0.01803259737789631,
-      "learning_rate": 8.78835231722059e-05,
-      "loss": 0.013,
+      "epoch": 0.7922705314009661,
+      "grad_norm": 0.18038594722747803,
+      "learning_rate": 3.166023166023166e-05,
+      "loss": 0.0517,
       "step": 164
     },
     {
-      "epoch": 1.5172413793103448,
-      "grad_norm": 0.015303310938179493,
-      "learning_rate": 8.767179481638303e-05,
-      "loss": 0.0124,
-      "step": 165
-    },
-    {
-      "epoch": 1.5172413793103448,
-      "eval_loss": 0.013795728795230389,
-      "eval_runtime": 7.5973,
-      "eval_samples_per_second": 6.581,
-      "eval_steps_per_second": 1.711,
+      "epoch": 0.7971014492753623,
+      "grad_norm": 0.11978429555892944,
+      "learning_rate": 3.185328185328185e-05,
+      "loss": 0.037,
       "step": 165
     },
     {
-      "epoch": 1.5264367816091955,
-      "grad_norm": 0.01512802392244339,
-      "learning_rate": 8.745849232544681e-05,
-      "loss": 0.0113,
+      "epoch": 0.8019323671497585,
+      "grad_norm": 0.13728344440460205,
+      "learning_rate": 3.204633204633205e-05,
+      "loss": 0.0346,
       "step": 166
     },
     {
-      "epoch": 1.535632183908046,
-      "grad_norm": 0.01602269895374775,
-      "learning_rate": 8.724362461235029e-05,
-      "loss": 0.0137,
+      "epoch": 0.8067632850241546,
+      "grad_norm": 0.173124760389328,
+      "learning_rate": 3.2239382239382244e-05,
+      "loss": 0.0512,
       "step": 167
     },
     {
-      "epoch": 1.5448275862068965,
-      "grad_norm": 0.015303592197597027,
-      "learning_rate": 8.702720065545024e-05,
-      "loss": 0.0117,
+      "epoch": 0.8115942028985508,
+      "grad_norm": 0.09541524946689606,
+      "learning_rate": 3.2432432432432436e-05,
+      "loss": 0.0369,
       "step": 168
     },
     {
-      "epoch": 1.5540229885057473,
-      "grad_norm": 0.014865975826978683,
-      "learning_rate": 8.680922949813178e-05,
-      "loss": 0.0122,
+      "epoch": 0.8164251207729468,
+      "grad_norm": 0.14173609018325806,
+      "learning_rate": 3.262548262548263e-05,
+      "loss": 0.0536,
       "step": 169
     },
     {
-      "epoch": 1.5632183908045976,
-      "grad_norm": 0.01662212237715721,
-      "learning_rate": 8.658972024843062e-05,
-      "loss": 0.0124,
-      "step": 170
-    },
-    {
-      "epoch": 1.5632183908045976,
-      "eval_loss": 0.013545939698815346,
-      "eval_runtime": 7.6043,
-      "eval_samples_per_second": 6.575,
-      "eval_steps_per_second": 1.71,
+      "epoch": 0.821256038647343,
+      "grad_norm": 0.14616850018501282,
+      "learning_rate": 3.281853281853282e-05,
+      "loss": 0.039,
       "step": 170
     },
     {
-      "epoch": 1.5724137931034483,
-      "grad_norm": 0.016805054619908333,
-      "learning_rate": 8.636868207865244e-05,
-      "loss": 0.0126,
+      "epoch": 0.8260869565217391,
+      "grad_norm": 0.12013251334428787,
+      "learning_rate": 3.3011583011583014e-05,
+      "loss": 0.0488,
       "step": 171
     },
     {
-      "epoch": 1.5816091954022988,
-      "grad_norm": 0.015283631160855293,
-      "learning_rate": 8.614612422498964e-05,
-      "loss": 0.0129,
+      "epoch": 0.8309178743961353,
+      "grad_norm": 0.1221630722284317,
+      "learning_rate": 3.3204633204633207e-05,
+      "loss": 0.0486,
       "step": 172
     },
     {
-      "epoch": 1.5908045977011493,
-      "grad_norm": 0.014891255646944046,
-      "learning_rate": 8.592205598713539e-05,
-      "loss": 0.0133,
+      "epoch": 0.8357487922705314,
+      "grad_norm": 0.1088973730802536,
+      "learning_rate": 3.33976833976834e-05,
+      "loss": 0.043,
       "step": 173
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.01469021663069725,
-      "learning_rate": 8.569648672789497e-05,
-      "loss": 0.0113,
+      "epoch": 0.8405797101449275,
+      "grad_norm": 0.11308977752923965,
+      "learning_rate": 3.359073359073359e-05,
+      "loss": 0.0478,
       "step": 174
     },
     {
-      "epoch": 1.6091954022988506,
-      "grad_norm": 0.02000586874783039,
-      "learning_rate": 8.546942587279465e-05,
-      "loss": 0.0138,
+      "epoch": 0.8454106280193237,
+      "grad_norm": 0.10720162838697433,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 0.0432,
       "step": 175
     },
     {
-      "epoch": 1.6091954022988506,
-      "eval_loss": 0.01319777499884367,
-      "eval_runtime": 7.6152,
-      "eval_samples_per_second": 6.566,
-      "eval_steps_per_second": 1.707,
+      "epoch": 0.8454106280193237,
+      "eval_loss": 0.04698673263192177,
+      "eval_runtime": 20.6051,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
       "step": 175
     },
     {
-      "epoch": 1.6183908045977011,
-      "grad_norm": 0.018022779375314713,
-      "learning_rate": 8.524088290968781e-05,
-      "loss": 0.0127,
+      "epoch": 0.8502415458937198,
+      "grad_norm": 0.1255665421485901,
+      "learning_rate": 3.397683397683398e-05,
+      "loss": 0.0444,
       "step": 176
     },
     {
-      "epoch": 1.6275862068965519,
-      "grad_norm": 0.01829381100833416,
-      "learning_rate": 8.501086738835843e-05,
-      "loss": 0.0131,
+      "epoch": 0.855072463768116,
+      "grad_norm": 0.108990378677845,
+      "learning_rate": 3.416988416988417e-05,
+      "loss": 0.0469,
       "step": 177
     },
     {
-      "epoch": 1.6367816091954022,
-      "grad_norm": 0.016279663890600204,
-      "learning_rate": 8.47793889201221e-05,
-      "loss": 0.0118,
+      "epoch": 0.8599033816425121,
+      "grad_norm": 0.10082966834306717,
+      "learning_rate": 3.436293436293436e-05,
+      "loss": 0.0508,
       "step": 178
     },
     {
-      "epoch": 1.645977011494253,
-      "grad_norm": 0.018548633903265,
-      "learning_rate": 8.45464571774244e-05,
-      "loss": 0.0149,
+      "epoch": 0.8647342995169082,
+      "grad_norm": 0.11663074791431427,
+      "learning_rate": 3.4555984555984555e-05,
+      "loss": 0.04,
       "step": 179
     },
     {
-      "epoch": 1.6551724137931034,
-      "grad_norm": 0.016604119911789894,
-      "learning_rate": 8.43120818934367e-05,
-      "loss": 0.0112,
-      "step": 180
-    },
-    {
-      "epoch": 1.6551724137931034,
-      "eval_loss": 0.013563334941864014,
-      "eval_runtime": 7.6136,
-      "eval_samples_per_second": 6.567,
-      "eval_steps_per_second": 1.707,
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.11602170765399933,
+      "learning_rate": 3.4749034749034754e-05,
+      "loss": 0.0382,
       "step": 180
     },
     {
-      "epoch": 1.664367816091954,
-      "grad_norm": 0.016890155151486397,
-      "learning_rate": 8.407627286164948e-05,
-      "loss": 0.0121,
+      "epoch": 0.8743961352657005,
+      "grad_norm": 0.10239856690168381,
+      "learning_rate": 3.4942084942084947e-05,
+      "loss": 0.0384,
       "step": 181
     },
     {
-      "epoch": 1.6735632183908047,
-      "grad_norm": 0.014763073064386845,
-      "learning_rate": 8.383903993546311e-05,
-      "loss": 0.0118,
+      "epoch": 0.8792270531400966,
+      "grad_norm": 0.1251702755689621,
+      "learning_rate": 3.513513513513514e-05,
+      "loss": 0.0557,
       "step": 182
     },
     {
-      "epoch": 1.6827586206896552,
-      "grad_norm": 0.016909824684262276,
-      "learning_rate": 8.360039302777612e-05,
-      "loss": 0.0123,
+      "epoch": 0.8840579710144928,
+      "grad_norm": 0.1245378628373146,
+      "learning_rate": 3.532818532818533e-05,
+      "loss": 0.0564,
       "step": 183
     },
     {
-      "epoch": 1.6919540229885057,
-      "grad_norm": 0.015340789221227169,
-      "learning_rate": 8.336034211057098e-05,
-      "loss": 0.0124,
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.11449820548295975,
+      "learning_rate": 3.5521235521235524e-05,
+      "loss": 0.0459,
       "step": 184
     },
     {
-      "epoch": 1.7011494252873565,
-      "grad_norm": 0.01562805287539959,
-      "learning_rate": 8.31188972144974e-05,
-      "loss": 0.0102,
-      "step": 185
-    },
-    {
-      "epoch": 1.7011494252873565,
-      "eval_loss": 0.013533576391637325,
-      "eval_runtime": 7.5972,
-      "eval_samples_per_second": 6.581,
-      "eval_steps_per_second": 1.711,
+      "epoch": 0.893719806763285,
+      "grad_norm": 0.11923693865537643,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.0382,
       "step": 185
     },
     {
-      "epoch": 1.7103448275862068,
-      "grad_norm": 0.0161032285541296,
-      "learning_rate": 8.28760684284532e-05,
-      "loss": 0.0104,
+      "epoch": 0.8985507246376812,
+      "grad_norm": 0.132978156208992,
+      "learning_rate": 3.590733590733591e-05,
+      "loss": 0.0572,
       "step": 186
     },
     {
-      "epoch": 1.7195402298850575,
-      "grad_norm": 0.017128756269812584,
-      "learning_rate": 8.263186589916273e-05,
-      "loss": 0.0129,
+      "epoch": 0.9033816425120773,
+      "grad_norm": 0.1146927997469902,
+      "learning_rate": 3.61003861003861e-05,
+      "loss": 0.034,
       "step": 187
     },
     {
-      "epoch": 1.728735632183908,
-      "grad_norm": 0.018114803358912468,
-      "learning_rate": 8.238629983075294e-05,
-      "loss": 0.0124,
+      "epoch": 0.9082125603864735,
+      "grad_norm": 0.1315777748823166,
+      "learning_rate": 3.6293436293436295e-05,
+      "loss": 0.0522,
       "step": 188
     },
     {
-      "epoch": 1.7379310344827585,
-      "grad_norm": 0.019191117957234383,
-      "learning_rate": 8.213938048432697e-05,
-      "loss": 0.0135,
+      "epoch": 0.9130434782608695,
+      "grad_norm": 0.10078331083059311,
+      "learning_rate": 3.648648648648649e-05,
+      "loss": 0.0361,
       "step": 189
     },
     {
-      "epoch": 1.7471264367816093,
-      "grad_norm": 0.01860605925321579,
-      "learning_rate": 8.18911181775353e-05,
-      "loss": 0.0135,
-      "step": 190
-    },
-    {
-      "epoch": 1.7471264367816093,
-      "eval_loss": 0.013264240697026253,
-      "eval_runtime": 7.5732,
-      "eval_samples_per_second": 6.602,
-      "eval_steps_per_second": 1.717,
+      "epoch": 0.9178743961352657,
+      "grad_norm": 0.12355741858482361,
+      "learning_rate": 3.667953667953668e-05,
+      "loss": 0.0568,
       "step": 190
     },
     {
-      "epoch": 1.7563218390804598,
-      "grad_norm": 0.016007546335458755,
-      "learning_rate": 8.164152328414476e-05,
-      "loss": 0.0118,
+      "epoch": 0.9227053140096618,
+      "grad_norm": 0.11051063239574432,
+      "learning_rate": 3.687258687258687e-05,
+      "loss": 0.053,
       "step": 191
     },
     {
-      "epoch": 1.7655172413793103,
-      "grad_norm": 0.017371613532304764,
-      "learning_rate": 8.139060623360493e-05,
-      "loss": 0.0126,
+      "epoch": 0.927536231884058,
+      "grad_norm": 0.12470270693302155,
+      "learning_rate": 3.7065637065637065e-05,
+      "loss": 0.0517,
       "step": 192
     },
     {
-      "epoch": 1.774712643678161,
-      "grad_norm": 0.017206458374857903,
-      "learning_rate": 8.113837751061246e-05,
-      "loss": 0.0103,
+      "epoch": 0.9323671497584541,
+      "grad_norm": 0.10947062075138092,
+      "learning_rate": 3.725868725868726e-05,
+      "loss": 0.037,
       "step": 193
     },
     {
-      "epoch": 1.7839080459770114,
-      "grad_norm": 0.016995050013065338,
-      "learning_rate": 8.088484765467286e-05,
-      "loss": 0.0111,
+      "epoch": 0.9371980676328503,
+      "grad_norm": 0.12240796536207199,
+      "learning_rate": 3.745173745173745e-05,
+      "loss": 0.0536,
       "step": 194
     },
     {
-      "epoch": 1.793103448275862,
-      "grad_norm": 0.013478915207087994,
-      "learning_rate": 8.063002725966015e-05,
-      "loss": 0.01,
-      "step": 195
-    },
-    {
-      "epoch": 1.793103448275862,
-      "eval_loss": 0.013534443452954292,
-      "eval_runtime": 7.5956,
-      "eval_samples_per_second": 6.583,
-      "eval_steps_per_second": 1.712,
+      "epoch": 0.9420289855072463,
+      "grad_norm": 0.12782225012779236,
+      "learning_rate": 3.764478764478765e-05,
+      "loss": 0.0519,
       "step": 195
     },
     {
-      "epoch": 1.8022988505747126,
-      "grad_norm": 0.021648980677127838,
-      "learning_rate": 8.037392697337418e-05,
-      "loss": 0.0143,
+      "epoch": 0.9468599033816425,
+      "grad_norm": 0.11264186352491379,
+      "learning_rate": 3.783783783783784e-05,
+      "loss": 0.043,
       "step": 196
     },
     {
-      "epoch": 1.8114942528735631,
-      "grad_norm": 0.015355809591710567,
-      "learning_rate": 8.011655749709575e-05,
-      "loss": 0.011,
+      "epoch": 0.9516908212560387,
+      "grad_norm": 0.10567909479141235,
+      "learning_rate": 3.8030888030888035e-05,
+      "loss": 0.0528,
       "step": 197
     },
     {
-      "epoch": 1.8206896551724139,
-      "grad_norm": 0.014543996192514896,
-      "learning_rate": 7.985792958513931e-05,
-      "loss": 0.0101,
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.11025624722242355,
+      "learning_rate": 3.822393822393823e-05,
+      "loss": 0.0423,
       "step": 198
     },
     {
-      "epoch": 1.8298850574712644,
-      "grad_norm": 0.01613461971282959,
-      "learning_rate": 7.95980540444038e-05,
-      "loss": 0.0124,
+      "epoch": 0.961352657004831,
+      "grad_norm": 0.12501060962677002,
+      "learning_rate": 3.841698841698842e-05,
+      "loss": 0.0466,
       "step": 199
     },
     {
-      "epoch": 1.839080459770115,
-      "grad_norm": 0.017040664330124855,
-      "learning_rate": 7.93369417339209e-05,
-      "loss": 0.0115,
+      "epoch": 0.966183574879227,
+      "grad_norm": 0.11633525043725967,
+      "learning_rate": 3.861003861003861e-05,
+      "loss": 0.0416,
       "step": 200
     },
     {
-      "epoch": 1.839080459770115,
-      "eval_loss": 0.01306515745818615,
-      "eval_runtime": 7.5974,
-      "eval_samples_per_second": 6.581,
-      "eval_steps_per_second": 1.711,
+      "epoch": 0.966183574879227,
+      "eval_loss": 0.04470054805278778,
+      "eval_runtime": 21.3093,
+      "eval_samples_per_second": 4.693,
+      "eval_steps_per_second": 0.141,
       "step": 200
     },
     {
-      "epoch": 1.8482758620689657,
-      "grad_norm": 0.01773899607360363,
-      "learning_rate": 7.907460356440133e-05,
-      "loss": 0.0133,
+      "epoch": 0.9710144927536232,
+      "grad_norm": 0.09972456842660904,
+      "learning_rate": 3.8803088803088805e-05,
+      "loss": 0.043,
       "step": 201
     },
     {
-      "epoch": 1.857471264367816,
-      "grad_norm": 0.012051481753587723,
-      "learning_rate": 7.881105049777901e-05,
-      "loss": 0.0093,
+      "epoch": 0.9758454106280193,
+      "grad_norm": 0.12026359885931015,
+      "learning_rate": 3.8996138996139e-05,
+      "loss": 0.0388,
       "step": 202
     },
     {
-      "epoch": 1.8666666666666667,
-      "grad_norm": 0.017261862754821777,
-      "learning_rate": 7.854629354675291e-05,
-      "loss": 0.0125,
+      "epoch": 0.9806763285024155,
+      "grad_norm": 0.12035425007343292,
+      "learning_rate": 3.918918918918919e-05,
+      "loss": 0.0609,
       "step": 203
     },
     {
-      "epoch": 1.8758620689655172,
-      "grad_norm": 0.01634056121110916,
-      "learning_rate": 7.828034377432693e-05,
-      "loss": 0.0111,
+      "epoch": 0.9855072463768116,
+      "grad_norm": 0.11870287358760834,
+      "learning_rate": 3.938223938223938e-05,
+      "loss": 0.0478,
       "step": 204
     },
     {
-      "epoch": 1.8850574712643677,
-      "grad_norm": 0.016423719003796577,
-      "learning_rate": 7.801321229334764e-05,
-      "loss": 0.0113,
-      "step": 205
-    },
-    {
-      "epoch": 1.8850574712643677,
-      "eval_loss": 0.012708608992397785,
-      "eval_runtime": 7.5908,
-      "eval_samples_per_second": 6.587,
-      "eval_steps_per_second": 1.713,
+      "epoch": 0.9903381642512077,
+      "grad_norm": 0.13712666928768158,
+      "learning_rate": 3.9575289575289576e-05,
+      "loss": 0.0497,
       "step": 205
     },
     {
-      "epoch": 1.8942528735632185,
-      "grad_norm": 0.015081810764968395,
-      "learning_rate": 7.774491026603985e-05,
-      "loss": 0.013,
+      "epoch": 0.9951690821256038,
+      "grad_norm": 0.1284405142068863,
+      "learning_rate": 3.976833976833977e-05,
+      "loss": 0.0457,
       "step": 206
     },
     {
-      "epoch": 1.903448275862069,
-      "grad_norm": 0.0127123286947608,
-      "learning_rate": 7.74754489035403e-05,
-      "loss": 0.0106,
+      "epoch": 1.0,
+      "grad_norm": 0.16791677474975586,
+      "learning_rate": 3.996138996138996e-05,
+      "loss": 0.0323,
       "step": 207
     },
     {
-      "epoch": 1.9126436781609195,
-      "grad_norm": 0.016098791733384132,
-      "learning_rate": 7.720483946542914e-05,
-      "loss": 0.0112,
+      "epoch": 1.0048309178743962,
+      "grad_norm": 0.15935100615024567,
+      "learning_rate": 4.015444015444015e-05,
+      "loss": 0.04,
       "step": 208
     },
     {
-      "epoch": 1.9218390804597703,
-      "grad_norm": 0.01502301823347807,
-      "learning_rate": 7.69330932592594e-05,
-      "loss": 0.0119,
+      "epoch": 1.0096618357487923,
+      "grad_norm": 0.09461764991283417,
+      "learning_rate": 4.034749034749035e-05,
+      "loss": 0.0431,
       "step": 209
     },
     {
-      "epoch": 1.9310344827586206,
-      "grad_norm": 0.016128163784742355,
-      "learning_rate": 7.666022164008457e-05,
-      "loss": 0.0107,
-      "step": 210
-    },
-    {
-      "epoch": 1.9310344827586206,
-      "eval_loss": 0.012785980477929115,
-      "eval_runtime": 7.5896,
-      "eval_samples_per_second": 6.588,
-      "eval_steps_per_second": 1.713,
+      "epoch": 1.0144927536231885,
+      "grad_norm": 0.15307606756687164,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 0.0393,
       "step": 210
     },
     {
-      "epoch": 1.9402298850574713,
-      "grad_norm": 0.014872719533741474,
-      "learning_rate": 7.63862360099841e-05,
-      "loss": 0.0106,
+      "epoch": 1.0193236714975846,
+      "grad_norm": 0.1187703013420105,
+      "learning_rate": 4.073359073359074e-05,
+      "loss": 0.0495,
       "step": 211
     },
     {
-      "epoch": 1.9494252873563218,
-      "grad_norm": 0.016305750235915184,
-      "learning_rate": 7.611114781758692e-05,
-      "loss": 0.0107,
+      "epoch": 1.0241545893719808,
+      "grad_norm": 0.14451833069324493,
+      "learning_rate": 4.092664092664093e-05,
+      "loss": 0.0375,
       "step": 212
     },
     {
-      "epoch": 1.9586206896551723,
-      "grad_norm": 0.014701612293720245,
-      "learning_rate": 7.583496855759316e-05,
-      "loss": 0.0124,
+      "epoch": 1.0289855072463767,
+      "grad_norm": 0.11559619009494781,
+      "learning_rate": 4.111969111969112e-05,
+      "loss": 0.0275,
       "step": 213
     },
     {
-      "epoch": 1.967816091954023,
-      "grad_norm": 0.0179608054459095,
-      "learning_rate": 7.555770977029367e-05,
-      "loss": 0.0115,
+      "epoch": 1.0338164251207729,
+      "grad_norm": 0.1330314427614212,
+      "learning_rate": 4.1312741312741316e-05,
+      "loss": 0.0515,
       "step": 214
     },
     {
-      "epoch": 1.9770114942528736,
-      "grad_norm": 0.016583839431405067,
-      "learning_rate": 7.527938304108795e-05,
-      "loss": 0.0122,
-      "step": 215
-    },
-    {
-      "epoch": 1.9770114942528736,
-      "eval_loss": 0.01276768371462822,
-      "eval_runtime": 7.6299,
-      "eval_samples_per_second": 6.553,
-      "eval_steps_per_second": 1.704,
+      "epoch": 1.038647342995169,
+      "grad_norm": 0.10810398310422897,
+      "learning_rate": 4.150579150579151e-05,
+      "loss": 0.0377,
       "step": 215
     },
     {
-      "epoch": 1.986206896551724,
-      "grad_norm": 0.014455633237957954,
-      "learning_rate": 7.500000000000001e-05,
-      "loss": 0.0109,
+      "epoch": 1.0434782608695652,
+      "grad_norm": 0.11305640637874603,
+      "learning_rate": 4.16988416988417e-05,
+      "loss": 0.0362,
       "step": 216
     },
     {
-      "epoch": 1.9954022988505749,
-      "grad_norm": 0.014771837741136551,
-      "learning_rate": 7.471957232119234e-05,
-      "loss": 0.0114,
+      "epoch": 1.0483091787439613,
+      "grad_norm": 0.10515987873077393,
+      "learning_rate": 4.189189189189189e-05,
+      "loss": 0.0379,
       "step": 217
     },
     {
-      "epoch": 2.004597701149425,
-      "grad_norm": 0.019322361797094345,
-      "learning_rate": 7.443811172247821e-05,
-      "loss": 0.0123,
+      "epoch": 1.0531400966183575,
+      "grad_norm": 0.11391875147819519,
+      "learning_rate": 4.2084942084942086e-05,
+      "loss": 0.0441,
       "step": 218
     },
     {
-      "epoch": 2.013793103448276,
-      "grad_norm": 0.011849606409668922,
-      "learning_rate": 7.415562996483192e-05,
-      "loss": 0.0081,
+      "epoch": 1.0579710144927537,
+      "grad_norm": 0.1047784611582756,
+      "learning_rate": 4.227799227799228e-05,
+      "loss": 0.046,
       "step": 219
     },
     {
-      "epoch": 2.0229885057471266,
-      "grad_norm": 0.01762513443827629,
-      "learning_rate": 7.387213885189746e-05,
-      "loss": 0.0099,
-      "step": 220
-    },
-    {
-      "epoch": 2.0229885057471266,
-      "eval_loss": 0.012783367186784744,
-      "eval_runtime": 7.6029,
-      "eval_samples_per_second": 6.576,
-      "eval_steps_per_second": 1.71,
+      "epoch": 1.0628019323671498,
+      "grad_norm": 0.13552096486091614,
+      "learning_rate": 4.247104247104247e-05,
+      "loss": 0.0638,
       "step": 220
     },
     {
-      "epoch": 2.032183908045977,
-      "grad_norm": 0.0132211335003376,
-      "learning_rate": 7.358765022949519e-05,
-      "loss": 0.0088,
+      "epoch": 1.067632850241546,
+      "grad_norm": 0.09388095885515213,
+      "learning_rate": 4.2664092664092664e-05,
+      "loss": 0.0558,
       "step": 221
     },
     {
-      "epoch": 2.0413793103448277,
-      "grad_norm": 0.01868424378335476,
-      "learning_rate": 7.330217598512695e-05,
-      "loss": 0.0112,
+      "epoch": 1.0724637681159421,
+      "grad_norm": 0.14217819273471832,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.0377,
       "step": 222
     },
     {
-      "epoch": 2.050574712643678,
-      "grad_norm": 0.016267532482743263,
-      "learning_rate": 7.30157280474793e-05,
-      "loss": 0.0092,
+      "epoch": 1.077294685990338,
+      "grad_norm": 0.0974343866109848,
+      "learning_rate": 4.305019305019305e-05,
+      "loss": 0.0531,
       "step": 223
     },
     {
-      "epoch": 2.0597701149425287,
-      "grad_norm": 0.014060255140066147,
-      "learning_rate": 7.272831838592503e-05,
-      "loss": 0.0091,
+      "epoch": 1.0821256038647342,
+      "grad_norm": 0.10802186280488968,
+      "learning_rate": 4.324324324324325e-05,
+      "loss": 0.0441,
       "step": 224
     },
     {
-      "epoch": 2.0689655172413794,
-      "grad_norm": 0.017770465463399887,
-      "learning_rate": 7.243995901002312e-05,
-      "loss": 0.0121,
+      "epoch": 1.0869565217391304,
+      "grad_norm": 0.11783885210752487,
+      "learning_rate": 4.343629343629344e-05,
+      "loss": 0.0441,
       "step": 225
     },
     {
-      "epoch": 2.0689655172413794,
-      "eval_loss": 0.012876266613602638,
-      "eval_runtime": 7.5829,
-      "eval_samples_per_second": 6.594,
-      "eval_steps_per_second": 1.714,
+      "epoch": 1.0869565217391304,
+      "eval_loss": 0.042829480022192,
+      "eval_runtime": 20.6021,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
       "step": 225
     },
     {
-      "epoch": 2.0781609195402297,
-      "grad_norm": 0.01353057473897934,
-      "learning_rate": 7.215066196901676e-05,
-      "loss": 0.0095,
+      "epoch": 1.0917874396135265,
+      "grad_norm": 0.09211745113134384,
+      "learning_rate": 4.3629343629343633e-05,
+      "loss": 0.0354,
       "step": 226
     },
     {
-      "epoch": 2.0873563218390805,
-      "grad_norm": 0.01702524535357952,
-      "learning_rate": 7.186043935133005e-05,
-      "loss": 0.0095,
+      "epoch": 1.0966183574879227,
+      "grad_norm": 0.10802285373210907,
+      "learning_rate": 4.3822393822393826e-05,
+      "loss": 0.0463,
       "step": 227
     },
     {
-      "epoch": 2.0965517241379312,
-      "grad_norm": 0.019207211211323738,
-      "learning_rate": 7.156930328406268e-05,
-      "loss": 0.0099,
+      "epoch": 1.1014492753623188,
+      "grad_norm": 0.11975615471601486,
+      "learning_rate": 4.401544401544402e-05,
+      "loss": 0.0358,
       "step": 228
     },
     {
-      "epoch": 2.1057471264367815,
-      "grad_norm": 0.017192739993333817,
-      "learning_rate": 7.127726593248337e-05,
-      "loss": 0.0108,
+      "epoch": 1.106280193236715,
+      "grad_norm": 0.10383973270654678,
+      "learning_rate": 4.420849420849421e-05,
+      "loss": 0.0513,
       "step": 229
     },
     {
-      "epoch": 2.1149425287356323,
-      "grad_norm": 0.01726214587688446,
-      "learning_rate": 7.098433949952146e-05,
-      "loss": 0.0103,
-      "step": 230
-    },
-    {
-      "epoch": 2.1149425287356323,
-      "eval_loss": 0.012764949351549149,
-      "eval_runtime": 7.5742,
-      "eval_samples_per_second": 6.601,
-      "eval_steps_per_second": 1.716,
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.09274217486381531,
+      "learning_rate": 4.4401544401544404e-05,
+      "loss": 0.0391,
       "step": 230
     },
     {
-      "epoch": 2.1241379310344826,
-      "grad_norm": 0.016892077401280403,
-      "learning_rate": 7.069053622525696e-05,
-      "loss": 0.0095,
+      "epoch": 1.1159420289855073,
+      "grad_norm": 0.13271774351596832,
+      "learning_rate": 4.4594594594594596e-05,
+      "loss": 0.0373,
       "step": 231
     },
     {
-      "epoch": 2.1333333333333333,
-      "grad_norm": 0.01737057976424694,
-      "learning_rate": 7.039586838640919e-05,
-      "loss": 0.0101,
+      "epoch": 1.1207729468599035,
+      "grad_norm": 0.12627987563610077,
+      "learning_rate": 4.478764478764479e-05,
+      "loss": 0.0436,
       "step": 232
     },
     {
-      "epoch": 2.142528735632184,
-      "grad_norm": 0.017956653609871864,
-      "learning_rate": 7.01003482958237e-05,
-      "loss": 0.009,
+      "epoch": 1.1256038647342996,
+      "grad_norm": 0.1234574168920517,
+      "learning_rate": 4.498069498069498e-05,
+      "loss": 0.0389,
       "step": 233
     },
     {
-      "epoch": 2.1517241379310343,
-      "grad_norm": 0.018064096570014954,
-      "learning_rate": 6.980398830195785e-05,
-      "loss": 0.0115,
+      "epoch": 1.1304347826086956,
+      "grad_norm": 0.1091989129781723,
+      "learning_rate": 4.5173745173745174e-05,
+      "loss": 0.0409,
       "step": 234
     },
     {
-      "epoch": 2.160919540229885,
-      "grad_norm": 0.017505567520856857,
-      "learning_rate": 6.950680078836474e-05,
-      "loss": 0.01,
-      "step": 235
-    },
-    {
-      "epoch": 2.160919540229885,
-      "eval_loss": 0.012695999816060066,
-      "eval_runtime": 7.6048,
-      "eval_samples_per_second": 6.575,
-      "eval_steps_per_second": 1.709,
+      "epoch": 1.1352657004830917,
+      "grad_norm": 0.10395840555429459,
+      "learning_rate": 4.536679536679537e-05,
+      "loss": 0.0357,
       "step": 235
     },
     {
-      "epoch": 2.170114942528736,
-      "grad_norm": 0.01884928159415722,
-      "learning_rate": 6.920879817317589e-05,
-      "loss": 0.0108,
+      "epoch": 1.1400966183574879,
+      "grad_norm": 0.10336270928382874,
+      "learning_rate": 4.555984555984556e-05,
+      "loss": 0.0431,
       "step": 236
     },
     {
-      "epoch": 2.179310344827586,
-      "grad_norm": 0.01585654355585575,
-      "learning_rate": 6.890999290858214e-05,
-      "loss": 0.0089,
+      "epoch": 1.144927536231884,
+      "grad_norm": 0.09441915899515152,
+      "learning_rate": 4.575289575289575e-05,
+      "loss": 0.0346,
       "step": 237
     },
     {
-      "epoch": 2.188505747126437,
-      "grad_norm": 0.015242693945765495,
-      "learning_rate": 6.861039748031351e-05,
-      "loss": 0.0094,
+      "epoch": 1.1497584541062802,
+      "grad_norm": 0.11620728671550751,
+      "learning_rate": 4.594594594594595e-05,
+      "loss": 0.0436,
       "step": 238
     },
     {
-      "epoch": 2.197701149425287,
-      "grad_norm": 0.01854494772851467,
-      "learning_rate": 6.83100244071174e-05,
-      "loss": 0.0099,
+      "epoch": 1.1545893719806763,
+      "grad_norm": 0.09966114163398743,
+      "learning_rate": 4.6138996138996144e-05,
+      "loss": 0.0377,
       "step": 239
     },
     {
-      "epoch": 2.206896551724138,
-      "grad_norm": 0.016470277681946754,
-      "learning_rate": 6.800888624023553e-05,
-      "loss": 0.0089,
-      "step": 240
-    },
-    {
-      "epoch": 2.206896551724138,
-      "eval_loss": 0.01267631258815527,
-      "eval_runtime": 7.5818,
-      "eval_samples_per_second": 6.595,
-      "eval_steps_per_second": 1.715,
+      "epoch": 1.1594202898550725,
+      "grad_norm": 0.07768991589546204,
+      "learning_rate": 4.6332046332046336e-05,
+      "loss": 0.0328,
       "step": 240
     },
     {
-      "epoch": 2.2160919540229886,
-      "grad_norm": 0.017559630796313286,
-      "learning_rate": 6.770699556287939e-05,
-      "loss": 0.0105,
+      "epoch": 1.1642512077294687,
+      "grad_norm": 0.09339828789234161,
+      "learning_rate": 4.652509652509653e-05,
+      "loss": 0.0363,
       "step": 241
     },
     {
-      "epoch": 2.225287356321839,
-      "grad_norm": 0.016117800027132034,
-      "learning_rate": 6.740436498970452e-05,
-      "loss": 0.0099,
+      "epoch": 1.1690821256038648,
+      "grad_norm": 0.11105544120073318,
+      "learning_rate": 4.671814671814672e-05,
+      "loss": 0.0392,
       "step": 242
     },
     {
-      "epoch": 2.2344827586206897,
-      "grad_norm": 0.01739688403904438,
-      "learning_rate": 6.710100716628344e-05,
-      "loss": 0.0099,
+      "epoch": 1.1739130434782608,
+      "grad_norm": 0.09607312828302383,
+      "learning_rate": 4.6911196911196914e-05,
+      "loss": 0.0404,
       "step": 243
     },
     {
-      "epoch": 2.2436781609195404,
-      "grad_norm": 0.01701669581234455,
-      "learning_rate": 6.679693476857711e-05,
-      "loss": 0.0085,
+      "epoch": 1.178743961352657,
+      "grad_norm": 0.1227753683924675,
+      "learning_rate": 4.710424710424711e-05,
+      "loss": 0.035,
       "step": 244
     },
     {
-      "epoch": 2.2528735632183907,
-      "grad_norm": 0.017626378685235977,
-      "learning_rate": 6.649216050240539e-05,
-      "loss": 0.0089,
-      "step": 245
-    },
-    {
-      "epoch": 2.2528735632183907,
-      "eval_loss": 0.012675927020609379,
-      "eval_runtime": 7.5971,
-      "eval_samples_per_second": 6.581,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.183574879227053,
+      "grad_norm": 0.09502430260181427,
+      "learning_rate": 4.72972972972973e-05,
+      "loss": 0.0381,
       "step": 245
     },
     {
-      "epoch": 2.2620689655172415,
-      "grad_norm": 0.0238157007843256,
-      "learning_rate": 6.618669710291606e-05,
-      "loss": 0.0092,
+      "epoch": 1.1884057971014492,
+      "grad_norm": 0.10500189661979675,
+      "learning_rate": 4.749034749034749e-05,
+      "loss": 0.0441,
       "step": 246
     },
     {
-      "epoch": 2.2712643678160918,
-      "grad_norm": 0.017967745661735535,
-      "learning_rate": 6.588055733405266e-05,
-      "loss": 0.0102,
+      "epoch": 1.1932367149758454,
+      "grad_norm": 0.10394671559333801,
+      "learning_rate": 4.7683397683397685e-05,
+      "loss": 0.0434,
       "step": 247
     },
     {
-      "epoch": 2.2804597701149425,
-      "grad_norm": 0.015404468402266502,
-      "learning_rate": 6.557375398802123e-05,
-      "loss": 0.0095,
+      "epoch": 1.1980676328502415,
+      "grad_norm": 0.14034156501293182,
+      "learning_rate": 4.787644787644788e-05,
+      "loss": 0.067,
       "step": 248
     },
     {
-      "epoch": 2.2896551724137932,
-      "grad_norm": 0.015446755103766918,
-      "learning_rate": 6.526629988475567e-05,
-      "loss": 0.0092,
+      "epoch": 1.2028985507246377,
+      "grad_norm": 0.11108700186014175,
+      "learning_rate": 4.806949806949807e-05,
+      "loss": 0.0399,
       "step": 249
     },
     {
-      "epoch": 2.2988505747126435,
-      "grad_norm": 0.016809450462460518,
-      "learning_rate": 6.495820787138209e-05,
-      "loss": 0.0105,
+      "epoch": 1.2077294685990339,
+      "grad_norm": 0.09620919078588486,
+      "learning_rate": 4.826254826254826e-05,
+      "loss": 0.0348,
       "step": 250
     },
     {
-      "epoch": 2.2988505747126435,
-      "eval_loss": 0.012465451844036579,
-      "eval_runtime": 7.6011,
-      "eval_samples_per_second": 6.578,
-      "eval_steps_per_second": 1.71,
+      "epoch": 1.2077294685990339,
+      "eval_loss": 0.040457550436258316,
+      "eval_runtime": 20.5825,
+      "eval_samples_per_second": 4.859,
+      "eval_steps_per_second": 0.146,
       "step": 250
     },
     {
-      "epoch": 2.3080459770114943,
-      "grad_norm": 0.015970921143889427,
-      "learning_rate": 6.464949082168204e-05,
-      "loss": 0.0093,
+      "epoch": 1.21256038647343,
+      "grad_norm": 0.09701947867870331,
+      "learning_rate": 4.8455598455598455e-05,
+      "loss": 0.0364,
       "step": 251
     },
     {
-      "epoch": 2.317241379310345,
-      "grad_norm": 0.017121300101280212,
-      "learning_rate": 6.434016163555452e-05,
-      "loss": 0.0102,
+      "epoch": 1.2173913043478262,
+      "grad_norm": 0.09098837524652481,
+      "learning_rate": 4.8648648648648654e-05,
+      "loss": 0.0315,
       "step": 252
     },
     {
-      "epoch": 2.3264367816091953,
-      "grad_norm": 0.01726561225950718,
-      "learning_rate": 6.403023323847695e-05,
-      "loss": 0.0093,
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.11047874391078949,
+      "learning_rate": 4.884169884169885e-05,
+      "loss": 0.0293,
       "step": 253
     },
     {
-      "epoch": 2.335632183908046,
-      "grad_norm": 0.014097567647695541,
-      "learning_rate": 6.371971858096508e-05,
-      "loss": 0.0096,
+      "epoch": 1.2270531400966185,
+      "grad_norm": 0.12440991401672363,
+      "learning_rate": 4.903474903474904e-05,
+      "loss": 0.0399,
       "step": 254
     },
     {
-      "epoch": 2.344827586206897,
-      "grad_norm": 0.015593893826007843,
-      "learning_rate": 6.340863063803188e-05,
-      "loss": 0.0093,
-      "step": 255
-    },
-    {
-      "epoch": 2.344827586206897,
-      "eval_loss": 0.012436559423804283,
-      "eval_runtime": 7.5957,
-      "eval_samples_per_second": 6.583,
-      "eval_steps_per_second": 1.712,
+      "epoch": 1.2318840579710144,
+      "grad_norm": 0.11009038239717484,
+      "learning_rate": 4.922779922779923e-05,
+      "loss": 0.0435,
       "step": 255
     },
     {
-      "epoch": 2.354022988505747,
-      "grad_norm": 0.01714129000902176,
-      "learning_rate": 6.30969824086453e-05,
-      "loss": 0.0107,
+      "epoch": 1.2367149758454106,
+      "grad_norm": 0.09897538274526596,
+      "learning_rate": 4.9420849420849425e-05,
+      "loss": 0.0402,
       "step": 256
     },
     {
-      "epoch": 2.363218390804598,
-      "grad_norm": 0.018002189695835114,
-      "learning_rate": 6.27847869151852e-05,
-      "loss": 0.0103,
+      "epoch": 1.2415458937198067,
+      "grad_norm": 0.10943715274333954,
+      "learning_rate": 4.961389961389962e-05,
+      "loss": 0.0365,
       "step": 257
     },
     {
-      "epoch": 2.372413793103448,
-      "grad_norm": 0.015969395637512207,
-      "learning_rate": 6.247205720289907e-05,
-      "loss": 0.0099,
+      "epoch": 1.2463768115942029,
+      "grad_norm": 0.11430627107620239,
+      "learning_rate": 4.980694980694981e-05,
+      "loss": 0.0395,
       "step": 258
     },
     {
-      "epoch": 2.381609195402299,
-      "grad_norm": 0.020000826567411423,
-      "learning_rate": 6.215880633935708e-05,
-      "loss": 0.0105,
+      "epoch": 1.251207729468599,
+      "grad_norm": 0.1264277547597885,
+      "learning_rate": 5e-05,
+      "loss": 0.0361,
       "step": 259
     },
     {
-      "epoch": 2.3908045977011496,
-      "grad_norm": 0.0170020442456007,
-      "learning_rate": 6.184504741390596e-05,
-      "loss": 0.0097,
-      "step": 260
-    },
-    {
-      "epoch": 2.3908045977011496,
-      "eval_loss": 0.01255771517753601,
-      "eval_runtime": 7.5967,
-      "eval_samples_per_second": 6.582,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.2560386473429952,
+      "grad_norm": 0.10544534027576447,
+      "learning_rate": 5.0193050193050195e-05,
+      "loss": 0.0316,
       "step": 260
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 0.018086500465869904,
-      "learning_rate": 6.153079353712201e-05,
-      "loss": 0.0116,
+      "epoch": 1.2608695652173914,
+      "grad_norm": 0.08700928092002869,
+      "learning_rate": 5.038610038610039e-05,
+      "loss": 0.0335,
       "step": 261
     },
     {
-      "epoch": 2.4091954022988507,
-      "grad_norm": 0.017995236441493034,
-      "learning_rate": 6.121605784026339e-05,
-      "loss": 0.0098,
+      "epoch": 1.2657004830917875,
+      "grad_norm": 0.10566674917936325,
+      "learning_rate": 5.057915057915058e-05,
+      "loss": 0.0362,
       "step": 262
     },
     {
-      "epoch": 2.418390804597701,
-      "grad_norm": 0.016703004017472267,
-      "learning_rate": 6.09008534747213e-05,
-      "loss": 0.0085,
+      "epoch": 1.2705314009661834,
+      "grad_norm": 0.10313791781663895,
+      "learning_rate": 5.077220077220077e-05,
+      "loss": 0.0457,
       "step": 263
     },
     {
-      "epoch": 2.4275862068965517,
-      "grad_norm": 0.01667882315814495,
-      "learning_rate": 6.058519361147055e-05,
-      "loss": 0.0096,
+      "epoch": 1.2753623188405796,
+      "grad_norm": 0.10589325428009033,
+      "learning_rate": 5.0965250965250965e-05,
+      "loss": 0.033,
       "step": 264
     },
     {
-      "epoch": 2.4367816091954024,
-      "grad_norm": 0.014494238421320915,
-      "learning_rate": 6.02690914405191e-05,
-      "loss": 0.0091,
-      "step": 265
-    },
-    {
-      "epoch": 2.4367816091954024,
-      "eval_loss": 0.0126058179885149,
-      "eval_runtime": 7.5992,
-      "eval_samples_per_second": 6.58,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.2801932367149758,
+      "grad_norm": 0.1091860756278038,
+      "learning_rate": 5.115830115830116e-05,
+      "loss": 0.0408,
       "step": 265
     },
     {
-      "epoch": 2.4459770114942527,
-      "grad_norm": 0.018936868757009506,
-      "learning_rate": 5.995256017035703e-05,
-      "loss": 0.0102,
+      "epoch": 1.285024154589372,
+      "grad_norm": 0.11298917233943939,
+      "learning_rate": 5.135135135135135e-05,
+      "loss": 0.0354,
       "step": 266
     },
     {
-      "epoch": 2.4551724137931035,
-      "grad_norm": 0.016619672998785973,
-      "learning_rate": 5.963561302740449e-05,
-      "loss": 0.0091,
+      "epoch": 1.289855072463768,
+      "grad_norm": 0.09899977594614029,
+      "learning_rate": 5.154440154440154e-05,
+      "loss": 0.0441,
       "step": 267
     },
     {
-      "epoch": 2.464367816091954,
-      "grad_norm": 0.01624908670783043,
-      "learning_rate": 5.9318263255459116e-05,
-      "loss": 0.0097,
+      "epoch": 1.2946859903381642,
+      "grad_norm": 0.09906386584043503,
+      "learning_rate": 5.1737451737451736e-05,
+      "loss": 0.043,
       "step": 268
     },
     {
-      "epoch": 2.4735632183908045,
-      "grad_norm": 0.015271931886672974,
-      "learning_rate": 5.900052411514257e-05,
-      "loss": 0.0095,
+      "epoch": 1.2995169082125604,
+      "grad_norm": 0.09934026002883911,
+      "learning_rate": 5.193050193050193e-05,
+      "loss": 0.0307,
       "step": 269
     },
     {
-      "epoch": 2.4827586206896552,
-      "grad_norm": 0.01610173098742962,
-      "learning_rate": 5.868240888334653e-05,
-      "loss": 0.0095,
-      "step": 270
-    },
-    {
-      "epoch": 2.4827586206896552,
-      "eval_loss": 0.01236729696393013,
-      "eval_runtime": 7.5958,
-      "eval_samples_per_second": 6.583,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.3043478260869565,
+      "grad_norm": 0.09088680893182755,
+      "learning_rate": 5.212355212355212e-05,
+      "loss": 0.036,
       "step": 270
     },
     {
-      "epoch": 2.491954022988506,
-      "grad_norm": 0.013876480981707573,
-      "learning_rate": 5.836393085267776e-05,
-      "loss": 0.0091,
+      "epoch": 1.3091787439613527,
+      "grad_norm": 0.10093280673027039,
+      "learning_rate": 5.2316602316602313e-05,
+      "loss": 0.0361,
       "step": 271
     },
     {
-      "epoch": 2.5011494252873563,
-      "grad_norm": 0.014515344053506851,
-      "learning_rate": 5.804510333090287e-05,
-      "loss": 0.0087,
+      "epoch": 1.3140096618357489,
+      "grad_norm": 0.12925419211387634,
+      "learning_rate": 5.2509652509652506e-05,
+      "loss": 0.0415,
       "step": 272
     },
     {
-      "epoch": 2.510344827586207,
-      "grad_norm": 0.015616660006344318,
-      "learning_rate": 5.772593964039203e-05,
-      "loss": 0.0088,
+      "epoch": 1.318840579710145,
+      "grad_norm": 0.10835471749305725,
+      "learning_rate": 5.27027027027027e-05,
+      "loss": 0.0412,
       "step": 273
     },
     {
-      "epoch": 2.5195402298850573,
-      "grad_norm": 0.016088545322418213,
-      "learning_rate": 5.740645311756245e-05,
-      "loss": 0.0095,
+      "epoch": 1.3236714975845412,
+      "grad_norm": 0.12303876876831055,
+      "learning_rate": 5.2895752895752905e-05,
+      "loss": 0.0489,
       "step": 274
     },
     {
-      "epoch": 2.528735632183908,
-      "grad_norm": 0.014027421362698078,
-      "learning_rate": 5.708665711232103e-05,
-      "loss": 0.0094,
+      "epoch": 1.3285024154589373,
+      "grad_norm": 0.09909823536872864,
+      "learning_rate": 5.30888030888031e-05,
+      "loss": 0.0355,
       "step": 275
     },
     {
-      "epoch": 2.528735632183908,
-      "eval_loss": 0.012253012508153915,
-      "eval_runtime": 7.5717,
-      "eval_samples_per_second": 6.604,
-      "eval_steps_per_second": 1.717,
+      "epoch": 1.3285024154589373,
+      "eval_loss": 0.03857467696070671,
+      "eval_runtime": 20.5983,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
       "step": 275
     },
     {
-      "epoch": 2.5379310344827584,
-      "grad_norm": 0.015265434980392456,
-      "learning_rate": 5.6766564987506566e-05,
-      "loss": 0.0102,
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.12137026339769363,
+      "learning_rate": 5.328185328185329e-05,
+      "loss": 0.0376,
       "step": 276
     },
     {
-      "epoch": 2.547126436781609,
-      "grad_norm": 0.01524539664387703,
-      "learning_rate": 5.644619011833133e-05,
-      "loss": 0.0089,
+      "epoch": 1.3381642512077294,
+      "grad_norm": 0.10434843599796295,
+      "learning_rate": 5.347490347490348e-05,
+      "loss": 0.0453,
       "step": 277
     },
     {
-      "epoch": 2.55632183908046,
-      "grad_norm": 0.01515972800552845,
-      "learning_rate": 5.6125545891822274e-05,
-      "loss": 0.009,
+      "epoch": 1.3429951690821256,
+      "grad_norm": 0.07658998668193817,
+      "learning_rate": 5.3667953667953675e-05,
+      "loss": 0.0256,
       "step": 278
     },
     {
-      "epoch": 2.56551724137931,
-      "grad_norm": 0.017095813527703285,
-      "learning_rate": 5.5804645706261514e-05,
-      "loss": 0.0103,
+      "epoch": 1.3478260869565217,
+      "grad_norm": 0.09582667052745819,
+      "learning_rate": 5.386100386100387e-05,
+      "loss": 0.0308,
       "step": 279
     },
     {
-      "epoch": 2.574712643678161,
-      "grad_norm": 0.014950582757592201,
-      "learning_rate": 5.548350297062659e-05,
-      "loss": 0.0092,
-      "step": 280
-    },
-    {
-      "epoch": 2.574712643678161,
-      "eval_loss": 0.011931383982300758,
-      "eval_runtime": 7.6218,
-      "eval_samples_per_second": 6.56,
-      "eval_steps_per_second": 1.706,
+      "epoch": 1.3526570048309179,
+      "grad_norm": 0.09777297079563141,
+      "learning_rate": 5.405405405405406e-05,
+      "loss": 0.0343,
       "step": 280
     },
     {
-      "epoch": 2.5839080459770116,
-      "grad_norm": 0.015228749252855778,
-      "learning_rate": 5.516213110403009e-05,
-      "loss": 0.0093,
+      "epoch": 1.357487922705314,
+      "grad_norm": 0.09060278534889221,
+      "learning_rate": 5.424710424710425e-05,
+      "loss": 0.0385,
       "step": 281
     },
     {
-      "epoch": 2.593103448275862,
-      "grad_norm": 0.020853353664278984,
-      "learning_rate": 5.484054353515896e-05,
-      "loss": 0.0116,
+      "epoch": 1.3623188405797102,
+      "grad_norm": 0.09172627329826355,
+      "learning_rate": 5.4440154440154445e-05,
+      "loss": 0.037,
       "step": 282
     },
     {
-      "epoch": 2.6022988505747127,
-      "grad_norm": 0.013651091605424881,
-      "learning_rate": 5.451875370171341e-05,
-      "loss": 0.0076,
+      "epoch": 1.3671497584541064,
+      "grad_norm": 0.10294190049171448,
+      "learning_rate": 5.463320463320464e-05,
+      "loss": 0.0386,
       "step": 283
     },
     {
-      "epoch": 2.6114942528735634,
-      "grad_norm": 0.015777967870235443,
-      "learning_rate": 5.419677504984534e-05,
-      "loss": 0.0088,
+      "epoch": 1.3719806763285023,
+      "grad_norm": 0.10812503099441528,
+      "learning_rate": 5.482625482625483e-05,
+      "loss": 0.045,
       "step": 284
     },
     {
-      "epoch": 2.6206896551724137,
-      "grad_norm": 0.014214356429874897,
-      "learning_rate": 5.387462103359655e-05,
-      "loss": 0.0084,
-      "step": 285
-    },
-    {
-      "epoch": 2.6206896551724137,
-      "eval_loss": 0.012128188274800777,
-      "eval_runtime": 7.6207,
-      "eval_samples_per_second": 6.561,
-      "eval_steps_per_second": 1.706,
+      "epoch": 1.3768115942028984,
+      "grad_norm": 0.07604746520519257,
+      "learning_rate": 5.501930501930502e-05,
+      "loss": 0.0308,
       "step": 285
     },
     {
-      "epoch": 2.6298850574712644,
-      "grad_norm": 0.02143423818051815,
-      "learning_rate": 5.355230511433651e-05,
-      "loss": 0.0109,
+      "epoch": 1.3816425120772946,
+      "grad_norm": 0.10711754858493805,
+      "learning_rate": 5.5212355212355216e-05,
+      "loss": 0.039,
       "step": 286
     },
     {
-      "epoch": 2.639080459770115,
-      "grad_norm": 0.016306888312101364,
-      "learning_rate": 5.32298407601999e-05,
-      "loss": 0.009,
+      "epoch": 1.3864734299516908,
+      "grad_norm": 0.14930835366249084,
+      "learning_rate": 5.540540540540541e-05,
+      "loss": 0.0309,
       "step": 287
     },
     {
-      "epoch": 2.6482758620689655,
-      "grad_norm": 0.015416650101542473,
-      "learning_rate": 5.290724144552379e-05,
-      "loss": 0.0091,
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.10966020077466965,
+      "learning_rate": 5.55984555984556e-05,
+      "loss": 0.0384,
       "step": 288
     },
     {
-      "epoch": 2.657471264367816,
-      "grad_norm": 0.018417850136756897,
-      "learning_rate": 5.258452065028473e-05,
-      "loss": 0.0096,
+      "epoch": 1.396135265700483,
+      "grad_norm": 0.10118502378463745,
+      "learning_rate": 5.5791505791505794e-05,
+      "loss": 0.0387,
       "step": 289
     },
     {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 0.016621043905615807,
-      "learning_rate": 5.226169185953532e-05,
-      "loss": 0.0098,
-      "step": 290
-    },
-    {
-      "epoch": 2.6666666666666665,
-      "eval_loss": 0.01195099763572216,
-      "eval_runtime": 7.5972,
-      "eval_samples_per_second": 6.581,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.4009661835748792,
+      "grad_norm": 0.10067062824964523,
+      "learning_rate": 5.5984555984555986e-05,
+      "loss": 0.0401,
       "step": 290
     },
     {
-      "epoch": 2.6758620689655173,
-      "grad_norm": 0.016165705397725105,
-      "learning_rate": 5.193876856284085e-05,
-      "loss": 0.0088,
+      "epoch": 1.4057971014492754,
+      "grad_norm": 0.08439848572015762,
+      "learning_rate": 5.617760617760618e-05,
+      "loss": 0.0307,
       "step": 291
     },
     {
-      "epoch": 2.6850574712643676,
-      "grad_norm": 0.015073074027895927,
-      "learning_rate": 5.1615764253715536e-05,
-      "loss": 0.0089,
+      "epoch": 1.4106280193236715,
+      "grad_norm": 0.09646830707788467,
+      "learning_rate": 5.637065637065637e-05,
+      "loss": 0.0391,
       "step": 292
     },
     {
-      "epoch": 2.6942528735632183,
-      "grad_norm": 0.019986115396022797,
-      "learning_rate": 5.129269242905882e-05,
-      "loss": 0.0104,
+      "epoch": 1.4154589371980677,
+      "grad_norm": 0.09401887655258179,
+      "learning_rate": 5.6563706563706564e-05,
+      "loss": 0.0384,
       "step": 293
     },
     {
-      "epoch": 2.703448275862069,
-      "grad_norm": 0.01431217696517706,
-      "learning_rate": 5.096956658859122e-05,
-      "loss": 0.0087,
+      "epoch": 1.4202898550724639,
+      "grad_norm": 0.10123006999492645,
+      "learning_rate": 5.6756756756756757e-05,
+      "loss": 0.038,
       "step": 294
     },
     {
-      "epoch": 2.7126436781609193,
-      "grad_norm": 0.0174038615077734,
-      "learning_rate": 5.064640023429043e-05,
-      "loss": 0.0097,
-      "step": 295
-    },
-    {
-      "epoch": 2.7126436781609193,
-      "eval_loss": 0.012155573815107346,
-      "eval_runtime": 7.6146,
-      "eval_samples_per_second": 6.566,
-      "eval_steps_per_second": 1.707,
+      "epoch": 1.42512077294686,
+      "grad_norm": 0.09332787990570068,
+      "learning_rate": 5.694980694980695e-05,
+      "loss": 0.0349,
       "step": 295
     },
     {
-      "epoch": 2.72183908045977,
-      "grad_norm": 0.01756209135055542,
-      "learning_rate": 5.0323206869826966e-05,
-      "loss": 0.0095,
+      "epoch": 1.4299516908212562,
+      "grad_norm": 0.07685629278421402,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 0.0322,
       "step": 296
     },
     {
-      "epoch": 2.731034482758621,
-      "grad_norm": 0.018789948895573616,
-      "learning_rate": 5e-05,
-      "loss": 0.0094,
+      "epoch": 1.434782608695652,
+      "grad_norm": 0.10600823909044266,
+      "learning_rate": 5.7335907335907334e-05,
+      "loss": 0.0452,
       "step": 297
     },
     {
-      "epoch": 2.740229885057471,
-      "grad_norm": 0.01564960367977619,
-      "learning_rate": 4.967679313017303e-05,
-      "loss": 0.0079,
+      "epoch": 1.4396135265700483,
+      "grad_norm": 0.12247564643621445,
+      "learning_rate": 5.752895752895753e-05,
+      "loss": 0.0398,
       "step": 298
     },
     {
-      "epoch": 2.749425287356322,
-      "grad_norm": 0.015516403131186962,
-      "learning_rate": 4.9353599765709584e-05,
-      "loss": 0.0088,
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.09121886640787125,
+      "learning_rate": 5.772200772200772e-05,
+      "loss": 0.0388,
       "step": 299
     },
     {
-      "epoch": 2.7586206896551726,
-      "grad_norm": 0.017091600224375725,
-      "learning_rate": 4.903043341140879e-05,
-      "loss": 0.0093,
+      "epoch": 1.4492753623188406,
+      "grad_norm": 0.13584840297698975,
+      "learning_rate": 5.791505791505791e-05,
+      "loss": 0.0379,
       "step": 300
     },
     {
-      "epoch": 2.7586206896551726,
-      "eval_loss": 0.012067421339452267,
-      "eval_runtime": 7.6195,
-      "eval_samples_per_second": 6.562,
-      "eval_steps_per_second": 1.706,
+      "epoch": 1.4492753623188406,
+      "eval_loss": 0.03581848368048668,
+      "eval_runtime": 20.6035,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
       "step": 300
     },
     {
-      "epoch": 2.767816091954023,
-      "grad_norm": 0.020157935097813606,
-      "learning_rate": 4.870730757094121e-05,
-      "loss": 0.0099,
+      "epoch": 1.4541062801932367,
+      "grad_norm": 0.09803396463394165,
+      "learning_rate": 5.8108108108108105e-05,
+      "loss": 0.0353,
       "step": 301
     },
     {
-      "epoch": 2.7770114942528736,
-      "grad_norm": 0.016472700983285904,
-      "learning_rate": 4.8384235746284476e-05,
-      "loss": 0.0077,
+      "epoch": 1.458937198067633,
+      "grad_norm": 0.10005538165569305,
+      "learning_rate": 5.83011583011583e-05,
+      "loss": 0.0377,
       "step": 302
     },
     {
-      "epoch": 2.7862068965517244,
-      "grad_norm": 0.016428852453827858,
-      "learning_rate": 4.806123143715916e-05,
-      "loss": 0.0084,
+      "epoch": 1.463768115942029,
+      "grad_norm": 0.08617591857910156,
+      "learning_rate": 5.8494208494208503e-05,
+      "loss": 0.0277,
       "step": 303
     },
     {
-      "epoch": 2.7954022988505747,
-      "grad_norm": 0.017473004758358,
-      "learning_rate": 4.7738308140464685e-05,
-      "loss": 0.0089,
+      "epoch": 1.4685990338164252,
+      "grad_norm": 0.10733260959386826,
+      "learning_rate": 5.8687258687258696e-05,
+      "loss": 0.0413,
       "step": 304
     },
     {
-      "epoch": 2.8045977011494254,
-      "grad_norm": 0.015140421688556671,
-      "learning_rate": 4.7415479349715275e-05,
-      "loss": 0.0096,
-      "step": 305
-    },
-    {
-      "epoch": 2.8045977011494254,
-      "eval_loss": 0.011938001029193401,
-      "eval_runtime": 7.5857,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 1.4734299516908211,
+      "grad_norm": 0.08133935183286667,
+      "learning_rate": 5.888030888030889e-05,
+      "loss": 0.0386,
       "step": 305
     },
     {
-      "epoch": 2.8137931034482757,
-      "grad_norm": 0.01662864163517952,
-      "learning_rate": 4.709275855447621e-05,
-      "loss": 0.0091,
+      "epoch": 1.4782608695652173,
+      "grad_norm": 0.10636088252067566,
+      "learning_rate": 5.907335907335908e-05,
+      "loss": 0.0358,
       "step": 306
     },
     {
-      "epoch": 2.8229885057471265,
-      "grad_norm": 0.019283581525087357,
-      "learning_rate": 4.677015923980011e-05,
-      "loss": 0.0101,
+      "epoch": 1.4830917874396135,
+      "grad_norm": 0.10020960122346878,
+      "learning_rate": 5.9266409266409274e-05,
+      "loss": 0.0301,
       "step": 307
     },
     {
-      "epoch": 2.8321839080459768,
-      "grad_norm": 0.016029009595513344,
-      "learning_rate": 4.6447694885663514e-05,
-      "loss": 0.009,
+      "epoch": 1.4879227053140096,
+      "grad_norm": 0.08966588973999023,
+      "learning_rate": 5.9459459459459466e-05,
+      "loss": 0.0418,
       "step": 308
     },
     {
-      "epoch": 2.8413793103448275,
-      "grad_norm": 0.020449141040444374,
-      "learning_rate": 4.612537896640346e-05,
-      "loss": 0.0089,
+      "epoch": 1.4927536231884058,
+      "grad_norm": 0.12198576331138611,
+      "learning_rate": 5.965250965250966e-05,
+      "loss": 0.0417,
       "step": 309
     },
     {
-      "epoch": 2.8505747126436782,
-      "grad_norm": 0.017856908962130547,
-      "learning_rate": 4.5803224950154656e-05,
-      "loss": 0.0097,
-      "step": 310
-    },
-    {
-      "epoch": 2.8505747126436782,
-      "eval_loss": 0.01165664754807949,
-      "eval_runtime": 7.5917,
-      "eval_samples_per_second": 6.586,
-      "eval_steps_per_second": 1.712,
+      "epoch": 1.497584541062802,
+      "grad_norm": 0.11839515715837479,
+      "learning_rate": 5.984555984555985e-05,
+      "loss": 0.0402,
       "step": 310
     },
     {
-      "epoch": 2.8597701149425285,
-      "grad_norm": 0.014575740322470665,
-      "learning_rate": 4.54812462982866e-05,
-      "loss": 0.0082,
+      "epoch": 1.502415458937198,
+      "grad_norm": 0.10999340564012527,
+      "learning_rate": 6.0038610038610044e-05,
+      "loss": 0.048,
       "step": 311
     },
     {
-      "epoch": 2.8689655172413793,
-      "grad_norm": 0.017482686787843704,
-      "learning_rate": 4.515945646484105e-05,
-      "loss": 0.0105,
+      "epoch": 1.5072463768115942,
+      "grad_norm": 0.10006649047136307,
+      "learning_rate": 6.023166023166024e-05,
+      "loss": 0.0349,
       "step": 312
     },
     {
-      "epoch": 2.87816091954023,
-      "grad_norm": 0.014415116980671883,
-      "learning_rate": 4.4837868895969936e-05,
-      "loss": 0.0085,
+      "epoch": 1.5120772946859904,
+      "grad_norm": 0.10847444832324982,
+      "learning_rate": 6.042471042471043e-05,
+      "loss": 0.0378,
       "step": 313
     },
     {
-      "epoch": 2.8873563218390803,
-      "grad_norm": 0.01374480128288269,
-      "learning_rate": 4.451649702937342e-05,
-      "loss": 0.008,
+      "epoch": 1.5169082125603865,
+      "grad_norm": 0.08905550837516785,
+      "learning_rate": 6.061776061776062e-05,
+      "loss": 0.0338,
       "step": 314
     },
     {
-      "epoch": 2.896551724137931,
-      "grad_norm": 0.015432983636856079,
-      "learning_rate": 4.4195354293738484e-05,
-      "loss": 0.0101,
-      "step": 315
-    },
-    {
-      "epoch": 2.896551724137931,
-      "eval_loss": 0.011835181154310703,
-      "eval_runtime": 7.5858,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 1.5217391304347827,
+      "grad_norm": 0.1286388784646988,
+      "learning_rate": 6.0810810810810814e-05,
+      "loss": 0.0286,
       "step": 315
     },
     {
-      "epoch": 2.905747126436782,
-      "grad_norm": 0.01638283021748066,
-      "learning_rate": 4.387445410817774e-05,
-      "loss": 0.0091,
+      "epoch": 1.5265700483091789,
+      "grad_norm": 0.0705726221203804,
+      "learning_rate": 6.100386100386101e-05,
+      "loss": 0.0254,
       "step": 316
     },
     {
-      "epoch": 2.914942528735632,
-      "grad_norm": 0.016213780269026756,
-      "learning_rate": 4.355380988166867e-05,
-      "loss": 0.0085,
+      "epoch": 1.531400966183575,
+      "grad_norm": 0.08916336297988892,
+      "learning_rate": 6.11969111969112e-05,
+      "loss": 0.0344,
       "step": 317
     },
     {
-      "epoch": 2.924137931034483,
-      "grad_norm": 0.012681382708251476,
-      "learning_rate": 4.323343501249346e-05,
-      "loss": 0.0089,
+      "epoch": 1.5362318840579712,
+      "grad_norm": 0.08876073360443115,
+      "learning_rate": 6.138996138996139e-05,
+      "loss": 0.0313,
       "step": 318
     },
     {
-      "epoch": 2.9333333333333336,
-      "grad_norm": 0.01257992722094059,
-      "learning_rate": 4.2913342887678985e-05,
-      "loss": 0.0088,
+      "epoch": 1.541062801932367,
+      "grad_norm": 0.07882793992757797,
+      "learning_rate": 6.158301158301159e-05,
+      "loss": 0.0475,
       "step": 319
     },
     {
-      "epoch": 2.942528735632184,
-      "grad_norm": 0.014623278751969337,
-      "learning_rate": 4.259354688243757e-05,
-      "loss": 0.0088,
-      "step": 320
-    },
-    {
-      "epoch": 2.942528735632184,
-      "eval_loss": 0.011810057796537876,
-      "eval_runtime": 7.5985,
-      "eval_samples_per_second": 6.58,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.5458937198067633,
+      "grad_norm": 0.09577280282974243,
+      "learning_rate": 6.177606177606178e-05,
+      "loss": 0.0373,
       "step": 320
     },
     {
-      "epoch": 2.9517241379310346,
-      "grad_norm": 0.016209494322538376,
-      "learning_rate": 4.227406035960798e-05,
-      "loss": 0.0093,
+      "epoch": 1.5507246376811594,
+      "grad_norm": 0.08283651620149612,
+      "learning_rate": 6.196911196911198e-05,
+      "loss": 0.0398,
       "step": 321
     },
     {
-      "epoch": 2.960919540229885,
-      "grad_norm": 0.013485006988048553,
-      "learning_rate": 4.195489666909713e-05,
-      "loss": 0.0081,
+      "epoch": 1.5555555555555556,
+      "grad_norm": 0.08609731495380402,
+      "learning_rate": 6.216216216216216e-05,
+      "loss": 0.0333,
       "step": 322
     },
     {
-      "epoch": 2.9701149425287356,
-      "grad_norm": 0.012734117917716503,
-      "learning_rate": 4.1636069147322246e-05,
-      "loss": 0.0072,
+      "epoch": 1.5603864734299517,
+      "grad_norm": 0.1029728502035141,
+      "learning_rate": 6.235521235521236e-05,
+      "loss": 0.0378,
       "step": 323
     },
     {
-      "epoch": 2.979310344827586,
-      "grad_norm": 0.02069373056292534,
-      "learning_rate": 4.131759111665349e-05,
-      "loss": 0.0094,
+      "epoch": 1.5652173913043477,
+      "grad_norm": 0.08994361758232117,
+      "learning_rate": 6.254826254826255e-05,
+      "loss": 0.0385,
       "step": 324
     },
     {
-      "epoch": 2.9885057471264367,
-      "grad_norm": 0.018109718337655067,
-      "learning_rate": 4.099947588485744e-05,
-      "loss": 0.0096,
+      "epoch": 1.5700483091787438,
+      "grad_norm": 0.09228026866912842,
+      "learning_rate": 6.274131274131275e-05,
+      "loss": 0.032,
       "step": 325
     },
     {
-      "epoch": 2.9885057471264367,
-      "eval_loss": 0.011837408877909184,
-      "eval_runtime": 7.6092,
-      "eval_samples_per_second": 6.571,
-      "eval_steps_per_second": 1.708,
+      "epoch": 1.5700483091787438,
+      "eval_loss": 0.03536108508706093,
+      "eval_runtime": 20.5807,
+      "eval_samples_per_second": 4.859,
+      "eval_steps_per_second": 0.146,
       "step": 325
     },
     {
-      "epoch": 2.9977011494252874,
-      "grad_norm": 0.02106163091957569,
-      "learning_rate": 4.06817367445409e-05,
-      "loss": 0.0093,
+      "epoch": 1.57487922705314,
+      "grad_norm": 0.10328676551580429,
+      "learning_rate": 6.293436293436293e-05,
+      "loss": 0.0358,
       "step": 326
     },
     {
-      "epoch": 3.0068965517241377,
-      "grad_norm": 0.013615073636174202,
-      "learning_rate": 4.036438697259551e-05,
-      "loss": 0.0072,
+      "epoch": 1.5797101449275361,
+      "grad_norm": 0.09661010652780533,
+      "learning_rate": 6.312741312741313e-05,
+      "loss": 0.0346,
       "step": 327
     },
     {
-      "epoch": 3.0160919540229885,
-      "grad_norm": 0.01178102008998394,
-      "learning_rate": 4.004743982964298e-05,
-      "loss": 0.0069,
+      "epoch": 1.5845410628019323,
+      "grad_norm": 0.09876400232315063,
+      "learning_rate": 6.332046332046332e-05,
+      "loss": 0.0348,
       "step": 328
     },
     {
-      "epoch": 3.025287356321839,
-      "grad_norm": 0.01662798412144184,
-      "learning_rate": 3.97309085594809e-05,
-      "loss": 0.0095,
+      "epoch": 1.5893719806763285,
+      "grad_norm": 0.10440344363451004,
+      "learning_rate": 6.351351351351352e-05,
+      "loss": 0.0345,
       "step": 329
     },
     {
-      "epoch": 3.0344827586206895,
-      "grad_norm": 0.016467638313770294,
-      "learning_rate": 3.941480638852948e-05,
-      "loss": 0.0078,
-      "step": 330
-    },
-    {
-      "epoch": 3.0344827586206895,
-      "eval_loss": 0.011860949918627739,
-      "eval_runtime": 7.5804,
-      "eval_samples_per_second": 6.596,
-      "eval_steps_per_second": 1.715,
+      "epoch": 1.5942028985507246,
+      "grad_norm": 0.09511513262987137,
+      "learning_rate": 6.37065637065637e-05,
+      "loss": 0.037,
       "step": 330
     },
     {
-      "epoch": 3.0436781609195402,
-      "grad_norm": 0.01591646671295166,
-      "learning_rate": 3.909914652527871e-05,
-      "loss": 0.0073,
+      "epoch": 1.5990338164251208,
+      "grad_norm": 0.1074519231915474,
+      "learning_rate": 6.38996138996139e-05,
+      "loss": 0.0469,
       "step": 331
     },
     {
-      "epoch": 3.052873563218391,
-      "grad_norm": 0.01757364347577095,
-      "learning_rate": 3.878394215973663e-05,
-      "loss": 0.0074,
+      "epoch": 1.603864734299517,
+      "grad_norm": 0.10466543585062027,
+      "learning_rate": 6.40926640926641e-05,
+      "loss": 0.0305,
       "step": 332
     },
     {
-      "epoch": 3.0620689655172413,
-      "grad_norm": 0.017451900988817215,
-      "learning_rate": 3.846920646287799e-05,
-      "loss": 0.0077,
+      "epoch": 1.608695652173913,
+      "grad_norm": 0.08107554912567139,
+      "learning_rate": 6.428571428571429e-05,
+      "loss": 0.0345,
       "step": 333
     },
     {
-      "epoch": 3.071264367816092,
-      "grad_norm": 0.01796400360763073,
-      "learning_rate": 3.815495258609404e-05,
-      "loss": 0.0077,
+      "epoch": 1.6135265700483092,
+      "grad_norm": 0.08928783237934113,
+      "learning_rate": 6.447876447876449e-05,
+      "loss": 0.0332,
       "step": 334
     },
     {
-      "epoch": 3.0804597701149423,
-      "grad_norm": 0.01757225953042507,
-      "learning_rate": 3.784119366064293e-05,
-      "loss": 0.0064,
-      "step": 335
-    },
-    {
-      "epoch": 3.0804597701149423,
-      "eval_loss": 0.01191374659538269,
-      "eval_runtime": 7.5955,
-      "eval_samples_per_second": 6.583,
-      "eval_steps_per_second": 1.712,
+      "epoch": 1.6183574879227054,
+      "grad_norm": 0.10237609595060349,
+      "learning_rate": 6.467181467181467e-05,
+      "loss": 0.0372,
       "step": 335
     },
     {
-      "epoch": 3.089655172413793,
-      "grad_norm": 0.01832747273147106,
-      "learning_rate": 3.752794279710094e-05,
-      "loss": 0.007,
+      "epoch": 1.6231884057971016,
+      "grad_norm": 0.08093827217817307,
+      "learning_rate": 6.486486486486487e-05,
+      "loss": 0.0365,
       "step": 336
     },
     {
-      "epoch": 3.098850574712644,
-      "grad_norm": 0.018249882385134697,
-      "learning_rate": 3.721521308481482e-05,
-      "loss": 0.0083,
+      "epoch": 1.6280193236714977,
+      "grad_norm": 0.08602860569953918,
+      "learning_rate": 6.505791505791506e-05,
+      "loss": 0.035,
       "step": 337
     },
     {
-      "epoch": 3.108045977011494,
-      "grad_norm": 0.017133980989456177,
-      "learning_rate": 3.6903017591354706e-05,
-      "loss": 0.0062,
+      "epoch": 1.6328502415458939,
+      "grad_norm": 0.09471640735864639,
+      "learning_rate": 6.525096525096526e-05,
+      "loss": 0.0422,
       "step": 338
     },
     {
-      "epoch": 3.117241379310345,
-      "grad_norm": 0.018890004605054855,
-      "learning_rate": 3.6591369361968124e-05,
-      "loss": 0.0068,
+      "epoch": 1.6376811594202898,
+      "grad_norm": 0.10974764078855515,
+      "learning_rate": 6.544401544401544e-05,
+      "loss": 0.0369,
       "step": 339
     },
     {
-      "epoch": 3.1264367816091956,
-      "grad_norm": 0.015423454344272614,
-      "learning_rate": 3.628028141903493e-05,
-      "loss": 0.0073,
-      "step": 340
-    },
-    {
-      "epoch": 3.1264367816091956,
-      "eval_loss": 0.01213847566395998,
-      "eval_runtime": 7.5868,
-      "eval_samples_per_second": 6.59,
-      "eval_steps_per_second": 1.713,
+      "epoch": 1.642512077294686,
+      "grad_norm": 0.08305064588785172,
+      "learning_rate": 6.563706563706564e-05,
+      "loss": 0.0508,
       "step": 340
     },
     {
-      "epoch": 3.135632183908046,
-      "grad_norm": 0.015354766510426998,
-      "learning_rate": 3.596976676152306e-05,
-      "loss": 0.0074,
+      "epoch": 1.6473429951690821,
+      "grad_norm": 0.09225677698850632,
+      "learning_rate": 6.583011583011583e-05,
+      "loss": 0.0392,
       "step": 341
     },
     {
-      "epoch": 3.1448275862068966,
-      "grad_norm": 0.018978066742420197,
-      "learning_rate": 3.5659838364445505e-05,
-      "loss": 0.0071,
+      "epoch": 1.6521739130434783,
+      "grad_norm": 0.10002700239419937,
+      "learning_rate": 6.602316602316603e-05,
+      "loss": 0.0376,
       "step": 342
     },
     {
-      "epoch": 3.154022988505747,
-      "grad_norm": 0.021498536691069603,
-      "learning_rate": 3.535050917831797e-05,
-      "loss": 0.0072,
+      "epoch": 1.6570048309178744,
+      "grad_norm": 0.064641572535038,
+      "learning_rate": 6.621621621621621e-05,
+      "loss": 0.0284,
       "step": 343
     },
     {
-      "epoch": 3.1632183908045977,
-      "grad_norm": 0.01662536896765232,
-      "learning_rate": 3.5041792128617927e-05,
-      "loss": 0.0063,
+      "epoch": 1.6618357487922706,
+      "grad_norm": 0.08822232484817505,
+      "learning_rate": 6.640926640926641e-05,
+      "loss": 0.0411,
       "step": 344
     },
     {
-      "epoch": 3.1724137931034484,
-      "grad_norm": 0.016221007332205772,
-      "learning_rate": 3.473370011524435e-05,
-      "loss": 0.0066,
-      "step": 345
-    },
-    {
-      "epoch": 3.1724137931034484,
-      "eval_loss": 0.012059099040925503,
-      "eval_runtime": 7.5892,
-      "eval_samples_per_second": 6.588,
-      "eval_steps_per_second": 1.713,
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.07955625653266907,
+      "learning_rate": 6.66023166023166e-05,
+      "loss": 0.0286,
       "step": 345
     },
     {
-      "epoch": 3.1816091954022987,
-      "grad_norm": 0.021475255489349365,
-      "learning_rate": 3.442624601197877e-05,
-      "loss": 0.0076,
+      "epoch": 1.6714975845410627,
+      "grad_norm": 0.08300414681434631,
+      "learning_rate": 6.67953667953668e-05,
+      "loss": 0.0248,
       "step": 346
     },
     {
-      "epoch": 3.1908045977011494,
-      "grad_norm": 0.015069660730659962,
-      "learning_rate": 3.4119442665947344e-05,
-      "loss": 0.0064,
+      "epoch": 1.6763285024154588,
+      "grad_norm": 0.08104908466339111,
+      "learning_rate": 6.6988416988417e-05,
+      "loss": 0.0342,
       "step": 347
     },
     {
-      "epoch": 3.2,
-      "grad_norm": 0.01803274266421795,
-      "learning_rate": 3.381330289708396e-05,
-      "loss": 0.0082,
+      "epoch": 1.681159420289855,
+      "grad_norm": 0.0685056522488594,
+      "learning_rate": 6.718146718146718e-05,
+      "loss": 0.0277,
       "step": 348
     },
     {
-      "epoch": 3.2091954022988505,
-      "grad_norm": 0.01855016127228737,
-      "learning_rate": 3.350783949759462e-05,
-      "loss": 0.0072,
+      "epoch": 1.6859903381642511,
+      "grad_norm": 0.09631809592247009,
+      "learning_rate": 6.737451737451738e-05,
+      "loss": 0.0342,
       "step": 349
     },
     {
-      "epoch": 3.218390804597701,
-      "grad_norm": 0.017640260979533195,
-      "learning_rate": 3.3203065231422904e-05,
-      "loss": 0.0067,
+      "epoch": 1.6908212560386473,
+      "grad_norm": 0.09193519502878189,
+      "learning_rate": 6.756756756756757e-05,
+      "loss": 0.0342,
       "step": 350
     },
     {
-      "epoch": 3.218390804597701,
-      "eval_loss": 0.011746257543563843,
-      "eval_runtime": 7.5783,
-      "eval_samples_per_second": 6.598,
-      "eval_steps_per_second": 1.715,
+      "epoch": 1.6908212560386473,
+      "eval_loss": 0.033477578312158585,
+      "eval_runtime": 20.5968,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
       "step": 350
     },
     {
-      "epoch": 3.227586206896552,
-      "grad_norm": 0.018065195530653,
-      "learning_rate": 3.289899283371657e-05,
-      "loss": 0.0082,
+      "epoch": 1.6956521739130435,
+      "grad_norm": 0.08512197434902191,
+      "learning_rate": 6.776061776061777e-05,
+      "loss": 0.0344,
       "step": 351
     },
     {
-      "epoch": 3.2367816091954023,
-      "grad_norm": 0.01586788520216942,
-      "learning_rate": 3.2595635010295475e-05,
-      "loss": 0.007,
+      "epoch": 1.7004830917874396,
+      "grad_norm": 0.11075153201818466,
+      "learning_rate": 6.795366795366795e-05,
+      "loss": 0.041,
       "step": 352
     },
     {
-      "epoch": 3.245977011494253,
-      "grad_norm": 0.01758764311671257,
-      "learning_rate": 3.2293004437120624e-05,
-      "loss": 0.0073,
+      "epoch": 1.7053140096618358,
+      "grad_norm": 0.0982983410358429,
+      "learning_rate": 6.814671814671815e-05,
+      "loss": 0.0319,
       "step": 353
     },
     {
-      "epoch": 3.2551724137931033,
-      "grad_norm": 0.018066026270389557,
-      "learning_rate": 3.199111375976449e-05,
-      "loss": 0.0075,
+      "epoch": 1.710144927536232,
+      "grad_norm": 0.0894618108868599,
+      "learning_rate": 6.833976833976834e-05,
+      "loss": 0.0422,
       "step": 354
     },
     {
-      "epoch": 3.264367816091954,
-      "grad_norm": 0.02090166136622429,
-      "learning_rate": 3.1689975592882603e-05,
-      "loss": 0.007,
-      "step": 355
-    },
-    {
-      "epoch": 3.264367816091954,
-      "eval_loss": 0.011750459671020508,
-      "eval_runtime": 7.5842,
-      "eval_samples_per_second": 6.593,
-      "eval_steps_per_second": 1.714,
+      "epoch": 1.714975845410628,
+      "grad_norm": 0.11245858669281006,
+      "learning_rate": 6.853281853281854e-05,
+      "loss": 0.0443,
       "step": 355
     },
     {
-      "epoch": 3.2735632183908048,
-      "grad_norm": 0.0185533594340086,
-      "learning_rate": 3.1389602519686515e-05,
-      "loss": 0.0078,
+      "epoch": 1.7198067632850242,
+      "grad_norm": 0.0929783433675766,
+      "learning_rate": 6.872586872586872e-05,
+      "loss": 0.0341,
       "step": 356
     },
     {
-      "epoch": 3.282758620689655,
-      "grad_norm": 0.018230192363262177,
-      "learning_rate": 3.109000709141788e-05,
-      "loss": 0.007,
+      "epoch": 1.7246376811594204,
+      "grad_norm": 0.07707154005765915,
+      "learning_rate": 6.891891891891892e-05,
+      "loss": 0.0301,
       "step": 357
     },
     {
-      "epoch": 3.291954022988506,
-      "grad_norm": 0.016010766848921776,
-      "learning_rate": 3.079120182682412e-05,
-      "loss": 0.0061,
+      "epoch": 1.7294685990338166,
+      "grad_norm": 0.08752428740262985,
+      "learning_rate": 6.911196911196911e-05,
+      "loss": 0.0341,
       "step": 358
     },
     {
-      "epoch": 3.301149425287356,
-      "grad_norm": 0.019986698403954506,
-      "learning_rate": 3.049319921163526e-05,
-      "loss": 0.0085,
+      "epoch": 1.7342995169082127,
+      "grad_norm": 0.11009437590837479,
+      "learning_rate": 6.930501930501931e-05,
+      "loss": 0.0393,
       "step": 359
     },
     {
-      "epoch": 3.310344827586207,
-      "grad_norm": 0.01619451493024826,
-      "learning_rate": 3.019601169804216e-05,
-      "loss": 0.0072,
-      "step": 360
-    },
-    {
-      "epoch": 3.310344827586207,
-      "eval_loss": 0.011605262756347656,
-      "eval_runtime": 7.5872,
-      "eval_samples_per_second": 6.59,
-      "eval_steps_per_second": 1.713,
+      "epoch": 1.7391304347826086,
+      "grad_norm": 0.09310301393270493,
+      "learning_rate": 6.949806949806951e-05,
+      "loss": 0.0334,
       "step": 360
     },
     {
-      "epoch": 3.3195402298850576,
-      "grad_norm": 0.017471255734562874,
-      "learning_rate": 2.9899651704176325e-05,
-      "loss": 0.0075,
+      "epoch": 1.7439613526570048,
+      "grad_norm": 0.10008851438760757,
+      "learning_rate": 6.96911196911197e-05,
+      "loss": 0.0332,
       "step": 361
     },
     {
-      "epoch": 3.328735632183908,
-      "grad_norm": 0.015944065526127815,
-      "learning_rate": 2.9604131613590824e-05,
-      "loss": 0.0059,
+      "epoch": 1.748792270531401,
+      "grad_norm": 0.09315995872020721,
+      "learning_rate": 6.988416988416989e-05,
+      "loss": 0.043,
       "step": 362
     },
     {
-      "epoch": 3.3379310344827586,
-      "grad_norm": 0.02053840085864067,
-      "learning_rate": 2.9309463774743046e-05,
-      "loss": 0.0068,
+      "epoch": 1.7536231884057971,
+      "grad_norm": 0.09456358104944229,
+      "learning_rate": 7.007722007722008e-05,
+      "loss": 0.0498,
       "step": 363
     },
     {
-      "epoch": 3.3471264367816094,
-      "grad_norm": 0.022107843309640884,
-      "learning_rate": 2.901566050047855e-05,
-      "loss": 0.0076,
+      "epoch": 1.7584541062801933,
+      "grad_norm": 0.07672219723463058,
+      "learning_rate": 7.027027027027028e-05,
+      "loss": 0.0352,
       "step": 364
     },
     {
-      "epoch": 3.3563218390804597,
-      "grad_norm": 0.022166235372424126,
-      "learning_rate": 2.872273406751664e-05,
-      "loss": 0.0074,
-      "step": 365
-    },
-    {
-      "epoch": 3.3563218390804597,
-      "eval_loss": 0.011744478717446327,
-      "eval_runtime": 7.5789,
-      "eval_samples_per_second": 6.597,
-      "eval_steps_per_second": 1.715,
+      "epoch": 1.7632850241545892,
+      "grad_norm": 0.07560185343027115,
+      "learning_rate": 7.046332046332046e-05,
+      "loss": 0.0346,
       "step": 365
     },
     {
-      "epoch": 3.3655172413793104,
-      "grad_norm": 0.018682653084397316,
-      "learning_rate": 2.8430696715937337e-05,
-      "loss": 0.007,
+      "epoch": 1.7681159420289854,
+      "grad_norm": 0.07019049674272537,
+      "learning_rate": 7.065637065637066e-05,
+      "loss": 0.0322,
       "step": 366
     },
     {
-      "epoch": 3.374712643678161,
-      "grad_norm": 0.023426858708262444,
-      "learning_rate": 2.8139560648669962e-05,
-      "loss": 0.0062,
+      "epoch": 1.7729468599033815,
+      "grad_norm": 0.0801636353135109,
+      "learning_rate": 7.084942084942085e-05,
+      "loss": 0.0277,
       "step": 367
     },
     {
-      "epoch": 3.3839080459770114,
-      "grad_norm": 0.019427569583058357,
-      "learning_rate": 2.7849338030983257e-05,
-      "loss": 0.0065,
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.08249375224113464,
+      "learning_rate": 7.104247104247105e-05,
+      "loss": 0.0523,
       "step": 368
     },
     {
-      "epoch": 3.393103448275862,
-      "grad_norm": 0.015961861237883568,
-      "learning_rate": 2.7560040989976892e-05,
-      "loss": 0.0062,
+      "epoch": 1.7826086956521738,
+      "grad_norm": 0.08103843778371811,
+      "learning_rate": 7.123552123552123e-05,
+      "loss": 0.0343,
       "step": 369
     },
     {
-      "epoch": 3.4022988505747125,
-      "grad_norm": 0.018800493329763412,
-      "learning_rate": 2.7271681614074973e-05,
-      "loss": 0.0067,
-      "step": 370
-    },
-    {
-      "epoch": 3.4022988505747125,
-      "eval_loss": 0.011655151844024658,
-      "eval_runtime": 7.5992,
-      "eval_samples_per_second": 6.58,
-      "eval_steps_per_second": 1.711,
+      "epoch": 1.78743961352657,
+      "grad_norm": 0.08398215472698212,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.032,
       "step": 370
     },
     {
-      "epoch": 3.4114942528735632,
-      "grad_norm": 0.02263784408569336,
-      "learning_rate": 2.6984271952520722e-05,
-      "loss": 0.0074,
+      "epoch": 1.7922705314009661,
+      "grad_norm": 0.07273257523775101,
+      "learning_rate": 7.162162162162162e-05,
+      "loss": 0.0315,
       "step": 371
     },
     {
-      "epoch": 3.420689655172414,
-      "grad_norm": 0.01691795513033867,
-      "learning_rate": 2.6697824014873075e-05,
-      "loss": 0.0069,
+      "epoch": 1.7971014492753623,
+      "grad_norm": 0.0843791663646698,
+      "learning_rate": 7.181467181467182e-05,
+      "loss": 0.0343,
       "step": 372
     },
     {
-      "epoch": 3.4298850574712643,
-      "grad_norm": 0.016610443592071533,
-      "learning_rate": 2.641234977050484e-05,
-      "loss": 0.007,
+      "epoch": 1.8019323671497585,
+      "grad_norm": 0.06816914677619934,
+      "learning_rate": 7.2007722007722e-05,
+      "loss": 0.0323,
       "step": 373
     },
     {
-      "epoch": 3.439080459770115,
-      "grad_norm": 0.013923544436693192,
-      "learning_rate": 2.612786114810255e-05,
-      "loss": 0.006,
+      "epoch": 1.8067632850241546,
+      "grad_norm": 0.0995233952999115,
+      "learning_rate": 7.22007722007722e-05,
+      "loss": 0.0252,
       "step": 374
     },
     {
-      "epoch": 3.4482758620689653,
-      "grad_norm": 0.016392454504966736,
-      "learning_rate": 2.5844370035168073e-05,
-      "loss": 0.0072,
+      "epoch": 1.8115942028985508,
+      "grad_norm": 0.07620427757501602,
+      "learning_rate": 7.23938223938224e-05,
+      "loss": 0.0318,
       "step": 375
     },
     {
-      "epoch": 3.4482758620689653,
-      "eval_loss": 0.01167826447635889,
-      "eval_runtime": 7.5952,
-      "eval_samples_per_second": 6.583,
-      "eval_steps_per_second": 1.712,
+      "epoch": 1.8115942028985508,
+      "eval_loss": 0.03241335600614548,
+      "eval_runtime": 20.5896,
+      "eval_samples_per_second": 4.857,
+      "eval_steps_per_second": 0.146,
       "step": 375
     },
     {
-      "epoch": 3.457471264367816,
-      "grad_norm": 0.017350174486637115,
-      "learning_rate": 2.5561888277521794e-05,
-      "loss": 0.0073,
+      "epoch": 1.816425120772947,
+      "grad_norm": 0.08072451502084732,
+      "learning_rate": 7.258687258687259e-05,
+      "loss": 0.0306,
       "step": 376
     },
     {
-      "epoch": 3.466666666666667,
-      "grad_norm": 0.016711924225091934,
-      "learning_rate": 2.528042767880766e-05,
-      "loss": 0.0063,
+      "epoch": 1.821256038647343,
+      "grad_norm": 0.08463121950626373,
+      "learning_rate": 7.277992277992279e-05,
+      "loss": 0.0246,
       "step": 377
     },
     {
-      "epoch": 3.475862068965517,
-      "grad_norm": 0.019175494089722633,
-      "learning_rate": 2.500000000000001e-05,
-      "loss": 0.0073,
+      "epoch": 1.8260869565217392,
+      "grad_norm": 0.08696448057889938,
+      "learning_rate": 7.297297297297297e-05,
+      "loss": 0.0391,
       "step": 378
     },
     {
-      "epoch": 3.485057471264368,
-      "grad_norm": 0.018161507323384285,
-      "learning_rate": 2.4720616958912053e-05,
-      "loss": 0.0077,
+      "epoch": 1.8309178743961354,
+      "grad_norm": 0.08394154906272888,
+      "learning_rate": 7.316602316602317e-05,
+      "loss": 0.0503,
       "step": 379
     },
     {
-      "epoch": 3.4942528735632186,
-      "grad_norm": 0.017984988167881966,
-      "learning_rate": 2.4442290229706344e-05,
-      "loss": 0.0069,
-      "step": 380
-    },
-    {
-      "epoch": 3.4942528735632186,
-      "eval_loss": 0.011730669066309929,
-      "eval_runtime": 7.5885,
-      "eval_samples_per_second": 6.589,
-      "eval_steps_per_second": 1.713,
+      "epoch": 1.8357487922705316,
+      "grad_norm": 0.09307282418012619,
+      "learning_rate": 7.335907335907336e-05,
+      "loss": 0.0338,
       "step": 380
     },
     {
-      "epoch": 3.503448275862069,
-      "grad_norm": 0.016771584749221802,
-      "learning_rate": 2.4165031442406855e-05,
-      "loss": 0.0066,
+      "epoch": 1.8405797101449275,
+      "grad_norm": 0.08928454667329788,
+      "learning_rate": 7.355212355212356e-05,
+      "loss": 0.0414,
       "step": 381
     },
     {
-      "epoch": 3.5126436781609196,
-      "grad_norm": 0.016384167596697807,
-      "learning_rate": 2.3888852182413085e-05,
-      "loss": 0.0068,
+      "epoch": 1.8454106280193237,
+      "grad_norm": 0.06912971287965775,
+      "learning_rate": 7.374517374517374e-05,
+      "loss": 0.0299,
       "step": 382
     },
     {
-      "epoch": 3.5218390804597703,
-      "grad_norm": 0.016403399407863617,
-      "learning_rate": 2.361376399001592e-05,
-      "loss": 0.0069,
+      "epoch": 1.8502415458937198,
+      "grad_norm": 0.07190634310245514,
+      "learning_rate": 7.393822393822394e-05,
+      "loss": 0.0293,
       "step": 383
     },
     {
-      "epoch": 3.5310344827586206,
-      "grad_norm": 0.017848780378699303,
-      "learning_rate": 2.333977835991545e-05,
-      "loss": 0.0077,
+      "epoch": 1.855072463768116,
+      "grad_norm": 0.08201923221349716,
+      "learning_rate": 7.413127413127413e-05,
+      "loss": 0.0378,
       "step": 384
     },
     {
-      "epoch": 3.5402298850574714,
-      "grad_norm": 0.02174249291419983,
-      "learning_rate": 2.3066906740740623e-05,
-      "loss": 0.0076,
-      "step": 385
-    },
-    {
-      "epoch": 3.5402298850574714,
-      "eval_loss": 0.011591249145567417,
-      "eval_runtime": 7.5857,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 1.8599033816425121,
+      "grad_norm": 0.09282411634922028,
+      "learning_rate": 7.432432432432433e-05,
+      "loss": 0.0318,
       "step": 385
     },
     {
-      "epoch": 3.5494252873563217,
-      "grad_norm": 0.018354855477809906,
-      "learning_rate": 2.2795160534570864e-05,
-      "loss": 0.0073,
+      "epoch": 1.864734299516908,
+      "grad_norm": 0.09017174690961838,
+      "learning_rate": 7.451737451737452e-05,
+      "loss": 0.0373,
       "step": 386
     },
     {
-      "epoch": 3.5586206896551724,
-      "grad_norm": 0.01478376891463995,
-      "learning_rate": 2.25245510964597e-05,
-      "loss": 0.0067,
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.09520550072193146,
+      "learning_rate": 7.471042471042471e-05,
+      "loss": 0.0374,
       "step": 387
     },
     {
-      "epoch": 3.5678160919540227,
-      "grad_norm": 0.020804837346076965,
-      "learning_rate": 2.225508973396016e-05,
-      "loss": 0.0066,
+      "epoch": 1.8743961352657004,
+      "grad_norm": 0.08649712800979614,
+      "learning_rate": 7.49034749034749e-05,
+      "loss": 0.038,
       "step": 388
     },
     {
-      "epoch": 3.5770114942528735,
-      "grad_norm": 0.017460566014051437,
-      "learning_rate": 2.198678770665238e-05,
-      "loss": 0.0062,
+      "epoch": 1.8792270531400965,
+      "grad_norm": 0.08308789879083633,
+      "learning_rate": 7.50965250965251e-05,
+      "loss": 0.0318,
       "step": 389
     },
     {
-      "epoch": 3.586206896551724,
-      "grad_norm": 0.01834852248430252,
-      "learning_rate": 2.171965622567308e-05,
-      "loss": 0.0068,
-      "step": 390
-    },
-    {
-      "epoch": 3.586206896551724,
-      "eval_loss": 0.011386090889573097,
-      "eval_runtime": 7.5843,
-      "eval_samples_per_second": 6.593,
-      "eval_steps_per_second": 1.714,
+      "epoch": 1.8840579710144927,
+      "grad_norm": 0.07252487540245056,
+      "learning_rate": 7.52895752895753e-05,
+      "loss": 0.0266,
       "step": 390
     },
     {
-      "epoch": 3.5954022988505745,
-      "grad_norm": 0.015800220891833305,
-      "learning_rate": 2.1453706453247087e-05,
-      "loss": 0.0064,
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.07101837545633316,
+      "learning_rate": 7.548262548262549e-05,
+      "loss": 0.0334,
       "step": 391
     },
     {
-      "epoch": 3.6045977011494252,
-      "grad_norm": 0.017407681792974472,
-      "learning_rate": 2.1188949502220983e-05,
-      "loss": 0.0073,
+      "epoch": 1.893719806763285,
+      "grad_norm": 0.07671766728162766,
+      "learning_rate": 7.567567567567568e-05,
+      "loss": 0.037,
       "step": 392
     },
     {
-      "epoch": 3.613793103448276,
-      "grad_norm": 0.01731077767908573,
-      "learning_rate": 2.0925396435598664e-05,
-      "loss": 0.0063,
+      "epoch": 1.8985507246376812,
+      "grad_norm": 0.0921880453824997,
+      "learning_rate": 7.586872586872587e-05,
+      "loss": 0.0349,
       "step": 393
     },
     {
-      "epoch": 3.6229885057471263,
-      "grad_norm": 0.01906469836831093,
-      "learning_rate": 2.066305826607911e-05,
-      "loss": 0.0077,
+      "epoch": 1.9033816425120773,
+      "grad_norm": 0.1014850065112114,
+      "learning_rate": 7.606177606177607e-05,
+      "loss": 0.0505,
       "step": 394
     },
     {
-      "epoch": 3.632183908045977,
-      "grad_norm": 0.017172574996948242,
-      "learning_rate": 2.0401945955596206e-05,
-      "loss": 0.0074,
-      "step": 395
-    },
-    {
-      "epoch": 3.632183908045977,
-      "eval_loss": 0.011461833491921425,
-      "eval_runtime": 7.6124,
-      "eval_samples_per_second": 6.568,
-      "eval_steps_per_second": 1.708,
+      "epoch": 1.9082125603864735,
+      "grad_norm": 0.07078011333942413,
+      "learning_rate": 7.625482625482626e-05,
+      "loss": 0.0276,
       "step": 395
     },
     {
-      "epoch": 3.6413793103448278,
-      "grad_norm": 0.019811954349279404,
-      "learning_rate": 2.0142070414860704e-05,
-      "loss": 0.0077,
+      "epoch": 1.9130434782608696,
+      "grad_norm": 0.07615656405687332,
+      "learning_rate": 7.644787644787645e-05,
+      "loss": 0.0291,
       "step": 396
     },
     {
-      "epoch": 3.650574712643678,
-      "grad_norm": 0.01807416044175625,
-      "learning_rate": 1.9883442502904283e-05,
-      "loss": 0.0075,
+      "epoch": 1.9178743961352658,
+      "grad_norm": 0.06996695697307587,
+      "learning_rate": 7.664092664092664e-05,
+      "loss": 0.026,
       "step": 397
     },
     {
-      "epoch": 3.659770114942529,
-      "grad_norm": 0.021201424300670624,
-      "learning_rate": 1.9626073026625818e-05,
-      "loss": 0.0077,
+      "epoch": 1.922705314009662,
+      "grad_norm": 0.06688771396875381,
+      "learning_rate": 7.683397683397684e-05,
+      "loss": 0.0321,
       "step": 398
     },
     {
-      "epoch": 3.6689655172413795,
-      "grad_norm": 0.01894940622150898,
-      "learning_rate": 1.936997274033986e-05,
-      "loss": 0.0062,
+      "epoch": 1.927536231884058,
+      "grad_norm": 0.06577587872743607,
+      "learning_rate": 7.702702702702703e-05,
+      "loss": 0.0288,
       "step": 399
     },
     {
-      "epoch": 3.67816091954023,
-      "grad_norm": 0.016950421035289764,
-      "learning_rate": 1.9115152345327152e-05,
-      "loss": 0.0065,
+      "epoch": 1.9323671497584543,
+      "grad_norm": 0.07597734034061432,
+      "learning_rate": 7.722007722007723e-05,
+      "loss": 0.031,
       "step": 400
     },
     {
-      "epoch": 3.67816091954023,
-      "eval_loss": 0.011437707580626011,
-      "eval_runtime": 7.6082,
-      "eval_samples_per_second": 6.572,
-      "eval_steps_per_second": 1.709,
+      "epoch": 1.9323671497584543,
+      "eval_loss": 0.031759362667798996,
+      "eval_runtime": 20.6062,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
       "step": 400
     },
     {
-      "epoch": 3.6873563218390806,
-      "grad_norm": 0.01757856458425522,
-      "learning_rate": 1.8861622489387555e-05,
-      "loss": 0.0062,
+      "epoch": 1.9371980676328504,
+      "grad_norm": 0.08483527600765228,
+      "learning_rate": 7.741312741312741e-05,
+      "loss": 0.0339,
       "step": 401
     },
     {
-      "epoch": 3.696551724137931,
-      "grad_norm": 0.020201781764626503,
-      "learning_rate": 1.8609393766395085e-05,
-      "loss": 0.0079,
+      "epoch": 1.9420289855072463,
+      "grad_norm": 0.06974739581346512,
+      "learning_rate": 7.760617760617761e-05,
+      "loss": 0.0313,
       "step": 402
     },
     {
-      "epoch": 3.7057471264367816,
-      "grad_norm": 0.015953106805682182,
-      "learning_rate": 1.835847671585526e-05,
-      "loss": 0.0068,
+      "epoch": 1.9468599033816425,
+      "grad_norm": 0.093897745013237,
+      "learning_rate": 7.779922779922781e-05,
+      "loss": 0.0332,
       "step": 403
     },
     {
-      "epoch": 3.714942528735632,
-      "grad_norm": 0.018131552264094353,
-      "learning_rate": 1.8108881822464696e-05,
-      "loss": 0.0066,
+      "epoch": 1.9516908212560387,
+      "grad_norm": 0.06541071832180023,
+      "learning_rate": 7.7992277992278e-05,
+      "loss": 0.0245,
       "step": 404
     },
     {
-      "epoch": 3.7241379310344827,
-      "grad_norm": 0.018745753914117813,
-      "learning_rate": 1.7860619515673033e-05,
-      "loss": 0.007,
-      "step": 405
-    },
-    {
-      "epoch": 3.7241379310344827,
-      "eval_loss": 0.011230139061808586,
-      "eval_runtime": 7.5892,
-      "eval_samples_per_second": 6.588,
-      "eval_steps_per_second": 1.713,
+      "epoch": 1.9565217391304348,
+      "grad_norm": 0.07860124856233597,
+      "learning_rate": 7.81853281853282e-05,
+      "loss": 0.0288,
       "step": 405
     },
     {
-      "epoch": 3.7333333333333334,
-      "grad_norm": 0.019100049510598183,
-      "learning_rate": 1.7613700169247056e-05,
-      "loss": 0.0078,
+      "epoch": 1.961352657004831,
+      "grad_norm": 0.09602763503789902,
+      "learning_rate": 7.837837837837838e-05,
+      "loss": 0.0374,
       "step": 406
     },
     {
-      "epoch": 3.7425287356321837,
-      "grad_norm": 0.018397876992821693,
-      "learning_rate": 1.7368134100837287e-05,
-      "loss": 0.0077,
+      "epoch": 1.966183574879227,
+      "grad_norm": 0.07784169167280197,
+      "learning_rate": 7.857142857142858e-05,
+      "loss": 0.0284,
       "step": 407
     },
     {
-      "epoch": 3.7517241379310344,
-      "grad_norm": 0.01846805214881897,
-      "learning_rate": 1.7123931571546827e-05,
-      "loss": 0.0073,
+      "epoch": 1.971014492753623,
+      "grad_norm": 0.08984600007534027,
+      "learning_rate": 7.876447876447877e-05,
+      "loss": 0.0325,
       "step": 408
     },
     {
-      "epoch": 3.760919540229885,
-      "grad_norm": 0.019604753702878952,
-      "learning_rate": 1.6881102785502616e-05,
-      "loss": 0.0081,
+      "epoch": 1.9758454106280192,
+      "grad_norm": 0.06867984682321548,
+      "learning_rate": 7.895752895752897e-05,
+      "loss": 0.028,
       "step": 409
     },
     {
-      "epoch": 3.7701149425287355,
-      "grad_norm": 0.019953811541199684,
-      "learning_rate": 1.6639657889429018e-05,
-      "loss": 0.0064,
-      "step": 410
-    },
-    {
-      "epoch": 3.7701149425287355,
-      "eval_loss": 0.01116408035159111,
-      "eval_runtime": 7.6015,
-      "eval_samples_per_second": 6.578,
-      "eval_steps_per_second": 1.71,
+      "epoch": 1.9806763285024154,
+      "grad_norm": 0.09378314763307571,
+      "learning_rate": 7.915057915057915e-05,
+      "loss": 0.0453,
       "step": 410
     },
     {
-      "epoch": 3.779310344827586,
-      "grad_norm": 0.018632952123880386,
-      "learning_rate": 1.639960697222388e-05,
-      "loss": 0.0063,
+      "epoch": 1.9855072463768115,
+      "grad_norm": 0.07811439037322998,
+      "learning_rate": 7.934362934362935e-05,
+      "loss": 0.0309,
       "step": 411
     },
     {
-      "epoch": 3.788505747126437,
-      "grad_norm": 0.018017446622252464,
-      "learning_rate": 1.6160960064536908e-05,
-      "loss": 0.0074,
+      "epoch": 1.9903381642512077,
+      "grad_norm": 0.059358660131692886,
+      "learning_rate": 7.953667953667954e-05,
+      "loss": 0.0226,
       "step": 412
     },
     {
-      "epoch": 3.7977011494252872,
-      "grad_norm": 0.016660140827298164,
-      "learning_rate": 1.592372713835055e-05,
-      "loss": 0.0068,
+      "epoch": 1.9951690821256038,
+      "grad_norm": 0.0751197412610054,
+      "learning_rate": 7.972972972972974e-05,
+      "loss": 0.0271,
       "step": 413
     },
     {
-      "epoch": 3.806896551724138,
-      "grad_norm": 0.016572022810578346,
-      "learning_rate": 1.5687918106563326e-05,
-      "loss": 0.007,
+      "epoch": 2.0,
+      "grad_norm": 0.14637447893619537,
+      "learning_rate": 7.992277992277992e-05,
+      "loss": 0.0192,
       "step": 414
     },
     {
-      "epoch": 3.8160919540229887,
-      "grad_norm": 0.017351387068629265,
-      "learning_rate": 1.545354282257562e-05,
-      "loss": 0.0073,
-      "step": 415
-    },
-    {
-      "epoch": 3.8160919540229887,
-      "eval_loss": 0.011113017797470093,
-      "eval_runtime": 7.5996,
-      "eval_samples_per_second": 6.579,
-      "eval_steps_per_second": 1.711,
+      "epoch": 2.004830917874396,
+      "grad_norm": 0.08730384707450867,
+      "learning_rate": 8.011583011583012e-05,
+      "loss": 0.0383,
       "step": 415
     },
     {
-      "epoch": 3.825287356321839,
-      "grad_norm": 0.017648274078965187,
-      "learning_rate": 1.52206110798779e-05,
-      "loss": 0.0058,
+      "epoch": 2.0096618357487923,
+      "grad_norm": 0.0848565399646759,
+      "learning_rate": 8.03088803088803e-05,
+      "loss": 0.0304,
       "step": 416
     },
     {
-      "epoch": 3.8344827586206898,
-      "grad_norm": 0.019093746319413185,
-      "learning_rate": 1.4989132611641576e-05,
-      "loss": 0.0071,
+      "epoch": 2.0144927536231885,
+      "grad_norm": 0.0779394656419754,
+      "learning_rate": 8.05019305019305e-05,
+      "loss": 0.0281,
       "step": 417
     },
     {
-      "epoch": 3.84367816091954,
-      "grad_norm": 0.0174701064825058,
-      "learning_rate": 1.4759117090312197e-05,
-      "loss": 0.0061,
+      "epoch": 2.0193236714975846,
+      "grad_norm": 0.09085428714752197,
+      "learning_rate": 8.06949806949807e-05,
+      "loss": 0.0282,
       "step": 418
     },
     {
-      "epoch": 3.852873563218391,
-      "grad_norm": 0.01850154623389244,
-      "learning_rate": 1.453057412720536e-05,
-      "loss": 0.0067,
+      "epoch": 2.024154589371981,
+      "grad_norm": 0.07310831546783447,
+      "learning_rate": 8.088803088803089e-05,
+      "loss": 0.0259,
       "step": 419
     },
     {
-      "epoch": 3.862068965517241,
-      "grad_norm": 0.01711280085146427,
-      "learning_rate": 1.4303513272105057e-05,
-      "loss": 0.0065,
-      "step": 420
-    },
-    {
-      "epoch": 3.862068965517241,
-      "eval_loss": 0.011265527456998825,
-      "eval_runtime": 7.5976,
-      "eval_samples_per_second": 6.581,
-      "eval_steps_per_second": 1.711,
+      "epoch": 2.028985507246377,
+      "grad_norm": 0.07102567702531815,
+      "learning_rate": 8.108108108108109e-05,
+      "loss": 0.0217,
       "step": 420
     },
     {
-      "epoch": 3.871264367816092,
-      "grad_norm": 0.019471535459160805,
-      "learning_rate": 1.4077944012864636e-05,
-      "loss": 0.0072,
+      "epoch": 2.033816425120773,
+      "grad_norm": 0.0762273445725441,
+      "learning_rate": 8.127413127413128e-05,
+      "loss": 0.0287,
       "step": 421
     },
     {
-      "epoch": 3.8804597701149426,
-      "grad_norm": 0.020510006695985794,
-      "learning_rate": 1.3853875775010355e-05,
-      "loss": 0.0078,
+      "epoch": 2.0386473429951693,
+      "grad_norm": 0.0700920894742012,
+      "learning_rate": 8.146718146718148e-05,
+      "loss": 0.0251,
       "step": 422
     },
     {
-      "epoch": 3.889655172413793,
-      "grad_norm": 0.021030457690358162,
-      "learning_rate": 1.3631317921347563e-05,
-      "loss": 0.007,
+      "epoch": 2.0434782608695654,
+      "grad_norm": 0.08108256012201309,
+      "learning_rate": 8.166023166023166e-05,
+      "loss": 0.0274,
       "step": 423
     },
     {
-      "epoch": 3.8988505747126436,
-      "grad_norm": 0.01627511717379093,
-      "learning_rate": 1.3410279751569399e-05,
-      "loss": 0.0066,
+      "epoch": 2.0483091787439616,
+      "grad_norm": 0.07796541601419449,
+      "learning_rate": 8.185328185328186e-05,
+      "loss": 0.0222,
       "step": 424
     },
     {
-      "epoch": 3.9080459770114944,
-      "grad_norm": 0.018043728545308113,
-      "learning_rate": 1.3190770501868243e-05,
-      "loss": 0.0069,
+      "epoch": 2.0531400966183573,
+      "grad_norm": 0.08583205938339233,
+      "learning_rate": 8.204633204633205e-05,
+      "loss": 0.0283,
       "step": 425
     },
     {
-      "epoch": 3.9080459770114944,
-      "eval_loss": 0.011119170114398003,
-      "eval_runtime": 7.5908,
-      "eval_samples_per_second": 6.587,
-      "eval_steps_per_second": 1.713,
+      "epoch": 2.0531400966183573,
+      "eval_loss": 0.03210844099521637,
+      "eval_runtime": 21.3092,
+      "eval_samples_per_second": 4.693,
+      "eval_steps_per_second": 0.141,
       "step": 425
     },
     {
-      "epoch": 3.9172413793103447,
-      "grad_norm": 0.01714397221803665,
-      "learning_rate": 1.297279934454978e-05,
-      "loss": 0.007,
+      "epoch": 2.0579710144927534,
+      "grad_norm": 0.0870027244091034,
+      "learning_rate": 8.223938223938225e-05,
+      "loss": 0.0316,
       "step": 426
     },
     {
-      "epoch": 3.9264367816091954,
-      "grad_norm": 0.017846615985035896,
-      "learning_rate": 1.2756375387649716e-05,
-      "loss": 0.007,
+      "epoch": 2.0628019323671496,
+      "grad_norm": 0.11122583597898483,
+      "learning_rate": 8.243243243243243e-05,
+      "loss": 0.0345,
       "step": 427
     },
     {
-      "epoch": 3.935632183908046,
-      "grad_norm": 0.01910008303821087,
-      "learning_rate": 1.25415076745532e-05,
-      "loss": 0.0069,
+      "epoch": 2.0676328502415457,
+      "grad_norm": 0.07087469846010208,
+      "learning_rate": 8.262548262548263e-05,
+      "loss": 0.0256,
       "step": 428
     },
     {
-      "epoch": 3.9448275862068964,
-      "grad_norm": 0.015933765098452568,
-      "learning_rate": 1.2328205183616965e-05,
-      "loss": 0.0064,
+      "epoch": 2.072463768115942,
+      "grad_norm": 0.07167962193489075,
+      "learning_rate": 8.281853281853282e-05,
+      "loss": 0.029,
       "step": 429
     },
     {
-      "epoch": 3.954022988505747,
-      "grad_norm": 0.018724463880062103,
-      "learning_rate": 1.2116476827794104e-05,
-      "loss": 0.0065,
-      "step": 430
-    },
-    {
-      "epoch": 3.954022988505747,
-      "eval_loss": 0.01107843779027462,
-      "eval_runtime": 7.5853,
-      "eval_samples_per_second": 6.592,
-      "eval_steps_per_second": 1.714,
+      "epoch": 2.077294685990338,
+      "grad_norm": 0.0908481627702713,
+      "learning_rate": 8.301158301158302e-05,
+      "loss": 0.0337,
       "step": 430
     },
     {
-      "epoch": 3.963218390804598,
-      "grad_norm": 0.018081670626997948,
-      "learning_rate": 1.1906331454261704e-05,
-      "loss": 0.006,
+      "epoch": 2.082125603864734,
+      "grad_norm": 0.14270740747451782,
+      "learning_rate": 8.32046332046332e-05,
+      "loss": 0.0605,
       "step": 431
     },
     {
-      "epoch": 3.972413793103448,
-      "grad_norm": 0.023170167580246925,
-      "learning_rate": 1.1697777844051105e-05,
-      "loss": 0.0071,
+      "epoch": 2.0869565217391304,
+      "grad_norm": 0.08899253606796265,
+      "learning_rate": 8.33976833976834e-05,
+      "loss": 0.0232,
       "step": 432
     },
     {
-      "epoch": 3.981609195402299,
-      "grad_norm": 0.017531078308820724,
-      "learning_rate": 1.1490824711681025e-05,
-      "loss": 0.0068,
+      "epoch": 2.0917874396135265,
+      "grad_norm": 0.08228383958339691,
+      "learning_rate": 8.35907335907336e-05,
+      "loss": 0.022,
       "step": 433
     },
     {
-      "epoch": 3.9908045977011493,
-      "grad_norm": 0.018324997276067734,
-      "learning_rate": 1.1285480704793377e-05,
-      "loss": 0.0068,
+      "epoch": 2.0966183574879227,
+      "grad_norm": 0.08265294134616852,
+      "learning_rate": 8.378378378378379e-05,
+      "loss": 0.026,
       "step": 434
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 0.020008638501167297,
-      "learning_rate": 1.1081754403791999e-05,
-      "loss": 0.0076,
-      "step": 435
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 0.0111029502004385,
-      "eval_runtime": 7.5917,
-      "eval_samples_per_second": 6.586,
-      "eval_steps_per_second": 1.712,
+      "epoch": 2.101449275362319,
+      "grad_norm": 0.08507855236530304,
+      "learning_rate": 8.397683397683399e-05,
+      "loss": 0.0286,
       "step": 435
     },
     {
-      "epoch": 4.00919540229885,
-      "grad_norm": 0.014917205087840557,
-      "learning_rate": 1.0879654321484012e-05,
-      "loss": 0.0051,
+      "epoch": 2.106280193236715,
+      "grad_norm": 0.09821953624486923,
+      "learning_rate": 8.416988416988417e-05,
+      "loss": 0.0233,
       "step": 436
     },
     {
-      "epoch": 4.0183908045977015,
-      "grad_norm": 0.014704613946378231,
-      "learning_rate": 1.0679188902724191e-05,
-      "loss": 0.0051,
+      "epoch": 2.111111111111111,
+      "grad_norm": 0.09826794266700745,
+      "learning_rate": 8.436293436293437e-05,
+      "loss": 0.0245,
       "step": 437
     },
     {
-      "epoch": 4.027586206896552,
-      "grad_norm": 0.01385421585291624,
-      "learning_rate": 1.0480366524062042e-05,
-      "loss": 0.0052,
+      "epoch": 2.1159420289855073,
+      "grad_norm": 0.06491506844758987,
+      "learning_rate": 8.455598455598456e-05,
+      "loss": 0.0238,
       "step": 438
     },
     {
-      "epoch": 4.036781609195402,
-      "grad_norm": 0.017096884548664093,
-      "learning_rate": 1.0283195493391823e-05,
-      "loss": 0.006,
+      "epoch": 2.1207729468599035,
+      "grad_norm": 0.09309545159339905,
+      "learning_rate": 8.474903474903476e-05,
+      "loss": 0.0244,
       "step": 439
     },
     {
-      "epoch": 4.045977011494253,
-      "grad_norm": 0.020865080878138542,
-      "learning_rate": 1.008768404960535e-05,
-      "loss": 0.0047,
-      "step": 440
-    },
-    {
-      "epoch": 4.045977011494253,
-      "eval_loss": 0.011515479534864426,
-      "eval_runtime": 7.5882,
-      "eval_samples_per_second": 6.589,
-      "eval_steps_per_second": 1.713,
+      "epoch": 2.1256038647342996,
+      "grad_norm": 0.07788459956645966,
+      "learning_rate": 8.494208494208494e-05,
+      "loss": 0.0316,
       "step": 440
     },
     {
-      "epoch": 4.055172413793104,
-      "grad_norm": 0.016257256269454956,
-      "learning_rate": 9.893840362247809e-06,
-      "loss": 0.0051,
+      "epoch": 2.130434782608696,
+      "grad_norm": 0.08980804681777954,
+      "learning_rate": 8.513513513513514e-05,
+      "loss": 0.035,
       "step": 441
     },
     {
-      "epoch": 4.064367816091954,
-      "grad_norm": 0.01711389049887657,
-      "learning_rate": 9.701672531176286e-06,
-      "loss": 0.0056,
+      "epoch": 2.135265700483092,
+      "grad_norm": 0.11862210929393768,
+      "learning_rate": 8.532818532818533e-05,
+      "loss": 0.0355,
       "step": 442
     },
     {
-      "epoch": 4.073563218390804,
-      "grad_norm": 0.018976813182234764,
-      "learning_rate": 9.511188586221376e-06,
-      "loss": 0.0056,
+      "epoch": 2.140096618357488,
+      "grad_norm": 0.09856075048446655,
+      "learning_rate": 8.552123552123553e-05,
+      "loss": 0.042,
       "step": 443
     },
     {
-      "epoch": 4.082758620689655,
-      "grad_norm": 0.016858583316206932,
-      "learning_rate": 9.322396486851626e-06,
-      "loss": 0.0051,
+      "epoch": 2.1449275362318843,
+      "grad_norm": 0.06776445358991623,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 0.0211,
       "step": 444
     },
     {
-      "epoch": 4.091954022988506,
-      "grad_norm": 0.016142934560775757,
-      "learning_rate": 9.135304121840976e-06,
-      "loss": 0.0053,
-      "step": 445
-    },
-    {
-      "epoch": 4.091954022988506,
-      "eval_loss": 0.01191615965217352,
-      "eval_runtime": 7.5902,
-      "eval_samples_per_second": 6.587,
-      "eval_steps_per_second": 1.713,
+      "epoch": 2.14975845410628,
+      "grad_norm": 0.0670533999800682,
+      "learning_rate": 8.590733590733591e-05,
+      "loss": 0.0216,
       "step": 445
     },
     {
-      "epoch": 4.101149425287356,
-      "grad_norm": 0.024709809571504593,
-      "learning_rate": 8.949919308939082e-06,
-      "loss": 0.0058,
+      "epoch": 2.154589371980676,
+      "grad_norm": 0.07002390176057816,
+      "learning_rate": 8.61003861003861e-05,
+      "loss": 0.0252,
       "step": 446
     },
     {
-      "epoch": 4.110344827586207,
-      "grad_norm": 0.015532772988080978,
-      "learning_rate": 8.766249794544662e-06,
-      "loss": 0.0052,
+      "epoch": 2.1594202898550723,
+      "grad_norm": 0.07812980562448502,
+      "learning_rate": 8.62934362934363e-05,
+      "loss": 0.0363,
       "step": 447
     },
     {
-      "epoch": 4.119540229885057,
-      "grad_norm": 0.015922829508781433,
-      "learning_rate": 8.584303253381847e-06,
-      "loss": 0.0057,
+      "epoch": 2.1642512077294684,
+      "grad_norm": 0.08270443975925446,
+      "learning_rate": 8.64864864864865e-05,
+      "loss": 0.0257,
       "step": 448
     },
     {
-      "epoch": 4.128735632183908,
-      "grad_norm": 0.024093857035040855,
-      "learning_rate": 8.404087288179424e-06,
-      "loss": 0.0052,
+      "epoch": 2.1690821256038646,
+      "grad_norm": 0.08274964988231659,
+      "learning_rate": 8.667953667953668e-05,
+      "loss": 0.033,
       "step": 449
     },
     {
-      "epoch": 4.137931034482759,
-      "grad_norm": 0.018626809120178223,
-      "learning_rate": 8.225609429353187e-06,
-      "loss": 0.0053,
+      "epoch": 2.1739130434782608,
+      "grad_norm": 0.0919279232621193,
+      "learning_rate": 8.687258687258688e-05,
+      "loss": 0.0275,
       "step": 450
     },
     {
-      "epoch": 4.137931034482759,
-      "eval_loss": 0.011974362656474113,
-      "eval_runtime": 7.6303,
-      "eval_samples_per_second": 6.553,
-      "eval_steps_per_second": 1.704,
+      "epoch": 2.1739130434782608,
+      "eval_loss": 0.03367118537425995,
+      "eval_runtime": 20.6662,
+      "eval_samples_per_second": 4.839,
+      "eval_steps_per_second": 0.145,
       "step": 450
     },
     {
-      "epoch": 4.147126436781609,
-      "grad_norm": 0.02421395108103752,
-      "learning_rate": 8.048877134691268e-06,
-      "loss": 0.0067,
+      "epoch": 2.178743961352657,
+      "grad_norm": 0.0769813284277916,
+      "learning_rate": 8.706563706563707e-05,
+      "loss": 0.0219,
       "step": 451
     },
     {
-      "epoch": 4.1563218390804595,
-      "grad_norm": 0.019452223554253578,
-      "learning_rate": 7.873897789042523e-06,
-      "loss": 0.0055,
+      "epoch": 2.183574879227053,
+      "grad_norm": 0.0803806260228157,
+      "learning_rate": 8.725868725868727e-05,
+      "loss": 0.0304,
       "step": 452
     },
     {
-      "epoch": 4.165517241379311,
-      "grad_norm": 0.01944682002067566,
-      "learning_rate": 7.700678704007947e-06,
-      "loss": 0.0051,
+      "epoch": 2.1884057971014492,
+      "grad_norm": 0.07121909409761429,
+      "learning_rate": 8.745173745173745e-05,
+      "loss": 0.0269,
       "step": 453
     },
     {
-      "epoch": 4.174712643678161,
-      "grad_norm": 0.0154283307492733,
-      "learning_rate": 7.529227117635135e-06,
-      "loss": 0.0053,
+      "epoch": 2.1932367149758454,
+      "grad_norm": 0.08506215363740921,
+      "learning_rate": 8.764478764478765e-05,
+      "loss": 0.0285,
       "step": 454
     },
     {
-      "epoch": 4.183908045977011,
-      "grad_norm": 0.02008649706840515,
-      "learning_rate": 7.35955019411585e-06,
-      "loss": 0.0055,
-      "step": 455
-    },
-    {
-      "epoch": 4.183908045977011,
-      "eval_loss": 0.01187676191329956,
-      "eval_runtime": 7.6617,
-      "eval_samples_per_second": 6.526,
-      "eval_steps_per_second": 1.697,
+      "epoch": 2.1980676328502415,
+      "grad_norm": 0.07936449348926544,
+      "learning_rate": 8.783783783783784e-05,
+      "loss": 0.0235,
       "step": 455
     },
     {
-      "epoch": 4.1931034482758625,
-      "grad_norm": 0.01813184656202793,
-      "learning_rate": 7.191655023486682e-06,
-      "loss": 0.0054,
+      "epoch": 2.2028985507246377,
+      "grad_norm": 0.084862120449543,
+      "learning_rate": 8.803088803088804e-05,
+      "loss": 0.0307,
       "step": 456
     },
     {
-      "epoch": 4.202298850574713,
-      "grad_norm": 0.019368577748537064,
-      "learning_rate": 7.02554862133275e-06,
-      "loss": 0.0067,
+      "epoch": 2.207729468599034,
+      "grad_norm": 0.08003883808851242,
+      "learning_rate": 8.822393822393822e-05,
+      "loss": 0.025,
       "step": 457
     },
     {
-      "epoch": 4.211494252873563,
-      "grad_norm": 0.02604977786540985,
-      "learning_rate": 6.861237928494579e-06,
-      "loss": 0.0044,
+      "epoch": 2.21256038647343,
+      "grad_norm": 0.07105076313018799,
+      "learning_rate": 8.841698841698842e-05,
+      "loss": 0.0225,
       "step": 458
     },
     {
-      "epoch": 4.220689655172414,
-      "grad_norm": 0.016761422157287598,
-      "learning_rate": 6.698729810778065e-06,
-      "loss": 0.0053,
+      "epoch": 2.217391304347826,
+      "grad_norm": 0.07181154191493988,
+      "learning_rate": 8.861003861003861e-05,
+      "loss": 0.0207,
       "step": 459
     },
     {
-      "epoch": 4.2298850574712645,
-      "grad_norm": 0.015887631103396416,
-      "learning_rate": 6.53803105866761e-06,
-      "loss": 0.0053,
-      "step": 460
-    },
-    {
-      "epoch": 4.2298850574712645,
-      "eval_loss": 0.011668752878904343,
-      "eval_runtime": 7.5746,
-      "eval_samples_per_second": 6.601,
-      "eval_steps_per_second": 1.716,
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.09786058962345123,
+      "learning_rate": 8.880308880308881e-05,
+      "loss": 0.0314,
       "step": 460
     },
     {
-      "epoch": 4.239080459770115,
-      "grad_norm": 0.020124757662415504,
-      "learning_rate": 6.379148387042316e-06,
-      "loss": 0.0045,
+      "epoch": 2.2270531400966185,
+      "grad_norm": 0.08781251311302185,
+      "learning_rate": 8.899613899613901e-05,
+      "loss": 0.0248,
       "step": 461
     },
     {
-      "epoch": 4.248275862068965,
-      "grad_norm": 0.015533823519945145,
-      "learning_rate": 6.222088434895462e-06,
-      "loss": 0.0045,
+      "epoch": 2.2318840579710146,
+      "grad_norm": 0.08720511198043823,
+      "learning_rate": 8.918918918918919e-05,
+      "loss": 0.029,
       "step": 462
     },
     {
-      "epoch": 4.257471264367816,
-      "grad_norm": 0.01983262039721012,
-      "learning_rate": 6.066857765057055e-06,
-      "loss": 0.0057,
+      "epoch": 2.236714975845411,
+      "grad_norm": 0.08457627147436142,
+      "learning_rate": 8.938223938223939e-05,
+      "loss": 0.0286,
       "step": 463
     },
     {
-      "epoch": 4.266666666666667,
-      "grad_norm": 0.01795038767158985,
-      "learning_rate": 5.9134628639196e-06,
-      "loss": 0.0058,
+      "epoch": 2.241545893719807,
+      "grad_norm": 0.07502002269029617,
+      "learning_rate": 8.957528957528958e-05,
+      "loss": 0.0265,
       "step": 464
     },
     {
-      "epoch": 4.275862068965517,
-      "grad_norm": 0.021382272243499756,
-      "learning_rate": 5.7619101411671095e-06,
-      "loss": 0.0053,
-      "step": 465
-    },
-    {
-      "epoch": 4.275862068965517,
-      "eval_loss": 0.011690051294863224,
-      "eval_runtime": 7.6002,
-      "eval_samples_per_second": 6.579,
-      "eval_steps_per_second": 1.71,
+      "epoch": 2.246376811594203,
+      "grad_norm": 0.06273361295461655,
+      "learning_rate": 8.976833976833978e-05,
+      "loss": 0.021,
       "step": 465
     },
     {
-      "epoch": 4.285057471264368,
-      "grad_norm": 0.019680311903357506,
-      "learning_rate": 5.6122059295072085e-06,
-      "loss": 0.0055,
+      "epoch": 2.2512077294685993,
+      "grad_norm": 0.0748235285282135,
+      "learning_rate": 8.996138996138996e-05,
+      "loss": 0.0291,
       "step": 466
     },
     {
-      "epoch": 4.294252873563218,
-      "grad_norm": 0.018579309806227684,
-      "learning_rate": 5.464356484406535e-06,
-      "loss": 0.0047,
+      "epoch": 2.2560386473429954,
+      "grad_norm": 0.06298107653856277,
+      "learning_rate": 9.015444015444016e-05,
+      "loss": 0.02,
       "step": 467
     },
     {
-      "epoch": 4.303448275862069,
-      "grad_norm": 0.020488621667027473,
-      "learning_rate": 5.318367983829392e-06,
-      "loss": 0.0061,
+      "epoch": 2.260869565217391,
+      "grad_norm": 0.08730265498161316,
+      "learning_rate": 9.034749034749035e-05,
+      "loss": 0.0311,
       "step": 468
     },
     {
-      "epoch": 4.31264367816092,
-      "grad_norm": 0.018219981342554092,
-      "learning_rate": 5.174246527979531e-06,
-      "loss": 0.005,
+      "epoch": 2.2657004830917873,
+      "grad_norm": 0.07982222735881805,
+      "learning_rate": 9.054054054054055e-05,
+      "loss": 0.031,
       "step": 469
     },
     {
-      "epoch": 4.32183908045977,
-      "grad_norm": 0.01655331440269947,
-      "learning_rate": 5.031998139045352e-06,
-      "loss": 0.0053,
-      "step": 470
-    },
-    {
-      "epoch": 4.32183908045977,
-      "eval_loss": 0.011675745248794556,
-      "eval_runtime": 7.6014,
-      "eval_samples_per_second": 6.578,
-      "eval_steps_per_second": 1.71,
+      "epoch": 2.2705314009661834,
+      "grad_norm": 0.09329807758331299,
+      "learning_rate": 9.073359073359073e-05,
+      "loss": 0.0309,
       "step": 470
     },
     {
-      "epoch": 4.3310344827586205,
-      "grad_norm": 0.01707630045711994,
-      "learning_rate": 4.891628760948114e-06,
-      "loss": 0.0047,
+      "epoch": 2.2753623188405796,
+      "grad_norm": 0.08422908186912537,
+      "learning_rate": 9.092664092664093e-05,
+      "loss": 0.0253,
       "step": 471
     },
     {
-      "epoch": 4.340229885057472,
-      "grad_norm": 0.015081142075359821,
-      "learning_rate": 4.7531442590937335e-06,
-      "loss": 0.0054,
+      "epoch": 2.2801932367149758,
+      "grad_norm": 0.08790479600429535,
+      "learning_rate": 9.111969111969112e-05,
+      "loss": 0.0236,
       "step": 472
     },
     {
-      "epoch": 4.349425287356322,
-      "grad_norm": 0.021828968077898026,
-      "learning_rate": 4.616550420127563e-06,
-      "loss": 0.0052,
+      "epoch": 2.285024154589372,
+      "grad_norm": 0.09036684781312943,
+      "learning_rate": 9.131274131274132e-05,
+      "loss": 0.0273,
       "step": 473
     },
     {
-      "epoch": 4.358620689655172,
-      "grad_norm": 0.01851847767829895,
-      "learning_rate": 4.4818529516926726e-06,
-      "loss": 0.006,
+      "epoch": 2.289855072463768,
+      "grad_norm": 0.08374612033367157,
+      "learning_rate": 9.15057915057915e-05,
+      "loss": 0.0325,
       "step": 474
     },
     {
-      "epoch": 4.3678160919540225,
-      "grad_norm": 0.01676092855632305,
-      "learning_rate": 4.349057482191299e-06,
-      "loss": 0.0058,
+      "epoch": 2.2946859903381642,
+      "grad_norm": 0.07355117052793503,
+      "learning_rate": 9.16988416988417e-05,
+      "loss": 0.026,
       "step": 475
     },
     {
-      "epoch": 4.3678160919540225,
-      "eval_loss": 0.011636043898761272,
-      "eval_runtime": 7.5917,
-      "eval_samples_per_second": 6.586,
-      "eval_steps_per_second": 1.712,
+      "epoch": 2.2946859903381642,
+      "eval_loss": 0.03135251998901367,
+      "eval_runtime": 20.6353,
+      "eval_samples_per_second": 4.846,
+      "eval_steps_per_second": 0.145,
       "step": 475
     },
     {
-      "epoch": 4.377011494252874,
-      "grad_norm": 0.018017876893281937,
-      "learning_rate": 4.218169560549706e-06,
-      "loss": 0.0055,
+      "epoch": 2.2995169082125604,
+      "grad_norm": 0.13940849900245667,
+      "learning_rate": 9.18918918918919e-05,
+      "loss": 0.0299,
       "step": 476
     },
     {
-      "epoch": 4.386206896551724,
-      "grad_norm": 0.018705466762185097,
-      "learning_rate": 4.089194655986306e-06,
-      "loss": 0.005,
+      "epoch": 2.3043478260869565,
+      "grad_norm": 0.09694752097129822,
+      "learning_rate": 9.208494208494209e-05,
+      "loss": 0.0342,
       "step": 477
     },
     {
-      "epoch": 4.395402298850574,
-      "grad_norm": 0.016500435769557953,
-      "learning_rate": 3.962138157783085e-06,
-      "loss": 0.0044,
+      "epoch": 2.3091787439613527,
+      "grad_norm": 0.07969164848327637,
+      "learning_rate": 9.227799227799229e-05,
+      "loss": 0.0265,
       "step": 478
     },
     {
-      "epoch": 4.4045977011494255,
-      "grad_norm": 0.016678597778081894,
-      "learning_rate": 3.837005375060482e-06,
-      "loss": 0.0049,
+      "epoch": 2.314009661835749,
+      "grad_norm": 0.07230686396360397,
+      "learning_rate": 9.247104247104247e-05,
+      "loss": 0.0255,
       "step": 479
     },
     {
-      "epoch": 4.413793103448276,
-      "grad_norm": 0.01764843426644802,
-      "learning_rate": 3.7138015365554833e-06,
-      "loss": 0.0053,
-      "step": 480
-    },
-    {
-      "epoch": 4.413793103448276,
-      "eval_loss": 0.011646818369626999,
-      "eval_runtime": 7.5865,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 2.318840579710145,
+      "grad_norm": 0.08269919455051422,
+      "learning_rate": 9.266409266409267e-05,
+      "loss": 0.0327,
       "step": 480
     },
     {
-      "epoch": 4.422988505747126,
-      "grad_norm": 0.018464326858520508,
-      "learning_rate": 3.5925317904031587e-06,
-      "loss": 0.0046,
+      "epoch": 2.323671497584541,
+      "grad_norm": 0.07120717316865921,
+      "learning_rate": 9.285714285714286e-05,
+      "loss": 0.0229,
       "step": 481
     },
     {
-      "epoch": 4.432183908045977,
-      "grad_norm": 0.016741527244448662,
-      "learning_rate": 3.4732012039215776e-06,
-      "loss": 0.0051,
+      "epoch": 2.3285024154589373,
+      "grad_norm": 0.0936795324087143,
+      "learning_rate": 9.305019305019306e-05,
+      "loss": 0.0261,
       "step": 482
     },
     {
-      "epoch": 4.441379310344828,
-      "grad_norm": 0.015406622551381588,
-      "learning_rate": 3.3558147633999728e-06,
-      "loss": 0.0051,
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.07093029469251633,
+      "learning_rate": 9.324324324324324e-05,
+      "loss": 0.0215,
       "step": 483
     },
     {
-      "epoch": 4.450574712643678,
-      "grad_norm": 0.018325865268707275,
-      "learning_rate": 3.2403773738905187e-06,
-      "loss": 0.0049,
+      "epoch": 2.3381642512077296,
+      "grad_norm": 0.0930793434381485,
+      "learning_rate": 9.343629343629344e-05,
+      "loss": 0.0285,
       "step": 484
     },
     {
-      "epoch": 4.459770114942529,
-      "grad_norm": 0.01707405038177967,
-      "learning_rate": 3.126893859003249e-06,
-      "loss": 0.0053,
-      "step": 485
-    },
-    {
-      "epoch": 4.459770114942529,
-      "eval_loss": 0.011750674806535244,
-      "eval_runtime": 7.5766,
-      "eval_samples_per_second": 6.599,
-      "eval_steps_per_second": 1.716,
+      "epoch": 2.342995169082126,
+      "grad_norm": 0.05888902395963669,
+      "learning_rate": 9.362934362934363e-05,
+      "loss": 0.0186,
       "step": 485
     },
     {
-      "epoch": 4.468965517241379,
-      "grad_norm": 0.01917904056608677,
-      "learning_rate": 3.0153689607045845e-06,
-      "loss": 0.0051,
+      "epoch": 2.3478260869565215,
+      "grad_norm": 0.07734473794698715,
+      "learning_rate": 9.382239382239383e-05,
+      "loss": 0.0326,
       "step": 486
     },
     {
-      "epoch": 4.47816091954023,
-      "grad_norm": 0.02045316807925701,
-      "learning_rate": 2.9058073391191375e-06,
-      "loss": 0.0054,
+      "epoch": 2.3526570048309177,
+      "grad_norm": 0.10177651047706604,
+      "learning_rate": 9.401544401544401e-05,
+      "loss": 0.0305,
       "step": 487
     },
     {
-      "epoch": 4.487356321839081,
-      "grad_norm": 0.017735499888658524,
-      "learning_rate": 2.798213572335001e-06,
-      "loss": 0.0061,
+      "epoch": 2.357487922705314,
+      "grad_norm": 0.08325226604938507,
+      "learning_rate": 9.420849420849421e-05,
+      "loss": 0.0372,
       "step": 488
     },
     {
-      "epoch": 4.496551724137931,
-      "grad_norm": 0.01793784834444523,
-      "learning_rate": 2.692592156212487e-06,
-      "loss": 0.0063,
+      "epoch": 2.36231884057971,
+      "grad_norm": 0.068690724670887,
+      "learning_rate": 9.44015444015444e-05,
+      "loss": 0.0352,
       "step": 489
     },
     {
-      "epoch": 4.505747126436781,
-      "grad_norm": 0.01671445369720459,
-      "learning_rate": 2.5889475041961765e-06,
-      "loss": 0.0051,
-      "step": 490
-    },
-    {
-      "epoch": 4.505747126436781,
-      "eval_loss": 0.011716877110302448,
-      "eval_runtime": 7.5908,
-      "eval_samples_per_second": 6.587,
-      "eval_steps_per_second": 1.713,
+      "epoch": 2.367149758454106,
+      "grad_norm": 0.08754605799913406,
+      "learning_rate": 9.45945945945946e-05,
+      "loss": 0.0269,
       "step": 490
     },
     {
-      "epoch": 4.514942528735633,
-      "grad_norm": 0.01827097497880459,
-      "learning_rate": 2.4872839471306084e-06,
-      "loss": 0.0052,
+      "epoch": 2.3719806763285023,
+      "grad_norm": 0.06412243843078613,
+      "learning_rate": 9.47876447876448e-05,
+      "loss": 0.0248,
       "step": 491
     },
     {
-      "epoch": 4.524137931034483,
-      "grad_norm": 0.018106255680322647,
-      "learning_rate": 2.3876057330792346e-06,
-      "loss": 0.0061,
+      "epoch": 2.3768115942028984,
+      "grad_norm": 0.07275252044200897,
+      "learning_rate": 9.498069498069498e-05,
+      "loss": 0.0279,
       "step": 492
     },
     {
-      "epoch": 4.533333333333333,
-      "grad_norm": 0.021528033539652824,
-      "learning_rate": 2.2899170271469428e-06,
-      "loss": 0.0057,
+      "epoch": 2.3816425120772946,
+      "grad_norm": 0.07856742292642593,
+      "learning_rate": 9.517374517374518e-05,
+      "loss": 0.0386,
       "step": 493
     },
     {
-      "epoch": 4.5425287356321835,
-      "grad_norm": 0.018092922866344452,
-      "learning_rate": 2.1942219113060212e-06,
-      "loss": 0.0045,
+      "epoch": 2.3864734299516908,
+      "grad_norm": 0.08813096582889557,
+      "learning_rate": 9.536679536679537e-05,
+      "loss": 0.0283,
       "step": 494
     },
     {
-      "epoch": 4.551724137931035,
-      "grad_norm": 0.018459059298038483,
-      "learning_rate": 2.100524384225555e-06,
-      "loss": 0.0053,
-      "step": 495
-    },
-    {
-      "epoch": 4.551724137931035,
-      "eval_loss": 0.011710132472217083,
-      "eval_runtime": 7.5789,
-      "eval_samples_per_second": 6.597,
-      "eval_steps_per_second": 1.715,
+      "epoch": 2.391304347826087,
+      "grad_norm": 0.07463379204273224,
+      "learning_rate": 9.555984555984557e-05,
+      "loss": 0.035,
       "step": 495
     },
     {
-      "epoch": 4.560919540229885,
-      "grad_norm": 0.018115947023034096,
-      "learning_rate": 2.0088283611044036e-06,
-      "loss": 0.005,
+      "epoch": 2.396135265700483,
+      "grad_norm": 0.0677594393491745,
+      "learning_rate": 9.575289575289575e-05,
+      "loss": 0.0245,
       "step": 496
     },
     {
-      "epoch": 4.570114942528735,
-      "grad_norm": 0.021283112466335297,
-      "learning_rate": 1.9191376735075427e-06,
-      "loss": 0.0051,
+      "epoch": 2.4009661835748792,
+      "grad_norm": 0.0777975544333458,
+      "learning_rate": 9.594594594594595e-05,
+      "loss": 0.0222,
       "step": 497
     },
     {
-      "epoch": 4.5793103448275865,
-      "grad_norm": 0.019610431045293808,
-      "learning_rate": 1.8314560692059835e-06,
-      "loss": 0.0051,
+      "epoch": 2.4057971014492754,
+      "grad_norm": 0.0731094554066658,
+      "learning_rate": 9.613899613899614e-05,
+      "loss": 0.0259,
       "step": 498
     },
     {
-      "epoch": 4.588505747126437,
-      "grad_norm": 0.02000650204718113,
-      "learning_rate": 1.7457872120201779e-06,
-      "loss": 0.0049,
+      "epoch": 2.4106280193236715,
+      "grad_norm": 0.07815240323543549,
+      "learning_rate": 9.633204633204634e-05,
+      "loss": 0.0246,
       "step": 499
     },
     {
-      "epoch": 4.597701149425287,
-      "grad_norm": 0.019521350041031837,
-      "learning_rate": 1.6621346816668992e-06,
-      "loss": 0.0059,
+      "epoch": 2.4154589371980677,
+      "grad_norm": 0.07070033997297287,
+      "learning_rate": 9.652509652509652e-05,
+      "loss": 0.0244,
       "step": 500
     },
     {
-      "epoch": 4.597701149425287,
-      "eval_loss": 0.01169579103589058,
-      "eval_runtime": 7.5788,
-      "eval_samples_per_second": 6.597,
-      "eval_steps_per_second": 1.715,
+      "epoch": 2.4154589371980677,
+      "eval_loss": 0.02849414572119713,
+      "eval_runtime": 20.5961,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
       "step": 500
     },
     {
-      "epoch": 4.606896551724138,
-      "grad_norm": 0.019491534680128098,
-      "learning_rate": 1.5805019736097104e-06,
-      "loss": 0.0058,
+      "epoch": 2.420289855072464,
+      "grad_norm": 0.07829859107732773,
+      "learning_rate": 9.671814671814672e-05,
+      "loss": 0.0208,
       "step": 501
     },
     {
-      "epoch": 4.6160919540229886,
-      "grad_norm": 0.01981920190155506,
-      "learning_rate": 1.5008924989128258e-06,
-      "loss": 0.0058,
+      "epoch": 2.42512077294686,
+      "grad_norm": 0.10270782560110092,
+      "learning_rate": 9.691119691119691e-05,
+      "loss": 0.0294,
       "step": 502
     },
     {
-      "epoch": 4.625287356321839,
-      "grad_norm": 0.01777651347219944,
-      "learning_rate": 1.4233095840986753e-06,
-      "loss": 0.0055,
+      "epoch": 2.429951690821256,
+      "grad_norm": 0.08430244773626328,
+      "learning_rate": 9.710424710424711e-05,
+      "loss": 0.0221,
       "step": 503
     },
     {
-      "epoch": 4.63448275862069,
-      "grad_norm": 0.019680196419358253,
-      "learning_rate": 1.3477564710088098e-06,
-      "loss": 0.0049,
+      "epoch": 2.4347826086956523,
+      "grad_norm": 0.07175758481025696,
+      "learning_rate": 9.729729729729731e-05,
+      "loss": 0.0251,
       "step": 504
     },
     {
-      "epoch": 4.64367816091954,
-      "grad_norm": 0.01789335533976555,
-      "learning_rate": 1.2742363166685034e-06,
-      "loss": 0.0055,
-      "step": 505
-    },
-    {
-      "epoch": 4.64367816091954,
-      "eval_loss": 0.011671243235468864,
-      "eval_runtime": 7.6173,
-      "eval_samples_per_second": 6.564,
-      "eval_steps_per_second": 1.707,
+      "epoch": 2.4396135265700485,
+      "grad_norm": 0.06120215356349945,
+      "learning_rate": 9.74903474903475e-05,
+      "loss": 0.0208,
       "step": 505
     },
     {
-      "epoch": 4.652873563218391,
-      "grad_norm": 0.01689436286687851,
-      "learning_rate": 1.2027521931548214e-06,
-      "loss": 0.0045,
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.07098649442195892,
+      "learning_rate": 9.76833976833977e-05,
+      "loss": 0.0225,
       "step": 506
     },
     {
-      "epoch": 4.662068965517241,
-      "grad_norm": 0.016485173255205154,
-      "learning_rate": 1.1333070874682216e-06,
-      "loss": 0.0055,
+      "epoch": 2.449275362318841,
+      "grad_norm": 0.07696057856082916,
+      "learning_rate": 9.787644787644788e-05,
+      "loss": 0.028,
       "step": 507
     },
     {
-      "epoch": 4.671264367816092,
-      "grad_norm": 0.02044127695262432,
-      "learning_rate": 1.0659039014077944e-06,
-      "loss": 0.0046,
+      "epoch": 2.454106280193237,
+      "grad_norm": 0.09887323528528214,
+      "learning_rate": 9.806949806949808e-05,
+      "loss": 0.0377,
       "step": 508
     },
     {
-      "epoch": 4.680459770114942,
-      "grad_norm": 0.019011957570910454,
-      "learning_rate": 1.0005454514499414e-06,
-      "loss": 0.0055,
+      "epoch": 2.4589371980676327,
+      "grad_norm": 0.06533748656511307,
+      "learning_rate": 9.826254826254826e-05,
+      "loss": 0.0252,
       "step": 509
     },
     {
-      "epoch": 4.689655172413794,
-      "grad_norm": 0.016083214432001114,
-      "learning_rate": 9.372344686307655e-07,
-      "loss": 0.0054,
-      "step": 510
-    },
-    {
-      "epoch": 4.689655172413794,
-      "eval_loss": 0.011643964797258377,
-      "eval_runtime": 7.5863,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 2.463768115942029,
+      "grad_norm": 0.06457691639661789,
+      "learning_rate": 9.845559845559846e-05,
+      "loss": 0.0316,
       "step": 510
     },
     {
-      "epoch": 4.698850574712644,
-      "grad_norm": 0.027418499812483788,
-      "learning_rate": 8.759735984318895e-07,
-      "loss": 0.0057,
+      "epoch": 2.468599033816425,
+      "grad_norm": 0.07615062594413757,
+      "learning_rate": 9.864864864864865e-05,
+      "loss": 0.0262,
       "step": 511
     },
     {
-      "epoch": 4.708045977011494,
-      "grad_norm": 0.016228538006544113,
-      "learning_rate": 8.167654006699443e-07,
-      "loss": 0.0054,
+      "epoch": 2.473429951690821,
+      "grad_norm": 0.05695481225848198,
+      "learning_rate": 9.884169884169885e-05,
+      "loss": 0.0232,
       "step": 512
     },
     {
-      "epoch": 4.7172413793103445,
-      "grad_norm": 0.02453775331377983,
-      "learning_rate": 7.596123493895991e-07,
-      "loss": 0.0059,
+      "epoch": 2.4782608695652173,
+      "grad_norm": 0.06162761151790619,
+      "learning_rate": 9.903474903474904e-05,
+      "loss": 0.0279,
       "step": 513
     },
     {
-      "epoch": 4.726436781609196,
-      "grad_norm": 0.019689669832587242,
-      "learning_rate": 7.04516832760177e-07,
-      "loss": 0.0062,
+      "epoch": 2.4830917874396135,
+      "grad_norm": 0.07245969772338867,
+      "learning_rate": 9.922779922779923e-05,
+      "loss": 0.0242,
       "step": 514
     },
     {
-      "epoch": 4.735632183908046,
-      "grad_norm": 0.017698440700769424,
-      "learning_rate": 6.514811529758747e-07,
-      "loss": 0.0055,
-      "step": 515
-    },
-    {
-      "epoch": 4.735632183908046,
-      "eval_loss": 0.011669199913740158,
-      "eval_runtime": 7.5858,
-      "eval_samples_per_second": 6.591,
-      "eval_steps_per_second": 1.714,
+      "epoch": 2.4879227053140096,
+      "grad_norm": 0.06011466681957245,
+      "learning_rate": 9.942084942084942e-05,
+      "loss": 0.0195,
       "step": 515
     },
     {
-      "epoch": 4.744827586206896,
-      "grad_norm": 0.0169072262942791,
-      "learning_rate": 6.005075261595494e-07,
-      "loss": 0.0053,
+      "epoch": 2.4927536231884058,
+      "grad_norm": 0.07303909212350845,
+      "learning_rate": 9.961389961389962e-05,
+      "loss": 0.0221,
       "step": 516
     },
     {
-      "epoch": 4.7540229885057474,
-      "grad_norm": 0.02021600864827633,
-      "learning_rate": 5.515980822701439e-07,
-      "loss": 0.0055,
+      "epoch": 2.497584541062802,
+      "grad_norm": 0.05043727904558182,
+      "learning_rate": 9.98069498069498e-05,
+      "loss": 0.0134,
       "step": 517
     },
     {
-      "epoch": 4.763218390804598,
-      "grad_norm": 0.021086223423480988,
-      "learning_rate": 5.047548650136513e-07,
-      "loss": 0.0053,
+      "epoch": 2.502415458937198,
+      "grad_norm": 0.06988222897052765,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
       "step": 518
     },
     {
-      "epoch": 4.772413793103448,
-      "grad_norm": 0.02129381150007248,
-      "learning_rate": 4.5997983175773417e-07,
-      "loss": 0.0054,
+      "epoch": 2.5072463768115942,
+      "grad_norm": 0.0786987692117691,
+      "learning_rate": 9.9999988623013e-05,
+      "loss": 0.0276,
       "step": 519
     },
     {
-      "epoch": 4.781609195402299,
-      "grad_norm": 0.018404850736260414,
-      "learning_rate": 4.1727485344994486e-07,
-      "loss": 0.0056,
-      "step": 520
-    },
-    {
-      "epoch": 4.781609195402299,
-      "eval_loss": 0.01162272784858942,
-      "eval_runtime": 7.644,
-      "eval_samples_per_second": 6.541,
-      "eval_steps_per_second": 1.701,
+      "epoch": 2.5120772946859904,
+      "grad_norm": 0.11683060973882675,
+      "learning_rate": 9.999995449205719e-05,
+      "loss": 0.0261,
       "step": 520
     },
     {
-      "epoch": 4.7908045977011495,
-      "grad_norm": 0.01828380487859249,
-      "learning_rate": 3.766417145395218e-07,
-      "loss": 0.0049,
+      "epoch": 2.5169082125603865,
+      "grad_norm": 0.08037126064300537,
+      "learning_rate": 9.999989760714809e-05,
+      "loss": 0.0255,
       "step": 521
     },
     {
-      "epoch": 4.8,
-      "grad_norm": 0.01730126328766346,
-      "learning_rate": 3.380821129028489e-07,
-      "loss": 0.0055,
+      "epoch": 2.5217391304347827,
+      "grad_norm": 0.05731765180826187,
+      "learning_rate": 9.999981796831159e-05,
+      "loss": 0.0201,
       "step": 522
     },
     {
-      "epoch": 4.809195402298851,
-      "grad_norm": 0.017449945211410522,
-      "learning_rate": 3.0159765977250673e-07,
-      "loss": 0.0057,
+      "epoch": 2.526570048309179,
+      "grad_norm": 0.07567796856164932,
+      "learning_rate": 9.999971557558395e-05,
+      "loss": 0.0262,
       "step": 523
     },
     {
-      "epoch": 4.818390804597701,
-      "grad_norm": 0.01770183816552162,
-      "learning_rate": 2.671898796699268e-07,
-      "loss": 0.005,
+      "epoch": 2.531400966183575,
+      "grad_norm": 0.07090573757886887,
+      "learning_rate": 9.999959042901174e-05,
+      "loss": 0.0248,
       "step": 524
     },
     {
-      "epoch": 4.827586206896552,
-      "grad_norm": 0.016674352809786797,
-      "learning_rate": 2.3486021034170857e-07,
-      "loss": 0.0048,
+      "epoch": 2.536231884057971,
+      "grad_norm": 0.08650020509958267,
+      "learning_rate": 9.999944252865192e-05,
+      "loss": 0.0281,
       "step": 525
     },
     {
-      "epoch": 4.827586206896552,
-      "eval_loss": 0.011633777059614658,
-      "eval_runtime": 7.591,
-      "eval_samples_per_second": 6.587,
-      "eval_steps_per_second": 1.713,
+      "epoch": 2.536231884057971,
+      "eval_loss": 0.02850218303501606,
+      "eval_runtime": 20.6152,
+      "eval_samples_per_second": 4.851,
+      "eval_steps_per_second": 0.146,
       "step": 525
     },
     {
-      "epoch": 4.836781609195402,
-      "grad_norm": 0.016963541507720947,
-      "learning_rate": 2.0461000269953456e-07,
-      "loss": 0.0061,
+      "epoch": 2.541062801932367,
+      "grad_norm": 0.0966835767030716,
+      "learning_rate": 9.999927187457181e-05,
+      "loss": 0.0385,
       "step": 526
     },
     {
-      "epoch": 4.845977011494253,
-      "grad_norm": 0.018301161006093025,
-      "learning_rate": 1.7644052076371542e-07,
-      "loss": 0.0054,
+      "epoch": 2.545893719806763,
+      "grad_norm": 0.06122938171029091,
+      "learning_rate": 9.999907846684906e-05,
+      "loss": 0.024,
       "step": 527
     },
     {
-      "epoch": 4.855172413793103,
-      "grad_norm": 0.018041977658867836,
-      "learning_rate": 1.503529416103988e-07,
-      "loss": 0.0048,
+      "epoch": 2.550724637681159,
+      "grad_norm": 0.06990386545658112,
+      "learning_rate": 9.999886230557167e-05,
+      "loss": 0.0259,
       "step": 528
     },
     {
-      "epoch": 4.864367816091954,
-      "grad_norm": 0.016023052856326103,
-      "learning_rate": 1.2634835532233657e-07,
-      "loss": 0.0053,
+      "epoch": 2.5555555555555554,
+      "grad_norm": 0.07949528098106384,
+      "learning_rate": 9.999862339083804e-05,
+      "loss": 0.0266,
       "step": 529
     },
     {
-      "epoch": 4.873563218390805,
-      "grad_norm": 0.01638129912316799,
-      "learning_rate": 1.044277649433989e-07,
-      "loss": 0.0049,
-      "step": 530
-    },
-    {
-      "epoch": 4.873563218390805,
-      "eval_loss": 0.01163394283503294,
-      "eval_runtime": 7.5775,
-      "eval_samples_per_second": 6.598,
-      "eval_steps_per_second": 1.716,
+      "epoch": 2.5603864734299515,
+      "grad_norm": 0.06846433132886887,
+      "learning_rate": 9.999836172275688e-05,
+      "loss": 0.0254,
       "step": 530
     },
     {
-      "epoch": 4.882758620689655,
-      "grad_norm": 0.02009713463485241,
-      "learning_rate": 8.459208643659122e-08,
-      "loss": 0.0047,
+      "epoch": 2.5652173913043477,
+      "grad_norm": 0.06829697638750076,
+      "learning_rate": 9.999807730144728e-05,
+      "loss": 0.0257,
       "step": 531
     },
     {
-      "epoch": 4.8919540229885055,
-      "grad_norm": 0.018126314505934715,
-      "learning_rate": 6.684214864584038e-08,
-      "loss": 0.0048,
+      "epoch": 2.570048309178744,
+      "grad_norm": 0.08349194377660751,
+      "learning_rate": 9.999777012703866e-05,
+      "loss": 0.0256,
       "step": 532
     },
     {
-      "epoch": 4.901149425287357,
-      "grad_norm": 0.014890230260789394,
-      "learning_rate": 5.11786932613223e-08,
-      "loss": 0.005,
+      "epoch": 2.57487922705314,
+      "grad_norm": 0.09388142079114914,
+      "learning_rate": 9.999744019967081e-05,
+      "loss": 0.0202,
       "step": 533
     },
     {
-      "epoch": 4.910344827586207,
-      "grad_norm": 0.016795996576547623,
-      "learning_rate": 3.760237478849793e-08,
-      "loss": 0.0044,
+      "epoch": 2.579710144927536,
+      "grad_norm": 0.06379712373018265,
+      "learning_rate": 9.999708751949389e-05,
+      "loss": 0.0293,
       "step": 534
     },
     {
-      "epoch": 4.919540229885057,
-      "grad_norm": 0.016731204465031624,
-      "learning_rate": 2.6113760520735108e-08,
-      "loss": 0.0043,
-      "step": 535
-    },
-    {
-      "epoch": 4.919540229885057,
-      "eval_loss": 0.011621917597949505,
-      "eval_runtime": 7.6095,
-      "eval_samples_per_second": 6.571,
-      "eval_steps_per_second": 1.708,
+      "epoch": 2.5845410628019323,
+      "grad_norm": 0.07250544428825378,
+      "learning_rate": 9.999671208666838e-05,
+      "loss": 0.0234,
       "step": 535
     },
     {
-      "epoch": 4.928735632183908,
-      "grad_norm": 0.02216600440442562,
-      "learning_rate": 1.6713330515627513e-08,
-      "loss": 0.0054,
+      "epoch": 2.5893719806763285,
+      "grad_norm": 0.073805071413517,
+      "learning_rate": 9.999631390136513e-05,
+      "loss": 0.0235,
       "step": 536
     },
     {
-      "epoch": 4.937931034482759,
-      "grad_norm": 0.026815466582775116,
-      "learning_rate": 9.401477574932926e-09,
-      "loss": 0.0048,
+      "epoch": 2.5942028985507246,
+      "grad_norm": 0.06304319947957993,
+      "learning_rate": 9.999589296376537e-05,
+      "loss": 0.0221,
       "step": 537
     },
     {
-      "epoch": 4.947126436781609,
-      "grad_norm": 0.01739557273685932,
-      "learning_rate": 4.178507228136397e-09,
-      "loss": 0.0051,
+      "epoch": 2.5990338164251208,
+      "grad_norm": 0.07283611595630646,
+      "learning_rate": 9.999544927406063e-05,
+      "loss": 0.0233,
       "step": 538
     },
     {
-      "epoch": 4.956321839080459,
-      "grad_norm": 0.017503153532743454,
-      "learning_rate": 1.0446377197104173e-09,
-      "loss": 0.0048,
+      "epoch": 2.603864734299517,
+      "grad_norm": 0.08080217987298965,
+      "learning_rate": 9.999498283245284e-05,
+      "loss": 0.0256,
       "step": 539
     },
     {
-      "epoch": 4.9655172413793105,
-      "grad_norm": 0.015828318893909454,
-      "learning_rate": 0.0,
-      "loss": 0.0046,
+      "epoch": 2.608695652173913,
+      "grad_norm": 0.07075383514165878,
+      "learning_rate": 9.999449363915427e-05,
+      "loss": 0.0231,
       "step": 540
     },
     {
-      "epoch": 4.9655172413793105,
-      "eval_loss": 0.011626788415014744,
-      "eval_runtime": 7.5888,
-      "eval_samples_per_second": 6.589,
-      "eval_steps_per_second": 1.713,
-      "step": 540
+      "epoch": 2.6135265700483092,
+      "grad_norm": 0.07120423018932343,
+      "learning_rate": 9.999398169438754e-05,
+      "loss": 0.0308,
+      "step": 541
     },
     {
-      "epoch": 4.9655172413793105,
-      "step": 540,
-      "total_flos": 1.0432767520256164e+18,
-      "train_loss": 0.011179022066709067,
-      "train_runtime": 8360.2497,
-      "train_samples_per_second": 2.081,
-      "train_steps_per_second": 0.065
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 540,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
-  "save_steps": 50,
-  "stateful_callbacks": {
+      "epoch": 2.6183574879227054,
+      "grad_norm": 0.09204115718603134,
+      "learning_rate": 9.999344699838562e-05,
+      "loss": 0.0204,
+      "step": 542
+    },
+    {
+      "epoch": 2.6231884057971016,
+      "grad_norm": 0.07161355018615723,
+      "learning_rate": 9.999288955139183e-05,
+      "loss": 0.0264,
+      "step": 543
+    },
+    {
+      "epoch": 2.6280193236714977,
+      "grad_norm": 0.07486795634031296,
+      "learning_rate": 9.999230935365989e-05,
+      "loss": 0.0272,
+      "step": 544
+    },
+    {
+      "epoch": 2.632850241545894,
+      "grad_norm": 0.06936871260404587,
+      "learning_rate": 9.999170640545378e-05,
+      "loss": 0.0252,
+      "step": 545
+    },
+    {
+      "epoch": 2.63768115942029,
+      "grad_norm": 0.07348743826150894,
+      "learning_rate": 9.999108070704795e-05,
+      "loss": 0.0301,
+      "step": 546
+    },
+    {
+      "epoch": 2.642512077294686,
+      "grad_norm": 0.07645001262426376,
+      "learning_rate": 9.99904322587271e-05,
+      "loss": 0.0271,
+      "step": 547
+    },
+    {
+      "epoch": 2.6473429951690823,
+      "grad_norm": 0.0847916230559349,
+      "learning_rate": 9.998976106078634e-05,
+      "loss": 0.0321,
+      "step": 548
+    },
+    {
+      "epoch": 2.6521739130434785,
+      "grad_norm": 0.0603368803858757,
+      "learning_rate": 9.99890671135311e-05,
+      "loss": 0.0188,
+      "step": 549
+    },
+    {
+      "epoch": 2.6570048309178746,
+      "grad_norm": 0.05718471109867096,
+      "learning_rate": 9.998835041727723e-05,
+      "loss": 0.0212,
+      "step": 550
+    },
+    {
+      "epoch": 2.6570048309178746,
+      "eval_loss": 0.02679239958524704,
+      "eval_runtime": 20.6052,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
+      "step": 550
+    },
+    {
+      "epoch": 2.661835748792271,
+      "grad_norm": 0.06785376369953156,
+      "learning_rate": 9.998761097235083e-05,
+      "loss": 0.0313,
+      "step": 551
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.11118228733539581,
+      "learning_rate": 9.998684877908844e-05,
+      "loss": 0.0486,
+      "step": 552
+    },
+    {
+      "epoch": 2.6714975845410627,
+      "grad_norm": 0.07705774903297424,
+      "learning_rate": 9.998606383783691e-05,
+      "loss": 0.0294,
+      "step": 553
+    },
+    {
+      "epoch": 2.676328502415459,
+      "grad_norm": 0.06260982900857925,
+      "learning_rate": 9.998525614895343e-05,
+      "loss": 0.019,
+      "step": 554
+    },
+    {
+      "epoch": 2.681159420289855,
+      "grad_norm": 0.07005689293146133,
+      "learning_rate": 9.99844257128056e-05,
+      "loss": 0.0253,
+      "step": 555
+    },
+    {
+      "epoch": 2.685990338164251,
+      "grad_norm": 0.06078100576996803,
+      "learning_rate": 9.99835725297713e-05,
+      "loss": 0.0209,
+      "step": 556
+    },
+    {
+      "epoch": 2.6908212560386473,
+      "grad_norm": 0.059335701167583466,
+      "learning_rate": 9.998269660023882e-05,
+      "loss": 0.0266,
+      "step": 557
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 0.06634645164012909,
+      "learning_rate": 9.998179792460676e-05,
+      "loss": 0.0301,
+      "step": 558
+    },
+    {
+      "epoch": 2.7004830917874396,
+      "grad_norm": 0.062110912054777145,
+      "learning_rate": 9.99808765032841e-05,
+      "loss": 0.0262,
+      "step": 559
+    },
+    {
+      "epoch": 2.7053140096618358,
+      "grad_norm": 0.07377223670482635,
+      "learning_rate": 9.997993233669014e-05,
+      "loss": 0.0225,
+      "step": 560
+    },
+    {
+      "epoch": 2.710144927536232,
+      "grad_norm": 0.05780097842216492,
+      "learning_rate": 9.997896542525459e-05,
+      "loss": 0.0184,
+      "step": 561
+    },
+    {
+      "epoch": 2.714975845410628,
+      "grad_norm": 0.050899870693683624,
+      "learning_rate": 9.997797576941744e-05,
+      "loss": 0.0182,
+      "step": 562
+    },
+    {
+      "epoch": 2.7198067632850242,
+      "grad_norm": 0.07264772057533264,
+      "learning_rate": 9.997696336962907e-05,
+      "loss": 0.0235,
+      "step": 563
+    },
+    {
+      "epoch": 2.7246376811594204,
+      "grad_norm": 0.0789547711610794,
+      "learning_rate": 9.997592822635021e-05,
+      "loss": 0.0294,
+      "step": 564
+    },
+    {
+      "epoch": 2.7294685990338166,
+      "grad_norm": 0.10101927071809769,
+      "learning_rate": 9.997487034005193e-05,
+      "loss": 0.0264,
+      "step": 565
+    },
+    {
+      "epoch": 2.7342995169082127,
+      "grad_norm": 0.059210047125816345,
+      "learning_rate": 9.997378971121564e-05,
+      "loss": 0.0182,
+      "step": 566
+    },
+    {
+      "epoch": 2.7391304347826084,
+      "grad_norm": 0.05791271850466728,
+      "learning_rate": 9.997268634033312e-05,
+      "loss": 0.0183,
+      "step": 567
+    },
+    {
+      "epoch": 2.7439613526570046,
+      "grad_norm": 0.07589870691299438,
+      "learning_rate": 9.99715602279065e-05,
+      "loss": 0.0256,
+      "step": 568
+    },
+    {
+      "epoch": 2.7487922705314007,
+      "grad_norm": 0.0665014311671257,
+      "learning_rate": 9.997041137444823e-05,
+      "loss": 0.0205,
+      "step": 569
+    },
+    {
+      "epoch": 2.753623188405797,
+      "grad_norm": 0.05872458219528198,
+      "learning_rate": 9.996923978048115e-05,
+      "loss": 0.0182,
+      "step": 570
+    },
+    {
+      "epoch": 2.758454106280193,
+      "grad_norm": 0.07963749021291733,
+      "learning_rate": 9.996804544653842e-05,
+      "loss": 0.0224,
+      "step": 571
+    },
+    {
+      "epoch": 2.763285024154589,
+      "grad_norm": 0.06978920102119446,
+      "learning_rate": 9.996682837316356e-05,
+      "loss": 0.0222,
+      "step": 572
+    },
+    {
+      "epoch": 2.7681159420289854,
+      "grad_norm": 0.06920437514781952,
+      "learning_rate": 9.996558856091043e-05,
+      "loss": 0.0233,
+      "step": 573
+    },
+    {
+      "epoch": 2.7729468599033815,
+      "grad_norm": 0.06745556741952896,
+      "learning_rate": 9.996432601034324e-05,
+      "loss": 0.0203,
+      "step": 574
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 0.06809904426336288,
+      "learning_rate": 9.996304072203657e-05,
+      "loss": 0.0221,
+      "step": 575
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "eval_loss": 0.02665121853351593,
+      "eval_runtime": 20.6302,
+      "eval_samples_per_second": 4.847,
+      "eval_steps_per_second": 0.145,
+      "step": 575
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.05648452416062355,
+      "learning_rate": 9.99617326965753e-05,
+      "loss": 0.019,
+      "step": 576
+    },
+    {
+      "epoch": 2.78743961352657,
+      "grad_norm": 0.06754475831985474,
+      "learning_rate": 9.996040193455472e-05,
+      "loss": 0.019,
+      "step": 577
+    },
+    {
+      "epoch": 2.792270531400966,
+      "grad_norm": 0.07781781256198883,
+      "learning_rate": 9.99590484365804e-05,
+      "loss": 0.0279,
+      "step": 578
+    },
+    {
+      "epoch": 2.7971014492753623,
+      "grad_norm": 0.06808023899793625,
+      "learning_rate": 9.995767220326829e-05,
+      "loss": 0.0188,
+      "step": 579
+    },
+    {
+      "epoch": 2.8019323671497585,
+      "grad_norm": 0.08858218789100647,
+      "learning_rate": 9.99562732352447e-05,
+      "loss": 0.0254,
+      "step": 580
+    },
+    {
+      "epoch": 2.8067632850241546,
+      "grad_norm": 0.0699530616402626,
+      "learning_rate": 9.995485153314628e-05,
+      "loss": 0.0229,
+      "step": 581
+    },
+    {
+      "epoch": 2.8115942028985508,
+      "grad_norm": 0.067716583609581,
+      "learning_rate": 9.995340709762002e-05,
+      "loss": 0.0248,
+      "step": 582
+    },
+    {
+      "epoch": 2.816425120772947,
+      "grad_norm": 0.06093629077076912,
+      "learning_rate": 9.995193992932321e-05,
+      "loss": 0.0201,
+      "step": 583
+    },
+    {
+      "epoch": 2.821256038647343,
+      "grad_norm": 0.061484821140766144,
+      "learning_rate": 9.995045002892358e-05,
+      "loss": 0.0279,
+      "step": 584
+    },
+    {
+      "epoch": 2.8260869565217392,
+      "grad_norm": 0.05432453379034996,
+      "learning_rate": 9.994893739709912e-05,
+      "loss": 0.022,
+      "step": 585
+    },
+    {
+      "epoch": 2.8309178743961354,
+      "grad_norm": 0.06907227635383606,
+      "learning_rate": 9.994740203453821e-05,
+      "loss": 0.0229,
+      "step": 586
+    },
+    {
+      "epoch": 2.8357487922705316,
+      "grad_norm": 0.08959613740444183,
+      "learning_rate": 9.994584394193957e-05,
+      "loss": 0.0293,
+      "step": 587
+    },
+    {
+      "epoch": 2.8405797101449277,
+      "grad_norm": 0.06915111094713211,
+      "learning_rate": 9.994426312001223e-05,
+      "loss": 0.0195,
+      "step": 588
+    },
+    {
+      "epoch": 2.845410628019324,
+      "grad_norm": 0.06990355998277664,
+      "learning_rate": 9.994265956947563e-05,
+      "loss": 0.0204,
+      "step": 589
+    },
+    {
+      "epoch": 2.85024154589372,
+      "grad_norm": 0.08454722166061401,
+      "learning_rate": 9.994103329105947e-05,
+      "loss": 0.024,
+      "step": 590
+    },
+    {
+      "epoch": 2.855072463768116,
+      "grad_norm": 0.11382115632295609,
+      "learning_rate": 9.993938428550387e-05,
+      "loss": 0.0297,
+      "step": 591
+    },
+    {
+      "epoch": 2.8599033816425123,
+      "grad_norm": 0.07049544900655746,
+      "learning_rate": 9.993771255355921e-05,
+      "loss": 0.0237,
+      "step": 592
+    },
+    {
+      "epoch": 2.864734299516908,
+      "grad_norm": 0.06032729148864746,
+      "learning_rate": 9.993601809598634e-05,
+      "loss": 0.0192,
+      "step": 593
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 0.0696573555469513,
+      "learning_rate": 9.99343009135563e-05,
+      "loss": 0.0233,
+      "step": 594
+    },
+    {
+      "epoch": 2.8743961352657004,
+      "grad_norm": 0.06839869916439056,
+      "learning_rate": 9.993256100705058e-05,
+      "loss": 0.0233,
+      "step": 595
+    },
+    {
+      "epoch": 2.8792270531400965,
+      "grad_norm": 0.061628565192222595,
+      "learning_rate": 9.993079837726096e-05,
+      "loss": 0.0228,
+      "step": 596
+    },
+    {
+      "epoch": 2.8840579710144927,
+      "grad_norm": 0.06356792151927948,
+      "learning_rate": 9.992901302498959e-05,
+      "loss": 0.0239,
+      "step": 597
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.06923532485961914,
+      "learning_rate": 9.992720495104895e-05,
+      "loss": 0.0264,
+      "step": 598
+    },
+    {
+      "epoch": 2.893719806763285,
+      "grad_norm": 0.07253517955541611,
+      "learning_rate": 9.992537415626183e-05,
+      "loss": 0.0218,
+      "step": 599
+    },
+    {
+      "epoch": 2.898550724637681,
+      "grad_norm": 0.07643329352140427,
+      "learning_rate": 9.992352064146142e-05,
+      "loss": 0.0225,
+      "step": 600
+    },
+    {
+      "epoch": 2.898550724637681,
+      "eval_loss": 0.026624999940395355,
+      "eval_runtime": 20.5987,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
+      "step": 600
+    },
+    {
+      "epoch": 2.9033816425120773,
+      "grad_norm": 0.055665720254182816,
+      "learning_rate": 9.992164440749119e-05,
+      "loss": 0.0198,
+      "step": 601
+    },
+    {
+      "epoch": 2.9082125603864735,
+      "grad_norm": 0.0848175436258316,
+      "learning_rate": 9.9919745455205e-05,
+      "loss": 0.0293,
+      "step": 602
+    },
+    {
+      "epoch": 2.9130434782608696,
+      "grad_norm": 0.06863771378993988,
+      "learning_rate": 9.991782378546702e-05,
+      "loss": 0.0237,
+      "step": 603
+    },
+    {
+      "epoch": 2.917874396135266,
+      "grad_norm": 0.07147162407636642,
+      "learning_rate": 9.991587939915173e-05,
+      "loss": 0.0246,
+      "step": 604
+    },
+    {
+      "epoch": 2.922705314009662,
+      "grad_norm": 0.07691524177789688,
+      "learning_rate": 9.991391229714401e-05,
+      "loss": 0.0251,
+      "step": 605
+    },
+    {
+      "epoch": 2.927536231884058,
+      "grad_norm": 0.06954370439052582,
+      "learning_rate": 9.991192248033908e-05,
+      "loss": 0.0251,
+      "step": 606
+    },
+    {
+      "epoch": 2.9323671497584543,
+      "grad_norm": 0.09142833948135376,
+      "learning_rate": 9.990990994964239e-05,
+      "loss": 0.0339,
+      "step": 607
+    },
+    {
+      "epoch": 2.9371980676328504,
+      "grad_norm": 0.09188190847635269,
+      "learning_rate": 9.990787470596985e-05,
+      "loss": 0.0228,
+      "step": 608
+    },
+    {
+      "epoch": 2.942028985507246,
+      "grad_norm": 0.07027522474527359,
+      "learning_rate": 9.990581675024763e-05,
+      "loss": 0.0232,
+      "step": 609
+    },
+    {
+      "epoch": 2.9468599033816423,
+      "grad_norm": 0.0655936598777771,
+      "learning_rate": 9.99037360834123e-05,
+      "loss": 0.0216,
+      "step": 610
+    },
+    {
+      "epoch": 2.9516908212560384,
+      "grad_norm": 0.062273845076560974,
+      "learning_rate": 9.99016327064107e-05,
+      "loss": 0.0189,
+      "step": 611
+    },
+    {
+      "epoch": 2.9565217391304346,
+      "grad_norm": 0.08527228981256485,
+      "learning_rate": 9.989950662020007e-05,
+      "loss": 0.0286,
+      "step": 612
+    },
+    {
+      "epoch": 2.9613526570048307,
+      "grad_norm": 0.06256254762411118,
+      "learning_rate": 9.98973578257479e-05,
+      "loss": 0.0204,
+      "step": 613
+    },
+    {
+      "epoch": 2.966183574879227,
+      "grad_norm": 0.08768539875745773,
+      "learning_rate": 9.989518632403208e-05,
+      "loss": 0.0325,
+      "step": 614
+    },
+    {
+      "epoch": 2.971014492753623,
+      "grad_norm": 0.07069282233715057,
+      "learning_rate": 9.989299211604082e-05,
+      "loss": 0.021,
+      "step": 615
+    },
+    {
+      "epoch": 2.975845410628019,
+      "grad_norm": 0.06434057652950287,
+      "learning_rate": 9.989077520277264e-05,
+      "loss": 0.0195,
+      "step": 616
+    },
+    {
+      "epoch": 2.9806763285024154,
+      "grad_norm": 0.07296837121248245,
+      "learning_rate": 9.988853558523646e-05,
+      "loss": 0.0254,
+      "step": 617
+    },
+    {
+      "epoch": 2.9855072463768115,
+      "grad_norm": 0.08000582456588745,
+      "learning_rate": 9.988627326445143e-05,
+      "loss": 0.0261,
+      "step": 618
+    },
+    {
+      "epoch": 2.9903381642512077,
+      "grad_norm": 0.08119505643844604,
+      "learning_rate": 9.988398824144714e-05,
+      "loss": 0.0316,
+      "step": 619
+    },
+    {
+      "epoch": 2.995169082125604,
+      "grad_norm": 0.053310543298721313,
+      "learning_rate": 9.98816805172634e-05,
+      "loss": 0.0216,
+      "step": 620
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.12208092957735062,
+      "learning_rate": 9.987935009295044e-05,
+      "loss": 0.016,
+      "step": 621
+    },
+    {
+      "epoch": 3.004830917874396,
+      "grad_norm": 0.06508877128362656,
+      "learning_rate": 9.987699696956878e-05,
+      "loss": 0.0244,
+      "step": 622
+    },
+    {
+      "epoch": 3.0096618357487923,
+      "grad_norm": 0.049330223351716995,
+      "learning_rate": 9.987462114818928e-05,
+      "loss": 0.0159,
+      "step": 623
+    },
+    {
+      "epoch": 3.0144927536231885,
+      "grad_norm": 0.06686580181121826,
+      "learning_rate": 9.987222262989315e-05,
+      "loss": 0.018,
+      "step": 624
+    },
+    {
+      "epoch": 3.0193236714975846,
+      "grad_norm": 0.061486225575208664,
+      "learning_rate": 9.986980141577187e-05,
+      "loss": 0.0264,
+      "step": 625
+    },
+    {
+      "epoch": 3.0193236714975846,
+      "eval_loss": 0.02915319800376892,
+      "eval_runtime": 20.6012,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 625
+    },
+    {
+      "epoch": 3.024154589371981,
+      "grad_norm": 0.06663984060287476,
+      "learning_rate": 9.98673575069273e-05,
+      "loss": 0.0181,
+      "step": 626
+    },
+    {
+      "epoch": 3.028985507246377,
+      "grad_norm": 0.06651222705841064,
+      "learning_rate": 9.98648909044716e-05,
+      "loss": 0.0183,
+      "step": 627
+    },
+    {
+      "epoch": 3.033816425120773,
+      "grad_norm": 0.06698988378047943,
+      "learning_rate": 9.986240160952732e-05,
+      "loss": 0.018,
+      "step": 628
+    },
+    {
+      "epoch": 3.0386473429951693,
+      "grad_norm": 0.08998807519674301,
+      "learning_rate": 9.985988962322721e-05,
+      "loss": 0.0198,
+      "step": 629
+    },
+    {
+      "epoch": 3.0434782608695654,
+      "grad_norm": 0.08781043440103531,
+      "learning_rate": 9.985735494671448e-05,
+      "loss": 0.0241,
+      "step": 630
+    },
+    {
+      "epoch": 3.0483091787439616,
+      "grad_norm": 0.06608462333679199,
+      "learning_rate": 9.985479758114259e-05,
+      "loss": 0.0139,
+      "step": 631
+    },
+    {
+      "epoch": 3.0531400966183573,
+      "grad_norm": 0.06002631038427353,
+      "learning_rate": 9.985221752767535e-05,
+      "loss": 0.0162,
+      "step": 632
+    },
+    {
+      "epoch": 3.0579710144927534,
+      "grad_norm": 0.05958976596593857,
+      "learning_rate": 9.984961478748688e-05,
+      "loss": 0.0185,
+      "step": 633
+    },
+    {
+      "epoch": 3.0628019323671496,
+      "grad_norm": 0.06394585967063904,
+      "learning_rate": 9.984698936176164e-05,
+      "loss": 0.0235,
+      "step": 634
+    },
+    {
+      "epoch": 3.0676328502415457,
+      "grad_norm": 0.07614199072122574,
+      "learning_rate": 9.984434125169441e-05,
+      "loss": 0.0202,
+      "step": 635
+    },
+    {
+      "epoch": 3.072463768115942,
+      "grad_norm": 0.06564124673604965,
+      "learning_rate": 9.98416704584903e-05,
+      "loss": 0.0165,
+      "step": 636
+    },
+    {
+      "epoch": 3.077294685990338,
+      "grad_norm": 0.06295037269592285,
+      "learning_rate": 9.983897698336471e-05,
+      "loss": 0.0193,
+      "step": 637
+    },
+    {
+      "epoch": 3.082125603864734,
+      "grad_norm": 0.06796164065599442,
+      "learning_rate": 9.98362608275434e-05,
+      "loss": 0.0185,
+      "step": 638
+    },
+    {
+      "epoch": 3.0869565217391304,
+      "grad_norm": 0.08391165733337402,
+      "learning_rate": 9.983352199226243e-05,
+      "loss": 0.0181,
+      "step": 639
+    },
+    {
+      "epoch": 3.0917874396135265,
+      "grad_norm": 0.062315862625837326,
+      "learning_rate": 9.98307604787682e-05,
+      "loss": 0.0164,
+      "step": 640
+    },
+    {
+      "epoch": 3.0966183574879227,
+      "grad_norm": 0.05803072825074196,
+      "learning_rate": 9.982797628831739e-05,
+      "loss": 0.0144,
+      "step": 641
+    },
+    {
+      "epoch": 3.101449275362319,
+      "grad_norm": 0.06700735539197922,
+      "learning_rate": 9.982516942217705e-05,
+      "loss": 0.0175,
+      "step": 642
+    },
+    {
+      "epoch": 3.106280193236715,
+      "grad_norm": 0.08924155682325363,
+      "learning_rate": 9.982233988162455e-05,
+      "loss": 0.0225,
+      "step": 643
+    },
+    {
+      "epoch": 3.111111111111111,
+      "grad_norm": 0.08006013929843903,
+      "learning_rate": 9.981948766794752e-05,
+      "loss": 0.016,
+      "step": 644
+    },
+    {
+      "epoch": 3.1159420289855073,
+      "grad_norm": 0.06664010137319565,
+      "learning_rate": 9.981661278244394e-05,
+      "loss": 0.0139,
+      "step": 645
+    },
+    {
+      "epoch": 3.1207729468599035,
+      "grad_norm": 0.062214843928813934,
+      "learning_rate": 9.981371522642212e-05,
+      "loss": 0.0143,
+      "step": 646
+    },
+    {
+      "epoch": 3.1256038647342996,
+      "grad_norm": 0.06204639747738838,
+      "learning_rate": 9.98107950012007e-05,
+      "loss": 0.0167,
+      "step": 647
+    },
+    {
+      "epoch": 3.130434782608696,
+      "grad_norm": 0.0846833735704422,
+      "learning_rate": 9.980785210810859e-05,
+      "loss": 0.0325,
+      "step": 648
+    },
+    {
+      "epoch": 3.135265700483092,
+      "grad_norm": 0.06417378038167953,
+      "learning_rate": 9.980488654848505e-05,
+      "loss": 0.0182,
+      "step": 649
+    },
+    {
+      "epoch": 3.140096618357488,
+      "grad_norm": 0.06934668123722076,
+      "learning_rate": 9.980189832367966e-05,
+      "loss": 0.0196,
+      "step": 650
+    },
+    {
+      "epoch": 3.140096618357488,
+      "eval_loss": 0.027954233810305595,
+      "eval_runtime": 21.3091,
+      "eval_samples_per_second": 4.693,
+      "eval_steps_per_second": 0.141,
+      "step": 650
+    },
+    {
+      "epoch": 3.1449275362318843,
+      "grad_norm": 0.13086830079555511,
+      "learning_rate": 9.979888743505225e-05,
+      "loss": 0.0146,
+      "step": 651
+    },
+    {
+      "epoch": 3.14975845410628,
+      "grad_norm": 0.06644698232412338,
+      "learning_rate": 9.979585388397308e-05,
+      "loss": 0.017,
+      "step": 652
+    },
+    {
+      "epoch": 3.154589371980676,
+      "grad_norm": 0.06060745194554329,
+      "learning_rate": 9.979279767182262e-05,
+      "loss": 0.0189,
+      "step": 653
+    },
+    {
+      "epoch": 3.1594202898550723,
+      "grad_norm": 0.05614025145769119,
+      "learning_rate": 9.978971879999169e-05,
+      "loss": 0.0154,
+      "step": 654
+    },
+    {
+      "epoch": 3.1642512077294684,
+      "grad_norm": 0.0934065654873848,
+      "learning_rate": 9.97866172698814e-05,
+      "loss": 0.0234,
+      "step": 655
+    },
+    {
+      "epoch": 3.1690821256038646,
+      "grad_norm": 0.08930619060993195,
+      "learning_rate": 9.978349308290325e-05,
+      "loss": 0.0178,
+      "step": 656
+    },
+    {
+      "epoch": 3.1739130434782608,
+      "grad_norm": 0.07199080288410187,
+      "learning_rate": 9.978034624047895e-05,
+      "loss": 0.0189,
+      "step": 657
+    },
+    {
+      "epoch": 3.178743961352657,
+      "grad_norm": 0.06786323338747025,
+      "learning_rate": 9.977717674404056e-05,
+      "loss": 0.0203,
+      "step": 658
+    },
+    {
+      "epoch": 3.183574879227053,
+      "grad_norm": 0.05707789584994316,
+      "learning_rate": 9.977398459503049e-05,
+      "loss": 0.0168,
+      "step": 659
+    },
+    {
+      "epoch": 3.1884057971014492,
+      "grad_norm": 0.05864753946661949,
+      "learning_rate": 9.977076979490138e-05,
+      "loss": 0.0145,
+      "step": 660
+    },
+    {
+      "epoch": 3.1932367149758454,
+      "grad_norm": 0.07442363351583481,
+      "learning_rate": 9.976753234511627e-05,
+      "loss": 0.0224,
+      "step": 661
+    },
+    {
+      "epoch": 3.1980676328502415,
+      "grad_norm": 0.05928374454379082,
+      "learning_rate": 9.97642722471484e-05,
+      "loss": 0.0151,
+      "step": 662
+    },
+    {
+      "epoch": 3.2028985507246377,
+      "grad_norm": 0.05867256596684456,
+      "learning_rate": 9.976098950248141e-05,
+      "loss": 0.0153,
+      "step": 663
+    },
+    {
+      "epoch": 3.207729468599034,
+      "grad_norm": 0.05609915778040886,
+      "learning_rate": 9.975768411260917e-05,
+      "loss": 0.017,
+      "step": 664
+    },
+    {
+      "epoch": 3.21256038647343,
+      "grad_norm": 0.08102008700370789,
+      "learning_rate": 9.975435607903596e-05,
+      "loss": 0.0224,
+      "step": 665
+    },
+    {
+      "epoch": 3.217391304347826,
+      "grad_norm": 0.04532697796821594,
+      "learning_rate": 9.975100540327624e-05,
+      "loss": 0.0127,
+      "step": 666
+    },
+    {
+      "epoch": 3.2222222222222223,
+      "grad_norm": 0.07046009600162506,
+      "learning_rate": 9.974763208685487e-05,
+      "loss": 0.0233,
+      "step": 667
+    },
+    {
+      "epoch": 3.2270531400966185,
+      "grad_norm": 0.059037089347839355,
+      "learning_rate": 9.974423613130697e-05,
+      "loss": 0.0155,
+      "step": 668
+    },
+    {
+      "epoch": 3.2318840579710146,
+      "grad_norm": 0.06592457741498947,
+      "learning_rate": 9.974081753817795e-05,
+      "loss": 0.0187,
+      "step": 669
+    },
+    {
+      "epoch": 3.236714975845411,
+      "grad_norm": 0.058792632073163986,
+      "learning_rate": 9.973737630902356e-05,
+      "loss": 0.0162,
+      "step": 670
+    },
+    {
+      "epoch": 3.241545893719807,
+      "grad_norm": 0.07621821016073227,
+      "learning_rate": 9.973391244540983e-05,
+      "loss": 0.0223,
+      "step": 671
+    },
+    {
+      "epoch": 3.246376811594203,
+      "grad_norm": 0.09599140286445618,
+      "learning_rate": 9.973042594891309e-05,
+      "loss": 0.0231,
+      "step": 672
+    },
+    {
+      "epoch": 3.2512077294685993,
+      "grad_norm": 0.06377045810222626,
+      "learning_rate": 9.972691682111997e-05,
+      "loss": 0.0178,
+      "step": 673
+    },
+    {
+      "epoch": 3.2560386473429954,
+      "grad_norm": 0.08155626058578491,
+      "learning_rate": 9.972338506362742e-05,
+      "loss": 0.0228,
+      "step": 674
+    },
+    {
+      "epoch": 3.260869565217391,
+      "grad_norm": 0.07435385137796402,
+      "learning_rate": 9.971983067804265e-05,
+      "loss": 0.0185,
+      "step": 675
+    },
+    {
+      "epoch": 3.260869565217391,
+      "eval_loss": 0.02637997642159462,
+      "eval_runtime": 20.6021,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 675
+    },
+    {
+      "epoch": 3.2657004830917873,
+      "grad_norm": 0.07467380911111832,
+      "learning_rate": 9.971625366598319e-05,
+      "loss": 0.0192,
+      "step": 676
+    },
+    {
+      "epoch": 3.2705314009661834,
+      "grad_norm": 0.07276886701583862,
+      "learning_rate": 9.971265402907688e-05,
+      "loss": 0.0203,
+      "step": 677
+    },
+    {
+      "epoch": 3.2753623188405796,
+      "grad_norm": 0.06594165414571762,
+      "learning_rate": 9.970903176896183e-05,
+      "loss": 0.0159,
+      "step": 678
+    },
+    {
+      "epoch": 3.2801932367149758,
+      "grad_norm": 0.06967869400978088,
+      "learning_rate": 9.970538688728644e-05,
+      "loss": 0.0171,
+      "step": 679
+    },
+    {
+      "epoch": 3.285024154589372,
+      "grad_norm": 0.07510808855295181,
+      "learning_rate": 9.970171938570946e-05,
+      "loss": 0.0174,
+      "step": 680
+    },
+    {
+      "epoch": 3.289855072463768,
+      "grad_norm": 0.06190778687596321,
+      "learning_rate": 9.969802926589986e-05,
+      "loss": 0.0171,
+      "step": 681
+    },
+    {
+      "epoch": 3.2946859903381642,
+      "grad_norm": 0.07371743023395538,
+      "learning_rate": 9.969431652953695e-05,
+      "loss": 0.0193,
+      "step": 682
+    },
+    {
+      "epoch": 3.2995169082125604,
+      "grad_norm": 0.14643917977809906,
+      "learning_rate": 9.969058117831034e-05,
+      "loss": 0.0229,
+      "step": 683
+    },
+    {
+      "epoch": 3.3043478260869565,
+      "grad_norm": 0.0634489580988884,
+      "learning_rate": 9.968682321391986e-05,
+      "loss": 0.0183,
+      "step": 684
+    },
+    {
+      "epoch": 3.3091787439613527,
+      "grad_norm": 0.07199045270681381,
+      "learning_rate": 9.968304263807574e-05,
+      "loss": 0.0167,
+      "step": 685
+    },
+    {
+      "epoch": 3.314009661835749,
+      "grad_norm": 0.05730137601494789,
+      "learning_rate": 9.96792394524984e-05,
+      "loss": 0.0165,
+      "step": 686
+    },
+    {
+      "epoch": 3.318840579710145,
+      "grad_norm": 0.07792101055383682,
+      "learning_rate": 9.967541365891863e-05,
+      "loss": 0.0243,
+      "step": 687
+    },
+    {
+      "epoch": 3.323671497584541,
+      "grad_norm": 0.06612062454223633,
+      "learning_rate": 9.967156525907743e-05,
+      "loss": 0.0177,
+      "step": 688
+    },
+    {
+      "epoch": 3.3285024154589373,
+      "grad_norm": 0.08418621122837067,
+      "learning_rate": 9.966769425472616e-05,
+      "loss": 0.0174,
+      "step": 689
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.08468875288963318,
+      "learning_rate": 9.966380064762642e-05,
+      "loss": 0.0152,
+      "step": 690
+    },
+    {
+      "epoch": 3.3381642512077296,
+      "grad_norm": 0.05754866451025009,
+      "learning_rate": 9.96598844395501e-05,
+      "loss": 0.0132,
+      "step": 691
+    },
+    {
+      "epoch": 3.342995169082126,
+      "grad_norm": 0.07585375756025314,
+      "learning_rate": 9.96559456322794e-05,
+      "loss": 0.0158,
+      "step": 692
+    },
+    {
+      "epoch": 3.3478260869565215,
+      "grad_norm": 0.08202559500932693,
+      "learning_rate": 9.96519842276068e-05,
+      "loss": 0.0208,
+      "step": 693
+    },
+    {
+      "epoch": 3.3526570048309177,
+      "grad_norm": 0.0768493190407753,
+      "learning_rate": 9.964800022733504e-05,
+      "loss": 0.024,
+      "step": 694
+    },
+    {
+      "epoch": 3.357487922705314,
+      "grad_norm": 0.09011416882276535,
+      "learning_rate": 9.964399363327716e-05,
+      "loss": 0.0229,
+      "step": 695
+    },
+    {
+      "epoch": 3.36231884057971,
+      "grad_norm": 0.06828708201646805,
+      "learning_rate": 9.963996444725647e-05,
+      "loss": 0.0216,
+      "step": 696
+    },
+    {
+      "epoch": 3.367149758454106,
+      "grad_norm": 0.09616364538669586,
+      "learning_rate": 9.96359126711066e-05,
+      "loss": 0.0213,
+      "step": 697
+    },
+    {
+      "epoch": 3.3719806763285023,
+      "grad_norm": 0.056025609374046326,
+      "learning_rate": 9.963183830667138e-05,
+      "loss": 0.0138,
+      "step": 698
+    },
+    {
+      "epoch": 3.3768115942028984,
+      "grad_norm": 0.052819810807704926,
+      "learning_rate": 9.9627741355805e-05,
+      "loss": 0.0145,
+      "step": 699
+    },
+    {
+      "epoch": 3.3816425120772946,
+      "grad_norm": 0.05790337547659874,
+      "learning_rate": 9.96236218203719e-05,
+      "loss": 0.0161,
+      "step": 700
+    },
+    {
+      "epoch": 3.3816425120772946,
+      "eval_loss": 0.024786872789263725,
+      "eval_runtime": 20.6335,
+      "eval_samples_per_second": 4.846,
+      "eval_steps_per_second": 0.145,
+      "step": 700
+    },
+    {
+      "epoch": 3.3864734299516908,
+      "grad_norm": 0.053379714488983154,
+      "learning_rate": 9.96194797022468e-05,
+      "loss": 0.0141,
+      "step": 701
+    },
+    {
+      "epoch": 3.391304347826087,
+      "grad_norm": 0.059687525033950806,
+      "learning_rate": 9.961531500331469e-05,
+      "loss": 0.0166,
+      "step": 702
+    },
+    {
+      "epoch": 3.396135265700483,
+      "grad_norm": 0.05600206181406975,
+      "learning_rate": 9.961112772547083e-05,
+      "loss": 0.0164,
+      "step": 703
+    },
+    {
+      "epoch": 3.4009661835748792,
+      "grad_norm": 0.07950199395418167,
+      "learning_rate": 9.960691787062076e-05,
+      "loss": 0.0169,
+      "step": 704
+    },
+    {
+      "epoch": 3.4057971014492754,
+      "grad_norm": 0.08069173246622086,
+      "learning_rate": 9.960268544068032e-05,
+      "loss": 0.025,
+      "step": 705
+    },
+    {
+      "epoch": 3.4106280193236715,
+      "grad_norm": 0.06130361557006836,
+      "learning_rate": 9.959843043757557e-05,
+      "loss": 0.0155,
+      "step": 706
+    },
+    {
+      "epoch": 3.4154589371980677,
+      "grad_norm": 0.06326994299888611,
+      "learning_rate": 9.959415286324289e-05,
+      "loss": 0.0141,
+      "step": 707
+    },
+    {
+      "epoch": 3.420289855072464,
+      "grad_norm": 0.07043007761240005,
+      "learning_rate": 9.958985271962896e-05,
+      "loss": 0.0143,
+      "step": 708
+    },
+    {
+      "epoch": 3.42512077294686,
+      "grad_norm": 0.09066811949014664,
+      "learning_rate": 9.958553000869061e-05,
+      "loss": 0.026,
+      "step": 709
+    },
+    {
+      "epoch": 3.429951690821256,
+      "grad_norm": 0.06408656388521194,
+      "learning_rate": 9.958118473239507e-05,
+      "loss": 0.0139,
+      "step": 710
+    },
+    {
+      "epoch": 3.4347826086956523,
+      "grad_norm": 0.06362169981002808,
+      "learning_rate": 9.957681689271977e-05,
+      "loss": 0.0166,
+      "step": 711
+    },
+    {
+      "epoch": 3.4396135265700485,
+      "grad_norm": 0.07092583179473877,
+      "learning_rate": 9.957242649165241e-05,
+      "loss": 0.0171,
+      "step": 712
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 0.07049217075109482,
+      "learning_rate": 9.956801353119099e-05,
+      "loss": 0.0176,
+      "step": 713
+    },
+    {
+      "epoch": 3.449275362318841,
+      "grad_norm": 0.06697899103164673,
+      "learning_rate": 9.956357801334375e-05,
+      "loss": 0.0182,
+      "step": 714
+    },
+    {
+      "epoch": 3.454106280193237,
+      "grad_norm": 0.06497643142938614,
+      "learning_rate": 9.955911994012923e-05,
+      "loss": 0.016,
+      "step": 715
+    },
+    {
+      "epoch": 3.4589371980676327,
+      "grad_norm": 0.06678501516580582,
+      "learning_rate": 9.955463931357616e-05,
+      "loss": 0.0229,
+      "step": 716
+    },
+    {
+      "epoch": 3.463768115942029,
+      "grad_norm": 0.04944721981883049,
+      "learning_rate": 9.955013613572362e-05,
+      "loss": 0.0177,
+      "step": 717
+    },
+    {
+      "epoch": 3.468599033816425,
+      "grad_norm": 0.054069582372903824,
+      "learning_rate": 9.954561040862088e-05,
+      "loss": 0.0165,
+      "step": 718
+    },
+    {
+      "epoch": 3.473429951690821,
+      "grad_norm": 0.07551441341638565,
+      "learning_rate": 9.954106213432755e-05,
+      "loss": 0.0286,
+      "step": 719
+    },
+    {
+      "epoch": 3.4782608695652173,
+      "grad_norm": 0.061375074088573456,
+      "learning_rate": 9.95364913149134e-05,
+      "loss": 0.0185,
+      "step": 720
+    },
+    {
+      "epoch": 3.4830917874396135,
+      "grad_norm": 0.060356441885232925,
+      "learning_rate": 9.953189795245857e-05,
+      "loss": 0.0168,
+      "step": 721
+    },
+    {
+      "epoch": 3.4879227053140096,
+      "grad_norm": 0.07686608284711838,
+      "learning_rate": 9.952728204905338e-05,
+      "loss": 0.0148,
+      "step": 722
+    },
+    {
+      "epoch": 3.4927536231884058,
+      "grad_norm": 0.05384129658341408,
+      "learning_rate": 9.952264360679844e-05,
+      "loss": 0.0171,
+      "step": 723
+    },
+    {
+      "epoch": 3.497584541062802,
+      "grad_norm": 0.05942288041114807,
+      "learning_rate": 9.951798262780458e-05,
+      "loss": 0.0172,
+      "step": 724
+    },
+    {
+      "epoch": 3.502415458937198,
+      "grad_norm": 0.07827403396368027,
+      "learning_rate": 9.951329911419298e-05,
+      "loss": 0.0186,
+      "step": 725
+    },
+    {
+      "epoch": 3.502415458937198,
+      "eval_loss": 0.02256963960826397,
+      "eval_runtime": 20.6118,
+      "eval_samples_per_second": 4.852,
+      "eval_steps_per_second": 0.146,
+      "step": 725
+    },
+    {
+      "epoch": 3.5072463768115942,
+      "grad_norm": 0.04993787035346031,
+      "learning_rate": 9.950859306809494e-05,
+      "loss": 0.0139,
+      "step": 726
+    },
+    {
+      "epoch": 3.5120772946859904,
+      "grad_norm": 0.07030709087848663,
+      "learning_rate": 9.950386449165212e-05,
+      "loss": 0.0166,
+      "step": 727
+    },
+    {
+      "epoch": 3.5169082125603865,
+      "grad_norm": 0.04969039931893349,
+      "learning_rate": 9.94991133870164e-05,
+      "loss": 0.012,
+      "step": 728
+    },
+    {
+      "epoch": 3.5217391304347827,
+      "grad_norm": 0.06086138263344765,
+      "learning_rate": 9.949433975634992e-05,
+      "loss": 0.019,
+      "step": 729
+    },
+    {
+      "epoch": 3.526570048309179,
+      "grad_norm": 0.060854848474264145,
+      "learning_rate": 9.948954360182503e-05,
+      "loss": 0.0144,
+      "step": 730
+    },
+    {
+      "epoch": 3.531400966183575,
+      "grad_norm": 0.06262096762657166,
+      "learning_rate": 9.948472492562438e-05,
+      "loss": 0.0164,
+      "step": 731
+    },
+    {
+      "epoch": 3.536231884057971,
+      "grad_norm": 0.093068428337574,
+      "learning_rate": 9.947988372994086e-05,
+      "loss": 0.0327,
+      "step": 732
+    },
+    {
+      "epoch": 3.541062801932367,
+      "grad_norm": 0.07284866273403168,
+      "learning_rate": 9.947502001697757e-05,
+      "loss": 0.0171,
+      "step": 733
+    },
+    {
+      "epoch": 3.545893719806763,
+      "grad_norm": 0.06965293735265732,
+      "learning_rate": 9.947013378894792e-05,
+      "loss": 0.0212,
+      "step": 734
+    },
+    {
+      "epoch": 3.550724637681159,
+      "grad_norm": 0.049892608076334,
+      "learning_rate": 9.946522504807551e-05,
+      "loss": 0.0135,
+      "step": 735
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 0.09122290462255478,
+      "learning_rate": 9.94602937965942e-05,
+      "loss": 0.0222,
+      "step": 736
+    },
+    {
+      "epoch": 3.5603864734299515,
+      "grad_norm": 0.055852945894002914,
+      "learning_rate": 9.945534003674812e-05,
+      "loss": 0.0201,
+      "step": 737
+    },
+    {
+      "epoch": 3.5652173913043477,
+      "grad_norm": 0.056373223662376404,
+      "learning_rate": 9.945036377079164e-05,
+      "loss": 0.0149,
+      "step": 738
+    },
+    {
+      "epoch": 3.570048309178744,
+      "grad_norm": 0.07455029338598251,
+      "learning_rate": 9.94453650009893e-05,
+      "loss": 0.0166,
+      "step": 739
+    },
+    {
+      "epoch": 3.57487922705314,
+      "grad_norm": 0.07053821533918381,
+      "learning_rate": 9.9440343729616e-05,
+      "loss": 0.0246,
+      "step": 740
+    },
+    {
+      "epoch": 3.579710144927536,
+      "grad_norm": 0.056464601308107376,
+      "learning_rate": 9.943529995895679e-05,
+      "loss": 0.0151,
+      "step": 741
+    },
+    {
+      "epoch": 3.5845410628019323,
+      "grad_norm": 0.0730137899518013,
+      "learning_rate": 9.943023369130698e-05,
+      "loss": 0.0172,
+      "step": 742
+    },
+    {
+      "epoch": 3.5893719806763285,
+      "grad_norm": 0.06213295832276344,
+      "learning_rate": 9.942514492897212e-05,
+      "loss": 0.0166,
+      "step": 743
+    },
+    {
+      "epoch": 3.5942028985507246,
+      "grad_norm": 0.07693560421466827,
+      "learning_rate": 9.942003367426803e-05,
+      "loss": 0.0184,
+      "step": 744
+    },
+    {
+      "epoch": 3.5990338164251208,
+      "grad_norm": 0.06450362503528595,
+      "learning_rate": 9.941489992952071e-05,
+      "loss": 0.0188,
+      "step": 745
+    },
+    {
+      "epoch": 3.603864734299517,
+      "grad_norm": 0.06144733726978302,
+      "learning_rate": 9.940974369706642e-05,
+      "loss": 0.0177,
+      "step": 746
+    },
+    {
+      "epoch": 3.608695652173913,
+      "grad_norm": 0.061011552810668945,
+      "learning_rate": 9.940456497925168e-05,
+      "loss": 0.0135,
+      "step": 747
+    },
+    {
+      "epoch": 3.6135265700483092,
+      "grad_norm": 0.05580676719546318,
+      "learning_rate": 9.939936377843321e-05,
+      "loss": 0.0131,
+      "step": 748
+    },
+    {
+      "epoch": 3.6183574879227054,
+      "grad_norm": 0.07480030506849289,
+      "learning_rate": 9.939414009697795e-05,
+      "loss": 0.018,
+      "step": 749
+    },
+    {
+      "epoch": 3.6231884057971016,
+      "grad_norm": 0.1004641205072403,
+      "learning_rate": 9.938889393726314e-05,
+      "loss": 0.0166,
+      "step": 750
+    },
+    {
+      "epoch": 3.6231884057971016,
+      "eval_loss": 0.021343432366847992,
+      "eval_runtime": 20.6071,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
+      "step": 750
+    },
+    {
+      "epoch": 3.6280193236714977,
+      "grad_norm": 0.07337497919797897,
+      "learning_rate": 9.938362530167613e-05,
+      "loss": 0.0187,
+      "step": 751
+    },
+    {
+      "epoch": 3.632850241545894,
+      "grad_norm": 0.06580359488725662,
+      "learning_rate": 9.937833419261462e-05,
+      "loss": 0.0131,
+      "step": 752
+    },
+    {
+      "epoch": 3.63768115942029,
+      "grad_norm": 0.07170310616493225,
+      "learning_rate": 9.937302061248646e-05,
+      "loss": 0.0182,
+      "step": 753
+    },
+    {
+      "epoch": 3.642512077294686,
+      "grad_norm": 0.07878907769918442,
+      "learning_rate": 9.936768456370977e-05,
+      "loss": 0.0165,
+      "step": 754
+    },
+    {
+      "epoch": 3.6473429951690823,
+      "grad_norm": 0.06542734801769257,
+      "learning_rate": 9.936232604871285e-05,
+      "loss": 0.0126,
+      "step": 755
+    },
+    {
+      "epoch": 3.6521739130434785,
+      "grad_norm": 0.057411663234233856,
+      "learning_rate": 9.935694506993427e-05,
+      "loss": 0.0136,
+      "step": 756
+    },
+    {
+      "epoch": 3.6570048309178746,
+      "grad_norm": 0.05752653628587723,
+      "learning_rate": 9.935154162982281e-05,
+      "loss": 0.0175,
+      "step": 757
+    },
+    {
+      "epoch": 3.661835748792271,
+      "grad_norm": 0.0549551360309124,
+      "learning_rate": 9.934611573083744e-05,
+      "loss": 0.0144,
+      "step": 758
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 0.06211842969059944,
+      "learning_rate": 9.934066737544741e-05,
+      "loss": 0.0159,
+      "step": 759
+    },
+    {
+      "epoch": 3.6714975845410627,
+      "grad_norm": 0.08960019052028656,
+      "learning_rate": 9.93351965661321e-05,
+      "loss": 0.0232,
+      "step": 760
+    },
+    {
+      "epoch": 3.676328502415459,
+      "grad_norm": 0.08052323758602142,
+      "learning_rate": 9.932970330538123e-05,
+      "loss": 0.0194,
+      "step": 761
+    },
+    {
+      "epoch": 3.681159420289855,
+      "grad_norm": 0.07267940789461136,
+      "learning_rate": 9.932418759569462e-05,
+      "loss": 0.022,
+      "step": 762
+    },
+    {
+      "epoch": 3.685990338164251,
+      "grad_norm": 0.04540833458304405,
+      "learning_rate": 9.931864943958238e-05,
+      "loss": 0.0146,
+      "step": 763
+    },
+    {
+      "epoch": 3.6908212560386473,
+      "grad_norm": 0.06606751680374146,
+      "learning_rate": 9.931308883956479e-05,
+      "loss": 0.0188,
+      "step": 764
+    },
+    {
+      "epoch": 3.6956521739130435,
+      "grad_norm": 0.0551118440926075,
+      "learning_rate": 9.930750579817239e-05,
+      "loss": 0.0155,
+      "step": 765
+    },
+    {
+      "epoch": 3.7004830917874396,
+      "grad_norm": 0.06126049533486366,
+      "learning_rate": 9.93019003179459e-05,
+      "loss": 0.0138,
+      "step": 766
+    },
+    {
+      "epoch": 3.7053140096618358,
+      "grad_norm": 0.06616178900003433,
+      "learning_rate": 9.929627240143625e-05,
+      "loss": 0.017,
+      "step": 767
+    },
+    {
+      "epoch": 3.710144927536232,
+      "grad_norm": 0.04692595824599266,
+      "learning_rate": 9.92906220512046e-05,
+      "loss": 0.0133,
+      "step": 768
+    },
+    {
+      "epoch": 3.714975845410628,
+      "grad_norm": 0.06590303778648376,
+      "learning_rate": 9.92849492698223e-05,
+      "loss": 0.0157,
+      "step": 769
+    },
+    {
+      "epoch": 3.7198067632850242,
+      "grad_norm": 0.0660870224237442,
+      "learning_rate": 9.927925405987093e-05,
+      "loss": 0.0165,
+      "step": 770
+    },
+    {
+      "epoch": 3.7246376811594204,
+      "grad_norm": 0.06835846602916718,
+      "learning_rate": 9.927353642394224e-05,
+      "loss": 0.0185,
+      "step": 771
+    },
+    {
+      "epoch": 3.7294685990338166,
+      "grad_norm": 0.06855834275484085,
+      "learning_rate": 9.926779636463824e-05,
+      "loss": 0.0157,
+      "step": 772
+    },
+    {
+      "epoch": 3.7342995169082127,
+      "grad_norm": 0.07278718799352646,
+      "learning_rate": 9.926203388457107e-05,
+      "loss": 0.0207,
+      "step": 773
+    },
+    {
+      "epoch": 3.7391304347826084,
+      "grad_norm": 0.0773162767291069,
+      "learning_rate": 9.925624898636317e-05,
+      "loss": 0.0186,
+      "step": 774
+    },
+    {
+      "epoch": 3.7439613526570046,
+      "grad_norm": 0.0571766160428524,
+      "learning_rate": 9.925044167264708e-05,
+      "loss": 0.0141,
+      "step": 775
+    },
+    {
+      "epoch": 3.7439613526570046,
+      "eval_loss": 0.021501345559954643,
+      "eval_runtime": 21.3287,
+      "eval_samples_per_second": 4.689,
+      "eval_steps_per_second": 0.141,
+      "step": 775
+    },
+    {
+      "epoch": 3.7487922705314007,
+      "grad_norm": 0.08599105477333069,
+      "learning_rate": 9.924461194606561e-05,
+      "loss": 0.0195,
+      "step": 776
+    },
+    {
+      "epoch": 3.753623188405797,
+      "grad_norm": 0.050466280430555344,
+      "learning_rate": 9.923875980927175e-05,
+      "loss": 0.0146,
+      "step": 777
+    },
+    {
+      "epoch": 3.758454106280193,
+      "grad_norm": 0.061095286160707474,
+      "learning_rate": 9.923288526492869e-05,
+      "loss": 0.0166,
+      "step": 778
+    },
+    {
+      "epoch": 3.763285024154589,
+      "grad_norm": 0.05484727770090103,
+      "learning_rate": 9.922698831570982e-05,
+      "loss": 0.0166,
+      "step": 779
+    },
+    {
+      "epoch": 3.7681159420289854,
+      "grad_norm": 0.06217224523425102,
+      "learning_rate": 9.92210689642987e-05,
+      "loss": 0.0218,
+      "step": 780
+    },
+    {
+      "epoch": 3.7729468599033815,
+      "grad_norm": 0.07372289896011353,
+      "learning_rate": 9.921512721338912e-05,
+      "loss": 0.019,
+      "step": 781
+    },
+    {
+      "epoch": 3.7777777777777777,
+      "grad_norm": 0.07227632403373718,
+      "learning_rate": 9.920916306568504e-05,
+      "loss": 0.0154,
+      "step": 782
+    },
+    {
+      "epoch": 3.782608695652174,
+      "grad_norm": 0.05255334824323654,
+      "learning_rate": 9.920317652390063e-05,
+      "loss": 0.0128,
+      "step": 783
+    },
+    {
+      "epoch": 3.78743961352657,
+      "grad_norm": 0.052771199494600296,
+      "learning_rate": 9.919716759076025e-05,
+      "loss": 0.0131,
+      "step": 784
+    },
+    {
+      "epoch": 3.792270531400966,
+      "grad_norm": 0.05544090270996094,
+      "learning_rate": 9.919113626899841e-05,
+      "loss": 0.0152,
+      "step": 785
+    },
+    {
+      "epoch": 3.7971014492753623,
+      "grad_norm": 0.058198217302560806,
+      "learning_rate": 9.918508256135988e-05,
+      "loss": 0.015,
+      "step": 786
+    },
+    {
+      "epoch": 3.8019323671497585,
+      "grad_norm": 0.058879077434539795,
+      "learning_rate": 9.917900647059955e-05,
+      "loss": 0.0156,
+      "step": 787
+    },
+    {
+      "epoch": 3.8067632850241546,
+      "grad_norm": 0.05986057594418526,
+      "learning_rate": 9.917290799948253e-05,
+      "loss": 0.0137,
+      "step": 788
+    },
+    {
+      "epoch": 3.8115942028985508,
+      "grad_norm": 0.07952949404716492,
+      "learning_rate": 9.916678715078411e-05,
+      "loss": 0.0173,
+      "step": 789
+    },
+    {
+      "epoch": 3.816425120772947,
+      "grad_norm": 0.07398118823766708,
+      "learning_rate": 9.916064392728979e-05,
+      "loss": 0.0186,
+      "step": 790
+    },
+    {
+      "epoch": 3.821256038647343,
+      "grad_norm": 0.0736398994922638,
+      "learning_rate": 9.915447833179519e-05,
+      "loss": 0.014,
+      "step": 791
+    },
+    {
+      "epoch": 3.8260869565217392,
+      "grad_norm": 0.08044742792844772,
+      "learning_rate": 9.914829036710614e-05,
+      "loss": 0.0162,
+      "step": 792
+    },
+    {
+      "epoch": 3.8309178743961354,
+      "grad_norm": 0.05634943023324013,
+      "learning_rate": 9.914208003603869e-05,
+      "loss": 0.0166,
+      "step": 793
+    },
+    {
+      "epoch": 3.8357487922705316,
+      "grad_norm": 0.06842391192913055,
+      "learning_rate": 9.913584734141901e-05,
+      "loss": 0.0125,
+      "step": 794
+    },
+    {
+      "epoch": 3.8405797101449277,
+      "grad_norm": 0.05727720633149147,
+      "learning_rate": 9.912959228608348e-05,
+      "loss": 0.0157,
+      "step": 795
+    },
+    {
+      "epoch": 3.845410628019324,
+      "grad_norm": 0.0616447813808918,
+      "learning_rate": 9.912331487287864e-05,
+      "loss": 0.0204,
+      "step": 796
+    },
+    {
+      "epoch": 3.85024154589372,
+      "grad_norm": 0.08196991682052612,
+      "learning_rate": 9.911701510466124e-05,
+      "loss": 0.0246,
+      "step": 797
+    },
+    {
+      "epoch": 3.855072463768116,
+      "grad_norm": 0.06222192198038101,
+      "learning_rate": 9.911069298429814e-05,
+      "loss": 0.0144,
+      "step": 798
+    },
+    {
+      "epoch": 3.8599033816425123,
+      "grad_norm": 0.1069980263710022,
+      "learning_rate": 9.910434851466642e-05,
+      "loss": 0.0194,
+      "step": 799
+    },
+    {
+      "epoch": 3.864734299516908,
+      "grad_norm": 0.06227678433060646,
+      "learning_rate": 9.90979816986533e-05,
+      "loss": 0.0186,
+      "step": 800
+    },
+    {
+      "epoch": 3.864734299516908,
+      "eval_loss": 0.021084513515233994,
+      "eval_runtime": 20.6111,
+      "eval_samples_per_second": 4.852,
+      "eval_steps_per_second": 0.146,
+      "step": 800
+    },
+    {
+      "epoch": 3.869565217391304,
+      "grad_norm": 0.06213577091693878,
+      "learning_rate": 9.909159253915623e-05,
+      "loss": 0.016,
+      "step": 801
+    },
+    {
+      "epoch": 3.8743961352657004,
+      "grad_norm": 0.061741989105939865,
+      "learning_rate": 9.908518103908274e-05,
+      "loss": 0.015,
+      "step": 802
+    },
+    {
+      "epoch": 3.8792270531400965,
+      "grad_norm": 0.05534172058105469,
+      "learning_rate": 9.907874720135061e-05,
+      "loss": 0.0138,
+      "step": 803
+    },
+    {
+      "epoch": 3.8840579710144927,
+      "grad_norm": 0.06317410618066788,
+      "learning_rate": 9.907229102888772e-05,
+      "loss": 0.0156,
+      "step": 804
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 0.06512167304754257,
+      "learning_rate": 9.906581252463216e-05,
+      "loss": 0.0202,
+      "step": 805
+    },
+    {
+      "epoch": 3.893719806763285,
+      "grad_norm": 0.08261583000421524,
+      "learning_rate": 9.905931169153215e-05,
+      "loss": 0.0151,
+      "step": 806
+    },
+    {
+      "epoch": 3.898550724637681,
+      "grad_norm": 0.0779329165816307,
+      "learning_rate": 9.905278853254609e-05,
+      "loss": 0.0253,
+      "step": 807
+    },
+    {
+      "epoch": 3.9033816425120773,
+      "grad_norm": 0.07737044990062714,
+      "learning_rate": 9.904624305064255e-05,
+      "loss": 0.0151,
+      "step": 808
+    },
+    {
+      "epoch": 3.9082125603864735,
+      "grad_norm": 0.07593659311532974,
+      "learning_rate": 9.903967524880022e-05,
+      "loss": 0.0204,
+      "step": 809
+    },
+    {
+      "epoch": 3.9130434782608696,
+      "grad_norm": 0.05290926992893219,
+      "learning_rate": 9.903308513000798e-05,
+      "loss": 0.0175,
+      "step": 810
+    },
+    {
+      "epoch": 3.917874396135266,
+      "grad_norm": 0.05973507836461067,
+      "learning_rate": 9.902647269726489e-05,
+      "loss": 0.0144,
+      "step": 811
+    },
+    {
+      "epoch": 3.922705314009662,
+      "grad_norm": 0.0560474619269371,
+      "learning_rate": 9.901983795358008e-05,
+      "loss": 0.0127,
+      "step": 812
+    },
+    {
+      "epoch": 3.927536231884058,
+      "grad_norm": 0.06078074499964714,
+      "learning_rate": 9.901318090197291e-05,
+      "loss": 0.0146,
+      "step": 813
+    },
+    {
+      "epoch": 3.9323671497584543,
+      "grad_norm": 0.06304194033145905,
+      "learning_rate": 9.900650154547286e-05,
+      "loss": 0.0179,
+      "step": 814
+    },
+    {
+      "epoch": 3.9371980676328504,
+      "grad_norm": 0.07986494153738022,
+      "learning_rate": 9.89997998871196e-05,
+      "loss": 0.0228,
+      "step": 815
+    },
+    {
+      "epoch": 3.942028985507246,
+      "grad_norm": 0.06479338556528091,
+      "learning_rate": 9.899307592996287e-05,
+      "loss": 0.0142,
+      "step": 816
+    },
+    {
+      "epoch": 3.9468599033816423,
+      "grad_norm": 0.06285912543535233,
+      "learning_rate": 9.898632967706264e-05,
+      "loss": 0.0166,
+      "step": 817
+    },
+    {
+      "epoch": 3.9516908212560384,
+      "grad_norm": 0.06384378671646118,
+      "learning_rate": 9.897956113148899e-05,
+      "loss": 0.0174,
+      "step": 818
+    },
+    {
+      "epoch": 3.9565217391304346,
+      "grad_norm": 0.06164379417896271,
+      "learning_rate": 9.897277029632212e-05,
+      "loss": 0.0143,
+      "step": 819
+    },
+    {
+      "epoch": 3.9613526570048307,
+      "grad_norm": 0.08341162651777267,
+      "learning_rate": 9.896595717465243e-05,
+      "loss": 0.0162,
+      "step": 820
+    },
+    {
+      "epoch": 3.966183574879227,
+      "grad_norm": 0.08315421640872955,
+      "learning_rate": 9.89591217695804e-05,
+      "loss": 0.0201,
+      "step": 821
+    },
+    {
+      "epoch": 3.971014492753623,
+      "grad_norm": 0.0733807235956192,
+      "learning_rate": 9.895226408421669e-05,
+      "loss": 0.0213,
+      "step": 822
+    },
+    {
+      "epoch": 3.975845410628019,
+      "grad_norm": 0.05566437914967537,
+      "learning_rate": 9.894538412168213e-05,
+      "loss": 0.0173,
+      "step": 823
+    },
+    {
+      "epoch": 3.9806763285024154,
+      "grad_norm": 0.0825580433011055,
+      "learning_rate": 9.89384818851076e-05,
+      "loss": 0.0231,
+      "step": 824
+    },
+    {
+      "epoch": 3.9855072463768115,
+      "grad_norm": 0.05818888917565346,
+      "learning_rate": 9.893155737763419e-05,
+      "loss": 0.0119,
+      "step": 825
+    },
+    {
+      "epoch": 3.9855072463768115,
+      "eval_loss": 0.020353205502033234,
+      "eval_runtime": 21.3607,
+      "eval_samples_per_second": 4.681,
+      "eval_steps_per_second": 0.14,
+      "step": 825
+    },
+    {
+      "epoch": 3.9903381642512077,
+      "grad_norm": 0.055032528936862946,
+      "learning_rate": 9.892461060241312e-05,
+      "loss": 0.0158,
+      "step": 826
+    },
+    {
+      "epoch": 3.995169082125604,
+      "grad_norm": 0.0609041228890419,
+      "learning_rate": 9.891764156260568e-05,
+      "loss": 0.0172,
+      "step": 827
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.07889176905155182,
+      "learning_rate": 9.891065026138338e-05,
+      "loss": 0.0125,
+      "step": 828
+    },
+    {
+      "epoch": 4.004830917874396,
+      "grad_norm": 0.04410771280527115,
+      "learning_rate": 9.890363670192776e-05,
+      "loss": 0.0131,
+      "step": 829
+    },
+    {
+      "epoch": 4.009661835748792,
+      "grad_norm": 0.05885075405240059,
+      "learning_rate": 9.889660088743063e-05,
+      "loss": 0.0163,
+      "step": 830
+    },
+    {
+      "epoch": 4.0144927536231885,
+      "grad_norm": 0.042296018451452255,
+      "learning_rate": 9.888954282109377e-05,
+      "loss": 0.0104,
+      "step": 831
+    },
+    {
+      "epoch": 4.019323671497585,
+      "grad_norm": 0.06296995282173157,
+      "learning_rate": 9.888246250612921e-05,
+      "loss": 0.0141,
+      "step": 832
+    },
+    {
+      "epoch": 4.024154589371981,
+      "grad_norm": 0.06166675686836243,
+      "learning_rate": 9.887535994575902e-05,
+      "loss": 0.0108,
+      "step": 833
+    },
+    {
+      "epoch": 4.028985507246377,
+      "grad_norm": 0.06275904923677444,
+      "learning_rate": 9.886823514321548e-05,
+      "loss": 0.0102,
+      "step": 834
+    },
+    {
+      "epoch": 4.033816425120773,
+      "grad_norm": 0.07666383683681488,
+      "learning_rate": 9.886108810174088e-05,
+      "loss": 0.0161,
+      "step": 835
+    },
+    {
+      "epoch": 4.038647342995169,
+      "grad_norm": 0.07572819292545319,
+      "learning_rate": 9.885391882458773e-05,
+      "loss": 0.0149,
+      "step": 836
+    },
+    {
+      "epoch": 4.043478260869565,
+      "grad_norm": 0.05219224467873573,
+      "learning_rate": 9.88467273150186e-05,
+      "loss": 0.0065,
+      "step": 837
+    },
+    {
+      "epoch": 4.048309178743962,
+      "grad_norm": 0.05272039771080017,
+      "learning_rate": 9.883951357630622e-05,
+      "loss": 0.009,
+      "step": 838
+    },
+    {
+      "epoch": 4.053140096618358,
+      "grad_norm": 0.058741066604852676,
+      "learning_rate": 9.88322776117334e-05,
+      "loss": 0.0121,
+      "step": 839
+    },
+    {
+      "epoch": 4.057971014492754,
+      "grad_norm": 0.0712701827287674,
+      "learning_rate": 9.882501942459308e-05,
+      "loss": 0.0107,
+      "step": 840
+    },
+    {
+      "epoch": 4.06280193236715,
+      "grad_norm": 0.05130418390035629,
+      "learning_rate": 9.881773901818832e-05,
+      "loss": 0.0102,
+      "step": 841
+    },
+    {
+      "epoch": 4.067632850241546,
+      "grad_norm": 0.07418301701545715,
+      "learning_rate": 9.881043639583227e-05,
+      "loss": 0.0114,
+      "step": 842
+    },
+    {
+      "epoch": 4.072463768115942,
+      "grad_norm": 0.0563097782433033,
+      "learning_rate": 9.880311156084823e-05,
+      "loss": 0.0101,
+      "step": 843
+    },
+    {
+      "epoch": 4.0772946859903385,
+      "grad_norm": 0.11274167895317078,
+      "learning_rate": 9.879576451656955e-05,
+      "loss": 0.0311,
+      "step": 844
+    },
+    {
+      "epoch": 4.082125603864735,
+      "grad_norm": 0.06734360754489899,
+      "learning_rate": 9.878839526633974e-05,
+      "loss": 0.0105,
+      "step": 845
+    },
+    {
+      "epoch": 4.086956521739131,
+      "grad_norm": 0.04483459144830704,
+      "learning_rate": 9.878100381351239e-05,
+      "loss": 0.0095,
+      "step": 846
+    },
+    {
+      "epoch": 4.091787439613527,
+      "grad_norm": 0.06002177298069,
+      "learning_rate": 9.877359016145117e-05,
+      "loss": 0.0105,
+      "step": 847
+    },
+    {
+      "epoch": 4.096618357487923,
+      "grad_norm": 0.059757571667432785,
+      "learning_rate": 9.876615431352994e-05,
+      "loss": 0.0116,
+      "step": 848
+    },
+    {
+      "epoch": 4.101449275362318,
+      "grad_norm": 0.07245704531669617,
+      "learning_rate": 9.875869627313255e-05,
+      "loss": 0.0111,
+      "step": 849
+    },
+    {
+      "epoch": 4.106280193236715,
+      "grad_norm": 0.05853939801454544,
+      "learning_rate": 9.875121604365302e-05,
+      "loss": 0.0097,
+      "step": 850
+    },
+    {
+      "epoch": 4.106280193236715,
+      "eval_loss": 0.02096910960972309,
+      "eval_runtime": 20.6071,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
+      "step": 850
+    },
+    {
+      "epoch": 4.111111111111111,
+      "grad_norm": 0.07421394437551498,
+      "learning_rate": 9.874371362849548e-05,
+      "loss": 0.0117,
+      "step": 851
+    },
+    {
+      "epoch": 4.115942028985507,
+      "grad_norm": 0.05691864714026451,
+      "learning_rate": 9.873618903107406e-05,
+      "loss": 0.0111,
+      "step": 852
+    },
+    {
+      "epoch": 4.120772946859903,
+      "grad_norm": 0.06150828301906586,
+      "learning_rate": 9.87286422548131e-05,
+      "loss": 0.0124,
+      "step": 853
+    },
+    {
+      "epoch": 4.125603864734299,
+      "grad_norm": 0.06017156317830086,
+      "learning_rate": 9.872107330314696e-05,
+      "loss": 0.0116,
+      "step": 854
+    },
+    {
+      "epoch": 4.130434782608695,
+      "grad_norm": 0.0666033923625946,
+      "learning_rate": 9.871348217952012e-05,
+      "loss": 0.0132,
+      "step": 855
+    },
+    {
+      "epoch": 4.1352657004830915,
+      "grad_norm": 0.07256283611059189,
+      "learning_rate": 9.870586888738715e-05,
+      "loss": 0.011,
+      "step": 856
+    },
+    {
+      "epoch": 4.140096618357488,
+      "grad_norm": 0.06847047805786133,
+      "learning_rate": 9.869823343021271e-05,
+      "loss": 0.0122,
+      "step": 857
+    },
+    {
+      "epoch": 4.144927536231884,
+      "grad_norm": 0.0449073351919651,
+      "learning_rate": 9.869057581147152e-05,
+      "loss": 0.0085,
+      "step": 858
+    },
+    {
+      "epoch": 4.14975845410628,
+      "grad_norm": 0.06358068436384201,
+      "learning_rate": 9.868289603464842e-05,
+      "loss": 0.0111,
+      "step": 859
+    },
+    {
+      "epoch": 4.154589371980676,
+      "grad_norm": 0.0689551904797554,
+      "learning_rate": 9.86751941032383e-05,
+      "loss": 0.0142,
+      "step": 860
+    },
+    {
+      "epoch": 4.159420289855072,
+      "grad_norm": 0.08959997445344925,
+      "learning_rate": 9.866747002074617e-05,
+      "loss": 0.015,
+      "step": 861
+    },
+    {
+      "epoch": 4.164251207729468,
+      "grad_norm": 0.059535712003707886,
+      "learning_rate": 9.865972379068711e-05,
+      "loss": 0.0132,
+      "step": 862
+    },
+    {
+      "epoch": 4.169082125603865,
+      "grad_norm": 0.047409236431121826,
+      "learning_rate": 9.865195541658623e-05,
+      "loss": 0.0098,
+      "step": 863
+    },
+    {
+      "epoch": 4.173913043478261,
+      "grad_norm": 0.05755442753434181,
+      "learning_rate": 9.86441649019788e-05,
+      "loss": 0.0143,
+      "step": 864
+    },
+    {
+      "epoch": 4.178743961352657,
+      "grad_norm": 0.05657627061009407,
+      "learning_rate": 9.86363522504101e-05,
+      "loss": 0.0114,
+      "step": 865
+    },
+    {
+      "epoch": 4.183574879227053,
+      "grad_norm": 0.05874384567141533,
+      "learning_rate": 9.862851746543554e-05,
+      "loss": 0.0112,
+      "step": 866
+    },
+    {
+      "epoch": 4.188405797101449,
+      "grad_norm": 0.057612430304288864,
+      "learning_rate": 9.862066055062051e-05,
+      "loss": 0.0102,
+      "step": 867
+    },
+    {
+      "epoch": 4.193236714975845,
+      "grad_norm": 0.07835967093706131,
+      "learning_rate": 9.861278150954059e-05,
+      "loss": 0.013,
+      "step": 868
+    },
+    {
+      "epoch": 4.1980676328502415,
+      "grad_norm": 0.07963374257087708,
+      "learning_rate": 9.860488034578132e-05,
+      "loss": 0.0118,
+      "step": 869
+    },
+    {
+      "epoch": 4.202898550724638,
+      "grad_norm": 0.06441879272460938,
+      "learning_rate": 9.85969570629384e-05,
+      "loss": 0.0139,
+      "step": 870
+    },
+    {
+      "epoch": 4.207729468599034,
+      "grad_norm": 0.061957623809576035,
+      "learning_rate": 9.858901166461754e-05,
+      "loss": 0.0115,
+      "step": 871
+    },
+    {
+      "epoch": 4.21256038647343,
+      "grad_norm": 0.1095745861530304,
+      "learning_rate": 9.85810441544345e-05,
+      "loss": 0.011,
+      "step": 872
+    },
+    {
+      "epoch": 4.217391304347826,
+      "grad_norm": 0.06752188503742218,
+      "learning_rate": 9.857305453601517e-05,
+      "loss": 0.0167,
+      "step": 873
+    },
+    {
+      "epoch": 4.222222222222222,
+      "grad_norm": 0.08148062229156494,
+      "learning_rate": 9.856504281299546e-05,
+      "loss": 0.0157,
+      "step": 874
+    },
+    {
+      "epoch": 4.2270531400966185,
+      "grad_norm": 0.050344184041023254,
+      "learning_rate": 9.85570089890213e-05,
+      "loss": 0.0095,
+      "step": 875
+    },
+    {
+      "epoch": 4.2270531400966185,
+      "eval_loss": 0.020448356866836548,
+      "eval_runtime": 20.6294,
+      "eval_samples_per_second": 4.847,
+      "eval_steps_per_second": 0.145,
+      "step": 875
+    },
+    {
+      "epoch": 4.231884057971015,
+      "grad_norm": 0.056328706443309784,
+      "learning_rate": 9.854895306774876e-05,
+      "loss": 0.0135,
+      "step": 876
+    },
+    {
+      "epoch": 4.236714975845411,
+      "grad_norm": 0.04534146934747696,
+      "learning_rate": 9.854087505284391e-05,
+      "loss": 0.0097,
+      "step": 877
+    },
+    {
+      "epoch": 4.241545893719807,
+      "grad_norm": 0.08734095841646194,
+      "learning_rate": 9.853277494798287e-05,
+      "loss": 0.0116,
+      "step": 878
+    },
+    {
+      "epoch": 4.246376811594203,
+      "grad_norm": 0.07698874920606613,
+      "learning_rate": 9.852465275685187e-05,
+      "loss": 0.0115,
+      "step": 879
+    },
+    {
+      "epoch": 4.251207729468599,
+      "grad_norm": 0.04827764257788658,
+      "learning_rate": 9.851650848314713e-05,
+      "loss": 0.0112,
+      "step": 880
+    },
+    {
+      "epoch": 4.256038647342995,
+      "grad_norm": 0.07417704910039902,
+      "learning_rate": 9.850834213057494e-05,
+      "loss": 0.0141,
+      "step": 881
+    },
+    {
+      "epoch": 4.260869565217392,
+      "grad_norm": 0.06426126509904861,
+      "learning_rate": 9.850015370285164e-05,
+      "loss": 0.0141,
+      "step": 882
+    },
+    {
+      "epoch": 4.265700483091788,
+      "grad_norm": 0.05702616274356842,
+      "learning_rate": 9.84919432037036e-05,
+      "loss": 0.0094,
+      "step": 883
+    },
+    {
+      "epoch": 4.270531400966184,
+      "grad_norm": 0.050962239503860474,
+      "learning_rate": 9.84837106368673e-05,
+      "loss": 0.0123,
+      "step": 884
+    },
+    {
+      "epoch": 4.27536231884058,
+      "grad_norm": 0.05947406217455864,
+      "learning_rate": 9.847545600608917e-05,
+      "loss": 0.0106,
+      "step": 885
+    },
+    {
+      "epoch": 4.280193236714976,
+      "grad_norm": 0.0490386001765728,
+      "learning_rate": 9.846717931512573e-05,
+      "loss": 0.0137,
+      "step": 886
+    },
+    {
+      "epoch": 4.285024154589372,
+      "grad_norm": 0.05706576630473137,
+      "learning_rate": 9.845888056774354e-05,
+      "loss": 0.0119,
+      "step": 887
+    },
+    {
+      "epoch": 4.2898550724637685,
+      "grad_norm": 0.056531455367803574,
+      "learning_rate": 9.845055976771919e-05,
+      "loss": 0.0118,
+      "step": 888
+    },
+    {
+      "epoch": 4.294685990338165,
+      "grad_norm": 0.060446158051490784,
+      "learning_rate": 9.844221691883929e-05,
+      "loss": 0.0138,
+      "step": 889
+    },
+    {
+      "epoch": 4.29951690821256,
+      "grad_norm": 0.056813742965459824,
+      "learning_rate": 9.843385202490051e-05,
+      "loss": 0.0111,
+      "step": 890
+    },
+    {
+      "epoch": 4.304347826086957,
+      "grad_norm": 0.07757464796304703,
+      "learning_rate": 9.842546508970955e-05,
+      "loss": 0.0117,
+      "step": 891
+    },
+    {
+      "epoch": 4.309178743961352,
+      "grad_norm": 0.07340128719806671,
+      "learning_rate": 9.841705611708311e-05,
+      "loss": 0.0188,
+      "step": 892
+    },
+    {
+      "epoch": 4.314009661835748,
+      "grad_norm": 0.07141716778278351,
+      "learning_rate": 9.840862511084798e-05,
+      "loss": 0.0136,
+      "step": 893
+    },
+    {
+      "epoch": 4.318840579710145,
+      "grad_norm": 0.07496558129787445,
+      "learning_rate": 9.840017207484089e-05,
+      "loss": 0.0135,
+      "step": 894
+    },
+    {
+      "epoch": 4.323671497584541,
+      "grad_norm": 0.11215018481016159,
+      "learning_rate": 9.839169701290868e-05,
+      "loss": 0.0123,
+      "step": 895
+    },
+    {
+      "epoch": 4.328502415458937,
+      "grad_norm": 0.07307403534650803,
+      "learning_rate": 9.838319992890816e-05,
+      "loss": 0.0121,
+      "step": 896
+    },
+    {
+      "epoch": 4.333333333333333,
+      "grad_norm": 0.06442231684923172,
+      "learning_rate": 9.837468082670617e-05,
+      "loss": 0.0122,
+      "step": 897
+    },
+    {
+      "epoch": 4.338164251207729,
+      "grad_norm": 0.07327039539813995,
+      "learning_rate": 9.836613971017961e-05,
+      "loss": 0.0131,
+      "step": 898
+    },
+    {
+      "epoch": 4.342995169082125,
+      "grad_norm": 0.06935934722423553,
+      "learning_rate": 9.835757658321533e-05,
+      "loss": 0.0121,
+      "step": 899
+    },
+    {
+      "epoch": 4.3478260869565215,
+      "grad_norm": 0.05545048043131828,
+      "learning_rate": 9.834899144971025e-05,
+      "loss": 0.0119,
+      "step": 900
+    },
+    {
+      "epoch": 4.3478260869565215,
+      "eval_loss": 0.020673103630542755,
+      "eval_runtime": 20.6013,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 900
+    },
+    {
+      "epoch": 4.352657004830918,
+      "grad_norm": 0.0609741285443306,
+      "learning_rate": 9.834038431357129e-05,
+      "loss": 0.0115,
+      "step": 901
+    },
+    {
+      "epoch": 4.357487922705314,
+      "grad_norm": 0.04639492928981781,
+      "learning_rate": 9.833175517871538e-05,
+      "loss": 0.0086,
+      "step": 902
+    },
+    {
+      "epoch": 4.36231884057971,
+      "grad_norm": 0.06929849088191986,
+      "learning_rate": 9.832310404906946e-05,
+      "loss": 0.0126,
+      "step": 903
+    },
+    {
+      "epoch": 4.367149758454106,
+      "grad_norm": 0.07134636491537094,
+      "learning_rate": 9.831443092857049e-05,
+      "loss": 0.0147,
+      "step": 904
+    },
+    {
+      "epoch": 4.371980676328502,
+      "grad_norm": 0.05646871030330658,
+      "learning_rate": 9.830573582116542e-05,
+      "loss": 0.0126,
+      "step": 905
+    },
+    {
+      "epoch": 4.3768115942028984,
+      "grad_norm": 0.05775350332260132,
+      "learning_rate": 9.829701873081122e-05,
+      "loss": 0.0114,
+      "step": 906
+    },
+    {
+      "epoch": 4.381642512077295,
+      "grad_norm": 0.07218906283378601,
+      "learning_rate": 9.828827966147485e-05,
+      "loss": 0.0173,
+      "step": 907
+    },
+    {
+      "epoch": 4.386473429951691,
+      "grad_norm": 0.04605533555150032,
+      "learning_rate": 9.827951861713329e-05,
+      "loss": 0.0127,
+      "step": 908
+    },
+    {
+      "epoch": 4.391304347826087,
+      "grad_norm": 1.5717793703079224,
+      "learning_rate": 9.827073560177351e-05,
+      "loss": 0.0215,
+      "step": 909
+    },
+    {
+      "epoch": 4.396135265700483,
+      "grad_norm": 0.05545201897621155,
+      "learning_rate": 9.826193061939249e-05,
+      "loss": 0.0182,
+      "step": 910
+    },
+    {
+      "epoch": 4.400966183574879,
+      "grad_norm": 0.07933639734983444,
+      "learning_rate": 9.825310367399716e-05,
+      "loss": 0.017,
+      "step": 911
+    },
+    {
+      "epoch": 4.405797101449275,
+      "grad_norm": 0.23300503194332123,
+      "learning_rate": 9.824425476960453e-05,
+      "loss": 0.0348,
+      "step": 912
+    },
+    {
+      "epoch": 4.4106280193236715,
+      "grad_norm": 0.08851808309555054,
+      "learning_rate": 9.823538391024151e-05,
+      "loss": 0.0244,
+      "step": 913
+    },
+    {
+      "epoch": 4.415458937198068,
+      "grad_norm": 0.23018939793109894,
+      "learning_rate": 9.822649109994508e-05,
+      "loss": 0.0213,
+      "step": 914
+    },
+    {
+      "epoch": 4.420289855072464,
+      "grad_norm": 0.06539275497198105,
+      "learning_rate": 9.821757634276217e-05,
+      "loss": 0.0216,
+      "step": 915
+    },
+    {
+      "epoch": 4.42512077294686,
+      "grad_norm": 0.08064073324203491,
+      "learning_rate": 9.820863964274969e-05,
+      "loss": 0.0235,
+      "step": 916
+    },
+    {
+      "epoch": 4.429951690821256,
+      "grad_norm": 0.09684203565120697,
+      "learning_rate": 9.819968100397455e-05,
+      "loss": 0.02,
+      "step": 917
+    },
+    {
+      "epoch": 4.434782608695652,
+      "grad_norm": 0.0828334391117096,
+      "learning_rate": 9.819070043051366e-05,
+      "loss": 0.0179,
+      "step": 918
+    },
+    {
+      "epoch": 4.4396135265700485,
+      "grad_norm": 0.08924815058708191,
+      "learning_rate": 9.818169792645388e-05,
+      "loss": 0.0188,
+      "step": 919
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.08673924207687378,
+      "learning_rate": 9.817267349589205e-05,
+      "loss": 0.0159,
+      "step": 920
+    },
+    {
+      "epoch": 4.449275362318841,
+      "grad_norm": 0.08189379423856735,
+      "learning_rate": 9.816362714293504e-05,
+      "loss": 0.0125,
+      "step": 921
+    },
+    {
+      "epoch": 4.454106280193237,
+      "grad_norm": 0.09734530746936798,
+      "learning_rate": 9.815455887169965e-05,
+      "loss": 0.0206,
+      "step": 922
+    },
+    {
+      "epoch": 4.458937198067633,
+      "grad_norm": 0.12248886376619339,
+      "learning_rate": 9.814546868631264e-05,
+      "loss": 0.0171,
+      "step": 923
+    },
+    {
+      "epoch": 4.463768115942029,
+      "grad_norm": 0.15595419704914093,
+      "learning_rate": 9.813635659091078e-05,
+      "loss": 0.0158,
+      "step": 924
+    },
+    {
+      "epoch": 4.468599033816425,
+      "grad_norm": 0.07428774982690811,
+      "learning_rate": 9.81272225896408e-05,
+      "loss": 0.0131,
+      "step": 925
+    },
+    {
+      "epoch": 4.468599033816425,
+      "eval_loss": 0.025713801383972168,
+      "eval_runtime": 21.3609,
+      "eval_samples_per_second": 4.681,
+      "eval_steps_per_second": 0.14,
+      "step": 925
+    },
+    {
+      "epoch": 4.473429951690822,
+      "grad_norm": 0.09882636368274689,
+      "learning_rate": 9.81180666866594e-05,
+      "loss": 0.0176,
+      "step": 926
+    },
+    {
+      "epoch": 4.478260869565218,
+      "grad_norm": 0.07712045311927795,
+      "learning_rate": 9.810888888613323e-05,
+      "loss": 0.014,
+      "step": 927
+    },
+    {
+      "epoch": 4.483091787439614,
+      "grad_norm": 0.10221463441848755,
+      "learning_rate": 9.809968919223893e-05,
+      "loss": 0.0204,
+      "step": 928
+    },
+    {
+      "epoch": 4.48792270531401,
+      "grad_norm": 0.07028807699680328,
+      "learning_rate": 9.809046760916309e-05,
+      "loss": 0.0123,
+      "step": 929
+    },
+    {
+      "epoch": 4.492753623188406,
+      "grad_norm": 0.08684271574020386,
+      "learning_rate": 9.808122414110228e-05,
+      "loss": 0.0147,
+      "step": 930
+    },
+    {
+      "epoch": 4.4975845410628015,
+      "grad_norm": 0.08106634020805359,
+      "learning_rate": 9.807195879226298e-05,
+      "loss": 0.0167,
+      "step": 931
+    },
+    {
+      "epoch": 4.5024154589371985,
+      "grad_norm": 0.10349582135677338,
+      "learning_rate": 9.806267156686169e-05,
+      "loss": 0.0139,
+      "step": 932
+    },
+    {
+      "epoch": 4.507246376811594,
+      "grad_norm": 0.08159704506397247,
+      "learning_rate": 9.80533624691248e-05,
+      "loss": 0.0131,
+      "step": 933
+    },
+    {
+      "epoch": 4.512077294685991,
+      "grad_norm": 0.07411141693592072,
+      "learning_rate": 9.804403150328872e-05,
+      "loss": 0.0145,
+      "step": 934
+    },
+    {
+      "epoch": 4.516908212560386,
+      "grad_norm": 0.08534754812717438,
+      "learning_rate": 9.803467867359977e-05,
+      "loss": 0.0164,
+      "step": 935
+    },
+    {
+      "epoch": 4.521739130434782,
+      "grad_norm": 0.07628322392702103,
+      "learning_rate": 9.802530398431424e-05,
+      "loss": 0.0154,
+      "step": 936
+    },
+    {
+      "epoch": 4.526570048309178,
+      "grad_norm": 0.06755129992961884,
+      "learning_rate": 9.801590743969835e-05,
+      "loss": 0.013,
+      "step": 937
+    },
+    {
+      "epoch": 4.531400966183575,
+      "grad_norm": 0.07023568451404572,
+      "learning_rate": 9.800648904402827e-05,
+      "loss": 0.0143,
+      "step": 938
+    },
+    {
+      "epoch": 4.536231884057971,
+      "grad_norm": 0.11462641507387161,
+      "learning_rate": 9.799704880159013e-05,
+      "loss": 0.0223,
+      "step": 939
+    },
+    {
+      "epoch": 4.541062801932367,
+      "grad_norm": 0.07971464842557907,
+      "learning_rate": 9.798758671668e-05,
+      "loss": 0.0144,
+      "step": 940
+    },
+    {
+      "epoch": 4.545893719806763,
+      "grad_norm": 0.07792872935533524,
+      "learning_rate": 9.797810279360385e-05,
+      "loss": 0.0141,
+      "step": 941
+    },
+    {
+      "epoch": 4.550724637681159,
+      "grad_norm": 0.06341885030269623,
+      "learning_rate": 9.796859703667762e-05,
+      "loss": 0.0141,
+      "step": 942
+    },
+    {
+      "epoch": 4.555555555555555,
+      "grad_norm": 0.0954955667257309,
+      "learning_rate": 9.795906945022721e-05,
+      "loss": 0.0171,
+      "step": 943
+    },
+    {
+      "epoch": 4.5603864734299515,
+      "grad_norm": 0.08577444404363632,
+      "learning_rate": 9.794952003858844e-05,
+      "loss": 0.0171,
+      "step": 944
+    },
+    {
+      "epoch": 4.565217391304348,
+      "grad_norm": 0.07980582863092422,
+      "learning_rate": 9.793994880610702e-05,
+      "loss": 0.0163,
+      "step": 945
+    },
+    {
+      "epoch": 4.570048309178744,
+      "grad_norm": 0.0770900696516037,
+      "learning_rate": 9.793035575713862e-05,
+      "loss": 0.0153,
+      "step": 946
+    },
+    {
+      "epoch": 4.57487922705314,
+      "grad_norm": 0.07175513356924057,
+      "learning_rate": 9.792074089604886e-05,
+      "loss": 0.0132,
+      "step": 947
+    },
+    {
+      "epoch": 4.579710144927536,
+      "grad_norm": 0.07475357502698898,
+      "learning_rate": 9.791110422721326e-05,
+      "loss": 0.0116,
+      "step": 948
+    },
+    {
+      "epoch": 4.584541062801932,
+      "grad_norm": 0.07414525747299194,
+      "learning_rate": 9.790144575501724e-05,
+      "loss": 0.0175,
+      "step": 949
+    },
+    {
+      "epoch": 4.5893719806763285,
+      "grad_norm": 0.058295838534832,
+      "learning_rate": 9.789176548385624e-05,
+      "loss": 0.0123,
+      "step": 950
+    },
+    {
+      "epoch": 4.5893719806763285,
+      "eval_loss": 0.02279841899871826,
+      "eval_runtime": 20.6131,
+      "eval_samples_per_second": 4.851,
+      "eval_steps_per_second": 0.146,
+      "step": 950
+    },
+    {
+      "epoch": 4.594202898550725,
+      "grad_norm": 0.061069000512361526,
+      "learning_rate": 9.78820634181355e-05,
+      "loss": 0.0099,
+      "step": 951
+    },
+    {
+      "epoch": 4.599033816425121,
+      "grad_norm": 0.0968296080827713,
+      "learning_rate": 9.787233956227023e-05,
+      "loss": 0.0177,
+      "step": 952
+    },
+    {
+      "epoch": 4.603864734299517,
+      "grad_norm": 0.07276859134435654,
+      "learning_rate": 9.786259392068559e-05,
+      "loss": 0.0124,
+      "step": 953
+    },
+    {
+      "epoch": 4.608695652173913,
+      "grad_norm": 0.08032967150211334,
+      "learning_rate": 9.78528264978166e-05,
+      "loss": 0.0121,
+      "step": 954
+    },
+    {
+      "epoch": 4.613526570048309,
+      "grad_norm": 0.09521745890378952,
+      "learning_rate": 9.78430372981082e-05,
+      "loss": 0.0122,
+      "step": 955
+    },
+    {
+      "epoch": 4.618357487922705,
+      "grad_norm": 0.08671263605356216,
+      "learning_rate": 9.783322632601529e-05,
+      "loss": 0.0132,
+      "step": 956
+    },
+    {
+      "epoch": 4.6231884057971016,
+      "grad_norm": 0.09144391119480133,
+      "learning_rate": 9.78233935860026e-05,
+      "loss": 0.0214,
+      "step": 957
+    },
+    {
+      "epoch": 4.628019323671498,
+      "grad_norm": 0.06860186159610748,
+      "learning_rate": 9.781353908254485e-05,
+      "loss": 0.0141,
+      "step": 958
+    },
+    {
+      "epoch": 4.632850241545894,
+      "grad_norm": 0.05559476092457771,
+      "learning_rate": 9.78036628201266e-05,
+      "loss": 0.0123,
+      "step": 959
+    },
+    {
+      "epoch": 4.63768115942029,
+      "grad_norm": 0.06625663489103317,
+      "learning_rate": 9.779376480324233e-05,
+      "loss": 0.0134,
+      "step": 960
+    },
+    {
+      "epoch": 4.642512077294686,
+      "grad_norm": 0.10225360840559006,
+      "learning_rate": 9.778384503639644e-05,
+      "loss": 0.0237,
+      "step": 961
+    },
+    {
+      "epoch": 4.647342995169082,
+      "grad_norm": 0.06415294110774994,
+      "learning_rate": 9.777390352410319e-05,
+      "loss": 0.0184,
+      "step": 962
+    },
+    {
+      "epoch": 4.6521739130434785,
+      "grad_norm": 0.057260725647211075,
+      "learning_rate": 9.776394027088678e-05,
+      "loss": 0.0131,
+      "step": 963
+    },
+    {
+      "epoch": 4.657004830917875,
+      "grad_norm": 0.07303762435913086,
+      "learning_rate": 9.775395528128128e-05,
+      "loss": 0.0139,
+      "step": 964
+    },
+    {
+      "epoch": 4.661835748792271,
+      "grad_norm": 0.059439219534397125,
+      "learning_rate": 9.774394855983065e-05,
+      "loss": 0.0125,
+      "step": 965
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 0.07815828174352646,
+      "learning_rate": 9.773392011108873e-05,
+      "loss": 0.015,
+      "step": 966
+    },
+    {
+      "epoch": 4.671497584541063,
+      "grad_norm": 0.0615699328482151,
+      "learning_rate": 9.772386993961928e-05,
+      "loss": 0.013,
+      "step": 967
+    },
+    {
+      "epoch": 4.676328502415459,
+      "grad_norm": 0.05981515347957611,
+      "learning_rate": 9.771379804999592e-05,
+      "loss": 0.0123,
+      "step": 968
+    },
+    {
+      "epoch": 4.681159420289855,
+      "grad_norm": 0.0661519393324852,
+      "learning_rate": 9.770370444680217e-05,
+      "loss": 0.0127,
+      "step": 969
+    },
+    {
+      "epoch": 4.685990338164252,
+      "grad_norm": 0.05894894897937775,
+      "learning_rate": 9.769358913463141e-05,
+      "loss": 0.0098,
+      "step": 970
+    },
+    {
+      "epoch": 4.690821256038648,
+      "grad_norm": 0.06429290771484375,
+      "learning_rate": 9.76834521180869e-05,
+      "loss": 0.0125,
+      "step": 971
+    },
+    {
+      "epoch": 4.695652173913043,
+      "grad_norm": 0.06996926665306091,
+      "learning_rate": 9.767329340178179e-05,
+      "loss": 0.0141,
+      "step": 972
+    },
+    {
+      "epoch": 4.70048309178744,
+      "grad_norm": 0.060817550867795944,
+      "learning_rate": 9.766311299033913e-05,
+      "loss": 0.0127,
+      "step": 973
+    },
+    {
+      "epoch": 4.705314009661835,
+      "grad_norm": 0.06457247585058212,
+      "learning_rate": 9.76529108883918e-05,
+      "loss": 0.0131,
+      "step": 974
+    },
+    {
+      "epoch": 4.710144927536232,
+      "grad_norm": 0.07117529958486557,
+      "learning_rate": 9.764268710058258e-05,
+      "loss": 0.0133,
+      "step": 975
+    },
+    {
+      "epoch": 4.710144927536232,
+      "eval_loss": 0.020405687391757965,
+      "eval_runtime": 20.6298,
+      "eval_samples_per_second": 4.847,
+      "eval_steps_per_second": 0.145,
+      "step": 975
+    },
+    {
+      "epoch": 4.714975845410628,
+      "grad_norm": 0.06411249190568924,
+      "learning_rate": 9.763244163156407e-05,
+      "loss": 0.0125,
+      "step": 976
+    },
+    {
+      "epoch": 4.719806763285024,
+      "grad_norm": 0.0879925787448883,
+      "learning_rate": 9.762217448599882e-05,
+      "loss": 0.0187,
+      "step": 977
+    },
+    {
+      "epoch": 4.72463768115942,
+      "grad_norm": 0.05720191076397896,
+      "learning_rate": 9.761188566855916e-05,
+      "loss": 0.0106,
+      "step": 978
+    },
+    {
+      "epoch": 4.729468599033816,
+      "grad_norm": 0.06640984117984772,
+      "learning_rate": 9.760157518392735e-05,
+      "loss": 0.0113,
+      "step": 979
+    },
+    {
+      "epoch": 4.734299516908212,
+      "grad_norm": 0.05869889631867409,
+      "learning_rate": 9.759124303679544e-05,
+      "loss": 0.011,
+      "step": 980
+    },
+    {
+      "epoch": 4.739130434782608,
+      "grad_norm": 0.052954427897930145,
+      "learning_rate": 9.758088923186541e-05,
+      "loss": 0.0115,
+      "step": 981
+    },
+    {
+      "epoch": 4.743961352657005,
+      "grad_norm": 0.049915559589862823,
+      "learning_rate": 9.757051377384906e-05,
+      "loss": 0.0114,
+      "step": 982
+    },
+    {
+      "epoch": 4.748792270531401,
+      "grad_norm": 0.07832711189985275,
+      "learning_rate": 9.756011666746805e-05,
+      "loss": 0.015,
+      "step": 983
+    },
+    {
+      "epoch": 4.753623188405797,
+      "grad_norm": 0.06998534500598907,
+      "learning_rate": 9.754969791745387e-05,
+      "loss": 0.0132,
+      "step": 984
+    },
+    {
+      "epoch": 4.758454106280193,
+      "grad_norm": 0.06319990009069443,
+      "learning_rate": 9.75392575285479e-05,
+      "loss": 0.0117,
+      "step": 985
+    },
+    {
+      "epoch": 4.763285024154589,
+      "grad_norm": 0.04931075870990753,
+      "learning_rate": 9.752879550550133e-05,
+      "loss": 0.0108,
+      "step": 986
+    },
+    {
+      "epoch": 4.768115942028985,
+      "grad_norm": 0.04907230660319328,
+      "learning_rate": 9.751831185307523e-05,
+      "loss": 0.0098,
+      "step": 987
+    },
+    {
+      "epoch": 4.7729468599033815,
+      "grad_norm": 0.06535334140062332,
+      "learning_rate": 9.750780657604051e-05,
+      "loss": 0.0144,
+      "step": 988
+    },
+    {
+      "epoch": 4.777777777777778,
+      "grad_norm": 0.08152740448713303,
+      "learning_rate": 9.749727967917785e-05,
+      "loss": 0.0126,
+      "step": 989
+    },
+    {
+      "epoch": 4.782608695652174,
+      "grad_norm": 0.055403295904397964,
+      "learning_rate": 9.748673116727787e-05,
+      "loss": 0.0134,
+      "step": 990
+    },
+    {
+      "epoch": 4.78743961352657,
+      "grad_norm": 0.0632065013051033,
+      "learning_rate": 9.747616104514097e-05,
+      "loss": 0.0123,
+      "step": 991
+    },
+    {
+      "epoch": 4.792270531400966,
+      "grad_norm": 0.058326445519924164,
+      "learning_rate": 9.74655693175774e-05,
+      "loss": 0.0119,
+      "step": 992
+    },
+    {
+      "epoch": 4.797101449275362,
+      "grad_norm": 0.04671332985162735,
+      "learning_rate": 9.745495598940722e-05,
+      "loss": 0.0094,
+      "step": 993
+    },
+    {
+      "epoch": 4.8019323671497585,
+      "grad_norm": 0.058310024440288544,
+      "learning_rate": 9.744432106546035e-05,
+      "loss": 0.0103,
+      "step": 994
+    },
+    {
+      "epoch": 4.806763285024155,
+      "grad_norm": 0.061691973358392715,
+      "learning_rate": 9.743366455057654e-05,
+      "loss": 0.0129,
+      "step": 995
+    },
+    {
+      "epoch": 4.811594202898551,
+      "grad_norm": 0.0760740339756012,
+      "learning_rate": 9.742298644960533e-05,
+      "loss": 0.0154,
+      "step": 996
+    },
+    {
+      "epoch": 4.816425120772947,
+      "grad_norm": 0.06041797623038292,
+      "learning_rate": 9.741228676740611e-05,
+      "loss": 0.013,
+      "step": 997
+    },
+    {
+      "epoch": 4.821256038647343,
+      "grad_norm": 0.06114066019654274,
+      "learning_rate": 9.74015655088481e-05,
+      "loss": 0.0069,
+      "step": 998
+    },
+    {
+      "epoch": 4.826086956521739,
+      "grad_norm": 0.11095196008682251,
+      "learning_rate": 9.739082267881029e-05,
+      "loss": 0.0128,
+      "step": 999
+    },
+    {
+      "epoch": 4.830917874396135,
+      "grad_norm": 0.058319080621004105,
+      "learning_rate": 9.738005828218155e-05,
+      "loss": 0.0115,
+      "step": 1000
+    },
+    {
+      "epoch": 4.830917874396135,
+      "eval_loss": 0.019139988347887993,
+      "eval_runtime": 20.6148,
+      "eval_samples_per_second": 4.851,
+      "eval_steps_per_second": 0.146,
+      "step": 1000
+    },
+    {
+      "epoch": 4.835748792270532,
+      "grad_norm": 0.06817466765642166,
+      "learning_rate": 9.736927232386055e-05,
+      "loss": 0.0156,
+      "step": 1001
+    },
+    {
+      "epoch": 4.840579710144928,
+      "grad_norm": 0.05912366136908531,
+      "learning_rate": 9.735846480875572e-05,
+      "loss": 0.0144,
+      "step": 1002
+    },
+    {
+      "epoch": 4.845410628019324,
+      "grad_norm": 0.08707557618618011,
+      "learning_rate": 9.734763574178536e-05,
+      "loss": 0.0128,
+      "step": 1003
+    },
+    {
+      "epoch": 4.85024154589372,
+      "grad_norm": 0.08672689646482468,
+      "learning_rate": 9.733678512787757e-05,
+      "loss": 0.0122,
+      "step": 1004
+    },
+    {
+      "epoch": 4.855072463768116,
+      "grad_norm": 0.06592126190662384,
+      "learning_rate": 9.732591297197021e-05,
+      "loss": 0.0163,
+      "step": 1005
+    },
+    {
+      "epoch": 4.859903381642512,
+      "grad_norm": 0.0610274113714695,
+      "learning_rate": 9.731501927901096e-05,
+      "loss": 0.0146,
+      "step": 1006
+    },
+    {
+      "epoch": 4.8647342995169085,
+      "grad_norm": 0.05984511226415634,
+      "learning_rate": 9.730410405395739e-05,
+      "loss": 0.0112,
+      "step": 1007
+    },
+    {
+      "epoch": 4.869565217391305,
+      "grad_norm": 0.04920889064669609,
+      "learning_rate": 9.729316730177674e-05,
+      "loss": 0.0091,
+      "step": 1008
+    },
+    {
+      "epoch": 4.874396135265701,
+      "grad_norm": 0.07654332369565964,
+      "learning_rate": 9.72822090274461e-05,
+      "loss": 0.012,
+      "step": 1009
+    },
+    {
+      "epoch": 4.879227053140097,
+      "grad_norm": 0.04907603561878204,
+      "learning_rate": 9.727122923595238e-05,
+      "loss": 0.0112,
+      "step": 1010
+    },
+    {
+      "epoch": 4.884057971014493,
+      "grad_norm": 0.04914016276597977,
+      "learning_rate": 9.726022793229224e-05,
+      "loss": 0.0105,
+      "step": 1011
+    },
+    {
+      "epoch": 4.888888888888889,
+      "grad_norm": 0.055787090212106705,
+      "learning_rate": 9.724920512147215e-05,
+      "loss": 0.0138,
+      "step": 1012
+    },
+    {
+      "epoch": 4.8937198067632846,
+      "grad_norm": 0.06109369918704033,
+      "learning_rate": 9.723816080850839e-05,
+      "loss": 0.0113,
+      "step": 1013
+    },
+    {
+      "epoch": 4.898550724637682,
+      "grad_norm": 0.0676606222987175,
+      "learning_rate": 9.722709499842695e-05,
+      "loss": 0.0123,
+      "step": 1014
+    },
+    {
+      "epoch": 4.903381642512077,
+      "grad_norm": 0.05014638230204582,
+      "learning_rate": 9.721600769626368e-05,
+      "loss": 0.013,
+      "step": 1015
+    },
+    {
+      "epoch": 4.908212560386474,
+      "grad_norm": 0.07281427085399628,
+      "learning_rate": 9.720489890706421e-05,
+      "loss": 0.0152,
+      "step": 1016
+    },
+    {
+      "epoch": 4.913043478260869,
+      "grad_norm": 0.08129395544528961,
+      "learning_rate": 9.71937686358839e-05,
+      "loss": 0.0194,
+      "step": 1017
+    },
+    {
+      "epoch": 4.917874396135265,
+      "grad_norm": 0.07277390360832214,
+      "learning_rate": 9.718261688778789e-05,
+      "loss": 0.0134,
+      "step": 1018
+    },
+    {
+      "epoch": 4.9227053140096615,
+      "grad_norm": 0.06831222772598267,
+      "learning_rate": 9.717144366785112e-05,
+      "loss": 0.0126,
+      "step": 1019
+    },
+    {
+      "epoch": 4.927536231884058,
+      "grad_norm": 0.059769704937934875,
+      "learning_rate": 9.716024898115832e-05,
+      "loss": 0.015,
+      "step": 1020
+    },
+    {
+      "epoch": 4.932367149758454,
+      "grad_norm": 0.07310938090085983,
+      "learning_rate": 9.714903283280393e-05,
+      "loss": 0.0125,
+      "step": 1021
+    },
+    {
+      "epoch": 4.93719806763285,
+      "grad_norm": 0.06537635624408722,
+      "learning_rate": 9.71377952278922e-05,
+      "loss": 0.0113,
+      "step": 1022
+    },
+    {
+      "epoch": 4.942028985507246,
+      "grad_norm": 0.07284176349639893,
+      "learning_rate": 9.712653617153715e-05,
+      "loss": 0.0131,
+      "step": 1023
+    },
+    {
+      "epoch": 4.946859903381642,
+      "grad_norm": 0.0574897825717926,
+      "learning_rate": 9.71152556688625e-05,
+      "loss": 0.0109,
+      "step": 1024
+    },
+    {
+      "epoch": 4.951690821256038,
+      "grad_norm": 0.07543810456991196,
+      "learning_rate": 9.710395372500183e-05,
+      "loss": 0.0152,
+      "step": 1025
+    },
+    {
+      "epoch": 4.951690821256038,
+      "eval_loss": 0.020091358572244644,
+      "eval_runtime": 20.6004,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 1025
+    },
+    {
+      "epoch": 4.956521739130435,
+      "grad_norm": 0.08204186707735062,
+      "learning_rate": 9.70926303450984e-05,
+      "loss": 0.0118,
+      "step": 1026
+    },
+    {
+      "epoch": 4.961352657004831,
+      "grad_norm": 0.06970630586147308,
+      "learning_rate": 9.708128553430524e-05,
+      "loss": 0.0109,
+      "step": 1027
+    },
+    {
+      "epoch": 4.966183574879227,
+      "grad_norm": 0.06959115713834763,
+      "learning_rate": 9.706991929778516e-05,
+      "loss": 0.0133,
+      "step": 1028
+    },
+    {
+      "epoch": 4.971014492753623,
+      "grad_norm": 0.05290956422686577,
+      "learning_rate": 9.705853164071065e-05,
+      "loss": 0.0113,
+      "step": 1029
+    },
+    {
+      "epoch": 4.975845410628019,
+      "grad_norm": 0.08411956578493118,
+      "learning_rate": 9.704712256826407e-05,
+      "loss": 0.0173,
+      "step": 1030
+    },
+    {
+      "epoch": 4.980676328502415,
+      "grad_norm": 0.05763190984725952,
+      "learning_rate": 9.703569208563741e-05,
+      "loss": 0.0105,
+      "step": 1031
+    },
+    {
+      "epoch": 4.9855072463768115,
+      "grad_norm": 0.07256191968917847,
+      "learning_rate": 9.702424019803247e-05,
+      "loss": 0.0163,
+      "step": 1032
+    },
+    {
+      "epoch": 4.990338164251208,
+      "grad_norm": 0.052556559443473816,
+      "learning_rate": 9.701276691066075e-05,
+      "loss": 0.0122,
+      "step": 1033
+    },
+    {
+      "epoch": 4.995169082125604,
+      "grad_norm": 0.06209128350019455,
+      "learning_rate": 9.700127222874351e-05,
+      "loss": 0.0155,
+      "step": 1034
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.07931708544492722,
+      "learning_rate": 9.698975615751176e-05,
+      "loss": 0.0092,
+      "step": 1035
+    },
+    {
+      "epoch": 5.004830917874396,
+      "grad_norm": 0.06323495507240295,
+      "learning_rate": 9.697821870220623e-05,
+      "loss": 0.0101,
+      "step": 1036
+    },
+    {
+      "epoch": 5.009661835748792,
+      "grad_norm": 0.057949360460042953,
+      "learning_rate": 9.696665986807734e-05,
+      "loss": 0.0108,
+      "step": 1037
+    },
+    {
+      "epoch": 5.0144927536231885,
+      "grad_norm": 0.04627874121069908,
+      "learning_rate": 9.695507966038533e-05,
+      "loss": 0.0079,
+      "step": 1038
+    },
+    {
+      "epoch": 5.019323671497585,
+      "grad_norm": 0.05395060405135155,
+      "learning_rate": 9.694347808440007e-05,
+      "loss": 0.0145,
+      "step": 1039
+    },
+    {
+      "epoch": 5.024154589371981,
+      "grad_norm": 0.0515243224799633,
+      "learning_rate": 9.693185514540122e-05,
+      "loss": 0.0089,
+      "step": 1040
+    },
+    {
+      "epoch": 5.028985507246377,
+      "grad_norm": 0.07171349972486496,
+      "learning_rate": 9.692021084867814e-05,
+      "loss": 0.0079,
+      "step": 1041
+    },
+    {
+      "epoch": 5.033816425120773,
+      "grad_norm": 0.0560278482735157,
+      "learning_rate": 9.690854519952995e-05,
+      "loss": 0.0081,
+      "step": 1042
+    },
+    {
+      "epoch": 5.038647342995169,
+      "grad_norm": 0.059097371995449066,
+      "learning_rate": 9.689685820326537e-05,
+      "loss": 0.0107,
+      "step": 1043
+    },
+    {
+      "epoch": 5.043478260869565,
+      "grad_norm": 0.055744510143995285,
+      "learning_rate": 9.688514986520296e-05,
+      "loss": 0.0087,
+      "step": 1044
+    },
+    {
+      "epoch": 5.048309178743962,
+      "grad_norm": 0.0558600053191185,
+      "learning_rate": 9.687342019067093e-05,
+      "loss": 0.0098,
+      "step": 1045
+    },
+    {
+      "epoch": 5.053140096618358,
+      "grad_norm": 0.06191682815551758,
+      "learning_rate": 9.686166918500723e-05,
+      "loss": 0.0116,
+      "step": 1046
+    },
+    {
+      "epoch": 5.057971014492754,
+      "grad_norm": 0.06881178915500641,
+      "learning_rate": 9.684989685355947e-05,
+      "loss": 0.0075,
+      "step": 1047
+    },
+    {
+      "epoch": 5.06280193236715,
+      "grad_norm": 0.05575813725590706,
+      "learning_rate": 9.683810320168504e-05,
+      "loss": 0.0069,
+      "step": 1048
+    },
+    {
+      "epoch": 5.067632850241546,
+      "grad_norm": 0.08796427398920059,
+      "learning_rate": 9.682628823475095e-05,
+      "loss": 0.0084,
+      "step": 1049
+    },
+    {
+      "epoch": 5.072463768115942,
+      "grad_norm": 0.04683411493897438,
+      "learning_rate": 9.681445195813399e-05,
+      "loss": 0.0075,
+      "step": 1050
+    },
+    {
+      "epoch": 5.072463768115942,
+      "eval_loss": 0.018816815689206123,
+      "eval_runtime": 20.6451,
+      "eval_samples_per_second": 4.844,
+      "eval_steps_per_second": 0.145,
+      "step": 1050
+    },
+    {
+      "epoch": 5.0772946859903385,
+      "grad_norm": 0.07573356479406357,
+      "learning_rate": 9.680259437722055e-05,
+      "loss": 0.0077,
+      "step": 1051
+    },
+    {
+      "epoch": 5.082125603864735,
+      "grad_norm": 0.056269437074661255,
+      "learning_rate": 9.679071549740682e-05,
+      "loss": 0.0093,
+      "step": 1052
+    },
+    {
+      "epoch": 5.086956521739131,
+      "grad_norm": 0.06449293345212936,
+      "learning_rate": 9.677881532409863e-05,
+      "loss": 0.0102,
+      "step": 1053
+    },
+    {
+      "epoch": 5.091787439613527,
+      "grad_norm": 0.047550760209560394,
+      "learning_rate": 9.676689386271147e-05,
+      "loss": 0.0076,
+      "step": 1054
+    },
+    {
+      "epoch": 5.096618357487923,
+      "grad_norm": 0.04935392364859581,
+      "learning_rate": 9.675495111867059e-05,
+      "loss": 0.0061,
+      "step": 1055
+    },
+    {
+      "epoch": 5.101449275362318,
+      "grad_norm": 0.05520419031381607,
+      "learning_rate": 9.674298709741087e-05,
+      "loss": 0.01,
+      "step": 1056
+    },
+    {
+      "epoch": 5.106280193236715,
+      "grad_norm": 0.06270143389701843,
+      "learning_rate": 9.673100180437689e-05,
+      "loss": 0.0083,
+      "step": 1057
+    },
+    {
+      "epoch": 5.111111111111111,
+      "grad_norm": 0.06183341518044472,
+      "learning_rate": 9.671899524502291e-05,
+      "loss": 0.006,
+      "step": 1058
+    },
+    {
+      "epoch": 5.115942028985507,
+      "grad_norm": 0.06538157910108566,
+      "learning_rate": 9.670696742481289e-05,
+      "loss": 0.0087,
+      "step": 1059
+    },
+    {
+      "epoch": 5.120772946859903,
+      "grad_norm": 0.06029185280203819,
+      "learning_rate": 9.66949183492204e-05,
+      "loss": 0.0111,
+      "step": 1060
+    },
+    {
+      "epoch": 5.125603864734299,
+      "grad_norm": 0.06412973999977112,
+      "learning_rate": 9.668284802372877e-05,
+      "loss": 0.0089,
+      "step": 1061
+    },
+    {
+      "epoch": 5.130434782608695,
+      "grad_norm": 0.07017388939857483,
+      "learning_rate": 9.667075645383095e-05,
+      "loss": 0.0088,
+      "step": 1062
+    },
+    {
+      "epoch": 5.1352657004830915,
+      "grad_norm": 0.07803435623645782,
+      "learning_rate": 9.665864364502955e-05,
+      "loss": 0.0095,
+      "step": 1063
+    },
+    {
+      "epoch": 5.140096618357488,
+      "grad_norm": 0.05609812214970589,
+      "learning_rate": 9.664650960283688e-05,
+      "loss": 0.0058,
+      "step": 1064
+    },
+    {
+      "epoch": 5.144927536231884,
+      "grad_norm": 0.052468396723270416,
+      "learning_rate": 9.663435433277486e-05,
+      "loss": 0.0087,
+      "step": 1065
+    },
+    {
+      "epoch": 5.14975845410628,
+      "grad_norm": 0.06423533707857132,
+      "learning_rate": 9.662217784037513e-05,
+      "loss": 0.008,
+      "step": 1066
+    },
+    {
+      "epoch": 5.154589371980676,
+      "grad_norm": 0.061567891389131546,
+      "learning_rate": 9.660998013117897e-05,
+      "loss": 0.0126,
+      "step": 1067
+    },
+    {
+      "epoch": 5.159420289855072,
+      "grad_norm": 0.053131893277168274,
+      "learning_rate": 9.659776121073729e-05,
+      "loss": 0.0077,
+      "step": 1068
+    },
+    {
+      "epoch": 5.164251207729468,
+      "grad_norm": 0.07207377254962921,
+      "learning_rate": 9.658552108461068e-05,
+      "loss": 0.0106,
+      "step": 1069
+    },
+    {
+      "epoch": 5.169082125603865,
+      "grad_norm": 0.0715598464012146,
+      "learning_rate": 9.657325975836935e-05,
+      "loss": 0.0113,
+      "step": 1070
+    },
+    {
+      "epoch": 5.173913043478261,
+      "grad_norm": 0.050884295254945755,
+      "learning_rate": 9.656097723759319e-05,
+      "loss": 0.0075,
+      "step": 1071
+    },
+    {
+      "epoch": 5.178743961352657,
+      "grad_norm": 0.049650318920612335,
+      "learning_rate": 9.654867352787174e-05,
+      "loss": 0.0092,
+      "step": 1072
+    },
+    {
+      "epoch": 5.183574879227053,
+      "grad_norm": 0.06503750383853912,
+      "learning_rate": 9.653634863480414e-05,
+      "loss": 0.0107,
+      "step": 1073
+    },
+    {
+      "epoch": 5.188405797101449,
+      "grad_norm": 0.05913107097148895,
+      "learning_rate": 9.652400256399922e-05,
+      "loss": 0.0088,
+      "step": 1074
+    },
+    {
+      "epoch": 5.193236714975845,
+      "grad_norm": 0.05056409910321236,
+      "learning_rate": 9.651163532107541e-05,
+      "loss": 0.0069,
+      "step": 1075
+    },
+    {
+      "epoch": 5.193236714975845,
+      "eval_loss": 0.016926271840929985,
+      "eval_runtime": 20.5906,
+      "eval_samples_per_second": 4.857,
+      "eval_steps_per_second": 0.146,
+      "step": 1075
+    },
+    {
+      "epoch": 5.1980676328502415,
+      "grad_norm": 0.056902553886175156,
+      "learning_rate": 9.649924691166078e-05,
+      "loss": 0.0087,
+      "step": 1076
+    },
+    {
+      "epoch": 5.202898550724638,
+      "grad_norm": 0.09051110595464706,
+      "learning_rate": 9.648683734139305e-05,
+      "loss": 0.0108,
+      "step": 1077
+    },
+    {
+      "epoch": 5.207729468599034,
+      "grad_norm": 0.06023527309298515,
+      "learning_rate": 9.647440661591958e-05,
+      "loss": 0.01,
+      "step": 1078
+    },
+    {
+      "epoch": 5.21256038647343,
+      "grad_norm": 0.10421982407569885,
+      "learning_rate": 9.64619547408973e-05,
+      "loss": 0.0136,
+      "step": 1079
+    },
+    {
+      "epoch": 5.217391304347826,
+      "grad_norm": 0.09369756281375885,
+      "learning_rate": 9.644948172199283e-05,
+      "loss": 0.0095,
+      "step": 1080
+    },
+    {
+      "epoch": 5.222222222222222,
+      "grad_norm": 0.062193822115659714,
+      "learning_rate": 9.643698756488238e-05,
+      "loss": 0.009,
+      "step": 1081
+    },
+    {
+      "epoch": 5.2270531400966185,
+      "grad_norm": 0.04095260426402092,
+      "learning_rate": 9.642447227525178e-05,
+      "loss": 0.0072,
+      "step": 1082
+    },
+    {
+      "epoch": 5.231884057971015,
+      "grad_norm": 0.06756602972745895,
+      "learning_rate": 9.641193585879646e-05,
+      "loss": 0.0097,
+      "step": 1083
+    },
+    {
+      "epoch": 5.236714975845411,
+      "grad_norm": 0.05496843904256821,
+      "learning_rate": 9.639937832122154e-05,
+      "loss": 0.0076,
+      "step": 1084
+    },
+    {
+      "epoch": 5.241545893719807,
+      "grad_norm": 0.055532056838274,
+      "learning_rate": 9.638679966824165e-05,
+      "loss": 0.0076,
+      "step": 1085
+    },
+    {
+      "epoch": 5.246376811594203,
+      "grad_norm": 0.04633563756942749,
+      "learning_rate": 9.637419990558111e-05,
+      "loss": 0.0066,
+      "step": 1086
+    },
+    {
+      "epoch": 5.251207729468599,
+      "grad_norm": 0.06235886737704277,
+      "learning_rate": 9.636157903897379e-05,
+      "loss": 0.0123,
+      "step": 1087
+    },
+    {
+      "epoch": 5.256038647342995,
+      "grad_norm": 0.06465509533882141,
+      "learning_rate": 9.63489370741632e-05,
+      "loss": 0.0127,
+      "step": 1088
+    },
+    {
+      "epoch": 5.260869565217392,
+      "grad_norm": 0.08732569962739944,
+      "learning_rate": 9.633627401690241e-05,
+      "loss": 0.013,
+      "step": 1089
+    },
+    {
+      "epoch": 5.265700483091788,
+      "grad_norm": 0.059913862496614456,
+      "learning_rate": 9.632358987295416e-05,
+      "loss": 0.009,
+      "step": 1090
+    },
+    {
+      "epoch": 5.270531400966184,
+      "grad_norm": 0.05865868553519249,
+      "learning_rate": 9.631088464809072e-05,
+      "loss": 0.008,
+      "step": 1091
+    },
+    {
+      "epoch": 5.27536231884058,
+      "grad_norm": 0.07305967807769775,
+      "learning_rate": 9.629815834809397e-05,
+      "loss": 0.0099,
+      "step": 1092
+    },
+    {
+      "epoch": 5.280193236714976,
+      "grad_norm": 0.060723256319761276,
+      "learning_rate": 9.628541097875541e-05,
+      "loss": 0.0105,
+      "step": 1093
+    },
+    {
+      "epoch": 5.285024154589372,
+      "grad_norm": 0.042938102036714554,
+      "learning_rate": 9.62726425458761e-05,
+      "loss": 0.0086,
+      "step": 1094
+    },
+    {
+      "epoch": 5.2898550724637685,
+      "grad_norm": 0.04853806272149086,
+      "learning_rate": 9.625985305526669e-05,
+      "loss": 0.0085,
+      "step": 1095
+    },
+    {
+      "epoch": 5.294685990338165,
+      "grad_norm": 0.056957464665174484,
+      "learning_rate": 9.62470425127474e-05,
+      "loss": 0.0077,
+      "step": 1096
+    },
+    {
+      "epoch": 5.29951690821256,
+      "grad_norm": 0.05829184129834175,
+      "learning_rate": 9.623421092414806e-05,
+      "loss": 0.0073,
+      "step": 1097
+    },
+    {
+      "epoch": 5.304347826086957,
+      "grad_norm": 0.08711043000221252,
+      "learning_rate": 9.622135829530807e-05,
+      "loss": 0.0073,
+      "step": 1098
+    },
+    {
+      "epoch": 5.309178743961352,
+      "grad_norm": 0.06256822496652603,
+      "learning_rate": 9.620848463207637e-05,
+      "loss": 0.0084,
+      "step": 1099
+    },
+    {
+      "epoch": 5.314009661835748,
+      "grad_norm": 0.04850585386157036,
+      "learning_rate": 9.619558994031152e-05,
+      "loss": 0.0073,
+      "step": 1100
+    },
+    {
+      "epoch": 5.314009661835748,
+      "eval_loss": 0.018239127472043037,
+      "eval_runtime": 20.6235,
+      "eval_samples_per_second": 4.849,
+      "eval_steps_per_second": 0.145,
+      "step": 1100
+    },
+    {
+      "epoch": 5.318840579710145,
+      "grad_norm": 0.05552690103650093,
+      "learning_rate": 9.618267422588164e-05,
+      "loss": 0.0093,
+      "step": 1101
+    },
+    {
+      "epoch": 5.323671497584541,
+      "grad_norm": 0.07265713810920715,
+      "learning_rate": 9.616973749466438e-05,
+      "loss": 0.0089,
+      "step": 1102
+    },
+    {
+      "epoch": 5.328502415458937,
+      "grad_norm": 0.06223799288272858,
+      "learning_rate": 9.615677975254701e-05,
+      "loss": 0.0079,
+      "step": 1103
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.06156628578901291,
+      "learning_rate": 9.614380100542631e-05,
+      "loss": 0.0098,
+      "step": 1104
+    },
+    {
+      "epoch": 5.338164251207729,
+      "grad_norm": 0.06282912194728851,
+      "learning_rate": 9.613080125920866e-05,
+      "loss": 0.0083,
+      "step": 1105
+    },
+    {
+      "epoch": 5.342995169082125,
+      "grad_norm": 0.052787791937589645,
+      "learning_rate": 9.611778051980995e-05,
+      "loss": 0.0089,
+      "step": 1106
+    },
+    {
+      "epoch": 5.3478260869565215,
+      "grad_norm": 0.05192113295197487,
+      "learning_rate": 9.610473879315567e-05,
+      "loss": 0.0076,
+      "step": 1107
+    },
+    {
+      "epoch": 5.352657004830918,
+      "grad_norm": 0.04756897687911987,
+      "learning_rate": 9.609167608518084e-05,
+      "loss": 0.0082,
+      "step": 1108
+    },
+    {
+      "epoch": 5.357487922705314,
+      "grad_norm": 0.060673873871564865,
+      "learning_rate": 9.607859240183003e-05,
+      "loss": 0.0087,
+      "step": 1109
+    },
+    {
+      "epoch": 5.36231884057971,
+      "grad_norm": 0.05874047428369522,
+      "learning_rate": 9.606548774905737e-05,
+      "loss": 0.0099,
+      "step": 1110
+    },
+    {
+      "epoch": 5.367149758454106,
+      "grad_norm": 0.05346541106700897,
+      "learning_rate": 9.605236213282649e-05,
+      "loss": 0.0078,
+      "step": 1111
+    },
+    {
+      "epoch": 5.371980676328502,
+      "grad_norm": 0.042168211191892624,
+      "learning_rate": 9.603921555911061e-05,
+      "loss": 0.0067,
+      "step": 1112
+    },
+    {
+      "epoch": 5.3768115942028984,
+      "grad_norm": 0.0657646432518959,
+      "learning_rate": 9.602604803389246e-05,
+      "loss": 0.0119,
+      "step": 1113
+    },
+    {
+      "epoch": 5.381642512077295,
+      "grad_norm": 0.060943853110075,
+      "learning_rate": 9.601285956316429e-05,
+      "loss": 0.0096,
+      "step": 1114
+    },
+    {
+      "epoch": 5.386473429951691,
+      "grad_norm": 0.07080808281898499,
+      "learning_rate": 9.599965015292794e-05,
+      "loss": 0.0101,
+      "step": 1115
+    },
+    {
+      "epoch": 5.391304347826087,
+      "grad_norm": 0.049159590154886246,
+      "learning_rate": 9.598641980919472e-05,
+      "loss": 0.0077,
+      "step": 1116
+    },
+    {
+      "epoch": 5.396135265700483,
+      "grad_norm": 0.04659260809421539,
+      "learning_rate": 9.59731685379855e-05,
+      "loss": 0.0067,
+      "step": 1117
+    },
+    {
+      "epoch": 5.400966183574879,
+      "grad_norm": 0.06035299971699715,
+      "learning_rate": 9.595989634533065e-05,
+      "loss": 0.0084,
+      "step": 1118
+    },
+    {
+      "epoch": 5.405797101449275,
+      "grad_norm": 0.0531546026468277,
+      "learning_rate": 9.594660323727006e-05,
+      "loss": 0.0092,
+      "step": 1119
+    },
+    {
+      "epoch": 5.4106280193236715,
+      "grad_norm": 0.06887521594762802,
+      "learning_rate": 9.593328921985319e-05,
+      "loss": 0.0108,
+      "step": 1120
+    },
+    {
+      "epoch": 5.415458937198068,
+      "grad_norm": 0.050380952656269073,
+      "learning_rate": 9.591995429913894e-05,
+      "loss": 0.0094,
+      "step": 1121
+    },
+    {
+      "epoch": 5.420289855072464,
+      "grad_norm": 0.06734389066696167,
+      "learning_rate": 9.590659848119577e-05,
+      "loss": 0.0073,
+      "step": 1122
+    },
+    {
+      "epoch": 5.42512077294686,
+      "grad_norm": 0.06262669712305069,
+      "learning_rate": 9.589322177210163e-05,
+      "loss": 0.0117,
+      "step": 1123
+    },
+    {
+      "epoch": 5.429951690821256,
+      "grad_norm": 0.05138638615608215,
+      "learning_rate": 9.587982417794401e-05,
+      "loss": 0.0067,
+      "step": 1124
+    },
+    {
+      "epoch": 5.434782608695652,
+      "grad_norm": 0.05120410770177841,
+      "learning_rate": 9.586640570481986e-05,
+      "loss": 0.0076,
+      "step": 1125
+    },
+    {
+      "epoch": 5.434782608695652,
+      "eval_loss": 0.01660805754363537,
+      "eval_runtime": 21.4666,
+      "eval_samples_per_second": 4.658,
+      "eval_steps_per_second": 0.14,
+      "step": 1125
+    },
+    {
+      "epoch": 5.4396135265700485,
+      "grad_norm": 0.04313327744603157,
+      "learning_rate": 9.585296635883567e-05,
+      "loss": 0.008,
+      "step": 1126
+    },
+    {
+      "epoch": 5.444444444444445,
+      "grad_norm": 0.07868213951587677,
+      "learning_rate": 9.583950614610738e-05,
+      "loss": 0.0076,
+      "step": 1127
+    },
+    {
+      "epoch": 5.449275362318841,
+      "grad_norm": 0.04285581782460213,
+      "learning_rate": 9.582602507276048e-05,
+      "loss": 0.0085,
+      "step": 1128
+    },
+    {
+      "epoch": 5.454106280193237,
+      "grad_norm": 0.062434256076812744,
+      "learning_rate": 9.581252314492995e-05,
+      "loss": 0.0094,
+      "step": 1129
+    },
+    {
+      "epoch": 5.458937198067633,
+      "grad_norm": 0.05328233167529106,
+      "learning_rate": 9.579900036876018e-05,
+      "loss": 0.0078,
+      "step": 1130
+    },
+    {
+      "epoch": 5.463768115942029,
+      "grad_norm": 0.08031158894300461,
+      "learning_rate": 9.578545675040516e-05,
+      "loss": 0.0084,
+      "step": 1131
+    },
+    {
+      "epoch": 5.468599033816425,
+      "grad_norm": 0.06782043725252151,
+      "learning_rate": 9.57718922960283e-05,
+      "loss": 0.0116,
+      "step": 1132
+    },
+    {
+      "epoch": 5.473429951690822,
+      "grad_norm": 0.06157178059220314,
+      "learning_rate": 9.575830701180248e-05,
+      "loss": 0.0107,
+      "step": 1133
+    },
+    {
+      "epoch": 5.478260869565218,
+      "grad_norm": 0.05720106512308121,
+      "learning_rate": 9.574470090391012e-05,
+      "loss": 0.0074,
+      "step": 1134
+    },
+    {
+      "epoch": 5.483091787439614,
+      "grad_norm": 0.06687933206558228,
+      "learning_rate": 9.573107397854307e-05,
+      "loss": 0.0081,
+      "step": 1135
+    },
+    {
+      "epoch": 5.48792270531401,
+      "grad_norm": 0.0551472008228302,
+      "learning_rate": 9.571742624190265e-05,
+      "loss": 0.0103,
+      "step": 1136
+    },
+    {
+      "epoch": 5.492753623188406,
+      "grad_norm": 0.05003214254975319,
+      "learning_rate": 9.570375770019968e-05,
+      "loss": 0.0072,
+      "step": 1137
+    },
+    {
+      "epoch": 5.4975845410628015,
+      "grad_norm": 0.047905758023262024,
+      "learning_rate": 9.569006835965443e-05,
+      "loss": 0.0072,
+      "step": 1138
+    },
+    {
+      "epoch": 5.5024154589371985,
+      "grad_norm": 0.04713456332683563,
+      "learning_rate": 9.567635822649663e-05,
+      "loss": 0.0089,
+      "step": 1139
+    },
+    {
+      "epoch": 5.507246376811594,
+      "grad_norm": 0.04646996036171913,
+      "learning_rate": 9.566262730696548e-05,
+      "loss": 0.007,
+      "step": 1140
+    },
+    {
+      "epoch": 5.512077294685991,
+      "grad_norm": 0.044961195439100266,
+      "learning_rate": 9.564887560730966e-05,
+      "loss": 0.0079,
+      "step": 1141
+    },
+    {
+      "epoch": 5.516908212560386,
+      "grad_norm": 0.051607292145490646,
+      "learning_rate": 9.563510313378725e-05,
+      "loss": 0.0089,
+      "step": 1142
+    },
+    {
+      "epoch": 5.521739130434782,
+      "grad_norm": 0.06743261218070984,
+      "learning_rate": 9.562130989266586e-05,
+      "loss": 0.0122,
+      "step": 1143
+    },
+    {
+      "epoch": 5.526570048309178,
+      "grad_norm": 0.08333702385425568,
+      "learning_rate": 9.560749589022249e-05,
+      "loss": 0.0095,
+      "step": 1144
+    },
+    {
+      "epoch": 5.531400966183575,
+      "grad_norm": 0.08595249801874161,
+      "learning_rate": 9.55936611327436e-05,
+      "loss": 0.01,
+      "step": 1145
+    },
+    {
+      "epoch": 5.536231884057971,
+      "grad_norm": 0.06153739243745804,
+      "learning_rate": 9.557980562652513e-05,
+      "loss": 0.0094,
+      "step": 1146
+    },
+    {
+      "epoch": 5.541062801932367,
+      "grad_norm": 0.08083317428827286,
+      "learning_rate": 9.556592937787241e-05,
+      "loss": 0.011,
+      "step": 1147
+    },
+    {
+      "epoch": 5.545893719806763,
+      "grad_norm": 0.07135039567947388,
+      "learning_rate": 9.555203239310024e-05,
+      "loss": 0.0094,
+      "step": 1148
+    },
+    {
+      "epoch": 5.550724637681159,
+      "grad_norm": 0.06862451136112213,
+      "learning_rate": 9.553811467853288e-05,
+      "loss": 0.009,
+      "step": 1149
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 0.052251044660806656,
+      "learning_rate": 9.552417624050398e-05,
+      "loss": 0.0084,
+      "step": 1150
+    },
+    {
+      "epoch": 5.555555555555555,
+      "eval_loss": 0.01725536584854126,
+      "eval_runtime": 20.6146,
+      "eval_samples_per_second": 4.851,
+      "eval_steps_per_second": 0.146,
+      "step": 1150
+    },
+    {
+      "epoch": 5.5603864734299515,
+      "grad_norm": 0.05517849326133728,
+      "learning_rate": 9.55102170853566e-05,
+      "loss": 0.0097,
+      "step": 1151
+    },
+    {
+      "epoch": 5.565217391304348,
+      "grad_norm": 0.04879816621541977,
+      "learning_rate": 9.549623721944334e-05,
+      "loss": 0.0097,
+      "step": 1152
+    },
+    {
+      "epoch": 5.570048309178744,
+      "grad_norm": 0.05694446712732315,
+      "learning_rate": 9.54822366491261e-05,
+      "loss": 0.0081,
+      "step": 1153
+    },
+    {
+      "epoch": 5.57487922705314,
+      "grad_norm": 0.04808945581316948,
+      "learning_rate": 9.546821538077626e-05,
+      "loss": 0.0085,
+      "step": 1154
+    },
+    {
+      "epoch": 5.579710144927536,
+      "grad_norm": 0.05540106073021889,
+      "learning_rate": 9.54541734207746e-05,
+      "loss": 0.0102,
+      "step": 1155
+    },
+    {
+      "epoch": 5.584541062801932,
+      "grad_norm": 0.052780650556087494,
+      "learning_rate": 9.544011077551136e-05,
+      "loss": 0.0088,
+      "step": 1156
+    },
+    {
+      "epoch": 5.5893719806763285,
+      "grad_norm": 0.06642550975084305,
+      "learning_rate": 9.542602745138614e-05,
+      "loss": 0.0087,
+      "step": 1157
+    },
+    {
+      "epoch": 5.594202898550725,
+      "grad_norm": 0.06861164420843124,
+      "learning_rate": 9.541192345480795e-05,
+      "loss": 0.0083,
+      "step": 1158
+    },
+    {
+      "epoch": 5.599033816425121,
+      "grad_norm": 0.05863761156797409,
+      "learning_rate": 9.539779879219528e-05,
+      "loss": 0.0101,
+      "step": 1159
+    },
+    {
+      "epoch": 5.603864734299517,
+      "grad_norm": 0.04983343183994293,
+      "learning_rate": 9.538365346997594e-05,
+      "loss": 0.0077,
+      "step": 1160
+    },
+    {
+      "epoch": 5.608695652173913,
+      "grad_norm": 0.06960771232843399,
+      "learning_rate": 9.536948749458718e-05,
+      "loss": 0.0156,
+      "step": 1161
+    },
+    {
+      "epoch": 5.613526570048309,
+      "grad_norm": 0.05016974359750748,
+      "learning_rate": 9.535530087247567e-05,
+      "loss": 0.0089,
+      "step": 1162
+    },
+    {
+      "epoch": 5.618357487922705,
+      "grad_norm": 0.04743446037173271,
+      "learning_rate": 9.53410936100974e-05,
+      "loss": 0.0062,
+      "step": 1163
+    },
+    {
+      "epoch": 5.6231884057971016,
+      "grad_norm": 0.05176220461726189,
+      "learning_rate": 9.532686571391784e-05,
+      "loss": 0.0075,
+      "step": 1164
+    },
+    {
+      "epoch": 5.628019323671498,
+      "grad_norm": 0.05748028680682182,
+      "learning_rate": 9.53126171904118e-05,
+      "loss": 0.0099,
+      "step": 1165
+    },
+    {
+      "epoch": 5.632850241545894,
+      "grad_norm": 0.05922771245241165,
+      "learning_rate": 9.529834804606351e-05,
+      "loss": 0.009,
+      "step": 1166
+    },
+    {
+      "epoch": 5.63768115942029,
+      "grad_norm": 0.0787665918469429,
+      "learning_rate": 9.528405828736656e-05,
+      "loss": 0.015,
+      "step": 1167
+    },
+    {
+      "epoch": 5.642512077294686,
+      "grad_norm": 0.05979447066783905,
+      "learning_rate": 9.52697479208239e-05,
+      "loss": 0.0092,
+      "step": 1168
+    },
+    {
+      "epoch": 5.647342995169082,
+      "grad_norm": 0.04697132483124733,
+      "learning_rate": 9.525541695294791e-05,
+      "loss": 0.008,
+      "step": 1169
+    },
+    {
+      "epoch": 5.6521739130434785,
+      "grad_norm": 0.06020793691277504,
+      "learning_rate": 9.52410653902603e-05,
+      "loss": 0.0133,
+      "step": 1170
+    },
+    {
+      "epoch": 5.657004830917875,
+      "grad_norm": 0.0417938232421875,
+      "learning_rate": 9.52266932392922e-05,
+      "loss": 0.0073,
+      "step": 1171
+    },
+    {
+      "epoch": 5.661835748792271,
+      "grad_norm": 0.041864678263664246,
+      "learning_rate": 9.521230050658405e-05,
+      "loss": 0.0061,
+      "step": 1172
+    },
+    {
+      "epoch": 5.666666666666667,
+      "grad_norm": 0.05600718781352043,
+      "learning_rate": 9.51978871986857e-05,
+      "loss": 0.0104,
+      "step": 1173
+    },
+    {
+      "epoch": 5.671497584541063,
+      "grad_norm": 0.09407051652669907,
+      "learning_rate": 9.518345332215637e-05,
+      "loss": 0.0089,
+      "step": 1174
+    },
+    {
+      "epoch": 5.676328502415459,
+      "grad_norm": 0.07216166704893112,
+      "learning_rate": 9.516899888356459e-05,
+      "loss": 0.0091,
+      "step": 1175
+    },
+    {
+      "epoch": 5.676328502415459,
+      "eval_loss": 0.017482729628682137,
+      "eval_runtime": 20.5935,
+      "eval_samples_per_second": 4.856,
+      "eval_steps_per_second": 0.146,
+      "step": 1175
+    },
+    {
+      "epoch": 5.681159420289855,
+      "grad_norm": 0.0422208234667778,
+      "learning_rate": 9.515452388948829e-05,
+      "loss": 0.0066,
+      "step": 1176
+    },
+    {
+      "epoch": 5.685990338164252,
+      "grad_norm": 0.06615504622459412,
+      "learning_rate": 9.514002834651474e-05,
+      "loss": 0.0099,
+      "step": 1177
+    },
+    {
+      "epoch": 5.690821256038648,
+      "grad_norm": 0.05418713763356209,
+      "learning_rate": 9.512551226124058e-05,
+      "loss": 0.0089,
+      "step": 1178
+    },
+    {
+      "epoch": 5.695652173913043,
+      "grad_norm": 0.060746192932128906,
+      "learning_rate": 9.511097564027175e-05,
+      "loss": 0.0132,
+      "step": 1179
+    },
+    {
+      "epoch": 5.70048309178744,
+      "grad_norm": 0.05127759650349617,
+      "learning_rate": 9.509641849022362e-05,
+      "loss": 0.0097,
+      "step": 1180
+    },
+    {
+      "epoch": 5.705314009661835,
+      "grad_norm": 0.05560661479830742,
+      "learning_rate": 9.508184081772081e-05,
+      "loss": 0.0098,
+      "step": 1181
+    },
+    {
+      "epoch": 5.710144927536232,
+      "grad_norm": 0.062068238854408264,
+      "learning_rate": 9.506724262939732e-05,
+      "loss": 0.0075,
+      "step": 1182
+    },
+    {
+      "epoch": 5.714975845410628,
+      "grad_norm": 0.061363585293293,
+      "learning_rate": 9.505262393189649e-05,
+      "loss": 0.0071,
+      "step": 1183
+    },
+    {
+      "epoch": 5.719806763285024,
+      "grad_norm": 0.05139882117509842,
+      "learning_rate": 9.5037984731871e-05,
+      "loss": 0.009,
+      "step": 1184
+    },
+    {
+      "epoch": 5.72463768115942,
+      "grad_norm": 0.05671796575188637,
+      "learning_rate": 9.502332503598283e-05,
+      "loss": 0.0091,
+      "step": 1185
+    },
+    {
+      "epoch": 5.729468599033816,
+      "grad_norm": 0.0627102181315422,
+      "learning_rate": 9.500864485090334e-05,
+      "loss": 0.0094,
+      "step": 1186
+    },
+    {
+      "epoch": 5.734299516908212,
+      "grad_norm": 0.0828758180141449,
+      "learning_rate": 9.499394418331315e-05,
+      "loss": 0.0132,
+      "step": 1187
+    },
+    {
+      "epoch": 5.739130434782608,
+      "grad_norm": 0.06312569975852966,
+      "learning_rate": 9.497922303990225e-05,
+      "loss": 0.0086,
+      "step": 1188
+    },
+    {
+      "epoch": 5.743961352657005,
+      "grad_norm": 0.04896373301744461,
+      "learning_rate": 9.496448142736991e-05,
+      "loss": 0.0064,
+      "step": 1189
+    },
+    {
+      "epoch": 5.748792270531401,
+      "grad_norm": 0.10162996500730515,
+      "learning_rate": 9.494971935242477e-05,
+      "loss": 0.0175,
+      "step": 1190
+    },
+    {
+      "epoch": 5.753623188405797,
+      "grad_norm": 0.055514682084321976,
+      "learning_rate": 9.493493682178472e-05,
+      "loss": 0.0101,
+      "step": 1191
+    },
+    {
+      "epoch": 5.758454106280193,
+      "grad_norm": 0.05568890646100044,
+      "learning_rate": 9.492013384217699e-05,
+      "loss": 0.0096,
+      "step": 1192
+    },
+    {
+      "epoch": 5.763285024154589,
+      "grad_norm": 0.05123963579535484,
+      "learning_rate": 9.49053104203381e-05,
+      "loss": 0.0106,
+      "step": 1193
+    },
+    {
+      "epoch": 5.768115942028985,
+      "grad_norm": 0.05242984741926193,
+      "learning_rate": 9.489046656301393e-05,
+      "loss": 0.0101,
+      "step": 1194
+    },
+    {
+      "epoch": 5.7729468599033815,
+      "grad_norm": 0.052701305598020554,
+      "learning_rate": 9.487560227695955e-05,
+      "loss": 0.0131,
+      "step": 1195
+    },
+    {
+      "epoch": 5.777777777777778,
+      "grad_norm": 0.0495404452085495,
+      "learning_rate": 9.486071756893944e-05,
+      "loss": 0.0075,
+      "step": 1196
+    },
+    {
+      "epoch": 5.782608695652174,
+      "grad_norm": 0.0440664142370224,
+      "learning_rate": 9.484581244572732e-05,
+      "loss": 0.0081,
+      "step": 1197
+    },
+    {
+      "epoch": 5.78743961352657,
+      "grad_norm": 0.04695248231291771,
+      "learning_rate": 9.48308869141062e-05,
+      "loss": 0.0084,
+      "step": 1198
+    },
+    {
+      "epoch": 5.792270531400966,
+      "grad_norm": 0.059293605387210846,
+      "learning_rate": 9.481594098086837e-05,
+      "loss": 0.011,
+      "step": 1199
+    },
+    {
+      "epoch": 5.797101449275362,
+      "grad_norm": 0.06034954637289047,
+      "learning_rate": 9.480097465281543e-05,
+      "loss": 0.0081,
+      "step": 1200
+    },
+    {
+      "epoch": 5.797101449275362,
+      "eval_loss": 0.01758287288248539,
+      "eval_runtime": 20.7195,
+      "eval_samples_per_second": 4.826,
+      "eval_steps_per_second": 0.145,
+      "step": 1200
+    },
+    {
+      "epoch": 5.8019323671497585,
+      "grad_norm": 0.05583087354898453,
+      "learning_rate": 9.478598793675825e-05,
+      "loss": 0.0082,
+      "step": 1201
+    },
+    {
+      "epoch": 5.806763285024155,
+      "grad_norm": 0.05717151612043381,
+      "learning_rate": 9.477098083951696e-05,
+      "loss": 0.0096,
+      "step": 1202
+    },
+    {
+      "epoch": 5.811594202898551,
+      "grad_norm": 0.06151145324110985,
+      "learning_rate": 9.475595336792102e-05,
+      "loss": 0.0072,
+      "step": 1203
+    },
+    {
+      "epoch": 5.816425120772947,
+      "grad_norm": 0.06331399083137512,
+      "learning_rate": 9.474090552880907e-05,
+      "loss": 0.0111,
+      "step": 1204
+    },
+    {
+      "epoch": 5.821256038647343,
+      "grad_norm": 0.05892903357744217,
+      "learning_rate": 9.472583732902914e-05,
+      "loss": 0.0086,
+      "step": 1205
+    },
+    {
+      "epoch": 5.826086956521739,
+      "grad_norm": 0.05928612872958183,
+      "learning_rate": 9.47107487754384e-05,
+      "loss": 0.0088,
+      "step": 1206
+    },
+    {
+      "epoch": 5.830917874396135,
+      "grad_norm": 0.06218522787094116,
+      "learning_rate": 9.469563987490335e-05,
+      "loss": 0.0072,
+      "step": 1207
+    },
+    {
+      "epoch": 5.835748792270532,
+      "grad_norm": 0.07384292781352997,
+      "learning_rate": 9.468051063429977e-05,
+      "loss": 0.0093,
+      "step": 1208
+    },
+    {
+      "epoch": 5.840579710144928,
+      "grad_norm": 0.05546915531158447,
+      "learning_rate": 9.466536106051265e-05,
+      "loss": 0.01,
+      "step": 1209
+    },
+    {
+      "epoch": 5.845410628019324,
+      "grad_norm": 0.05841489136219025,
+      "learning_rate": 9.465019116043625e-05,
+      "loss": 0.01,
+      "step": 1210
+    },
+    {
+      "epoch": 5.85024154589372,
+      "grad_norm": 0.049476198852062225,
+      "learning_rate": 9.463500094097406e-05,
+      "loss": 0.0094,
+      "step": 1211
+    },
+    {
+      "epoch": 5.855072463768116,
+      "grad_norm": 0.04447973519563675,
+      "learning_rate": 9.461979040903888e-05,
+      "loss": 0.0067,
+      "step": 1212
+    },
+    {
+      "epoch": 5.859903381642512,
+      "grad_norm": 0.048743024468421936,
+      "learning_rate": 9.460455957155267e-05,
+      "loss": 0.0075,
+      "step": 1213
+    },
+    {
+      "epoch": 5.8647342995169085,
+      "grad_norm": 0.05664145573973656,
+      "learning_rate": 9.458930843544671e-05,
+      "loss": 0.0104,
+      "step": 1214
+    },
+    {
+      "epoch": 5.869565217391305,
+      "grad_norm": 0.06544919312000275,
+      "learning_rate": 9.457403700766144e-05,
+      "loss": 0.0119,
+      "step": 1215
+    },
+    {
+      "epoch": 5.874396135265701,
+      "grad_norm": 0.06714371591806412,
+      "learning_rate": 9.455874529514661e-05,
+      "loss": 0.0121,
+      "step": 1216
+    },
+    {
+      "epoch": 5.879227053140097,
+      "grad_norm": 0.04535643011331558,
+      "learning_rate": 9.454343330486113e-05,
+      "loss": 0.0083,
+      "step": 1217
+    },
+    {
+      "epoch": 5.884057971014493,
+      "grad_norm": 0.05146961286664009,
+      "learning_rate": 9.45281010437732e-05,
+      "loss": 0.0084,
+      "step": 1218
+    },
+    {
+      "epoch": 5.888888888888889,
+      "grad_norm": 0.0556597076356411,
+      "learning_rate": 9.451274851886021e-05,
+      "loss": 0.0098,
+      "step": 1219
+    },
+    {
+      "epoch": 5.8937198067632846,
+      "grad_norm": 0.053466394543647766,
+      "learning_rate": 9.449737573710876e-05,
+      "loss": 0.0079,
+      "step": 1220
+    },
+    {
+      "epoch": 5.898550724637682,
+      "grad_norm": 0.06228194385766983,
+      "learning_rate": 9.448198270551472e-05,
+      "loss": 0.0085,
+      "step": 1221
+    },
+    {
+      "epoch": 5.903381642512077,
+      "grad_norm": 0.05365237966179848,
+      "learning_rate": 9.446656943108312e-05,
+      "loss": 0.0079,
+      "step": 1222
+    },
+    {
+      "epoch": 5.908212560386474,
+      "grad_norm": 0.04714881628751755,
+      "learning_rate": 9.445113592082823e-05,
+      "loss": 0.0069,
+      "step": 1223
+    },
+    {
+      "epoch": 5.913043478260869,
+      "grad_norm": 0.047249406576156616,
+      "learning_rate": 9.443568218177353e-05,
+      "loss": 0.0076,
+      "step": 1224
+    },
+    {
+      "epoch": 5.917874396135265,
+      "grad_norm": 0.06869522482156754,
+      "learning_rate": 9.442020822095168e-05,
+      "loss": 0.0071,
+      "step": 1225
+    },
+    {
+      "epoch": 5.917874396135265,
+      "eval_loss": 0.01747715473175049,
+      "eval_runtime": 21.3272,
+      "eval_samples_per_second": 4.689,
+      "eval_steps_per_second": 0.141,
+      "step": 1225
+    },
+    {
+      "epoch": 5.9227053140096615,
+      "grad_norm": 0.04264363646507263,
+      "learning_rate": 9.440471404540459e-05,
+      "loss": 0.0063,
+      "step": 1226
+    },
+    {
+      "epoch": 5.927536231884058,
+      "grad_norm": 0.06179254874587059,
+      "learning_rate": 9.438919966218332e-05,
+      "loss": 0.0068,
+      "step": 1227
+    },
+    {
+      "epoch": 5.932367149758454,
+      "grad_norm": 0.05508408322930336,
+      "learning_rate": 9.437366507834816e-05,
+      "loss": 0.0082,
+      "step": 1228
+    },
+    {
+      "epoch": 5.93719806763285,
+      "grad_norm": 0.06304654479026794,
+      "learning_rate": 9.435811030096857e-05,
+      "loss": 0.0105,
+      "step": 1229
+    },
+    {
+      "epoch": 5.942028985507246,
+      "grad_norm": 0.06207441911101341,
+      "learning_rate": 9.434253533712322e-05,
+      "loss": 0.0105,
+      "step": 1230
+    },
+    {
+      "epoch": 5.946859903381642,
+      "grad_norm": 0.06103399395942688,
+      "learning_rate": 9.432694019389997e-05,
+      "loss": 0.0084,
+      "step": 1231
+    },
+    {
+      "epoch": 5.951690821256038,
+      "grad_norm": 0.05756254121661186,
+      "learning_rate": 9.43113248783958e-05,
+      "loss": 0.0097,
+      "step": 1232
+    },
+    {
+      "epoch": 5.956521739130435,
+      "grad_norm": 0.055301081389188766,
+      "learning_rate": 9.429568939771698e-05,
+      "loss": 0.0081,
+      "step": 1233
+    },
+    {
+      "epoch": 5.961352657004831,
+      "grad_norm": 0.05765131488442421,
+      "learning_rate": 9.428003375897885e-05,
+      "loss": 0.0096,
+      "step": 1234
+    },
+    {
+      "epoch": 5.966183574879227,
+      "grad_norm": 0.0458630807697773,
+      "learning_rate": 9.426435796930599e-05,
+      "loss": 0.0089,
+      "step": 1235
+    },
+    {
+      "epoch": 5.971014492753623,
+      "grad_norm": 0.053680840879678726,
+      "learning_rate": 9.424866203583213e-05,
+      "loss": 0.0096,
+      "step": 1236
+    },
+    {
+      "epoch": 5.975845410628019,
+      "grad_norm": 0.04151618480682373,
+      "learning_rate": 9.423294596570016e-05,
+      "loss": 0.0088,
+      "step": 1237
+    },
+    {
+      "epoch": 5.980676328502415,
+      "grad_norm": 0.04098724573850632,
+      "learning_rate": 9.421720976606217e-05,
+      "loss": 0.0079,
+      "step": 1238
+    },
+    {
+      "epoch": 5.9855072463768115,
+      "grad_norm": 0.03668149933218956,
+      "learning_rate": 9.420145344407932e-05,
+      "loss": 0.0072,
+      "step": 1239
+    },
+    {
+      "epoch": 5.990338164251208,
+      "grad_norm": 0.044391755014657974,
+      "learning_rate": 9.418567700692205e-05,
+      "loss": 0.0105,
+      "step": 1240
+    },
+    {
+      "epoch": 5.995169082125604,
+      "grad_norm": 0.05425712838768959,
+      "learning_rate": 9.416988046176987e-05,
+      "loss": 0.0116,
+      "step": 1241
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.09103017300367355,
+      "learning_rate": 9.415406381581146e-05,
+      "loss": 0.006,
+      "step": 1242
+    },
+    {
+      "epoch": 6.004830917874396,
+      "grad_norm": 0.043098460882902145,
+      "learning_rate": 9.413822707624465e-05,
+      "loss": 0.006,
+      "step": 1243
+    },
+    {
+      "epoch": 6.009661835748792,
+      "grad_norm": 0.041810572147369385,
+      "learning_rate": 9.412237025027642e-05,
+      "loss": 0.0067,
+      "step": 1244
+    },
+    {
+      "epoch": 6.0144927536231885,
+      "grad_norm": 0.03212606534361839,
+      "learning_rate": 9.410649334512289e-05,
+      "loss": 0.0046,
+      "step": 1245
+    },
+    {
+      "epoch": 6.019323671497585,
+      "grad_norm": 0.04054751247167587,
+      "learning_rate": 9.40905963680093e-05,
+      "loss": 0.0068,
+      "step": 1246
+    },
+    {
+      "epoch": 6.024154589371981,
+      "grad_norm": 0.0570932999253273,
+      "learning_rate": 9.407467932617006e-05,
+      "loss": 0.006,
+      "step": 1247
+    },
+    {
+      "epoch": 6.028985507246377,
+      "grad_norm": 0.03890130668878555,
+      "learning_rate": 9.405874222684867e-05,
+      "loss": 0.0053,
+      "step": 1248
+    },
+    {
+      "epoch": 6.033816425120773,
+      "grad_norm": 0.07312671095132828,
+      "learning_rate": 9.404278507729777e-05,
+      "loss": 0.0088,
+      "step": 1249
+    },
+    {
+      "epoch": 6.038647342995169,
+      "grad_norm": 0.047085270285606384,
+      "learning_rate": 9.402680788477916e-05,
+      "loss": 0.0058,
+      "step": 1250
+    },
+    {
+      "epoch": 6.038647342995169,
+      "eval_loss": 0.018731381744146347,
+      "eval_runtime": 20.6014,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 1250
+    },
+    {
+      "epoch": 6.043478260869565,
+      "grad_norm": 0.04745680093765259,
+      "learning_rate": 9.401081065656372e-05,
+      "loss": 0.0056,
+      "step": 1251
+    },
+    {
+      "epoch": 6.048309178743962,
+      "grad_norm": 0.05071539804339409,
+      "learning_rate": 9.399479339993144e-05,
+      "loss": 0.0091,
+      "step": 1252
+    },
+    {
+      "epoch": 6.053140096618358,
+      "grad_norm": 0.05788218230009079,
+      "learning_rate": 9.397875612217148e-05,
+      "loss": 0.0053,
+      "step": 1253
+    },
+    {
+      "epoch": 6.057971014492754,
+      "grad_norm": 0.04517432302236557,
+      "learning_rate": 9.396269883058206e-05,
+      "loss": 0.0055,
+      "step": 1254
+    },
+    {
+      "epoch": 6.06280193236715,
+      "grad_norm": 0.04658951237797737,
+      "learning_rate": 9.394662153247051e-05,
+      "loss": 0.0064,
+      "step": 1255
+    },
+    {
+      "epoch": 6.067632850241546,
+      "grad_norm": 0.06327016651630402,
+      "learning_rate": 9.393052423515328e-05,
+      "loss": 0.0072,
+      "step": 1256
+    },
+    {
+      "epoch": 6.072463768115942,
+      "grad_norm": 0.04202742502093315,
+      "learning_rate": 9.391440694595595e-05,
+      "loss": 0.0054,
+      "step": 1257
+    },
+    {
+      "epoch": 6.0772946859903385,
+      "grad_norm": 0.04662294313311577,
+      "learning_rate": 9.389826967221314e-05,
+      "loss": 0.0049,
+      "step": 1258
+    },
+    {
+      "epoch": 6.082125603864735,
+      "grad_norm": 0.06586906313896179,
+      "learning_rate": 9.38821124212686e-05,
+      "loss": 0.0054,
+      "step": 1259
+    },
+    {
+      "epoch": 6.086956521739131,
+      "grad_norm": 0.03706370294094086,
+      "learning_rate": 9.386593520047516e-05,
+      "loss": 0.0041,
+      "step": 1260
+    },
+    {
+      "epoch": 6.091787439613527,
+      "grad_norm": 0.03640469163656235,
+      "learning_rate": 9.384973801719475e-05,
+      "loss": 0.0053,
+      "step": 1261
+    },
+    {
+      "epoch": 6.096618357487923,
+      "grad_norm": 0.05960666388273239,
+      "learning_rate": 9.383352087879838e-05,
+      "loss": 0.0062,
+      "step": 1262
+    },
+    {
+      "epoch": 6.101449275362318,
+      "grad_norm": 0.08039825409650803,
+      "learning_rate": 9.38172837926661e-05,
+      "loss": 0.0122,
+      "step": 1263
+    },
+    {
+      "epoch": 6.106280193236715,
+      "grad_norm": 0.061855610460042953,
+      "learning_rate": 9.380102676618712e-05,
+      "loss": 0.0072,
+      "step": 1264
+    },
+    {
+      "epoch": 6.111111111111111,
+      "grad_norm": 0.0521203950047493,
+      "learning_rate": 9.378474980675966e-05,
+      "loss": 0.0068,
+      "step": 1265
+    },
+    {
+      "epoch": 6.115942028985507,
+      "grad_norm": 0.056785743683576584,
+      "learning_rate": 9.376845292179103e-05,
+      "loss": 0.0065,
+      "step": 1266
+    },
+    {
+      "epoch": 6.120772946859903,
+      "grad_norm": 0.050204869359731674,
+      "learning_rate": 9.37521361186976e-05,
+      "loss": 0.0045,
+      "step": 1267
+    },
+    {
+      "epoch": 6.125603864734299,
+      "grad_norm": 0.06700023263692856,
+      "learning_rate": 9.373579940490484e-05,
+      "loss": 0.0095,
+      "step": 1268
+    },
+    {
+      "epoch": 6.130434782608695,
+      "grad_norm": 0.043674830347299576,
+      "learning_rate": 9.371944278784723e-05,
+      "loss": 0.0054,
+      "step": 1269
+    },
+    {
+      "epoch": 6.1352657004830915,
+      "grad_norm": 0.05406811088323593,
+      "learning_rate": 9.370306627496832e-05,
+      "loss": 0.0079,
+      "step": 1270
+    },
+    {
+      "epoch": 6.140096618357488,
+      "grad_norm": 0.04308624565601349,
+      "learning_rate": 9.368666987372074e-05,
+      "loss": 0.0057,
+      "step": 1271
+    },
+    {
+      "epoch": 6.144927536231884,
+      "grad_norm": 0.03334224224090576,
+      "learning_rate": 9.367025359156618e-05,
+      "loss": 0.0051,
+      "step": 1272
+    },
+    {
+      "epoch": 6.14975845410628,
+      "grad_norm": 0.0489167757332325,
+      "learning_rate": 9.36538174359753e-05,
+      "loss": 0.005,
+      "step": 1273
+    },
+    {
+      "epoch": 6.154589371980676,
+      "grad_norm": 0.06193888559937477,
+      "learning_rate": 9.363736141442791e-05,
+      "loss": 0.0074,
+      "step": 1274
+    },
+    {
+      "epoch": 6.159420289855072,
+      "grad_norm": 0.05088794603943825,
+      "learning_rate": 9.362088553441277e-05,
+      "loss": 0.0081,
+      "step": 1275
+    },
+    {
+      "epoch": 6.159420289855072,
+      "eval_loss": 0.016518954187631607,
+      "eval_runtime": 20.607,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
+      "step": 1275
+    },
+    {
+      "epoch": 6.164251207729468,
+      "grad_norm": 0.04955621808767319,
+      "learning_rate": 9.360438980342774e-05,
+      "loss": 0.006,
+      "step": 1276
+    },
+    {
+      "epoch": 6.169082125603865,
+      "grad_norm": 0.0462108850479126,
+      "learning_rate": 9.358787422897968e-05,
+      "loss": 0.0055,
+      "step": 1277
+    },
+    {
+      "epoch": 6.173913043478261,
+      "grad_norm": 0.05635419487953186,
+      "learning_rate": 9.357133881858448e-05,
+      "loss": 0.0064,
+      "step": 1278
+    },
+    {
+      "epoch": 6.178743961352657,
+      "grad_norm": 0.046862367540597916,
+      "learning_rate": 9.355478357976706e-05,
+      "loss": 0.0059,
+      "step": 1279
+    },
+    {
+      "epoch": 6.183574879227053,
+      "grad_norm": 0.04464318975806236,
+      "learning_rate": 9.35382085200614e-05,
+      "loss": 0.0053,
+      "step": 1280
+    },
+    {
+      "epoch": 6.188405797101449,
+      "grad_norm": 0.04266749694943428,
+      "learning_rate": 9.352161364701044e-05,
+      "loss": 0.0057,
+      "step": 1281
+    },
+    {
+      "epoch": 6.193236714975845,
+      "grad_norm": 0.04351837933063507,
+      "learning_rate": 9.35049989681662e-05,
+      "loss": 0.0065,
+      "step": 1282
+    },
+    {
+      "epoch": 6.1980676328502415,
+      "grad_norm": 0.03657788783311844,
+      "learning_rate": 9.348836449108962e-05,
+      "loss": 0.0043,
+      "step": 1283
+    },
+    {
+      "epoch": 6.202898550724638,
+      "grad_norm": 0.04788932949304581,
+      "learning_rate": 9.347171022335077e-05,
+      "loss": 0.0051,
+      "step": 1284
+    },
+    {
+      "epoch": 6.207729468599034,
+      "grad_norm": 0.06068427860736847,
+      "learning_rate": 9.345503617252864e-05,
+      "loss": 0.0061,
+      "step": 1285
+    },
+    {
+      "epoch": 6.21256038647343,
+      "grad_norm": 0.05637691169977188,
+      "learning_rate": 9.343834234621124e-05,
+      "loss": 0.0072,
+      "step": 1286
+    },
+    {
+      "epoch": 6.217391304347826,
+      "grad_norm": 0.043992042541503906,
+      "learning_rate": 9.342162875199562e-05,
+      "loss": 0.0057,
+      "step": 1287
+    },
+    {
+      "epoch": 6.222222222222222,
+      "grad_norm": 0.04882000759243965,
+      "learning_rate": 9.340489539748775e-05,
+      "loss": 0.0069,
+      "step": 1288
+    },
+    {
+      "epoch": 6.2270531400966185,
+      "grad_norm": 0.047943197190761566,
+      "learning_rate": 9.338814229030268e-05,
+      "loss": 0.0054,
+      "step": 1289
+    },
+    {
+      "epoch": 6.231884057971015,
+      "grad_norm": 0.05657848343253136,
+      "learning_rate": 9.337136943806437e-05,
+      "loss": 0.0072,
+      "step": 1290
+    },
+    {
+      "epoch": 6.236714975845411,
+      "grad_norm": 0.049666404724121094,
+      "learning_rate": 9.335457684840581e-05,
+      "loss": 0.0058,
+      "step": 1291
+    },
+    {
+      "epoch": 6.241545893719807,
+      "grad_norm": 0.0505865253508091,
+      "learning_rate": 9.333776452896897e-05,
+      "loss": 0.0064,
+      "step": 1292
+    },
+    {
+      "epoch": 6.246376811594203,
+      "grad_norm": 0.07791588455438614,
+      "learning_rate": 9.332093248740479e-05,
+      "loss": 0.0058,
+      "step": 1293
+    },
+    {
+      "epoch": 6.251207729468599,
+      "grad_norm": 0.03757460042834282,
+      "learning_rate": 9.330408073137319e-05,
+      "loss": 0.0039,
+      "step": 1294
+    },
+    {
+      "epoch": 6.256038647342995,
+      "grad_norm": 0.05717682093381882,
+      "learning_rate": 9.328720926854305e-05,
+      "loss": 0.0052,
+      "step": 1295
+    },
+    {
+      "epoch": 6.260869565217392,
+      "grad_norm": 0.05694163963198662,
+      "learning_rate": 9.327031810659224e-05,
+      "loss": 0.008,
+      "step": 1296
+    },
+    {
+      "epoch": 6.265700483091788,
+      "grad_norm": 0.06574101746082306,
+      "learning_rate": 9.325340725320755e-05,
+      "loss": 0.0061,
+      "step": 1297
+    },
+    {
+      "epoch": 6.270531400966184,
+      "grad_norm": 0.03355651721358299,
+      "learning_rate": 9.323647671608479e-05,
+      "loss": 0.0041,
+      "step": 1298
+    },
+    {
+      "epoch": 6.27536231884058,
+      "grad_norm": 0.047428570687770844,
+      "learning_rate": 9.321952650292871e-05,
+      "loss": 0.0058,
+      "step": 1299
+    },
+    {
+      "epoch": 6.280193236714976,
+      "grad_norm": 0.046640850603580475,
+      "learning_rate": 9.320255662145298e-05,
+      "loss": 0.0057,
+      "step": 1300
+    },
+    {
+      "epoch": 6.280193236714976,
+      "eval_loss": 0.017070595175027847,
+      "eval_runtime": 20.6176,
+      "eval_samples_per_second": 4.85,
+      "eval_steps_per_second": 0.146,
+      "step": 1300
+    },
+    {
+      "epoch": 6.285024154589372,
+      "grad_norm": 0.06039797514677048,
+      "learning_rate": 9.318556707938025e-05,
+      "loss": 0.0087,
+      "step": 1301
+    },
+    {
+      "epoch": 6.2898550724637685,
+      "grad_norm": 0.04693787172436714,
+      "learning_rate": 9.316855788444212e-05,
+      "loss": 0.0063,
+      "step": 1302
+    },
+    {
+      "epoch": 6.294685990338165,
+      "grad_norm": 0.048959773033857346,
+      "learning_rate": 9.315152904437911e-05,
+      "loss": 0.0067,
+      "step": 1303
+    },
+    {
+      "epoch": 6.29951690821256,
+      "grad_norm": 0.09099757671356201,
+      "learning_rate": 9.313448056694071e-05,
+      "loss": 0.0063,
+      "step": 1304
+    },
+    {
+      "epoch": 6.304347826086957,
+      "grad_norm": 0.04881768301129341,
+      "learning_rate": 9.311741245988535e-05,
+      "loss": 0.007,
+      "step": 1305
+    },
+    {
+      "epoch": 6.309178743961352,
+      "grad_norm": 0.044358160346746445,
+      "learning_rate": 9.310032473098033e-05,
+      "loss": 0.0058,
+      "step": 1306
+    },
+    {
+      "epoch": 6.314009661835748,
+      "grad_norm": 0.04900408908724785,
+      "learning_rate": 9.308321738800196e-05,
+      "loss": 0.0049,
+      "step": 1307
+    },
+    {
+      "epoch": 6.318840579710145,
+      "grad_norm": 0.03707882761955261,
+      "learning_rate": 9.306609043873542e-05,
+      "loss": 0.0047,
+      "step": 1308
+    },
+    {
+      "epoch": 6.323671497584541,
+      "grad_norm": 0.0487239807844162,
+      "learning_rate": 9.304894389097486e-05,
+      "loss": 0.0071,
+      "step": 1309
+    },
+    {
+      "epoch": 6.328502415458937,
+      "grad_norm": 0.06111222505569458,
+      "learning_rate": 9.303177775252327e-05,
+      "loss": 0.0087,
+      "step": 1310
+    },
+    {
+      "epoch": 6.333333333333333,
+      "grad_norm": 0.03556101396679878,
+      "learning_rate": 9.301459203119267e-05,
+      "loss": 0.0052,
+      "step": 1311
+    },
+    {
+      "epoch": 6.338164251207729,
+      "grad_norm": 0.06353302299976349,
+      "learning_rate": 9.299738673480391e-05,
+      "loss": 0.0063,
+      "step": 1312
+    },
+    {
+      "epoch": 6.342995169082125,
+      "grad_norm": 0.09204978495836258,
+      "learning_rate": 9.298016187118675e-05,
+      "loss": 0.0081,
+      "step": 1313
+    },
+    {
+      "epoch": 6.3478260869565215,
+      "grad_norm": 0.0807780846953392,
+      "learning_rate": 9.296291744817987e-05,
+      "loss": 0.0065,
+      "step": 1314
+    },
+    {
+      "epoch": 6.352657004830918,
+      "grad_norm": 0.06295431405305862,
+      "learning_rate": 9.294565347363085e-05,
+      "loss": 0.0062,
+      "step": 1315
+    },
+    {
+      "epoch": 6.357487922705314,
+      "grad_norm": 0.046747080981731415,
+      "learning_rate": 9.29283699553962e-05,
+      "loss": 0.006,
+      "step": 1316
+    },
+    {
+      "epoch": 6.36231884057971,
+      "grad_norm": 0.04669469594955444,
+      "learning_rate": 9.291106690134128e-05,
+      "loss": 0.0068,
+      "step": 1317
+    },
+    {
+      "epoch": 6.367149758454106,
+      "grad_norm": 0.049420833587646484,
+      "learning_rate": 9.289374431934035e-05,
+      "loss": 0.0063,
+      "step": 1318
+    },
+    {
+      "epoch": 6.371980676328502,
+      "grad_norm": 0.07315026968717575,
+      "learning_rate": 9.287640221727658e-05,
+      "loss": 0.0079,
+      "step": 1319
+    },
+    {
+      "epoch": 6.3768115942028984,
+      "grad_norm": 0.044721487909555435,
+      "learning_rate": 9.285904060304198e-05,
+      "loss": 0.0064,
+      "step": 1320
+    },
+    {
+      "epoch": 6.381642512077295,
+      "grad_norm": 0.045167844742536545,
+      "learning_rate": 9.284165948453747e-05,
+      "loss": 0.006,
+      "step": 1321
+    },
+    {
+      "epoch": 6.386473429951691,
+      "grad_norm": 0.045468948781490326,
+      "learning_rate": 9.282425886967284e-05,
+      "loss": 0.0066,
+      "step": 1322
+    },
+    {
+      "epoch": 6.391304347826087,
+      "grad_norm": 0.03726964443922043,
+      "learning_rate": 9.280683876636677e-05,
+      "loss": 0.0057,
+      "step": 1323
+    },
+    {
+      "epoch": 6.396135265700483,
+      "grad_norm": 0.054976094514131546,
+      "learning_rate": 9.278939918254677e-05,
+      "loss": 0.0064,
+      "step": 1324
+    },
+    {
+      "epoch": 6.400966183574879,
+      "grad_norm": 0.059892453253269196,
+      "learning_rate": 9.277194012614925e-05,
+      "loss": 0.0068,
+      "step": 1325
+    },
+    {
+      "epoch": 6.400966183574879,
+      "eval_loss": 0.016453834250569344,
+      "eval_runtime": 21.3612,
+      "eval_samples_per_second": 4.681,
+      "eval_steps_per_second": 0.14,
+      "step": 1325
+    },
+    {
+      "epoch": 6.405797101449275,
+      "grad_norm": 0.048259880393743515,
+      "learning_rate": 9.275446160511946e-05,
+      "loss": 0.0078,
+      "step": 1326
+    },
+    {
+      "epoch": 6.4106280193236715,
+      "grad_norm": 0.06093078851699829,
+      "learning_rate": 9.273696362741151e-05,
+      "loss": 0.0078,
+      "step": 1327
+    },
+    {
+      "epoch": 6.415458937198068,
+      "grad_norm": 0.048800837248563766,
+      "learning_rate": 9.271944620098841e-05,
+      "loss": 0.0071,
+      "step": 1328
+    },
+    {
+      "epoch": 6.420289855072464,
+      "grad_norm": 0.056117165833711624,
+      "learning_rate": 9.270190933382192e-05,
+      "loss": 0.0071,
+      "step": 1329
+    },
+    {
+      "epoch": 6.42512077294686,
+      "grad_norm": 0.06087968125939369,
+      "learning_rate": 9.268435303389275e-05,
+      "loss": 0.0081,
+      "step": 1330
+    },
+    {
+      "epoch": 6.429951690821256,
+      "grad_norm": 0.040126699954271317,
+      "learning_rate": 9.266677730919043e-05,
+      "loss": 0.0048,
+      "step": 1331
+    },
+    {
+      "epoch": 6.434782608695652,
+      "grad_norm": 0.04188450053334236,
+      "learning_rate": 9.264918216771325e-05,
+      "loss": 0.0055,
+      "step": 1332
+    },
+    {
+      "epoch": 6.4396135265700485,
+      "grad_norm": 0.04577482119202614,
+      "learning_rate": 9.263156761746844e-05,
+      "loss": 0.0061,
+      "step": 1333
+    },
+    {
+      "epoch": 6.444444444444445,
+      "grad_norm": 0.05482233315706253,
+      "learning_rate": 9.261393366647202e-05,
+      "loss": 0.0055,
+      "step": 1334
+    },
+    {
+      "epoch": 6.449275362318841,
+      "grad_norm": 0.04913961887359619,
+      "learning_rate": 9.259628032274882e-05,
+      "loss": 0.0055,
+      "step": 1335
+    },
+    {
+      "epoch": 6.454106280193237,
+      "grad_norm": 0.0711643248796463,
+      "learning_rate": 9.257860759433252e-05,
+      "loss": 0.0063,
+      "step": 1336
+    },
+    {
+      "epoch": 6.458937198067633,
+      "grad_norm": 0.040503520518541336,
+      "learning_rate": 9.256091548926565e-05,
+      "loss": 0.0049,
+      "step": 1337
+    },
+    {
+      "epoch": 6.463768115942029,
+      "grad_norm": 0.04341737926006317,
+      "learning_rate": 9.254320401559947e-05,
+      "loss": 0.005,
+      "step": 1338
+    },
+    {
+      "epoch": 6.468599033816425,
+      "grad_norm": 0.04081616923213005,
+      "learning_rate": 9.252547318139415e-05,
+      "loss": 0.0054,
+      "step": 1339
+    },
+    {
+      "epoch": 6.473429951690822,
+      "grad_norm": 0.059148360043764114,
+      "learning_rate": 9.250772299471859e-05,
+      "loss": 0.0062,
+      "step": 1340
+    },
+    {
+      "epoch": 6.478260869565218,
+      "grad_norm": 0.07044517248868942,
+      "learning_rate": 9.248995346365058e-05,
+      "loss": 0.0076,
+      "step": 1341
+    },
+    {
+      "epoch": 6.483091787439614,
+      "grad_norm": 0.0680868849158287,
+      "learning_rate": 9.247216459627665e-05,
+      "loss": 0.0059,
+      "step": 1342
+    },
+    {
+      "epoch": 6.48792270531401,
+      "grad_norm": 0.04491333290934563,
+      "learning_rate": 9.245435640069213e-05,
+      "loss": 0.0058,
+      "step": 1343
+    },
+    {
+      "epoch": 6.492753623188406,
+      "grad_norm": 0.06955177336931229,
+      "learning_rate": 9.243652888500119e-05,
+      "loss": 0.011,
+      "step": 1344
+    },
+    {
+      "epoch": 6.4975845410628015,
+      "grad_norm": 0.054043661803007126,
+      "learning_rate": 9.241868205731676e-05,
+      "loss": 0.0053,
+      "step": 1345
+    },
+    {
+      "epoch": 6.5024154589371985,
+      "grad_norm": 0.04067247360944748,
+      "learning_rate": 9.240081592576056e-05,
+      "loss": 0.0052,
+      "step": 1346
+    },
+    {
+      "epoch": 6.507246376811594,
+      "grad_norm": 0.04347148910164833,
+      "learning_rate": 9.23829304984631e-05,
+      "loss": 0.0064,
+      "step": 1347
+    },
+    {
+      "epoch": 6.512077294685991,
+      "grad_norm": 0.046369269490242004,
+      "learning_rate": 9.236502578356368e-05,
+      "loss": 0.0063,
+      "step": 1348
+    },
+    {
+      "epoch": 6.516908212560386,
+      "grad_norm": 0.048870883882045746,
+      "learning_rate": 9.234710178921035e-05,
+      "loss": 0.0059,
+      "step": 1349
+    },
+    {
+      "epoch": 6.521739130434782,
+      "grad_norm": 0.04518553987145424,
+      "learning_rate": 9.232915852355996e-05,
+      "loss": 0.0059,
+      "step": 1350
+    },
+    {
+      "epoch": 6.521739130434782,
+      "eval_loss": 0.01632525771856308,
+      "eval_runtime": 20.6147,
+      "eval_samples_per_second": 4.851,
+      "eval_steps_per_second": 0.146,
+      "step": 1350
+    },
+    {
+      "epoch": 6.526570048309178,
+      "grad_norm": 0.050386469811201096,
+      "learning_rate": 9.231119599477814e-05,
+      "loss": 0.0071,
+      "step": 1351
+    },
+    {
+      "epoch": 6.531400966183575,
+      "grad_norm": 0.03906089812517166,
+      "learning_rate": 9.229321421103924e-05,
+      "loss": 0.0053,
+      "step": 1352
+    },
+    {
+      "epoch": 6.536231884057971,
+      "grad_norm": 0.056843534111976624,
+      "learning_rate": 9.227521318052641e-05,
+      "loss": 0.0087,
+      "step": 1353
+    },
+    {
+      "epoch": 6.541062801932367,
+      "grad_norm": 0.05382772162556648,
+      "learning_rate": 9.225719291143156e-05,
+      "loss": 0.0061,
+      "step": 1354
+    },
+    {
+      "epoch": 6.545893719806763,
+      "grad_norm": 0.04173978790640831,
+      "learning_rate": 9.223915341195535e-05,
+      "loss": 0.0054,
+      "step": 1355
+    },
+    {
+      "epoch": 6.550724637681159,
+      "grad_norm": 0.055689819157123566,
+      "learning_rate": 9.222109469030716e-05,
+      "loss": 0.0059,
+      "step": 1356
+    },
+    {
+      "epoch": 6.555555555555555,
+      "grad_norm": 0.04744328185915947,
+      "learning_rate": 9.220301675470518e-05,
+      "loss": 0.0076,
+      "step": 1357
+    },
+    {
+      "epoch": 6.5603864734299515,
+      "grad_norm": 0.0510600246489048,
+      "learning_rate": 9.218491961337626e-05,
+      "loss": 0.0062,
+      "step": 1358
+    },
+    {
+      "epoch": 6.565217391304348,
+      "grad_norm": 0.041862208396196365,
+      "learning_rate": 9.216680327455608e-05,
+      "loss": 0.0056,
+      "step": 1359
+    },
+    {
+      "epoch": 6.570048309178744,
+      "grad_norm": 0.0443810373544693,
+      "learning_rate": 9.214866774648899e-05,
+      "loss": 0.0055,
+      "step": 1360
+    },
+    {
+      "epoch": 6.57487922705314,
+      "grad_norm": 0.05620473250746727,
+      "learning_rate": 9.21305130374281e-05,
+      "loss": 0.0073,
+      "step": 1361
+    },
+    {
+      "epoch": 6.579710144927536,
+      "grad_norm": 0.04563259333372116,
+      "learning_rate": 9.211233915563526e-05,
+      "loss": 0.0064,
+      "step": 1362
+    },
+    {
+      "epoch": 6.584541062801932,
+      "grad_norm": 0.05491187795996666,
+      "learning_rate": 9.209414610938101e-05,
+      "loss": 0.0077,
+      "step": 1363
+    },
+    {
+      "epoch": 6.5893719806763285,
+      "grad_norm": 0.055255644023418427,
+      "learning_rate": 9.207593390694466e-05,
+      "loss": 0.0071,
+      "step": 1364
+    },
+    {
+      "epoch": 6.594202898550725,
+      "grad_norm": 0.05295202136039734,
+      "learning_rate": 9.205770255661417e-05,
+      "loss": 0.0067,
+      "step": 1365
+    },
+    {
+      "epoch": 6.599033816425121,
+      "grad_norm": 0.07129408419132233,
+      "learning_rate": 9.20394520666863e-05,
+      "loss": 0.0099,
+      "step": 1366
+    },
+    {
+      "epoch": 6.603864734299517,
+      "grad_norm": 0.04434022307395935,
+      "learning_rate": 9.202118244546643e-05,
+      "loss": 0.0057,
+      "step": 1367
+    },
+    {
+      "epoch": 6.608695652173913,
+      "grad_norm": 0.050070665776729584,
+      "learning_rate": 9.200289370126871e-05,
+      "loss": 0.0057,
+      "step": 1368
+    },
+    {
+      "epoch": 6.613526570048309,
+      "grad_norm": 0.050253286957740784,
+      "learning_rate": 9.198458584241597e-05,
+      "loss": 0.0065,
+      "step": 1369
+    },
+    {
+      "epoch": 6.618357487922705,
+      "grad_norm": 0.05804944410920143,
+      "learning_rate": 9.196625887723975e-05,
+      "loss": 0.0058,
+      "step": 1370
+    },
+    {
+      "epoch": 6.6231884057971016,
+      "grad_norm": 0.05418354645371437,
+      "learning_rate": 9.194791281408026e-05,
+      "loss": 0.0065,
+      "step": 1371
+    },
+    {
+      "epoch": 6.628019323671498,
+      "grad_norm": 0.03469512239098549,
+      "learning_rate": 9.192954766128643e-05,
+      "loss": 0.0044,
+      "step": 1372
+    },
+    {
+      "epoch": 6.632850241545894,
+      "grad_norm": 0.044930633157491684,
+      "learning_rate": 9.191116342721586e-05,
+      "loss": 0.0055,
+      "step": 1373
+    },
+    {
+      "epoch": 6.63768115942029,
+      "grad_norm": 0.05017385631799698,
+      "learning_rate": 9.189276012023483e-05,
+      "loss": 0.006,
+      "step": 1374
+    },
+    {
+      "epoch": 6.642512077294686,
+      "grad_norm": 0.05084605887532234,
+      "learning_rate": 9.187433774871832e-05,
+      "loss": 0.0057,
+      "step": 1375
+    },
+    {
+      "epoch": 6.642512077294686,
+      "eval_loss": 0.01513008400797844,
+      "eval_runtime": 20.6668,
+      "eval_samples_per_second": 4.839,
+      "eval_steps_per_second": 0.145,
+      "step": 1375
+    },
+    {
+      "epoch": 6.647342995169082,
+      "grad_norm": 0.04787183925509453,
+      "learning_rate": 9.185589632104997e-05,
+      "loss": 0.0063,
+      "step": 1376
+    },
+    {
+      "epoch": 6.6521739130434785,
+      "grad_norm": 0.05482997000217438,
+      "learning_rate": 9.183743584562208e-05,
+      "loss": 0.0067,
+      "step": 1377
+    },
+    {
+      "epoch": 6.657004830917875,
+      "grad_norm": 0.0503399483859539,
+      "learning_rate": 9.181895633083564e-05,
+      "loss": 0.0068,
+      "step": 1378
+    },
+    {
+      "epoch": 6.661835748792271,
+      "grad_norm": 0.052327338606119156,
+      "learning_rate": 9.180045778510031e-05,
+      "loss": 0.0063,
+      "step": 1379
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 0.05098320543766022,
+      "learning_rate": 9.178194021683441e-05,
+      "loss": 0.005,
+      "step": 1380
+    },
+    {
+      "epoch": 6.671497584541063,
+      "grad_norm": 0.060967762023210526,
+      "learning_rate": 9.176340363446488e-05,
+      "loss": 0.0053,
+      "step": 1381
+    },
+    {
+      "epoch": 6.676328502415459,
+      "grad_norm": 0.04176538065075874,
+      "learning_rate": 9.174484804642733e-05,
+      "loss": 0.005,
+      "step": 1382
+    },
+    {
+      "epoch": 6.681159420289855,
+      "grad_norm": 0.04020035266876221,
+      "learning_rate": 9.172627346116606e-05,
+      "loss": 0.0071,
+      "step": 1383
+    },
+    {
+      "epoch": 6.685990338164252,
+      "grad_norm": 0.0703640952706337,
+      "learning_rate": 9.170767988713396e-05,
+      "loss": 0.008,
+      "step": 1384
+    },
+    {
+      "epoch": 6.690821256038648,
+      "grad_norm": 0.049804870039224625,
+      "learning_rate": 9.168906733279259e-05,
+      "loss": 0.0075,
+      "step": 1385
+    },
+    {
+      "epoch": 6.695652173913043,
+      "grad_norm": 0.04536246880888939,
+      "learning_rate": 9.167043580661215e-05,
+      "loss": 0.0054,
+      "step": 1386
+    },
+    {
+      "epoch": 6.70048309178744,
+      "grad_norm": 0.03730194270610809,
+      "learning_rate": 9.165178531707144e-05,
+      "loss": 0.005,
+      "step": 1387
+    },
+    {
+      "epoch": 6.705314009661835,
+      "grad_norm": 0.045735448598861694,
+      "learning_rate": 9.163311587265793e-05,
+      "loss": 0.0061,
+      "step": 1388
+    },
+    {
+      "epoch": 6.710144927536232,
+      "grad_norm": 0.04150770232081413,
+      "learning_rate": 9.161442748186771e-05,
+      "loss": 0.0051,
+      "step": 1389
+    },
+    {
+      "epoch": 6.714975845410628,
+      "grad_norm": 0.04654907062649727,
+      "learning_rate": 9.159572015320548e-05,
+      "loss": 0.0061,
+      "step": 1390
+    },
+    {
+      "epoch": 6.719806763285024,
+      "grad_norm": 0.06273826211690903,
+      "learning_rate": 9.157699389518456e-05,
+      "loss": 0.0083,
+      "step": 1391
+    },
+    {
+      "epoch": 6.72463768115942,
+      "grad_norm": 0.050027329474687576,
+      "learning_rate": 9.155824871632688e-05,
+      "loss": 0.0061,
+      "step": 1392
+    },
+    {
+      "epoch": 6.729468599033816,
+      "grad_norm": 0.04607383534312248,
+      "learning_rate": 9.153948462516299e-05,
+      "loss": 0.0053,
+      "step": 1393
+    },
+    {
+      "epoch": 6.734299516908212,
+      "grad_norm": 0.044204093515872955,
+      "learning_rate": 9.152070163023203e-05,
+      "loss": 0.0047,
+      "step": 1394
+    },
+    {
+      "epoch": 6.739130434782608,
+      "grad_norm": 0.043753284960985184,
+      "learning_rate": 9.150189974008179e-05,
+      "loss": 0.0067,
+      "step": 1395
+    },
+    {
+      "epoch": 6.743961352657005,
+      "grad_norm": 0.052429623901844025,
+      "learning_rate": 9.148307896326857e-05,
+      "loss": 0.007,
+      "step": 1396
+    },
+    {
+      "epoch": 6.748792270531401,
+      "grad_norm": 0.06474307924509048,
+      "learning_rate": 9.14642393083574e-05,
+      "loss": 0.0086,
+      "step": 1397
+    },
+    {
+      "epoch": 6.753623188405797,
+      "grad_norm": 0.0789802297949791,
+      "learning_rate": 9.144538078392173e-05,
+      "loss": 0.0082,
+      "step": 1398
+    },
+    {
+      "epoch": 6.758454106280193,
+      "grad_norm": 0.0541478730738163,
+      "learning_rate": 9.142650339854374e-05,
+      "loss": 0.0058,
+      "step": 1399
+    },
+    {
+      "epoch": 6.763285024154589,
+      "grad_norm": 0.04348256438970566,
+      "learning_rate": 9.140760716081414e-05,
+      "loss": 0.0061,
+      "step": 1400
+    },
+    {
+      "epoch": 6.763285024154589,
+      "eval_loss": 0.016424350440502167,
+      "eval_runtime": 20.611,
+      "eval_samples_per_second": 4.852,
+      "eval_steps_per_second": 0.146,
+      "step": 1400
+    },
+    {
+      "epoch": 6.768115942028985,
+      "grad_norm": 0.06769420206546783,
+      "learning_rate": 9.13886920793322e-05,
+      "loss": 0.0064,
+      "step": 1401
+    },
+    {
+      "epoch": 6.7729468599033815,
+      "grad_norm": 0.03817087411880493,
+      "learning_rate": 9.136975816270579e-05,
+      "loss": 0.005,
+      "step": 1402
+    },
+    {
+      "epoch": 6.777777777777778,
+      "grad_norm": 0.03812692314386368,
+      "learning_rate": 9.135080541955136e-05,
+      "loss": 0.0057,
+      "step": 1403
+    },
+    {
+      "epoch": 6.782608695652174,
+      "grad_norm": 0.050985757261514664,
+      "learning_rate": 9.133183385849392e-05,
+      "loss": 0.0066,
+      "step": 1404
+    },
+    {
+      "epoch": 6.78743961352657,
+      "grad_norm": 0.06294248253107071,
+      "learning_rate": 9.1312843488167e-05,
+      "loss": 0.0068,
+      "step": 1405
+    },
+    {
+      "epoch": 6.792270531400966,
+      "grad_norm": 0.05425696820020676,
+      "learning_rate": 9.129383431721276e-05,
+      "loss": 0.0064,
+      "step": 1406
+    },
+    {
+      "epoch": 6.797101449275362,
+      "grad_norm": 0.05892189219594002,
+      "learning_rate": 9.12748063542819e-05,
+      "loss": 0.0063,
+      "step": 1407
+    },
+    {
+      "epoch": 6.8019323671497585,
+      "grad_norm": 0.04670066386461258,
+      "learning_rate": 9.125575960803361e-05,
+      "loss": 0.0061,
+      "step": 1408
+    },
+    {
+      "epoch": 6.806763285024155,
+      "grad_norm": 0.0500483438372612,
+      "learning_rate": 9.123669408713569e-05,
+      "loss": 0.0057,
+      "step": 1409
+    },
+    {
+      "epoch": 6.811594202898551,
+      "grad_norm": 0.06111598014831543,
+      "learning_rate": 9.121760980026449e-05,
+      "loss": 0.0059,
+      "step": 1410
+    },
+    {
+      "epoch": 6.816425120772947,
+      "grad_norm": 0.040104810148477554,
+      "learning_rate": 9.119850675610486e-05,
+      "loss": 0.005,
+      "step": 1411
+    },
+    {
+      "epoch": 6.821256038647343,
+      "grad_norm": 0.058813292533159256,
+      "learning_rate": 9.11793849633502e-05,
+      "loss": 0.0076,
+      "step": 1412
+    },
+    {
+      "epoch": 6.826086956521739,
+      "grad_norm": 0.04925285279750824,
+      "learning_rate": 9.116024443070243e-05,
+      "loss": 0.0059,
+      "step": 1413
+    },
+    {
+      "epoch": 6.830917874396135,
+      "grad_norm": 0.05609273910522461,
+      "learning_rate": 9.114108516687205e-05,
+      "loss": 0.0068,
+      "step": 1414
+    },
+    {
+      "epoch": 6.835748792270532,
+      "grad_norm": 0.06939279288053513,
+      "learning_rate": 9.112190718057802e-05,
+      "loss": 0.0071,
+      "step": 1415
+    },
+    {
+      "epoch": 6.840579710144928,
+      "grad_norm": 0.057253647595644,
+      "learning_rate": 9.110271048054787e-05,
+      "loss": 0.0063,
+      "step": 1416
+    },
+    {
+      "epoch": 6.845410628019324,
+      "grad_norm": 0.0675768330693245,
+      "learning_rate": 9.10834950755176e-05,
+      "loss": 0.0079,
+      "step": 1417
+    },
+    {
+      "epoch": 6.85024154589372,
+      "grad_norm": 0.06414294987916946,
+      "learning_rate": 9.106426097423178e-05,
+      "loss": 0.0076,
+      "step": 1418
+    },
+    {
+      "epoch": 6.855072463768116,
+      "grad_norm": 0.05850673466920853,
+      "learning_rate": 9.104500818544342e-05,
+      "loss": 0.0088,
+      "step": 1419
+    },
+    {
+      "epoch": 6.859903381642512,
+      "grad_norm": 0.08548006415367126,
+      "learning_rate": 9.10257367179141e-05,
+      "loss": 0.0106,
+      "step": 1420
+    },
+    {
+      "epoch": 6.8647342995169085,
+      "grad_norm": 0.04931001737713814,
+      "learning_rate": 9.100644658041382e-05,
+      "loss": 0.0078,
+      "step": 1421
+    },
+    {
+      "epoch": 6.869565217391305,
+      "grad_norm": 0.0717255249619484,
+      "learning_rate": 9.098713778172119e-05,
+      "loss": 0.0074,
+      "step": 1422
+    },
+    {
+      "epoch": 6.874396135265701,
+      "grad_norm": 0.045794110745191574,
+      "learning_rate": 9.09678103306232e-05,
+      "loss": 0.0067,
+      "step": 1423
+    },
+    {
+      "epoch": 6.879227053140097,
+      "grad_norm": 0.04047313332557678,
+      "learning_rate": 9.094846423591539e-05,
+      "loss": 0.006,
+      "step": 1424
+    },
+    {
+      "epoch": 6.884057971014493,
+      "grad_norm": 0.0553482323884964,
+      "learning_rate": 9.092909950640179e-05,
+      "loss": 0.006,
+      "step": 1425
+    },
+    {
+      "epoch": 6.884057971014493,
+      "eval_loss": 0.015602950006723404,
+      "eval_runtime": 21.3308,
+      "eval_samples_per_second": 4.688,
+      "eval_steps_per_second": 0.141,
+      "step": 1425
+    },
+    {
+      "epoch": 6.888888888888889,
+      "grad_norm": 0.044621117413043976,
+      "learning_rate": 9.090971615089486e-05,
+      "loss": 0.0064,
+      "step": 1426
+    },
+    {
+      "epoch": 6.8937198067632846,
+      "grad_norm": 0.05033578351140022,
+      "learning_rate": 9.089031417821558e-05,
+      "loss": 0.0066,
+      "step": 1427
+    },
+    {
+      "epoch": 6.898550724637682,
+      "grad_norm": 0.03769382834434509,
+      "learning_rate": 9.08708935971934e-05,
+      "loss": 0.005,
+      "step": 1428
+    },
+    {
+      "epoch": 6.903381642512077,
+      "grad_norm": 0.04533964768052101,
+      "learning_rate": 9.085145441666622e-05,
+      "loss": 0.0064,
+      "step": 1429
+    },
+    {
+      "epoch": 6.908212560386474,
+      "grad_norm": 0.04200972989201546,
+      "learning_rate": 9.08319966454804e-05,
+      "loss": 0.0052,
+      "step": 1430
+    },
+    {
+      "epoch": 6.913043478260869,
+      "grad_norm": 0.05959639698266983,
+      "learning_rate": 9.081252029249078e-05,
+      "loss": 0.0069,
+      "step": 1431
+    },
+    {
+      "epoch": 6.917874396135265,
+      "grad_norm": 0.05127401277422905,
+      "learning_rate": 9.079302536656068e-05,
+      "loss": 0.0065,
+      "step": 1432
+    },
+    {
+      "epoch": 6.9227053140096615,
+      "grad_norm": 0.05631054937839508,
+      "learning_rate": 9.07735118765618e-05,
+      "loss": 0.0089,
+      "step": 1433
+    },
+    {
+      "epoch": 6.927536231884058,
+      "grad_norm": 0.0575186163187027,
+      "learning_rate": 9.075397983137434e-05,
+      "loss": 0.0087,
+      "step": 1434
+    },
+    {
+      "epoch": 6.932367149758454,
+      "grad_norm": 0.055179186165332794,
+      "learning_rate": 9.073442923988694e-05,
+      "loss": 0.0091,
+      "step": 1435
+    },
+    {
+      "epoch": 6.93719806763285,
+      "grad_norm": 0.04829880967736244,
+      "learning_rate": 9.071486011099665e-05,
+      "loss": 0.006,
+      "step": 1436
+    },
+    {
+      "epoch": 6.942028985507246,
+      "grad_norm": 0.05416601151227951,
+      "learning_rate": 9.069527245360902e-05,
+      "loss": 0.0055,
+      "step": 1437
+    },
+    {
+      "epoch": 6.946859903381642,
+      "grad_norm": 0.045596893876791,
+      "learning_rate": 9.067566627663796e-05,
+      "loss": 0.0068,
+      "step": 1438
+    },
+    {
+      "epoch": 6.951690821256038,
+      "grad_norm": 0.04718053713440895,
+      "learning_rate": 9.065604158900585e-05,
+      "loss": 0.0061,
+      "step": 1439
+    },
+    {
+      "epoch": 6.956521739130435,
+      "grad_norm": 0.04946717619895935,
+      "learning_rate": 9.063639839964347e-05,
+      "loss": 0.0061,
+      "step": 1440
+    },
+    {
+      "epoch": 6.961352657004831,
+      "grad_norm": 0.06101173162460327,
+      "learning_rate": 9.061673671749005e-05,
+      "loss": 0.0083,
+      "step": 1441
+    },
+    {
+      "epoch": 6.966183574879227,
+      "grad_norm": 0.05813223123550415,
+      "learning_rate": 9.05970565514932e-05,
+      "loss": 0.0077,
+      "step": 1442
+    },
+    {
+      "epoch": 6.971014492753623,
+      "grad_norm": 0.04281511530280113,
+      "learning_rate": 9.057735791060897e-05,
+      "loss": 0.0043,
+      "step": 1443
+    },
+    {
+      "epoch": 6.975845410628019,
+      "grad_norm": 0.04492162540555,
+      "learning_rate": 9.055764080380182e-05,
+      "loss": 0.006,
+      "step": 1444
+    },
+    {
+      "epoch": 6.980676328502415,
+      "grad_norm": 0.03525848686695099,
+      "learning_rate": 9.053790524004459e-05,
+      "loss": 0.0055,
+      "step": 1445
+    },
+    {
+      "epoch": 6.9855072463768115,
+      "grad_norm": 0.0547606386244297,
+      "learning_rate": 9.051815122831851e-05,
+      "loss": 0.0055,
+      "step": 1446
+    },
+    {
+      "epoch": 6.990338164251208,
+      "grad_norm": 0.05137596279382706,
+      "learning_rate": 9.049837877761325e-05,
+      "loss": 0.0072,
+      "step": 1447
+    },
+    {
+      "epoch": 6.995169082125604,
+      "grad_norm": 0.08006022125482559,
+      "learning_rate": 9.047858789692684e-05,
+      "loss": 0.0068,
+      "step": 1448
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0994105115532875,
+      "learning_rate": 9.045877859526573e-05,
+      "loss": 0.0079,
+      "step": 1449
+    },
+    {
+      "epoch": 7.004830917874396,
+      "grad_norm": 0.049069393426179886,
+      "learning_rate": 9.043895088164467e-05,
+      "loss": 0.0062,
+      "step": 1450
+    },
+    {
+      "epoch": 7.004830917874396,
+      "eval_loss": 0.016072051599621773,
+      "eval_runtime": 20.6059,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
+      "step": 1450
+    },
+    {
+      "epoch": 7.009661835748792,
+      "grad_norm": 0.03314821794629097,
+      "learning_rate": 9.041910476508688e-05,
+      "loss": 0.0038,
+      "step": 1451
+    },
+    {
+      "epoch": 7.0144927536231885,
+      "grad_norm": 0.047244373708963394,
+      "learning_rate": 9.039924025462392e-05,
+      "loss": 0.0046,
+      "step": 1452
+    },
+    {
+      "epoch": 7.019323671497585,
+      "grad_norm": 0.0837038904428482,
+      "learning_rate": 9.037935735929572e-05,
+      "loss": 0.0075,
+      "step": 1453
+    },
+    {
+      "epoch": 7.024154589371981,
+      "grad_norm": 0.031166687607765198,
+      "learning_rate": 9.035945608815056e-05,
+      "loss": 0.0045,
+      "step": 1454
+    },
+    {
+      "epoch": 7.028985507246377,
+      "grad_norm": 0.04692283272743225,
+      "learning_rate": 9.033953645024514e-05,
+      "loss": 0.0049,
+      "step": 1455
+    },
+    {
+      "epoch": 7.033816425120773,
+      "grad_norm": 0.057650938630104065,
+      "learning_rate": 9.031959845464443e-05,
+      "loss": 0.0042,
+      "step": 1456
+    },
+    {
+      "epoch": 7.038647342995169,
+      "grad_norm": 0.04346553981304169,
+      "learning_rate": 9.029964211042184e-05,
+      "loss": 0.0045,
+      "step": 1457
+    },
+    {
+      "epoch": 7.043478260869565,
+      "grad_norm": 0.04347117245197296,
+      "learning_rate": 9.027966742665906e-05,
+      "loss": 0.004,
+      "step": 1458
+    },
+    {
+      "epoch": 7.048309178743962,
+      "grad_norm": 0.07791293412446976,
+      "learning_rate": 9.025967441244621e-05,
+      "loss": 0.0046,
+      "step": 1459
+    },
+    {
+      "epoch": 7.053140096618358,
+      "grad_norm": 0.05497191846370697,
+      "learning_rate": 9.023966307688164e-05,
+      "loss": 0.0046,
+      "step": 1460
+    },
+    {
+      "epoch": 7.057971014492754,
+      "grad_norm": 0.07032640278339386,
+      "learning_rate": 9.021963342907213e-05,
+      "loss": 0.0066,
+      "step": 1461
+    },
+    {
+      "epoch": 7.06280193236715,
+      "grad_norm": 0.055110178887844086,
+      "learning_rate": 9.019958547813277e-05,
+      "loss": 0.0071,
+      "step": 1462
+    },
+    {
+      "epoch": 7.067632850241546,
+      "grad_norm": 0.05101725459098816,
+      "learning_rate": 9.017951923318694e-05,
+      "loss": 0.0056,
+      "step": 1463
+    },
+    {
+      "epoch": 7.072463768115942,
+      "grad_norm": 0.041964564472436905,
+      "learning_rate": 9.015943470336643e-05,
+      "loss": 0.0053,
+      "step": 1464
+    },
+    {
+      "epoch": 7.0772946859903385,
+      "grad_norm": 0.07983453571796417,
+      "learning_rate": 9.013933189781124e-05,
+      "loss": 0.0059,
+      "step": 1465
+    },
+    {
+      "epoch": 7.082125603864735,
+      "grad_norm": 0.05337269604206085,
+      "learning_rate": 9.011921082566977e-05,
+      "loss": 0.006,
+      "step": 1466
+    },
+    {
+      "epoch": 7.086956521739131,
+      "grad_norm": 0.05293073132634163,
+      "learning_rate": 9.009907149609869e-05,
+      "loss": 0.0059,
+      "step": 1467
+    },
+    {
+      "epoch": 7.091787439613527,
+      "grad_norm": 0.04248853400349617,
+      "learning_rate": 9.007891391826304e-05,
+      "loss": 0.0045,
+      "step": 1468
+    },
+    {
+      "epoch": 7.096618357487923,
+      "grad_norm": 0.04298196732997894,
+      "learning_rate": 9.005873810133606e-05,
+      "loss": 0.0046,
+      "step": 1469
+    },
+    {
+      "epoch": 7.101449275362318,
+      "grad_norm": 0.03962076082825661,
+      "learning_rate": 9.003854405449939e-05,
+      "loss": 0.0049,
+      "step": 1470
+    },
+    {
+      "epoch": 7.106280193236715,
+      "grad_norm": 0.04526727646589279,
+      "learning_rate": 9.001833178694292e-05,
+      "loss": 0.0046,
+      "step": 1471
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 0.04316214099526405,
+      "learning_rate": 8.999810130786484e-05,
+      "loss": 0.0063,
+      "step": 1472
+    },
+    {
+      "epoch": 7.115942028985507,
+      "grad_norm": 0.044104162603616714,
+      "learning_rate": 8.99778526264716e-05,
+      "loss": 0.0066,
+      "step": 1473
+    },
+    {
+      "epoch": 7.120772946859903,
+      "grad_norm": 0.08153866231441498,
+      "learning_rate": 8.9957585751978e-05,
+      "loss": 0.0049,
+      "step": 1474
+    },
+    {
+      "epoch": 7.125603864734299,
+      "grad_norm": 0.04352913051843643,
+      "learning_rate": 8.993730069360706e-05,
+      "loss": 0.006,
+      "step": 1475
+    },
+    {
+      "epoch": 7.125603864734299,
+      "eval_loss": 0.017760971561074257,
+      "eval_runtime": 20.5955,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
+      "step": 1475
+    },
+    {
+      "epoch": 7.130434782608695,
+      "grad_norm": 0.04243316128849983,
+      "learning_rate": 8.991699746059007e-05,
+      "loss": 0.0046,
+      "step": 1476
+    },
+    {
+      "epoch": 7.1352657004830915,
+      "grad_norm": 0.04733670502901077,
+      "learning_rate": 8.989667606216668e-05,
+      "loss": 0.0049,
+      "step": 1477
+    },
+    {
+      "epoch": 7.140096618357488,
+      "grad_norm": 0.04068618267774582,
+      "learning_rate": 8.987633650758466e-05,
+      "loss": 0.0048,
+      "step": 1478
+    },
+    {
+      "epoch": 7.144927536231884,
+      "grad_norm": 0.04422817751765251,
+      "learning_rate": 8.985597880610018e-05,
+      "loss": 0.0039,
+      "step": 1479
+    },
+    {
+      "epoch": 7.14975845410628,
+      "grad_norm": 0.04335469380021095,
+      "learning_rate": 8.983560296697757e-05,
+      "loss": 0.0045,
+      "step": 1480
+    },
+    {
+      "epoch": 7.154589371980676,
+      "grad_norm": 0.03592665493488312,
+      "learning_rate": 8.981520899948952e-05,
+      "loss": 0.0035,
+      "step": 1481
+    },
+    {
+      "epoch": 7.159420289855072,
+      "grad_norm": 0.056261006742715836,
+      "learning_rate": 8.979479691291683e-05,
+      "loss": 0.0045,
+      "step": 1482
+    },
+    {
+      "epoch": 7.164251207729468,
+      "grad_norm": 0.06332986056804657,
+      "learning_rate": 8.977436671654869e-05,
+      "loss": 0.0066,
+      "step": 1483
+    },
+    {
+      "epoch": 7.169082125603865,
+      "grad_norm": 0.038734789937734604,
+      "learning_rate": 8.97539184196824e-05,
+      "loss": 0.0032,
+      "step": 1484
+    },
+    {
+      "epoch": 7.173913043478261,
+      "grad_norm": 0.05872436612844467,
+      "learning_rate": 8.973345203162362e-05,
+      "loss": 0.0059,
+      "step": 1485
+    },
+    {
+      "epoch": 7.178743961352657,
+      "grad_norm": 0.06064346432685852,
+      "learning_rate": 8.971296756168615e-05,
+      "loss": 0.0067,
+      "step": 1486
+    },
+    {
+      "epoch": 7.183574879227053,
+      "grad_norm": 0.04277418926358223,
+      "learning_rate": 8.969246501919204e-05,
+      "loss": 0.0057,
+      "step": 1487
+    },
+    {
+      "epoch": 7.188405797101449,
+      "grad_norm": 0.027813758701086044,
+      "learning_rate": 8.967194441347161e-05,
+      "loss": 0.0028,
+      "step": 1488
+    },
+    {
+      "epoch": 7.193236714975845,
+      "grad_norm": 0.0485023558139801,
+      "learning_rate": 8.965140575386336e-05,
+      "loss": 0.0044,
+      "step": 1489
+    },
+    {
+      "epoch": 7.1980676328502415,
+      "grad_norm": 0.04239838570356369,
+      "learning_rate": 8.963084904971397e-05,
+      "loss": 0.0043,
+      "step": 1490
+    },
+    {
+      "epoch": 7.202898550724638,
+      "grad_norm": 0.043228596448898315,
+      "learning_rate": 8.961027431037843e-05,
+      "loss": 0.0047,
+      "step": 1491
+    },
+    {
+      "epoch": 7.207729468599034,
+      "grad_norm": 0.06608793139457703,
+      "learning_rate": 8.958968154521985e-05,
+      "loss": 0.0053,
+      "step": 1492
+    },
+    {
+      "epoch": 7.21256038647343,
+      "grad_norm": 0.045232873409986496,
+      "learning_rate": 8.956907076360958e-05,
+      "loss": 0.0043,
+      "step": 1493
+    },
+    {
+      "epoch": 7.217391304347826,
+      "grad_norm": 0.057870227843523026,
+      "learning_rate": 8.954844197492719e-05,
+      "loss": 0.0047,
+      "step": 1494
+    },
+    {
+      "epoch": 7.222222222222222,
+      "grad_norm": 0.03591402992606163,
+      "learning_rate": 8.952779518856038e-05,
+      "loss": 0.0038,
+      "step": 1495
+    },
+    {
+      "epoch": 7.2270531400966185,
+      "grad_norm": 0.05448876693844795,
+      "learning_rate": 8.950713041390511e-05,
+      "loss": 0.0053,
+      "step": 1496
+    },
+    {
+      "epoch": 7.231884057971015,
+      "grad_norm": 0.044634390622377396,
+      "learning_rate": 8.948644766036546e-05,
+      "loss": 0.0042,
+      "step": 1497
+    },
+    {
+      "epoch": 7.236714975845411,
+      "grad_norm": 0.04010568931698799,
+      "learning_rate": 8.946574693735375e-05,
+      "loss": 0.0048,
+      "step": 1498
+    },
+    {
+      "epoch": 7.241545893719807,
+      "grad_norm": 0.07772307097911835,
+      "learning_rate": 8.944502825429045e-05,
+      "loss": 0.0058,
+      "step": 1499
+    },
+    {
+      "epoch": 7.246376811594203,
+      "grad_norm": 0.06972262263298035,
+      "learning_rate": 8.942429162060421e-05,
+      "loss": 0.0059,
+      "step": 1500
+    },
+    {
+      "epoch": 7.246376811594203,
+      "eval_loss": 0.01691284403204918,
+      "eval_runtime": 20.613,
+      "eval_samples_per_second": 4.851,
+      "eval_steps_per_second": 0.146,
+      "step": 1500
+    },
+    {
+      "epoch": 7.251207729468599,
+      "grad_norm": 0.04732871428132057,
+      "learning_rate": 8.940353704573186e-05,
+      "loss": 0.0053,
+      "step": 1501
+    },
+    {
+      "epoch": 7.256038647342995,
+      "grad_norm": 0.03620045259594917,
+      "learning_rate": 8.938276453911834e-05,
+      "loss": 0.0042,
+      "step": 1502
+    },
+    {
+      "epoch": 7.260869565217392,
+      "grad_norm": 0.06282836198806763,
+      "learning_rate": 8.936197411021683e-05,
+      "loss": 0.0065,
+      "step": 1503
+    },
+    {
+      "epoch": 7.265700483091788,
+      "grad_norm": 0.057784438133239746,
+      "learning_rate": 8.934116576848861e-05,
+      "loss": 0.0049,
+      "step": 1504
+    },
+    {
+      "epoch": 7.270531400966184,
+      "grad_norm": 0.04168645292520523,
+      "learning_rate": 8.932033952340314e-05,
+      "loss": 0.0062,
+      "step": 1505
+    },
+    {
+      "epoch": 7.27536231884058,
+      "grad_norm": 0.06200752034783363,
+      "learning_rate": 8.9299495384438e-05,
+      "loss": 0.0062,
+      "step": 1506
+    },
+    {
+      "epoch": 7.280193236714976,
+      "grad_norm": 0.03737403079867363,
+      "learning_rate": 8.927863336107895e-05,
+      "loss": 0.004,
+      "step": 1507
+    },
+    {
+      "epoch": 7.285024154589372,
+      "grad_norm": 0.0492960587143898,
+      "learning_rate": 8.925775346281988e-05,
+      "loss": 0.0051,
+      "step": 1508
+    },
+    {
+      "epoch": 7.2898550724637685,
+      "grad_norm": 0.046553079038858414,
+      "learning_rate": 8.923685569916276e-05,
+      "loss": 0.0061,
+      "step": 1509
+    },
+    {
+      "epoch": 7.294685990338165,
+      "grad_norm": 0.0506129264831543,
+      "learning_rate": 8.921594007961774e-05,
+      "loss": 0.0049,
+      "step": 1510
+    },
+    {
+      "epoch": 7.29951690821256,
+      "grad_norm": 0.05896652489900589,
+      "learning_rate": 8.919500661370313e-05,
+      "loss": 0.0046,
+      "step": 1511
+    },
+    {
+      "epoch": 7.304347826086957,
+      "grad_norm": 0.058990173041820526,
+      "learning_rate": 8.917405531094529e-05,
+      "loss": 0.0052,
+      "step": 1512
+    },
+    {
+      "epoch": 7.309178743961352,
+      "grad_norm": 0.05548791214823723,
+      "learning_rate": 8.915308618087874e-05,
+      "loss": 0.0058,
+      "step": 1513
+    },
+    {
+      "epoch": 7.314009661835748,
+      "grad_norm": 0.03932204470038414,
+      "learning_rate": 8.913209923304608e-05,
+      "loss": 0.0053,
+      "step": 1514
+    },
+    {
+      "epoch": 7.318840579710145,
+      "grad_norm": 0.05363031104207039,
+      "learning_rate": 8.911109447699806e-05,
+      "loss": 0.0055,
+      "step": 1515
+    },
+    {
+      "epoch": 7.323671497584541,
+      "grad_norm": 0.030770711600780487,
+      "learning_rate": 8.909007192229352e-05,
+      "loss": 0.0036,
+      "step": 1516
+    },
+    {
+      "epoch": 7.328502415458937,
+      "grad_norm": 0.06688834726810455,
+      "learning_rate": 8.906903157849937e-05,
+      "loss": 0.0066,
+      "step": 1517
+    },
+    {
+      "epoch": 7.333333333333333,
+      "grad_norm": 0.04870985448360443,
+      "learning_rate": 8.904797345519065e-05,
+      "loss": 0.0047,
+      "step": 1518
+    },
+    {
+      "epoch": 7.338164251207729,
+      "grad_norm": 0.037944402545690536,
+      "learning_rate": 8.902689756195049e-05,
+      "loss": 0.004,
+      "step": 1519
+    },
+    {
+      "epoch": 7.342995169082125,
+      "grad_norm": 0.05561332404613495,
+      "learning_rate": 8.900580390837007e-05,
+      "loss": 0.007,
+      "step": 1520
+    },
+    {
+      "epoch": 7.3478260869565215,
+      "grad_norm": 0.05865646153688431,
+      "learning_rate": 8.89846925040487e-05,
+      "loss": 0.0051,
+      "step": 1521
+    },
+    {
+      "epoch": 7.352657004830918,
+      "grad_norm": 0.03414580225944519,
+      "learning_rate": 8.896356335859375e-05,
+      "loss": 0.004,
+      "step": 1522
+    },
+    {
+      "epoch": 7.357487922705314,
+      "grad_norm": 0.055078282952308655,
+      "learning_rate": 8.894241648162064e-05,
+      "loss": 0.0045,
+      "step": 1523
+    },
+    {
+      "epoch": 7.36231884057971,
+      "grad_norm": 0.044683605432510376,
+      "learning_rate": 8.89212518827529e-05,
+      "loss": 0.0048,
+      "step": 1524
+    },
+    {
+      "epoch": 7.367149758454106,
+      "grad_norm": 0.04428521916270256,
+      "learning_rate": 8.89000695716221e-05,
+      "loss": 0.0043,
+      "step": 1525
+    },
+    {
+      "epoch": 7.367149758454106,
+      "eval_loss": 0.01745455525815487,
+      "eval_runtime": 20.6235,
+      "eval_samples_per_second": 4.849,
+      "eval_steps_per_second": 0.145,
+      "step": 1525
+    },
+    {
+      "epoch": 7.371980676328502,
+      "grad_norm": 0.06754156947135925,
+      "learning_rate": 8.887886955786786e-05,
+      "loss": 0.0048,
+      "step": 1526
+    },
+    {
+      "epoch": 7.3768115942028984,
+      "grad_norm": 0.05913345888257027,
+      "learning_rate": 8.885765185113789e-05,
+      "loss": 0.0066,
+      "step": 1527
+    },
+    {
+      "epoch": 7.381642512077295,
+      "grad_norm": 0.06382418423891068,
+      "learning_rate": 8.883641646108793e-05,
+      "loss": 0.0069,
+      "step": 1528
+    },
+    {
+      "epoch": 7.386473429951691,
+      "grad_norm": 0.06488420814275742,
+      "learning_rate": 8.881516339738176e-05,
+      "loss": 0.0067,
+      "step": 1529
+    },
+    {
+      "epoch": 7.391304347826087,
+      "grad_norm": 0.04919671267271042,
+      "learning_rate": 8.879389266969122e-05,
+      "loss": 0.0044,
+      "step": 1530
+    },
+    {
+      "epoch": 7.396135265700483,
+      "grad_norm": 0.040717657655477524,
+      "learning_rate": 8.877260428769619e-05,
+      "loss": 0.0038,
+      "step": 1531
+    },
+    {
+      "epoch": 7.400966183574879,
+      "grad_norm": 0.06025019288063049,
+      "learning_rate": 8.875129826108456e-05,
+      "loss": 0.0049,
+      "step": 1532
+    },
+    {
+      "epoch": 7.405797101449275,
+      "grad_norm": 0.06267979741096497,
+      "learning_rate": 8.872997459955226e-05,
+      "loss": 0.0056,
+      "step": 1533
+    },
+    {
+      "epoch": 7.4106280193236715,
+      "grad_norm": 0.05493423342704773,
+      "learning_rate": 8.870863331280327e-05,
+      "loss": 0.0056,
+      "step": 1534
+    },
+    {
+      "epoch": 7.415458937198068,
+      "grad_norm": 0.046555474400520325,
+      "learning_rate": 8.868727441054958e-05,
+      "loss": 0.0049,
+      "step": 1535
+    },
+    {
+      "epoch": 7.420289855072464,
+      "grad_norm": 0.06376277655363083,
+      "learning_rate": 8.866589790251118e-05,
+      "loss": 0.0059,
+      "step": 1536
+    },
+    {
+      "epoch": 7.42512077294686,
+      "grad_norm": 0.05157245695590973,
+      "learning_rate": 8.864450379841604e-05,
+      "loss": 0.006,
+      "step": 1537
+    },
+    {
+      "epoch": 7.429951690821256,
+      "grad_norm": 0.054448019713163376,
+      "learning_rate": 8.862309210800024e-05,
+      "loss": 0.0057,
+      "step": 1538
+    },
+    {
+      "epoch": 7.434782608695652,
+      "grad_norm": 0.05123317614197731,
+      "learning_rate": 8.860166284100776e-05,
+      "loss": 0.0057,
+      "step": 1539
+    },
+    {
+      "epoch": 7.4396135265700485,
+      "grad_norm": 0.04118344932794571,
+      "learning_rate": 8.858021600719065e-05,
+      "loss": 0.0046,
+      "step": 1540
+    },
+    {
+      "epoch": 7.444444444444445,
+      "grad_norm": 0.05112331360578537,
+      "learning_rate": 8.855875161630887e-05,
+      "loss": 0.0051,
+      "step": 1541
+    },
+    {
+      "epoch": 7.449275362318841,
+      "grad_norm": 0.07118396461009979,
+      "learning_rate": 8.853726967813048e-05,
+      "loss": 0.0064,
+      "step": 1542
+    },
+    {
+      "epoch": 7.454106280193237,
+      "grad_norm": 0.04440603032708168,
+      "learning_rate": 8.851577020243144e-05,
+      "loss": 0.005,
+      "step": 1543
+    },
+    {
+      "epoch": 7.458937198067633,
+      "grad_norm": 0.033633433282375336,
+      "learning_rate": 8.849425319899574e-05,
+      "loss": 0.0039,
+      "step": 1544
+    },
+    {
+      "epoch": 7.463768115942029,
+      "grad_norm": 0.05634831264615059,
+      "learning_rate": 8.847271867761532e-05,
+      "loss": 0.0051,
+      "step": 1545
+    },
+    {
+      "epoch": 7.468599033816425,
+      "grad_norm": 0.05170667916536331,
+      "learning_rate": 8.845116664809007e-05,
+      "loss": 0.0057,
+      "step": 1546
+    },
+    {
+      "epoch": 7.473429951690822,
+      "grad_norm": 0.05650138854980469,
+      "learning_rate": 8.842959712022794e-05,
+      "loss": 0.0051,
+      "step": 1547
+    },
+    {
+      "epoch": 7.478260869565218,
+      "grad_norm": 0.05898016318678856,
+      "learning_rate": 8.84080101038447e-05,
+      "loss": 0.0057,
+      "step": 1548
+    },
+    {
+      "epoch": 7.483091787439614,
+      "grad_norm": 0.0635242760181427,
+      "learning_rate": 8.838640560876422e-05,
+      "loss": 0.0046,
+      "step": 1549
+    },
+    {
+      "epoch": 7.48792270531401,
+      "grad_norm": 0.08860024064779282,
+      "learning_rate": 8.836478364481824e-05,
+      "loss": 0.0049,
+      "step": 1550
+    },
+    {
+      "epoch": 7.48792270531401,
+      "eval_loss": 0.017847729846835136,
+      "eval_runtime": 20.6294,
+      "eval_samples_per_second": 4.847,
+      "eval_steps_per_second": 0.145,
+      "step": 1550
+    },
+    {
+      "epoch": 7.492753623188406,
+      "grad_norm": 0.06238357722759247,
+      "learning_rate": 8.834314422184648e-05,
+      "loss": 0.008,
+      "step": 1551
+    },
+    {
+      "epoch": 7.4975845410628015,
+      "grad_norm": 0.03494458273053169,
+      "learning_rate": 8.832148734969657e-05,
+      "loss": 0.0052,
+      "step": 1552
+    },
+    {
+      "epoch": 7.5024154589371985,
+      "grad_norm": 0.044305965304374695,
+      "learning_rate": 8.829981303822415e-05,
+      "loss": 0.0047,
+      "step": 1553
+    },
+    {
+      "epoch": 7.507246376811594,
+      "grad_norm": 0.0723017156124115,
+      "learning_rate": 8.827812129729271e-05,
+      "loss": 0.0069,
+      "step": 1554
+    },
+    {
+      "epoch": 7.512077294685991,
+      "grad_norm": 0.049968816339969635,
+      "learning_rate": 8.825641213677375e-05,
+      "loss": 0.0042,
+      "step": 1555
+    },
+    {
+      "epoch": 7.516908212560386,
+      "grad_norm": 0.05750703811645508,
+      "learning_rate": 8.823468556654666e-05,
+      "loss": 0.0067,
+      "step": 1556
+    },
+    {
+      "epoch": 7.521739130434782,
+      "grad_norm": 0.07065118849277496,
+      "learning_rate": 8.821294159649874e-05,
+      "loss": 0.0067,
+      "step": 1557
+    },
+    {
+      "epoch": 7.526570048309178,
+      "grad_norm": 0.046665143221616745,
+      "learning_rate": 8.819118023652525e-05,
+      "loss": 0.0042,
+      "step": 1558
+    },
+    {
+      "epoch": 7.531400966183575,
+      "grad_norm": 0.06773227453231812,
+      "learning_rate": 8.81694014965293e-05,
+      "loss": 0.0058,
+      "step": 1559
+    },
+    {
+      "epoch": 7.536231884057971,
+      "grad_norm": 0.04640137404203415,
+      "learning_rate": 8.814760538642198e-05,
+      "loss": 0.0056,
+      "step": 1560
+    },
+    {
+      "epoch": 7.541062801932367,
+      "grad_norm": 0.04420126974582672,
+      "learning_rate": 8.812579191612224e-05,
+      "loss": 0.0052,
+      "step": 1561
+    },
+    {
+      "epoch": 7.545893719806763,
+      "grad_norm": 0.05039765685796738,
+      "learning_rate": 8.810396109555695e-05,
+      "loss": 0.0054,
+      "step": 1562
+    },
+    {
+      "epoch": 7.550724637681159,
+      "grad_norm": 0.050329990684986115,
+      "learning_rate": 8.808211293466084e-05,
+      "loss": 0.0051,
+      "step": 1563
+    },
+    {
+      "epoch": 7.555555555555555,
+      "grad_norm": 0.05938250944018364,
+      "learning_rate": 8.80602474433766e-05,
+      "loss": 0.0063,
+      "step": 1564
+    },
+    {
+      "epoch": 7.5603864734299515,
+      "grad_norm": 0.041535988450050354,
+      "learning_rate": 8.803836463165476e-05,
+      "loss": 0.0053,
+      "step": 1565
+    },
+    {
+      "epoch": 7.565217391304348,
+      "grad_norm": 0.051995351910591125,
+      "learning_rate": 8.801646450945371e-05,
+      "loss": 0.0043,
+      "step": 1566
+    },
+    {
+      "epoch": 7.570048309178744,
+      "grad_norm": 0.04369938373565674,
+      "learning_rate": 8.799454708673977e-05,
+      "loss": 0.0048,
+      "step": 1567
+    },
+    {
+      "epoch": 7.57487922705314,
+      "grad_norm": 0.04908826947212219,
+      "learning_rate": 8.79726123734871e-05,
+      "loss": 0.0053,
+      "step": 1568
+    },
+    {
+      "epoch": 7.579710144927536,
+      "grad_norm": 0.05574946478009224,
+      "learning_rate": 8.795066037967775e-05,
+      "loss": 0.0067,
+      "step": 1569
+    },
+    {
+      "epoch": 7.584541062801932,
+      "grad_norm": 0.04000121355056763,
+      "learning_rate": 8.792869111530161e-05,
+      "loss": 0.0047,
+      "step": 1570
+    },
+    {
+      "epoch": 7.5893719806763285,
+      "grad_norm": 0.060283225029706955,
+      "learning_rate": 8.790670459035645e-05,
+      "loss": 0.0072,
+      "step": 1571
+    },
+    {
+      "epoch": 7.594202898550725,
+      "grad_norm": 0.06155461445450783,
+      "learning_rate": 8.788470081484787e-05,
+      "loss": 0.0069,
+      "step": 1572
+    },
+    {
+      "epoch": 7.599033816425121,
+      "grad_norm": 0.05512505769729614,
+      "learning_rate": 8.786267979878934e-05,
+      "loss": 0.0057,
+      "step": 1573
+    },
+    {
+      "epoch": 7.603864734299517,
+      "grad_norm": 0.037634339183568954,
+      "learning_rate": 8.78406415522022e-05,
+      "loss": 0.0042,
+      "step": 1574
+    },
+    {
+      "epoch": 7.608695652173913,
+      "grad_norm": 0.05426057055592537,
+      "learning_rate": 8.78185860851156e-05,
+      "loss": 0.0058,
+      "step": 1575
+    },
+    {
+      "epoch": 7.608695652173913,
+      "eval_loss": 0.015573016367852688,
+      "eval_runtime": 20.5607,
+      "eval_samples_per_second": 4.864,
+      "eval_steps_per_second": 0.146,
+      "step": 1575
+    },
+    {
+      "epoch": 7.613526570048309,
+      "grad_norm": 0.07868023961782455,
+      "learning_rate": 8.779651340756647e-05,
+      "loss": 0.0053,
+      "step": 1576
+    },
+    {
+      "epoch": 7.618357487922705,
+      "grad_norm": 0.06837618350982666,
+      "learning_rate": 8.77744235295997e-05,
+      "loss": 0.0044,
+      "step": 1577
+    },
+    {
+      "epoch": 7.6231884057971016,
+      "grad_norm": 0.05811638385057449,
+      "learning_rate": 8.775231646126791e-05,
+      "loss": 0.0053,
+      "step": 1578
+    },
+    {
+      "epoch": 7.628019323671498,
+      "grad_norm": 0.03706006333231926,
+      "learning_rate": 8.77301922126316e-05,
+      "loss": 0.0047,
+      "step": 1579
+    },
+    {
+      "epoch": 7.632850241545894,
+      "grad_norm": 0.044062525033950806,
+      "learning_rate": 8.770805079375902e-05,
+      "loss": 0.0052,
+      "step": 1580
+    },
+    {
+      "epoch": 7.63768115942029,
+      "grad_norm": 0.05137691646814346,
+      "learning_rate": 8.768589221472632e-05,
+      "loss": 0.0051,
+      "step": 1581
+    },
+    {
+      "epoch": 7.642512077294686,
+      "grad_norm": 0.04805343598127365,
+      "learning_rate": 8.766371648561738e-05,
+      "loss": 0.0053,
+      "step": 1582
+    },
+    {
+      "epoch": 7.647342995169082,
+      "grad_norm": 0.046713635325431824,
+      "learning_rate": 8.764152361652392e-05,
+      "loss": 0.0054,
+      "step": 1583
+    },
+    {
+      "epoch": 7.6521739130434785,
+      "grad_norm": 0.04009406641125679,
+      "learning_rate": 8.761931361754547e-05,
+      "loss": 0.005,
+      "step": 1584
+    },
+    {
+      "epoch": 7.657004830917875,
+      "grad_norm": 0.055948395282030106,
+      "learning_rate": 8.759708649878935e-05,
+      "loss": 0.005,
+      "step": 1585
+    },
+    {
+      "epoch": 7.661835748792271,
+      "grad_norm": 0.052176788449287415,
+      "learning_rate": 8.757484227037068e-05,
+      "loss": 0.0061,
+      "step": 1586
+    },
+    {
+      "epoch": 7.666666666666667,
+      "grad_norm": 0.05058382824063301,
+      "learning_rate": 8.75525809424123e-05,
+      "loss": 0.0047,
+      "step": 1587
+    },
+    {
+      "epoch": 7.671497584541063,
+      "grad_norm": 0.03293626755475998,
+      "learning_rate": 8.753030252504494e-05,
+      "loss": 0.0045,
+      "step": 1588
+    },
+    {
+      "epoch": 7.676328502415459,
+      "grad_norm": 0.038541220128536224,
+      "learning_rate": 8.750800702840702e-05,
+      "loss": 0.0045,
+      "step": 1589
+    },
+    {
+      "epoch": 7.681159420289855,
+      "grad_norm": 0.04410076513886452,
+      "learning_rate": 8.748569446264479e-05,
+      "loss": 0.0047,
+      "step": 1590
+    },
+    {
+      "epoch": 7.685990338164252,
+      "grad_norm": 0.050878047943115234,
+      "learning_rate": 8.74633648379122e-05,
+      "loss": 0.0051,
+      "step": 1591
+    },
+    {
+      "epoch": 7.690821256038648,
+      "grad_norm": 0.0634491965174675,
+      "learning_rate": 8.744101816437103e-05,
+      "loss": 0.0058,
+      "step": 1592
+    },
+    {
+      "epoch": 7.695652173913043,
+      "grad_norm": 0.06506568938493729,
+      "learning_rate": 8.741865445219077e-05,
+      "loss": 0.0043,
+      "step": 1593
+    },
+    {
+      "epoch": 7.70048309178744,
+      "grad_norm": 0.043520476669073105,
+      "learning_rate": 8.739627371154872e-05,
+      "loss": 0.0044,
+      "step": 1594
+    },
+    {
+      "epoch": 7.705314009661835,
+      "grad_norm": 0.04149644076824188,
+      "learning_rate": 8.73738759526299e-05,
+      "loss": 0.0045,
+      "step": 1595
+    },
+    {
+      "epoch": 7.710144927536232,
+      "grad_norm": 0.04817346855998039,
+      "learning_rate": 8.735146118562703e-05,
+      "loss": 0.0054,
+      "step": 1596
+    },
+    {
+      "epoch": 7.714975845410628,
+      "grad_norm": 0.07699626684188843,
+      "learning_rate": 8.732902942074061e-05,
+      "loss": 0.0074,
+      "step": 1597
+    },
+    {
+      "epoch": 7.719806763285024,
+      "grad_norm": 0.06528923660516739,
+      "learning_rate": 8.730658066817893e-05,
+      "loss": 0.0059,
+      "step": 1598
+    },
+    {
+      "epoch": 7.72463768115942,
+      "grad_norm": 0.052303560078144073,
+      "learning_rate": 8.728411493815791e-05,
+      "loss": 0.0058,
+      "step": 1599
+    },
+    {
+      "epoch": 7.729468599033816,
+      "grad_norm": 0.05582994595170021,
+      "learning_rate": 8.726163224090125e-05,
+      "loss": 0.0062,
+      "step": 1600
+    },
+    {
+      "epoch": 7.729468599033816,
+      "eval_loss": 0.015776341781020164,
+      "eval_runtime": 20.5969,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
+      "step": 1600
+    },
+    {
+      "epoch": 7.734299516908212,
+      "grad_norm": 0.0504007451236248,
+      "learning_rate": 8.723913258664038e-05,
+      "loss": 0.0061,
+      "step": 1601
+    },
+    {
+      "epoch": 7.739130434782608,
+      "grad_norm": 0.04513921961188316,
+      "learning_rate": 8.721661598561442e-05,
+      "loss": 0.0048,
+      "step": 1602
+    },
+    {
+      "epoch": 7.743961352657005,
+      "grad_norm": 0.050863321870565414,
+      "learning_rate": 8.719408244807021e-05,
+      "loss": 0.0058,
+      "step": 1603
+    },
+    {
+      "epoch": 7.748792270531401,
+      "grad_norm": 0.034483492374420166,
+      "learning_rate": 8.717153198426232e-05,
+      "loss": 0.0046,
+      "step": 1604
+    },
+    {
+      "epoch": 7.753623188405797,
+      "grad_norm": 0.05862283334136009,
+      "learning_rate": 8.714896460445297e-05,
+      "loss": 0.0061,
+      "step": 1605
+    },
+    {
+      "epoch": 7.758454106280193,
+      "grad_norm": 0.041233327239751816,
+      "learning_rate": 8.712638031891212e-05,
+      "loss": 0.0052,
+      "step": 1606
+    },
+    {
+      "epoch": 7.763285024154589,
+      "grad_norm": 0.051079899072647095,
+      "learning_rate": 8.710377913791746e-05,
+      "loss": 0.0062,
+      "step": 1607
+    },
+    {
+      "epoch": 7.768115942028985,
+      "grad_norm": 0.05466492846608162,
+      "learning_rate": 8.708116107175426e-05,
+      "loss": 0.0047,
+      "step": 1608
+    },
+    {
+      "epoch": 7.7729468599033815,
+      "grad_norm": 0.03811094909906387,
+      "learning_rate": 8.705852613071557e-05,
+      "loss": 0.0044,
+      "step": 1609
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 0.06326041370630264,
+      "learning_rate": 8.703587432510207e-05,
+      "loss": 0.0061,
+      "step": 1610
+    },
+    {
+      "epoch": 7.782608695652174,
+      "grad_norm": 0.04274899885058403,
+      "learning_rate": 8.701320566522217e-05,
+      "loss": 0.0038,
+      "step": 1611
+    },
+    {
+      "epoch": 7.78743961352657,
+      "grad_norm": 0.06019265204668045,
+      "learning_rate": 8.699052016139188e-05,
+      "loss": 0.007,
+      "step": 1612
+    },
+    {
+      "epoch": 7.792270531400966,
+      "grad_norm": 0.04409782588481903,
+      "learning_rate": 8.696781782393491e-05,
+      "loss": 0.0044,
+      "step": 1613
+    },
+    {
+      "epoch": 7.797101449275362,
+      "grad_norm": 0.035827215760946274,
+      "learning_rate": 8.694509866318263e-05,
+      "loss": 0.004,
+      "step": 1614
+    },
+    {
+      "epoch": 7.8019323671497585,
+      "grad_norm": 0.04697452858090401,
+      "learning_rate": 8.692236268947408e-05,
+      "loss": 0.0043,
+      "step": 1615
+    },
+    {
+      "epoch": 7.806763285024155,
+      "grad_norm": 0.056868743151426315,
+      "learning_rate": 8.68996099131559e-05,
+      "loss": 0.0059,
+      "step": 1616
+    },
+    {
+      "epoch": 7.811594202898551,
+      "grad_norm": 0.04871121421456337,
+      "learning_rate": 8.687684034458245e-05,
+      "loss": 0.0068,
+      "step": 1617
+    },
+    {
+      "epoch": 7.816425120772947,
+      "grad_norm": 0.04334891214966774,
+      "learning_rate": 8.685405399411568e-05,
+      "loss": 0.0048,
+      "step": 1618
+    },
+    {
+      "epoch": 7.821256038647343,
+      "grad_norm": 0.08146171271800995,
+      "learning_rate": 8.683125087212518e-05,
+      "loss": 0.0048,
+      "step": 1619
+    },
+    {
+      "epoch": 7.826086956521739,
+      "grad_norm": 0.04689481481909752,
+      "learning_rate": 8.680843098898819e-05,
+      "loss": 0.0055,
+      "step": 1620
+    },
+    {
+      "epoch": 7.830917874396135,
+      "grad_norm": 0.0372406505048275,
+      "learning_rate": 8.678559435508958e-05,
+      "loss": 0.0054,
+      "step": 1621
+    },
+    {
+      "epoch": 7.835748792270532,
+      "grad_norm": 0.045953404158353806,
+      "learning_rate": 8.67627409808218e-05,
+      "loss": 0.0042,
+      "step": 1622
+    },
+    {
+      "epoch": 7.840579710144928,
+      "grad_norm": 0.04336479306221008,
+      "learning_rate": 8.6739870876585e-05,
+      "loss": 0.0056,
+      "step": 1623
+    },
+    {
+      "epoch": 7.845410628019324,
+      "grad_norm": 0.055914826691150665,
+      "learning_rate": 8.671698405278685e-05,
+      "loss": 0.0058,
+      "step": 1624
+    },
+    {
+      "epoch": 7.85024154589372,
+      "grad_norm": 0.05467041954398155,
+      "learning_rate": 8.66940805198427e-05,
+      "loss": 0.0045,
+      "step": 1625
+    },
+    {
+      "epoch": 7.85024154589372,
+      "eval_loss": 0.015103131532669067,
+      "eval_runtime": 21.329,
+      "eval_samples_per_second": 4.688,
+      "eval_steps_per_second": 0.141,
+      "step": 1625
+    },
+    {
+      "epoch": 7.855072463768116,
+      "grad_norm": 0.042717594653367996,
+      "learning_rate": 8.667116028817548e-05,
+      "loss": 0.0056,
+      "step": 1626
+    },
+    {
+      "epoch": 7.859903381642512,
+      "grad_norm": 0.0534902848303318,
+      "learning_rate": 8.66482233682157e-05,
+      "loss": 0.0055,
+      "step": 1627
+    },
+    {
+      "epoch": 7.8647342995169085,
+      "grad_norm": 0.040888525545597076,
+      "learning_rate": 8.662526977040148e-05,
+      "loss": 0.0046,
+      "step": 1628
+    },
+    {
+      "epoch": 7.869565217391305,
+      "grad_norm": 0.04883519187569618,
+      "learning_rate": 8.660229950517857e-05,
+      "loss": 0.005,
+      "step": 1629
+    },
+    {
+      "epoch": 7.874396135265701,
+      "grad_norm": 0.039643142372369766,
+      "learning_rate": 8.657931258300023e-05,
+      "loss": 0.0041,
+      "step": 1630
+    },
+    {
+      "epoch": 7.879227053140097,
+      "grad_norm": 0.039371222257614136,
+      "learning_rate": 8.655630901432734e-05,
+      "loss": 0.0048,
+      "step": 1631
+    },
+    {
+      "epoch": 7.884057971014493,
+      "grad_norm": 0.04320335388183594,
+      "learning_rate": 8.653328880962836e-05,
+      "loss": 0.0038,
+      "step": 1632
+    },
+    {
+      "epoch": 7.888888888888889,
+      "grad_norm": 0.06078319251537323,
+      "learning_rate": 8.651025197937931e-05,
+      "loss": 0.0051,
+      "step": 1633
+    },
+    {
+      "epoch": 7.8937198067632846,
+      "grad_norm": 0.05101476237177849,
+      "learning_rate": 8.648719853406377e-05,
+      "loss": 0.0055,
+      "step": 1634
+    },
+    {
+      "epoch": 7.898550724637682,
+      "grad_norm": 0.06241091340780258,
+      "learning_rate": 8.646412848417292e-05,
+      "loss": 0.0052,
+      "step": 1635
+    },
+    {
+      "epoch": 7.903381642512077,
+      "grad_norm": 0.0694604367017746,
+      "learning_rate": 8.644104184020542e-05,
+      "loss": 0.007,
+      "step": 1636
+    },
+    {
+      "epoch": 7.908212560386474,
+      "grad_norm": 0.04248025640845299,
+      "learning_rate": 8.641793861266757e-05,
+      "loss": 0.0054,
+      "step": 1637
+    },
+    {
+      "epoch": 7.913043478260869,
+      "grad_norm": 0.04426901042461395,
+      "learning_rate": 8.639481881207314e-05,
+      "loss": 0.0056,
+      "step": 1638
+    },
+    {
+      "epoch": 7.917874396135265,
+      "grad_norm": 0.05379689857363701,
+      "learning_rate": 8.637168244894351e-05,
+      "loss": 0.0052,
+      "step": 1639
+    },
+    {
+      "epoch": 7.9227053140096615,
+      "grad_norm": 0.05331415683031082,
+      "learning_rate": 8.634852953380756e-05,
+      "loss": 0.0063,
+      "step": 1640
+    },
+    {
+      "epoch": 7.927536231884058,
+      "grad_norm": 0.05386782065033913,
+      "learning_rate": 8.632536007720167e-05,
+      "loss": 0.0059,
+      "step": 1641
+    },
+    {
+      "epoch": 7.932367149758454,
+      "grad_norm": 0.07617385685443878,
+      "learning_rate": 8.630217408966983e-05,
+      "loss": 0.0062,
+      "step": 1642
+    },
+    {
+      "epoch": 7.93719806763285,
+      "grad_norm": 0.05009270831942558,
+      "learning_rate": 8.627897158176346e-05,
+      "loss": 0.0044,
+      "step": 1643
+    },
+    {
+      "epoch": 7.942028985507246,
+      "grad_norm": 0.043669749051332474,
+      "learning_rate": 8.62557525640416e-05,
+      "loss": 0.0054,
+      "step": 1644
+    },
+    {
+      "epoch": 7.946859903381642,
+      "grad_norm": 0.04212132841348648,
+      "learning_rate": 8.623251704707071e-05,
+      "loss": 0.0041,
+      "step": 1645
+    },
+    {
+      "epoch": 7.951690821256038,
+      "grad_norm": 0.04619299992918968,
+      "learning_rate": 8.62092650414248e-05,
+      "loss": 0.0049,
+      "step": 1646
+    },
+    {
+      "epoch": 7.956521739130435,
+      "grad_norm": 0.04855784773826599,
+      "learning_rate": 8.61859965576854e-05,
+      "loss": 0.0051,
+      "step": 1647
+    },
+    {
+      "epoch": 7.961352657004831,
+      "grad_norm": 0.051754970103502274,
+      "learning_rate": 8.616271160644149e-05,
+      "loss": 0.0061,
+      "step": 1648
+    },
+    {
+      "epoch": 7.966183574879227,
+      "grad_norm": 0.04847295209765434,
+      "learning_rate": 8.613941019828961e-05,
+      "loss": 0.0053,
+      "step": 1649
+    },
+    {
+      "epoch": 7.971014492753623,
+      "grad_norm": 0.04961397871375084,
+      "learning_rate": 8.611609234383374e-05,
+      "loss": 0.0054,
+      "step": 1650
+    },
+    {
+      "epoch": 7.971014492753623,
+      "eval_loss": 0.014968699775636196,
+      "eval_runtime": 20.6206,
+      "eval_samples_per_second": 4.85,
+      "eval_steps_per_second": 0.145,
+      "step": 1650
+    },
+    {
+      "epoch": 7.975845410628019,
+      "grad_norm": 0.09579728543758392,
+      "learning_rate": 8.609275805368532e-05,
+      "loss": 0.0063,
+      "step": 1651
+    },
+    {
+      "epoch": 7.980676328502415,
+      "grad_norm": 0.03453390672802925,
+      "learning_rate": 8.606940733846335e-05,
+      "loss": 0.0037,
+      "step": 1652
+    },
+    {
+      "epoch": 7.9855072463768115,
+      "grad_norm": 0.03772159665822983,
+      "learning_rate": 8.604604020879427e-05,
+      "loss": 0.004,
+      "step": 1653
+    },
+    {
+      "epoch": 7.990338164251208,
+      "grad_norm": 0.04900532588362694,
+      "learning_rate": 8.602265667531193e-05,
+      "loss": 0.0064,
+      "step": 1654
+    },
+    {
+      "epoch": 7.995169082125604,
+      "grad_norm": 0.049145378172397614,
+      "learning_rate": 8.599925674865774e-05,
+      "loss": 0.0057,
+      "step": 1655
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.08164945989847183,
+      "learning_rate": 8.597584043948053e-05,
+      "loss": 0.0057,
+      "step": 1656
+    },
+    {
+      "epoch": 8.004830917874395,
+      "grad_norm": 0.05199301242828369,
+      "learning_rate": 8.595240775843653e-05,
+      "loss": 0.004,
+      "step": 1657
+    },
+    {
+      "epoch": 8.009661835748792,
+      "grad_norm": 0.0596538744866848,
+      "learning_rate": 8.592895871618953e-05,
+      "loss": 0.0046,
+      "step": 1658
+    },
+    {
+      "epoch": 8.014492753623188,
+      "grad_norm": 0.0621766597032547,
+      "learning_rate": 8.590549332341068e-05,
+      "loss": 0.0041,
+      "step": 1659
+    },
+    {
+      "epoch": 8.019323671497585,
+      "grad_norm": 0.03637773171067238,
+      "learning_rate": 8.58820115907786e-05,
+      "loss": 0.0032,
+      "step": 1660
+    },
+    {
+      "epoch": 8.02415458937198,
+      "grad_norm": 0.04378081485629082,
+      "learning_rate": 8.585851352897935e-05,
+      "loss": 0.0041,
+      "step": 1661
+    },
+    {
+      "epoch": 8.028985507246377,
+      "grad_norm": 0.04154336825013161,
+      "learning_rate": 8.58349991487064e-05,
+      "loss": 0.0036,
+      "step": 1662
+    },
+    {
+      "epoch": 8.033816425120772,
+      "grad_norm": 0.04366371035575867,
+      "learning_rate": 8.581146846066071e-05,
+      "loss": 0.0044,
+      "step": 1663
+    },
+    {
+      "epoch": 8.03864734299517,
+      "grad_norm": 0.04134861007332802,
+      "learning_rate": 8.578792147555055e-05,
+      "loss": 0.0047,
+      "step": 1664
+    },
+    {
+      "epoch": 8.043478260869565,
+      "grad_norm": 0.051076147705316544,
+      "learning_rate": 8.576435820409171e-05,
+      "loss": 0.0039,
+      "step": 1665
+    },
+    {
+      "epoch": 8.048309178743962,
+      "grad_norm": 0.04088986665010452,
+      "learning_rate": 8.574077865700734e-05,
+      "loss": 0.0044,
+      "step": 1666
+    },
+    {
+      "epoch": 8.053140096618357,
+      "grad_norm": 0.056600384414196014,
+      "learning_rate": 8.5717182845028e-05,
+      "loss": 0.0036,
+      "step": 1667
+    },
+    {
+      "epoch": 8.057971014492754,
+      "grad_norm": 0.05603436380624771,
+      "learning_rate": 8.569357077889167e-05,
+      "loss": 0.0043,
+      "step": 1668
+    },
+    {
+      "epoch": 8.06280193236715,
+      "grad_norm": 0.04432399570941925,
+      "learning_rate": 8.566994246934372e-05,
+      "loss": 0.0041,
+      "step": 1669
+    },
+    {
+      "epoch": 8.067632850241546,
+      "grad_norm": 0.04413706064224243,
+      "learning_rate": 8.564629792713691e-05,
+      "loss": 0.0039,
+      "step": 1670
+    },
+    {
+      "epoch": 8.072463768115941,
+      "grad_norm": 0.04078343138098717,
+      "learning_rate": 8.562263716303138e-05,
+      "loss": 0.0038,
+      "step": 1671
+    },
+    {
+      "epoch": 8.077294685990339,
+      "grad_norm": 0.05788585543632507,
+      "learning_rate": 8.559896018779463e-05,
+      "loss": 0.0044,
+      "step": 1672
+    },
+    {
+      "epoch": 8.082125603864734,
+      "grad_norm": 0.04225607216358185,
+      "learning_rate": 8.557526701220162e-05,
+      "loss": 0.0043,
+      "step": 1673
+    },
+    {
+      "epoch": 8.08695652173913,
+      "grad_norm": 0.0805094763636589,
+      "learning_rate": 8.55515576470346e-05,
+      "loss": 0.0049,
+      "step": 1674
+    },
+    {
+      "epoch": 8.091787439613526,
+      "grad_norm": 0.03907567262649536,
+      "learning_rate": 8.552783210308321e-05,
+      "loss": 0.0042,
+      "step": 1675
+    },
+    {
+      "epoch": 8.091787439613526,
+      "eval_loss": 0.01567777618765831,
+      "eval_runtime": 20.6168,
+      "eval_samples_per_second": 4.85,
+      "eval_steps_per_second": 0.146,
+      "step": 1675
+    },
+    {
+      "epoch": 8.096618357487923,
+      "grad_norm": 0.0445052869617939,
+      "learning_rate": 8.550409039114447e-05,
+      "loss": 0.0036,
+      "step": 1676
+    },
+    {
+      "epoch": 8.101449275362318,
+      "grad_norm": 0.043486401438713074,
+      "learning_rate": 8.548033252202274e-05,
+      "loss": 0.0043,
+      "step": 1677
+    },
+    {
+      "epoch": 8.106280193236715,
+      "grad_norm": 0.03254947438836098,
+      "learning_rate": 8.545655850652972e-05,
+      "loss": 0.0038,
+      "step": 1678
+    },
+    {
+      "epoch": 8.11111111111111,
+      "grad_norm": 0.04751577973365784,
+      "learning_rate": 8.543276835548452e-05,
+      "loss": 0.0036,
+      "step": 1679
+    },
+    {
+      "epoch": 8.115942028985508,
+      "grad_norm": 0.042737722396850586,
+      "learning_rate": 8.540896207971352e-05,
+      "loss": 0.0042,
+      "step": 1680
+    },
+    {
+      "epoch": 8.120772946859903,
+      "grad_norm": 0.03660351410508156,
+      "learning_rate": 8.538513969005047e-05,
+      "loss": 0.0038,
+      "step": 1681
+    },
+    {
+      "epoch": 8.1256038647343,
+      "grad_norm": 0.0444486103951931,
+      "learning_rate": 8.536130119733647e-05,
+      "loss": 0.0042,
+      "step": 1682
+    },
+    {
+      "epoch": 8.130434782608695,
+      "grad_norm": 0.036316633224487305,
+      "learning_rate": 8.53374466124199e-05,
+      "loss": 0.003,
+      "step": 1683
+    },
+    {
+      "epoch": 8.135265700483092,
+      "grad_norm": 0.0428357869386673,
+      "learning_rate": 8.53135759461565e-05,
+      "loss": 0.0036,
+      "step": 1684
+    },
+    {
+      "epoch": 8.140096618357488,
+      "grad_norm": 0.03688013181090355,
+      "learning_rate": 8.528968920940933e-05,
+      "loss": 0.0038,
+      "step": 1685
+    },
+    {
+      "epoch": 8.144927536231885,
+      "grad_norm": 0.041358936578035355,
+      "learning_rate": 8.526578641304874e-05,
+      "loss": 0.0037,
+      "step": 1686
+    },
+    {
+      "epoch": 8.14975845410628,
+      "grad_norm": 0.047944411635398865,
+      "learning_rate": 8.524186756795242e-05,
+      "loss": 0.0036,
+      "step": 1687
+    },
+    {
+      "epoch": 8.154589371980677,
+      "grad_norm": 0.09513163566589355,
+      "learning_rate": 8.521793268500535e-05,
+      "loss": 0.0061,
+      "step": 1688
+    },
+    {
+      "epoch": 8.159420289855072,
+      "grad_norm": 0.025504419580101967,
+      "learning_rate": 8.519398177509978e-05,
+      "loss": 0.0029,
+      "step": 1689
+    },
+    {
+      "epoch": 8.16425120772947,
+      "grad_norm": 0.05210709571838379,
+      "learning_rate": 8.51700148491353e-05,
+      "loss": 0.0034,
+      "step": 1690
+    },
+    {
+      "epoch": 8.169082125603865,
+      "grad_norm": 0.044690266251564026,
+      "learning_rate": 8.514603191801873e-05,
+      "loss": 0.0038,
+      "step": 1691
+    },
+    {
+      "epoch": 8.173913043478262,
+      "grad_norm": 0.04786112904548645,
+      "learning_rate": 8.512203299266425e-05,
+      "loss": 0.0046,
+      "step": 1692
+    },
+    {
+      "epoch": 8.178743961352657,
+      "grad_norm": 0.03577842935919762,
+      "learning_rate": 8.509801808399326e-05,
+      "loss": 0.0038,
+      "step": 1693
+    },
+    {
+      "epoch": 8.183574879227054,
+      "grad_norm": 0.07160219550132751,
+      "learning_rate": 8.507398720293447e-05,
+      "loss": 0.0064,
+      "step": 1694
+    },
+    {
+      "epoch": 8.18840579710145,
+      "grad_norm": 0.07394543290138245,
+      "learning_rate": 8.504994036042381e-05,
+      "loss": 0.0054,
+      "step": 1695
+    },
+    {
+      "epoch": 8.193236714975846,
+      "grad_norm": 0.034880660474300385,
+      "learning_rate": 8.502587756740452e-05,
+      "loss": 0.0034,
+      "step": 1696
+    },
+    {
+      "epoch": 8.198067632850242,
+      "grad_norm": 0.0353291779756546,
+      "learning_rate": 8.500179883482709e-05,
+      "loss": 0.0041,
+      "step": 1697
+    },
+    {
+      "epoch": 8.202898550724637,
+      "grad_norm": 0.03237657994031906,
+      "learning_rate": 8.497770417364925e-05,
+      "loss": 0.0035,
+      "step": 1698
+    },
+    {
+      "epoch": 8.207729468599034,
+      "grad_norm": 0.04432433471083641,
+      "learning_rate": 8.495359359483599e-05,
+      "loss": 0.0057,
+      "step": 1699
+    },
+    {
+      "epoch": 8.21256038647343,
+      "grad_norm": 0.037772729992866516,
+      "learning_rate": 8.492946710935953e-05,
+      "loss": 0.0039,
+      "step": 1700
+    },
+    {
+      "epoch": 8.21256038647343,
+      "eval_loss": 0.015713628381490707,
+      "eval_runtime": 20.6015,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 1700
+    },
+    {
+      "epoch": 8.217391304347826,
+      "grad_norm": 0.03376033529639244,
+      "learning_rate": 8.490532472819937e-05,
+      "loss": 0.0036,
+      "step": 1701
+    },
+    {
+      "epoch": 8.222222222222221,
+      "grad_norm": 0.03267105296254158,
+      "learning_rate": 8.488116646234215e-05,
+      "loss": 0.003,
+      "step": 1702
+    },
+    {
+      "epoch": 8.227053140096618,
+      "grad_norm": 0.029169436544179916,
+      "learning_rate": 8.485699232278186e-05,
+      "loss": 0.0028,
+      "step": 1703
+    },
+    {
+      "epoch": 8.231884057971014,
+      "grad_norm": 0.04096829146146774,
+      "learning_rate": 8.483280232051962e-05,
+      "loss": 0.0048,
+      "step": 1704
+    },
+    {
+      "epoch": 8.23671497584541,
+      "grad_norm": 0.08265417814254761,
+      "learning_rate": 8.480859646656382e-05,
+      "loss": 0.0051,
+      "step": 1705
+    },
+    {
+      "epoch": 8.241545893719806,
+      "grad_norm": 0.043714456260204315,
+      "learning_rate": 8.478437477193006e-05,
+      "loss": 0.0045,
+      "step": 1706
+    },
+    {
+      "epoch": 8.246376811594203,
+      "grad_norm": 0.03147174045443535,
+      "learning_rate": 8.476013724764112e-05,
+      "loss": 0.0033,
+      "step": 1707
+    },
+    {
+      "epoch": 8.251207729468598,
+      "grad_norm": 0.04156283661723137,
+      "learning_rate": 8.4735883904727e-05,
+      "loss": 0.0042,
+      "step": 1708
+    },
+    {
+      "epoch": 8.256038647342995,
+      "grad_norm": 0.05636567994952202,
+      "learning_rate": 8.471161475422487e-05,
+      "loss": 0.0046,
+      "step": 1709
+    },
+    {
+      "epoch": 8.26086956521739,
+      "grad_norm": 0.04402732104063034,
+      "learning_rate": 8.468732980717918e-05,
+      "loss": 0.0042,
+      "step": 1710
+    },
+    {
+      "epoch": 8.265700483091788,
+      "grad_norm": 0.03477048873901367,
+      "learning_rate": 8.466302907464147e-05,
+      "loss": 0.0035,
+      "step": 1711
+    },
+    {
+      "epoch": 8.270531400966183,
+      "grad_norm": 0.0782352164387703,
+      "learning_rate": 8.463871256767053e-05,
+      "loss": 0.0044,
+      "step": 1712
+    },
+    {
+      "epoch": 8.27536231884058,
+      "grad_norm": 0.04401532560586929,
+      "learning_rate": 8.461438029733228e-05,
+      "loss": 0.004,
+      "step": 1713
+    },
+    {
+      "epoch": 8.280193236714975,
+      "grad_norm": 0.04359148442745209,
+      "learning_rate": 8.459003227469985e-05,
+      "loss": 0.0057,
+      "step": 1714
+    },
+    {
+      "epoch": 8.285024154589372,
+      "grad_norm": 0.03307943418622017,
+      "learning_rate": 8.456566851085354e-05,
+      "loss": 0.003,
+      "step": 1715
+    },
+    {
+      "epoch": 8.289855072463768,
+      "grad_norm": 0.04660947248339653,
+      "learning_rate": 8.454128901688077e-05,
+      "loss": 0.0036,
+      "step": 1716
+    },
+    {
+      "epoch": 8.294685990338165,
+      "grad_norm": 0.05288444831967354,
+      "learning_rate": 8.451689380387616e-05,
+      "loss": 0.0047,
+      "step": 1717
+    },
+    {
+      "epoch": 8.29951690821256,
+      "grad_norm": 0.03572872653603554,
+      "learning_rate": 8.449248288294145e-05,
+      "loss": 0.004,
+      "step": 1718
+    },
+    {
+      "epoch": 8.304347826086957,
+      "grad_norm": 0.04837818816304207,
+      "learning_rate": 8.446805626518559e-05,
+      "loss": 0.0047,
+      "step": 1719
+    },
+    {
+      "epoch": 8.309178743961352,
+      "grad_norm": 0.05197962373495102,
+      "learning_rate": 8.444361396172462e-05,
+      "loss": 0.0037,
+      "step": 1720
+    },
+    {
+      "epoch": 8.31400966183575,
+      "grad_norm": 0.04145919159054756,
+      "learning_rate": 8.44191559836817e-05,
+      "loss": 0.0043,
+      "step": 1721
+    },
+    {
+      "epoch": 8.318840579710145,
+      "grad_norm": 0.03252958506345749,
+      "learning_rate": 8.439468234218721e-05,
+      "loss": 0.0038,
+      "step": 1722
+    },
+    {
+      "epoch": 8.323671497584542,
+      "grad_norm": 0.058449309319257736,
+      "learning_rate": 8.437019304837855e-05,
+      "loss": 0.0066,
+      "step": 1723
+    },
+    {
+      "epoch": 8.328502415458937,
+      "grad_norm": 0.041896555572748184,
+      "learning_rate": 8.43456881134003e-05,
+      "loss": 0.004,
+      "step": 1724
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 0.047328148037195206,
+      "learning_rate": 8.43211675484042e-05,
+      "loss": 0.0046,
+      "step": 1725
+    },
+    {
+      "epoch": 8.333333333333334,
+      "eval_loss": 0.016981441527605057,
+      "eval_runtime": 20.6049,
+      "eval_samples_per_second": 4.853,
+      "eval_steps_per_second": 0.146,
+      "step": 1725
+    },
+    {
+      "epoch": 8.33816425120773,
+      "grad_norm": 0.02978409081697464,
+      "learning_rate": 8.4296631364549e-05,
+      "loss": 0.0035,
+      "step": 1726
+    },
+    {
+      "epoch": 8.342995169082126,
+      "grad_norm": 0.043455787003040314,
+      "learning_rate": 8.427207957300062e-05,
+      "loss": 0.0046,
+      "step": 1727
+    },
+    {
+      "epoch": 8.347826086956522,
+      "grad_norm": 0.03001406230032444,
+      "learning_rate": 8.424751218493213e-05,
+      "loss": 0.0029,
+      "step": 1728
+    },
+    {
+      "epoch": 8.352657004830919,
+      "grad_norm": 0.03560645505785942,
+      "learning_rate": 8.422292921152358e-05,
+      "loss": 0.0043,
+      "step": 1729
+    },
+    {
+      "epoch": 8.357487922705314,
+      "grad_norm": 0.052797071635723114,
+      "learning_rate": 8.419833066396222e-05,
+      "loss": 0.0053,
+      "step": 1730
+    },
+    {
+      "epoch": 8.36231884057971,
+      "grad_norm": 0.06399412453174591,
+      "learning_rate": 8.417371655344232e-05,
+      "loss": 0.0062,
+      "step": 1731
+    },
+    {
+      "epoch": 8.367149758454106,
+      "grad_norm": 0.039564114063978195,
+      "learning_rate": 8.414908689116528e-05,
+      "loss": 0.0058,
+      "step": 1732
+    },
+    {
+      "epoch": 8.371980676328503,
+      "grad_norm": 0.04812881350517273,
+      "learning_rate": 8.412444168833953e-05,
+      "loss": 0.0046,
+      "step": 1733
+    },
+    {
+      "epoch": 8.376811594202898,
+      "grad_norm": 0.03506116196513176,
+      "learning_rate": 8.40997809561806e-05,
+      "loss": 0.0032,
+      "step": 1734
+    },
+    {
+      "epoch": 8.381642512077295,
+      "grad_norm": 0.03764081746339798,
+      "learning_rate": 8.40751047059111e-05,
+      "loss": 0.0033,
+      "step": 1735
+    },
+    {
+      "epoch": 8.38647342995169,
+      "grad_norm": 0.03246073052287102,
+      "learning_rate": 8.405041294876066e-05,
+      "loss": 0.0043,
+      "step": 1736
+    },
+    {
+      "epoch": 8.391304347826088,
+      "grad_norm": 0.03924914821982384,
+      "learning_rate": 8.402570569596601e-05,
+      "loss": 0.0043,
+      "step": 1737
+    },
+    {
+      "epoch": 8.396135265700483,
+      "grad_norm": 0.05545099079608917,
+      "learning_rate": 8.400098295877092e-05,
+      "loss": 0.0045,
+      "step": 1738
+    },
+    {
+      "epoch": 8.40096618357488,
+      "grad_norm": 0.04152775928378105,
+      "learning_rate": 8.397624474842617e-05,
+      "loss": 0.0036,
+      "step": 1739
+    },
+    {
+      "epoch": 8.405797101449275,
+      "grad_norm": 0.05080555006861687,
+      "learning_rate": 8.395149107618965e-05,
+      "loss": 0.0031,
+      "step": 1740
+    },
+    {
+      "epoch": 8.41062801932367,
+      "grad_norm": 0.03303583338856697,
+      "learning_rate": 8.392672195332622e-05,
+      "loss": 0.0042,
+      "step": 1741
+    },
+    {
+      "epoch": 8.415458937198068,
+      "grad_norm": 0.0483396053314209,
+      "learning_rate": 8.39019373911078e-05,
+      "loss": 0.004,
+      "step": 1742
+    },
+    {
+      "epoch": 8.420289855072463,
+      "grad_norm": 0.06342915445566177,
+      "learning_rate": 8.387713740081335e-05,
+      "loss": 0.0048,
+      "step": 1743
+    },
+    {
+      "epoch": 8.42512077294686,
+      "grad_norm": 0.034104350954294205,
+      "learning_rate": 8.385232199372885e-05,
+      "loss": 0.0036,
+      "step": 1744
+    },
+    {
+      "epoch": 8.429951690821255,
+      "grad_norm": 0.06683149188756943,
+      "learning_rate": 8.382749118114725e-05,
+      "loss": 0.0051,
+      "step": 1745
+    },
+    {
+      "epoch": 8.434782608695652,
+      "grad_norm": 0.041379477828741074,
+      "learning_rate": 8.380264497436856e-05,
+      "loss": 0.0041,
+      "step": 1746
+    },
+    {
+      "epoch": 8.439613526570048,
+      "grad_norm": 0.044812314212322235,
+      "learning_rate": 8.377778338469977e-05,
+      "loss": 0.0042,
+      "step": 1747
+    },
+    {
+      "epoch": 8.444444444444445,
+      "grad_norm": 0.04295221343636513,
+      "learning_rate": 8.37529064234549e-05,
+      "loss": 0.0035,
+      "step": 1748
+    },
+    {
+      "epoch": 8.44927536231884,
+      "grad_norm": 0.03582591563463211,
+      "learning_rate": 8.372801410195492e-05,
+      "loss": 0.0031,
+      "step": 1749
+    },
+    {
+      "epoch": 8.454106280193237,
+      "grad_norm": 0.030500952154397964,
+      "learning_rate": 8.370310643152782e-05,
+      "loss": 0.0025,
+      "step": 1750
+    },
+    {
+      "epoch": 8.454106280193237,
+      "eval_loss": 0.015351679176092148,
+      "eval_runtime": 20.6217,
+      "eval_samples_per_second": 4.849,
+      "eval_steps_per_second": 0.145,
+      "step": 1750
+    },
+    {
+      "epoch": 8.458937198067632,
+      "grad_norm": 0.08864534646272659,
+      "learning_rate": 8.36781834235086e-05,
+      "loss": 0.0089,
+      "step": 1751
+    },
+    {
+      "epoch": 8.46376811594203,
+      "grad_norm": 0.049155447632074356,
+      "learning_rate": 8.365324508923915e-05,
+      "loss": 0.0038,
+      "step": 1752
+    },
+    {
+      "epoch": 8.468599033816425,
+      "grad_norm": 0.039812907576560974,
+      "learning_rate": 8.362829144006846e-05,
+      "loss": 0.0044,
+      "step": 1753
+    },
+    {
+      "epoch": 8.473429951690822,
+      "grad_norm": 0.038564134389162064,
+      "learning_rate": 8.360332248735237e-05,
+      "loss": 0.0036,
+      "step": 1754
+    },
+    {
+      "epoch": 8.478260869565217,
+      "grad_norm": 0.05134737864136696,
+      "learning_rate": 8.357833824245376e-05,
+      "loss": 0.0045,
+      "step": 1755
+    },
+    {
+      "epoch": 8.483091787439614,
+      "grad_norm": 0.032198935747146606,
+      "learning_rate": 8.355333871674246e-05,
+      "loss": 0.0031,
+      "step": 1756
+    },
+    {
+      "epoch": 8.48792270531401,
+      "grad_norm": 0.03891000896692276,
+      "learning_rate": 8.352832392159523e-05,
+      "loss": 0.0035,
+      "step": 1757
+    },
+    {
+      "epoch": 8.492753623188406,
+      "grad_norm": 0.03442859277129173,
+      "learning_rate": 8.350329386839577e-05,
+      "loss": 0.0037,
+      "step": 1758
+    },
+    {
+      "epoch": 8.497584541062801,
+      "grad_norm": 0.04633500054478645,
+      "learning_rate": 8.347824856853476e-05,
+      "loss": 0.0044,
+      "step": 1759
+    },
+    {
+      "epoch": 8.502415458937199,
+      "grad_norm": 0.08129475265741348,
+      "learning_rate": 8.345318803340982e-05,
+      "loss": 0.0078,
+      "step": 1760
+    },
+    {
+      "epoch": 8.507246376811594,
+      "grad_norm": 0.05236010625958443,
+      "learning_rate": 8.342811227442548e-05,
+      "loss": 0.0068,
+      "step": 1761
+    },
+    {
+      "epoch": 8.51207729468599,
+      "grad_norm": 0.04416520893573761,
+      "learning_rate": 8.340302130299318e-05,
+      "loss": 0.0036,
+      "step": 1762
+    },
+    {
+      "epoch": 8.516908212560386,
+      "grad_norm": 0.04746413603425026,
+      "learning_rate": 8.33779151305313e-05,
+      "loss": 0.0038,
+      "step": 1763
+    },
+    {
+      "epoch": 8.521739130434783,
+      "grad_norm": 0.05367453023791313,
+      "learning_rate": 8.33527937684652e-05,
+      "loss": 0.0044,
+      "step": 1764
+    },
+    {
+      "epoch": 8.526570048309178,
+      "grad_norm": 0.04992407187819481,
+      "learning_rate": 8.332765722822702e-05,
+      "loss": 0.0046,
+      "step": 1765
+    },
+    {
+      "epoch": 8.531400966183575,
+      "grad_norm": 0.04729165509343147,
+      "learning_rate": 8.330250552125594e-05,
+      "loss": 0.0058,
+      "step": 1766
+    },
+    {
+      "epoch": 8.53623188405797,
+      "grad_norm": 0.03023175336420536,
+      "learning_rate": 8.327733865899799e-05,
+      "loss": 0.0035,
+      "step": 1767
+    },
+    {
+      "epoch": 8.541062801932368,
+      "grad_norm": 0.05886807292699814,
+      "learning_rate": 8.325215665290604e-05,
+      "loss": 0.0051,
+      "step": 1768
+    },
+    {
+      "epoch": 8.545893719806763,
+      "grad_norm": 0.055516134947538376,
+      "learning_rate": 8.322695951443995e-05,
+      "loss": 0.0043,
+      "step": 1769
+    },
+    {
+      "epoch": 8.55072463768116,
+      "grad_norm": 0.05037660151720047,
+      "learning_rate": 8.32017472550664e-05,
+      "loss": 0.0043,
+      "step": 1770
+    },
+    {
+      "epoch": 8.555555555555555,
+      "grad_norm": 0.0338997058570385,
+      "learning_rate": 8.317651988625898e-05,
+      "loss": 0.0026,
+      "step": 1771
+    },
+    {
+      "epoch": 8.560386473429952,
+      "grad_norm": 0.04957183822989464,
+      "learning_rate": 8.315127741949816e-05,
+      "loss": 0.0052,
+      "step": 1772
+    },
+    {
+      "epoch": 8.565217391304348,
+      "grad_norm": 0.03905091434717178,
+      "learning_rate": 8.312601986627124e-05,
+      "loss": 0.0034,
+      "step": 1773
+    },
+    {
+      "epoch": 8.570048309178745,
+      "grad_norm": 0.04605364799499512,
+      "learning_rate": 8.310074723807242e-05,
+      "loss": 0.005,
+      "step": 1774
+    },
+    {
+      "epoch": 8.57487922705314,
+      "grad_norm": 0.04077398031949997,
+      "learning_rate": 8.307545954640278e-05,
+      "loss": 0.0047,
+      "step": 1775
+    },
+    {
+      "epoch": 8.57487922705314,
+      "eval_loss": 0.015577097423374653,
+      "eval_runtime": 20.6393,
+      "eval_samples_per_second": 4.845,
+      "eval_steps_per_second": 0.145,
+      "step": 1775
+    },
+    {
+      "epoch": 8.579710144927537,
+      "grad_norm": 0.04953567311167717,
+      "learning_rate": 8.30501568027702e-05,
+      "loss": 0.0045,
+      "step": 1776
+    },
+    {
+      "epoch": 8.584541062801932,
+      "grad_norm": 0.0420534722507,
+      "learning_rate": 8.302483901868947e-05,
+      "loss": 0.0038,
+      "step": 1777
+    },
+    {
+      "epoch": 8.58937198067633,
+      "grad_norm": 0.03566531836986542,
+      "learning_rate": 8.299950620568216e-05,
+      "loss": 0.0044,
+      "step": 1778
+    },
+    {
+      "epoch": 8.594202898550725,
+      "grad_norm": 0.03538971394300461,
+      "learning_rate": 8.297415837527673e-05,
+      "loss": 0.0036,
+      "step": 1779
+    },
+    {
+      "epoch": 8.59903381642512,
+      "grad_norm": 0.05101785808801651,
+      "learning_rate": 8.294879553900847e-05,
+      "loss": 0.0045,
+      "step": 1780
+    },
+    {
+      "epoch": 8.603864734299517,
+      "grad_norm": 0.03426271304488182,
+      "learning_rate": 8.292341770841945e-05,
+      "loss": 0.003,
+      "step": 1781
+    },
+    {
+      "epoch": 8.608695652173914,
+      "grad_norm": 0.06078261137008667,
+      "learning_rate": 8.289802489505865e-05,
+      "loss": 0.0058,
+      "step": 1782
+    },
+    {
+      "epoch": 8.61352657004831,
+      "grad_norm": 0.04777718707919121,
+      "learning_rate": 8.287261711048178e-05,
+      "loss": 0.0047,
+      "step": 1783
+    },
+    {
+      "epoch": 8.618357487922705,
+      "grad_norm": 0.036562662571668625,
+      "learning_rate": 8.284719436625141e-05,
+      "loss": 0.0032,
+      "step": 1784
+    },
+    {
+      "epoch": 8.623188405797102,
+      "grad_norm": 0.037714749574661255,
+      "learning_rate": 8.282175667393691e-05,
+      "loss": 0.0044,
+      "step": 1785
+    },
+    {
+      "epoch": 8.628019323671497,
+      "grad_norm": 0.03402454033493996,
+      "learning_rate": 8.279630404511447e-05,
+      "loss": 0.003,
+      "step": 1786
+    },
+    {
+      "epoch": 8.632850241545894,
+      "grad_norm": 0.03453662991523743,
+      "learning_rate": 8.277083649136703e-05,
+      "loss": 0.0039,
+      "step": 1787
+    },
+    {
+      "epoch": 8.63768115942029,
+      "grad_norm": 0.03423767164349556,
+      "learning_rate": 8.274535402428436e-05,
+      "loss": 0.0041,
+      "step": 1788
+    },
+    {
+      "epoch": 8.642512077294686,
+      "grad_norm": 0.0426657572388649,
+      "learning_rate": 8.271985665546303e-05,
+      "loss": 0.004,
+      "step": 1789
+    },
+    {
+      "epoch": 8.647342995169081,
+      "grad_norm": 0.05961960554122925,
+      "learning_rate": 8.269434439650634e-05,
+      "loss": 0.0051,
+      "step": 1790
+    },
+    {
+      "epoch": 8.652173913043478,
+      "grad_norm": 0.055078715085983276,
+      "learning_rate": 8.266881725902439e-05,
+      "loss": 0.0054,
+      "step": 1791
+    },
+    {
+      "epoch": 8.657004830917874,
+      "grad_norm": 0.03182747960090637,
+      "learning_rate": 8.264327525463408e-05,
+      "loss": 0.0038,
+      "step": 1792
+    },
+    {
+      "epoch": 8.66183574879227,
+      "grad_norm": 0.028650034219026566,
+      "learning_rate": 8.261771839495907e-05,
+      "loss": 0.0037,
+      "step": 1793
+    },
+    {
+      "epoch": 8.666666666666666,
+      "grad_norm": 0.028739208355545998,
+      "learning_rate": 8.259214669162972e-05,
+      "loss": 0.003,
+      "step": 1794
+    },
+    {
+      "epoch": 8.671497584541063,
+      "grad_norm": 0.04689854383468628,
+      "learning_rate": 8.256656015628321e-05,
+      "loss": 0.0036,
+      "step": 1795
+    },
+    {
+      "epoch": 8.676328502415458,
+      "grad_norm": 0.04869701340794563,
+      "learning_rate": 8.254095880056345e-05,
+      "loss": 0.0037,
+      "step": 1796
+    },
+    {
+      "epoch": 8.681159420289855,
+      "grad_norm": 0.04003773257136345,
+      "learning_rate": 8.251534263612106e-05,
+      "loss": 0.0031,
+      "step": 1797
+    },
+    {
+      "epoch": 8.68599033816425,
+      "grad_norm": 0.0333455391228199,
+      "learning_rate": 8.248971167461348e-05,
+      "loss": 0.0037,
+      "step": 1798
+    },
+    {
+      "epoch": 8.690821256038648,
+      "grad_norm": 0.04959764704108238,
+      "learning_rate": 8.24640659277048e-05,
+      "loss": 0.0035,
+      "step": 1799
+    },
+    {
+      "epoch": 8.695652173913043,
+      "grad_norm": 0.046579279005527496,
+      "learning_rate": 8.243840540706588e-05,
+      "loss": 0.0044,
+      "step": 1800
+    },
+    {
+      "epoch": 8.695652173913043,
+      "eval_loss": 0.0165646243840456,
+      "eval_runtime": 20.6036,
+      "eval_samples_per_second": 4.854,
+      "eval_steps_per_second": 0.146,
+      "step": 1800
+    },
+    {
+      "epoch": 8.70048309178744,
+      "grad_norm": 0.038630325347185135,
+      "learning_rate": 8.24127301243743e-05,
+      "loss": 0.0033,
+      "step": 1801
+    },
+    {
+      "epoch": 8.705314009661835,
+      "grad_norm": 0.04081215336918831,
+      "learning_rate": 8.238704009131437e-05,
+      "loss": 0.0033,
+      "step": 1802
+    },
+    {
+      "epoch": 8.710144927536232,
+      "grad_norm": 0.05585318058729172,
+      "learning_rate": 8.236133531957706e-05,
+      "loss": 0.0042,
+      "step": 1803
+    },
+    {
+      "epoch": 8.714975845410628,
+      "grad_norm": 0.05531182512640953,
+      "learning_rate": 8.233561582086012e-05,
+      "loss": 0.0051,
+      "step": 1804
+    },
+    {
+      "epoch": 8.719806763285025,
+      "grad_norm": 0.04835759848356247,
+      "learning_rate": 8.230988160686793e-05,
+      "loss": 0.0032,
+      "step": 1805
+    },
+    {
+      "epoch": 8.72463768115942,
+      "grad_norm": 0.07098240405321121,
+      "learning_rate": 8.228413268931162e-05,
+      "loss": 0.0028,
+      "step": 1806
+    },
+    {
+      "epoch": 8.729468599033817,
+      "grad_norm": 0.06070776656270027,
+      "learning_rate": 8.225836907990902e-05,
+      "loss": 0.0055,
+      "step": 1807
+    },
+    {
+      "epoch": 8.734299516908212,
+      "grad_norm": 0.04384101182222366,
+      "learning_rate": 8.223259079038459e-05,
+      "loss": 0.0038,
+      "step": 1808
+    },
+    {
+      "epoch": 8.73913043478261,
+      "grad_norm": 0.06258191913366318,
+      "learning_rate": 8.22067978324695e-05,
+      "loss": 0.0043,
+      "step": 1809
+    },
+    {
+      "epoch": 8.743961352657005,
+      "grad_norm": 0.07775823026895523,
+      "learning_rate": 8.218099021790162e-05,
+      "loss": 0.0038,
+      "step": 1810
+    },
+    {
+      "epoch": 8.748792270531402,
+      "grad_norm": 0.05379994586110115,
+      "learning_rate": 8.215516795842542e-05,
+      "loss": 0.004,
+      "step": 1811
+    },
+    {
+      "epoch": 8.753623188405797,
+      "grad_norm": 0.04109995439648628,
+      "learning_rate": 8.212933106579213e-05,
+      "loss": 0.004,
+      "step": 1812
+    },
+    {
+      "epoch": 8.758454106280194,
+      "grad_norm": 0.06280244141817093,
+      "learning_rate": 8.210347955175956e-05,
+      "loss": 0.0042,
+      "step": 1813
+    },
+    {
+      "epoch": 8.76328502415459,
+      "grad_norm": 0.04878249764442444,
+      "learning_rate": 8.207761342809224e-05,
+      "loss": 0.0045,
+      "step": 1814
+    },
+    {
+      "epoch": 8.768115942028986,
+      "grad_norm": 0.03682022541761398,
+      "learning_rate": 8.205173270656125e-05,
+      "loss": 0.0041,
+      "step": 1815
+    },
+    {
+      "epoch": 8.772946859903382,
+      "grad_norm": 0.058330096304416656,
+      "learning_rate": 8.202583739894443e-05,
+      "loss": 0.0055,
+      "step": 1816
+    },
+    {
+      "epoch": 8.777777777777779,
+      "grad_norm": 0.051801662892103195,
+      "learning_rate": 8.199992751702618e-05,
+      "loss": 0.0048,
+      "step": 1817
+    },
+    {
+      "epoch": 8.782608695652174,
+      "grad_norm": 0.04357374832034111,
+      "learning_rate": 8.197400307259757e-05,
+      "loss": 0.0048,
+      "step": 1818
+    },
+    {
+      "epoch": 8.78743961352657,
+      "grad_norm": 0.04432510957121849,
+      "learning_rate": 8.194806407745625e-05,
+      "loss": 0.0043,
+      "step": 1819
+    },
+    {
+      "epoch": 8.792270531400966,
+      "grad_norm": 0.05476441606879234,
+      "learning_rate": 8.192211054340657e-05,
+      "loss": 0.0048,
+      "step": 1820
+    },
+    {
+      "epoch": 8.797101449275363,
+      "grad_norm": 0.05567046254873276,
+      "learning_rate": 8.189614248225942e-05,
+      "loss": 0.0055,
+      "step": 1821
+    },
+    {
+      "epoch": 8.801932367149758,
+      "grad_norm": 0.05372453108429909,
+      "learning_rate": 8.187015990583234e-05,
+      "loss": 0.0043,
+      "step": 1822
+    },
+    {
+      "epoch": 8.806763285024154,
+      "grad_norm": 0.03940294682979584,
+      "learning_rate": 8.184416282594946e-05,
+      "loss": 0.0038,
+      "step": 1823
+    },
+    {
+      "epoch": 8.81159420289855,
+      "grad_norm": 0.04525710269808769,
+      "learning_rate": 8.181815125444152e-05,
+      "loss": 0.0047,
+      "step": 1824
+    },
+    {
+      "epoch": 8.816425120772946,
+      "grad_norm": 0.03286072239279747,
+      "learning_rate": 8.179212520314586e-05,
+      "loss": 0.0031,
+      "step": 1825
+    },
+    {
+      "epoch": 8.816425120772946,
+      "eval_loss": 0.017208324745297432,
+      "eval_runtime": 20.6517,
+      "eval_samples_per_second": 4.842,
+      "eval_steps_per_second": 0.145,
+      "step": 1825
+    },
+    {
+      "epoch": 8.821256038647343,
+      "grad_norm": 0.1259443759918213,
+      "learning_rate": 8.17660846839064e-05,
+      "loss": 0.0043,
+      "step": 1826
+    },
+    {
+      "epoch": 8.826086956521738,
+      "grad_norm": 0.0417412668466568,
+      "learning_rate": 8.174002970857364e-05,
+      "loss": 0.004,
+      "step": 1827
+    },
+    {
+      "epoch": 8.830917874396135,
+      "grad_norm": 0.03266125172376633,
+      "learning_rate": 8.171396028900468e-05,
+      "loss": 0.0045,
+      "step": 1828
+    },
+    {
+      "epoch": 8.83574879227053,
+      "grad_norm": 0.05389874801039696,
+      "learning_rate": 8.168787643706315e-05,
+      "loss": 0.0046,
+      "step": 1829
+    },
+    {
+      "epoch": 8.840579710144928,
+      "grad_norm": 0.06979716569185257,
+      "learning_rate": 8.16617781646193e-05,
+      "loss": 0.0036,
+      "step": 1830
+    },
+    {
+      "epoch": 8.845410628019323,
+      "grad_norm": 0.049693189561367035,
+      "learning_rate": 8.163566548354991e-05,
+      "loss": 0.0049,
+      "step": 1831
+    },
+    {
+      "epoch": 8.85024154589372,
+      "grad_norm": 0.05725079029798508,
+      "learning_rate": 8.160953840573834e-05,
+      "loss": 0.0048,
+      "step": 1832
+    },
+    {
+      "epoch": 8.855072463768115,
+      "grad_norm": 0.044530972838401794,
+      "learning_rate": 8.158339694307446e-05,
+      "loss": 0.0044,
+      "step": 1833
+    },
+    {
+      "epoch": 8.859903381642512,
+      "grad_norm": 0.03541207313537598,
+      "learning_rate": 8.155724110745473e-05,
+      "loss": 0.0051,
+      "step": 1834
+    },
+    {
+      "epoch": 8.864734299516908,
+      "grad_norm": 0.058717984706163406,
+      "learning_rate": 8.153107091078212e-05,
+      "loss": 0.0045,
+      "step": 1835
+    },
+    {
+      "epoch": 8.869565217391305,
+      "grad_norm": 0.061354346573352814,
+      "learning_rate": 8.150488636496618e-05,
+      "loss": 0.0047,
+      "step": 1836
+    },
+    {
+      "epoch": 8.8743961352657,
+      "grad_norm": 0.037550557404756546,
+      "learning_rate": 8.14786874819229e-05,
+      "loss": 0.0052,
+      "step": 1837
+    },
+    {
+      "epoch": 8.879227053140097,
+      "grad_norm": 0.045513689517974854,
+      "learning_rate": 8.145247427357494e-05,
+      "loss": 0.0041,
+      "step": 1838
+    },
+    {
+      "epoch": 8.884057971014492,
+      "grad_norm": 0.05298766493797302,
+      "learning_rate": 8.142624675185134e-05,
+      "loss": 0.005,
+      "step": 1839
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.03655264526605606,
+      "learning_rate": 8.140000492868768e-05,
+      "loss": 0.0048,
+      "step": 1840
+    },
+    {
+      "epoch": 8.893719806763285,
+      "grad_norm": 0.044772639870643616,
+      "learning_rate": 8.137374881602614e-05,
+      "loss": 0.0035,
+      "step": 1841
+    },
+    {
+      "epoch": 8.898550724637682,
+      "grad_norm": 0.04357172176241875,
+      "learning_rate": 8.134747842581529e-05,
+      "loss": 0.0057,
+      "step": 1842
+    },
+    {
+      "epoch": 8.903381642512077,
+      "grad_norm": 0.07336288690567017,
+      "learning_rate": 8.132119377001027e-05,
+      "loss": 0.0057,
+      "step": 1843
+    },
+    {
+      "epoch": 8.908212560386474,
+      "grad_norm": 0.04212936386466026,
+      "learning_rate": 8.129489486057267e-05,
+      "loss": 0.0036,
+      "step": 1844
+    },
+    {
+      "epoch": 8.91304347826087,
+      "grad_norm": 0.04403834789991379,
+      "learning_rate": 8.12685817094706e-05,
+      "loss": 0.0034,
+      "step": 1845
+    },
+    {
+      "epoch": 8.917874396135266,
+      "grad_norm": 0.062063757330179214,
+      "learning_rate": 8.124225432867863e-05,
+      "loss": 0.0041,
+      "step": 1846
+    },
+    {
+      "epoch": 8.922705314009661,
+      "grad_norm": 0.05694074556231499,
+      "learning_rate": 8.121591273017782e-05,
+      "loss": 0.0044,
+      "step": 1847
+    },
+    {
+      "epoch": 8.927536231884059,
+      "grad_norm": 0.06622301787137985,
+      "learning_rate": 8.118955692595567e-05,
+      "loss": 0.0048,
+      "step": 1848
+    },
+    {
+      "epoch": 8.932367149758454,
+      "grad_norm": 0.04630272090435028,
+      "learning_rate": 8.116318692800618e-05,
+      "loss": 0.0044,
+      "step": 1849
+    },
+    {
+      "epoch": 8.93719806763285,
+      "grad_norm": 0.040268830955028534,
+      "learning_rate": 8.113680274832978e-05,
+      "loss": 0.0029,
+      "step": 1850
+    },
+    {
+      "epoch": 8.93719806763285,
+      "eval_loss": 0.016717983409762383,
+      "eval_runtime": 20.5962,
+      "eval_samples_per_second": 4.855,
+      "eval_steps_per_second": 0.146,
+      "step": 1850
+    },
+    {
+      "epoch": 8.942028985507246,
+      "grad_norm": 0.038945265114307404,
+      "learning_rate": 8.111040439893338e-05,
+      "loss": 0.0038,
+      "step": 1851
+    },
+    {
+      "epoch": 8.946859903381643,
+      "grad_norm": 0.05721553415060043,
+      "learning_rate": 8.108399189183034e-05,
+      "loss": 0.0038,
+      "step": 1852
+    },
+    {
+      "epoch": 8.951690821256038,
+      "grad_norm": 0.03646910563111305,
+      "learning_rate": 8.105756523904045e-05,
+      "loss": 0.0029,
+      "step": 1853
+    },
+    {
+      "epoch": 8.956521739130435,
+      "grad_norm": 0.04455795884132385,
+      "learning_rate": 8.10311244525899e-05,
+      "loss": 0.0038,
+      "step": 1854
+    },
+    {
+      "epoch": 8.96135265700483,
+      "grad_norm": 0.054944783449172974,
+      "learning_rate": 8.100466954451139e-05,
+      "loss": 0.0044,
+      "step": 1855
+    },
+    {
+      "epoch": 8.966183574879228,
+      "grad_norm": 0.0795806273818016,
+      "learning_rate": 8.097820052684397e-05,
+      "loss": 0.005,
+      "step": 1856
+    },
+    {
+      "epoch": 8.971014492753623,
+      "grad_norm": 0.08476272970438004,
+      "learning_rate": 8.095171741163318e-05,
+      "loss": 0.0043,
+      "step": 1857
+    },
+    {
+      "epoch": 8.97584541062802,
+      "grad_norm": 0.06524492055177689,
+      "learning_rate": 8.092522021093091e-05,
+      "loss": 0.0051,
+      "step": 1858
+    },
+    {
+      "epoch": 8.980676328502415,
+      "grad_norm": 0.03843354806303978,
+      "learning_rate": 8.089870893679554e-05,
+      "loss": 0.0036,
+      "step": 1859
+    },
+    {
+      "epoch": 8.985507246376812,
+      "grad_norm": 0.04924914613366127,
+      "learning_rate": 8.087218360129176e-05,
+      "loss": 0.0047,
+      "step": 1860
+    },
+    {
+      "epoch": 8.990338164251208,
+      "grad_norm": 0.06344044208526611,
+      "learning_rate": 8.084564421649073e-05,
+      "loss": 0.0052,
+      "step": 1861
+    },
+    {
+      "epoch": 8.995169082125603,
+      "grad_norm": 0.05186523124575615,
+      "learning_rate": 8.081909079446998e-05,
+      "loss": 0.0059,
+      "step": 1862
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.04302655905485153,
+      "learning_rate": 8.079252334731342e-05,
+      "loss": 0.0028,
+      "step": 1863
+    },
+    {
+      "epoch": 9.004830917874395,
+      "grad_norm": 0.04055224359035492,
+      "learning_rate": 8.076594188711135e-05,
+      "loss": 0.0032,
+      "step": 1864
+    },
+    {
+      "epoch": 9.009661835748792,
+      "grad_norm": 0.036405082792043686,
+      "learning_rate": 8.073934642596044e-05,
+      "loss": 0.0038,
+      "step": 1865
+    },
+    {
+      "epoch": 9.014492753623188,
+      "grad_norm": 0.043045781552791595,
+      "learning_rate": 8.071273697596374e-05,
+      "loss": 0.0035,
+      "step": 1866
+    },
+    {
+      "epoch": 9.019323671497585,
+      "grad_norm": 0.036786459386348724,
+      "learning_rate": 8.068611354923068e-05,
+      "loss": 0.0036,
+      "step": 1867
+    },
+    {
+      "epoch": 9.02415458937198,
+      "grad_norm": 0.03571231663227081,
+      "learning_rate": 8.065947615787703e-05,
+      "loss": 0.0035,
+      "step": 1868
+    },
+    {
+      "epoch": 9.028985507246377,
+      "grad_norm": 0.043411120772361755,
+      "learning_rate": 8.063282481402491e-05,
+      "loss": 0.0038,
+      "step": 1869
+    },
+    {
+      "epoch": 9.033816425120772,
+      "grad_norm": 0.03864371031522751,
+      "learning_rate": 8.060615952980281e-05,
+      "loss": 0.0044,
+      "step": 1870
+    },
+    {
+      "epoch": 9.03864734299517,
+      "grad_norm": 0.06767242401838303,
+      "learning_rate": 8.057948031734555e-05,
+      "loss": 0.0049,
+      "step": 1871
+    },
+    {
+      "epoch": 9.043478260869565,
+      "grad_norm": 0.03778842091560364,
+      "learning_rate": 8.055278718879429e-05,
+      "loss": 0.0032,
+      "step": 1872
+    },
+    {
+      "epoch": 9.048309178743962,
+      "grad_norm": 0.0641743466258049,
+      "learning_rate": 8.052608015629654e-05,
+      "loss": 0.0072,
+      "step": 1873
+    },
+    {
+      "epoch": 9.053140096618357,
+      "grad_norm": 0.03386566415429115,
+      "learning_rate": 8.049935923200608e-05,
+      "loss": 0.0031,
+      "step": 1874
+    },
+    {
+      "epoch": 9.057971014492754,
+      "grad_norm": 0.03833288326859474,
+      "learning_rate": 8.04726244280831e-05,
+      "loss": 0.0032,
+      "step": 1875
+    },
+    {
+      "epoch": 9.057971014492754,
+      "eval_loss": 0.016934776678681374,
+      "eval_runtime": 20.5938,
+      "eval_samples_per_second": 4.856,
+      "eval_steps_per_second": 0.146,
+      "step": 1875
+    },
+    {
+      "epoch": 9.06280193236715,
+      "grad_norm": 0.03243975341320038,
+      "learning_rate": 8.044587575669406e-05,
+      "loss": 0.0031,
+      "step": 1876
+    },
+    {
+      "epoch": 9.067632850241546,
+      "grad_norm": 0.04720361903309822,
+      "learning_rate": 8.04191132300117e-05,
+      "loss": 0.0043,
+      "step": 1877
+    },
+    {
+      "epoch": 9.072463768115941,
+      "grad_norm": 0.04073271527886391,
+      "learning_rate": 8.039233686021512e-05,
+      "loss": 0.0037,
+      "step": 1878
+    },
+    {
+      "epoch": 9.077294685990339,
+      "grad_norm": 0.03657795116305351,
+      "learning_rate": 8.036554665948968e-05,
+      "loss": 0.0025,
+      "step": 1879
+    },
+    {
+      "epoch": 9.082125603864734,
+      "grad_norm": 0.05789383128285408,
+      "learning_rate": 8.033874264002707e-05,
+      "loss": 0.0037,
+      "step": 1880
+    },
+    {
+      "epoch": 9.08695652173913,
+      "grad_norm": 0.08801286667585373,
+      "learning_rate": 8.031192481402523e-05,
+      "loss": 0.0031,
+      "step": 1881
+    },
+    {
+      "epoch": 9.091787439613526,
+      "grad_norm": 0.039658673107624054,
+      "learning_rate": 8.028509319368842e-05,
+      "loss": 0.0029,
+      "step": 1882
+    },
+    {
+      "epoch": 9.096618357487923,
+      "grad_norm": 0.03780438005924225,
+      "learning_rate": 8.025824779122714e-05,
+      "loss": 0.0032,
+      "step": 1883
+    },
+    {
+      "epoch": 9.101449275362318,
+      "grad_norm": 0.029846245422959328,
+      "learning_rate": 8.02313886188582e-05,
+      "loss": 0.0025,
+      "step": 1884
+    },
+    {
+      "epoch": 9.106280193236715,
+      "grad_norm": 0.03359315171837807,
+      "learning_rate": 8.020451568880465e-05,
+      "loss": 0.0025,
+      "step": 1885
+    },
+    {
+      "epoch": 9.11111111111111,
+      "grad_norm": 0.06273707747459412,
+      "learning_rate": 8.017762901329582e-05,
+      "loss": 0.0037,
+      "step": 1886
+    },
+    {
+      "epoch": 9.115942028985508,
+      "grad_norm": 0.05208655446767807,
+      "learning_rate": 8.015072860456727e-05,
+      "loss": 0.004,
+      "step": 1887
+    },
+    {
+      "epoch": 9.120772946859903,
+      "grad_norm": 0.05353975296020508,
+      "learning_rate": 8.012381447486081e-05,
+      "loss": 0.0034,
+      "step": 1888
+    },
+    {
+      "epoch": 9.1256038647343,
+      "grad_norm": 0.05432965233922005,
+      "learning_rate": 8.009688663642454e-05,
+      "loss": 0.0031,
+      "step": 1889
+    },
+    {
+      "epoch": 9.130434782608695,
+      "grad_norm": 0.04310327395796776,
+      "learning_rate": 8.006994510151275e-05,
+      "loss": 0.0028,
+      "step": 1890
+    },
+    {
+      "epoch": 9.135265700483092,
+      "grad_norm": 0.044074833393096924,
+      "learning_rate": 8.004298988238598e-05,
+      "loss": 0.0031,
+      "step": 1891
+    },
+    {
+      "epoch": 9.140096618357488,
+      "grad_norm": 0.044513918459415436,
+      "learning_rate": 8.001602099131103e-05,
+      "loss": 0.0032,
+      "step": 1892
+    },
+    {
+      "epoch": 9.144927536231885,
+      "grad_norm": 0.051814496517181396,
+      "learning_rate": 7.998903844056083e-05,
+      "loss": 0.0053,
+      "step": 1893
+    },
+    {
+      "epoch": 9.14975845410628,
+      "grad_norm": 0.02605430781841278,
+      "learning_rate": 7.996204224241462e-05,
+      "loss": 0.0028,
+      "step": 1894
+    },
+    {
+      "epoch": 9.154589371980677,
+      "grad_norm": 0.033382847905159,
+      "learning_rate": 7.993503240915781e-05,
+      "loss": 0.0029,
+      "step": 1895
+    },
+    {
+      "epoch": 9.159420289855072,
+      "grad_norm": 0.05876319855451584,
+      "learning_rate": 7.990800895308202e-05,
+      "loss": 0.0045,
+      "step": 1896
+    },
+    {
+      "epoch": 9.16425120772947,
+      "grad_norm": 0.061334945261478424,
+      "learning_rate": 7.988097188648507e-05,
+      "loss": 0.0028,
+      "step": 1897
+    },
+    {
+      "epoch": 9.169082125603865,
+      "grad_norm": 0.06805699318647385,
+      "learning_rate": 7.985392122167098e-05,
+      "loss": 0.0044,
+      "step": 1898
+    },
+    {
+      "epoch": 9.173913043478262,
+      "grad_norm": 0.02976112812757492,
+      "learning_rate": 7.982685697094995e-05,
+      "loss": 0.003,
+      "step": 1899
+    },
+    {
+      "epoch": 9.178743961352657,
+      "grad_norm": 0.0250865388661623,
+      "learning_rate": 7.979977914663836e-05,
+      "loss": 0.0036,
+      "step": 1900
+    },
+    {
+      "epoch": 9.178743961352657,
+      "eval_loss": 0.01666688360273838,
+      "eval_runtime": 20.6487,
+      "eval_samples_per_second": 4.843,
+      "eval_steps_per_second": 0.145,
+      "step": 1900
+    },
+    {
+      "epoch": 9.178743961352657,
+      "step": 1900,
+      "total_flos": 7.164017972028703e+18,
+      "train_loss": 0.021690809104444556,
+      "train_runtime": 50941.7483,
+      "train_samples_per_second": 4.858,
+      "train_steps_per_second": 0.102
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5175,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 25,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 5
+      }
+    },
     "TrainerControl": {
       "args": {
         "should_epoch_stop": false,
@@ -4679,8 +13952,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.0432767520256164e+18,
-  "train_batch_size": 4,
+  "total_flos": 7.164017972028703e+18,
+  "train_batch_size": 48,
   "trial_name": null,
   "trial_params": null
 }