diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,4685 +1,7553 @@
 {
-  "best_metric": 0.014902754686772823,
-  "best_model_checkpoint": "/home/paperspace/Data/models/akoul_whitehorseliquidity_25c/llm3br256/checkpoint-400",
-  "epoch": 5.0,
-  "eval_steps": 5,
-  "global_step": 540,
+  "best_metric": 0.0008496911614201963,
+  "best_model_checkpoint": "/home/paperspace/Data/models/akoul_whitehorseliquidity_25c/llm3br256/checkpoint-900",
+  "epoch": 3.313131313131313,
+  "eval_steps": 25,
+  "global_step": 1025,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.009259259259259259,
-      "grad_norm": 0.29716095328330994,
-      "learning_rate": 1.8518518518518519e-06,
-      "loss": 0.1002,
+      "epoch": 0.0032323232323232323,
+      "grad_norm": 0.03562889248132706,
+      "learning_rate": 1.2936610608020701e-07,
+      "loss": 0.0088,
       "step": 1
     },
     {
-      "epoch": 0.018518518518518517,
-      "grad_norm": 0.2648535370826721,
-      "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.0936,
+      "epoch": 0.006464646464646465,
+      "grad_norm": 0.03539334237575531,
+      "learning_rate": 2.5873221216041403e-07,
+      "loss": 0.0098,
       "step": 2
     },
     {
-      "epoch": 0.027777777777777776,
-      "grad_norm": 0.24819649755954742,
-      "learning_rate": 5.555555555555556e-06,
-      "loss": 0.0898,
+      "epoch": 0.009696969696969697,
+      "grad_norm": 0.03888387605547905,
+      "learning_rate": 3.8809831824062096e-07,
+      "loss": 0.0103,
       "step": 3
     },
     {
-      "epoch": 0.037037037037037035,
-      "grad_norm": 0.23442289233207703,
-      "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.087,
+      "epoch": 0.01292929292929293,
+      "grad_norm": 0.03283314406871796,
+      "learning_rate": 5.174644243208281e-07,
+      "loss": 0.0123,
       "step": 4
     },
     {
-      "epoch": 0.046296296296296294,
-      "grad_norm": 0.26300737261772156,
-      "learning_rate": 9.259259259259259e-06,
-      "loss": 0.0904,
-      "step": 5
-    },
-    {
-      "epoch": 0.046296296296296294,
-      "eval_loss": 0.0950983464717865,
-      "eval_runtime": 11.9584,
-      "eval_samples_per_second": 4.181,
-      "eval_steps_per_second": 1.087,
+      "epoch": 0.01616161616161616,
+      "grad_norm": 0.033392902463674545,
+      "learning_rate": 6.468305304010349e-07,
+      "loss": 0.0104,
       "step": 5
     },
     {
-      "epoch": 0.05555555555555555,
-      "grad_norm": 0.18399731814861298,
-      "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.0805,
+      "epoch": 0.019393939393939394,
+      "grad_norm": 0.041360873728990555,
+      "learning_rate": 7.761966364812419e-07,
+      "loss": 0.0135,
       "step": 6
     },
     {
-      "epoch": 0.06481481481481481,
-      "grad_norm": 0.19827856123447418,
-      "learning_rate": 1.2962962962962962e-05,
-      "loss": 0.0782,
+      "epoch": 0.022626262626262626,
+      "grad_norm": 0.03297152742743492,
+      "learning_rate": 9.055627425614489e-07,
+      "loss": 0.01,
       "step": 7
     },
     {
-      "epoch": 0.07407407407407407,
-      "grad_norm": 0.13050280511379242,
-      "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.0636,
+      "epoch": 0.02585858585858586,
+      "grad_norm": 0.03385542333126068,
+      "learning_rate": 1.0349288486416561e-06,
+      "loss": 0.0092,
       "step": 8
     },
     {
-      "epoch": 0.08333333333333333,
-      "grad_norm": 0.12110771238803864,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.056,
+      "epoch": 0.02909090909090909,
+      "grad_norm": 0.038460493087768555,
+      "learning_rate": 1.164294954721863e-06,
+      "loss": 0.0132,
       "step": 9
     },
     {
-      "epoch": 0.09259259259259259,
-      "grad_norm": 0.1111820638179779,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.053,
-      "step": 10
-    },
-    {
-      "epoch": 0.09259259259259259,
-      "eval_loss": 0.04887561500072479,
-      "eval_runtime": 9.1057,
-      "eval_samples_per_second": 5.491,
-      "eval_steps_per_second": 1.428,
+      "epoch": 0.03232323232323232,
+      "grad_norm": 0.030043406412005424,
+      "learning_rate": 1.2936610608020699e-06,
+      "loss": 0.0088,
       "step": 10
     },
     {
-      "epoch": 0.10185185185185185,
-      "grad_norm": 0.0779903382062912,
-      "learning_rate": 2.037037037037037e-05,
-      "loss": 0.0538,
+      "epoch": 0.035555555555555556,
+      "grad_norm": 0.039398446679115295,
+      "learning_rate": 1.423027166882277e-06,
+      "loss": 0.0118,
       "step": 11
     },
     {
-      "epoch": 0.1111111111111111,
-      "grad_norm": 0.08193033933639526,
-      "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.0398,
+      "epoch": 0.03878787878787879,
+      "grad_norm": 0.03809528797864914,
+      "learning_rate": 1.5523932729624839e-06,
+      "loss": 0.0146,
       "step": 12
     },
     {
-      "epoch": 0.12037037037037036,
-      "grad_norm": 0.0821649506688118,
-      "learning_rate": 2.4074074074074074e-05,
-      "loss": 0.0473,
+      "epoch": 0.04202020202020202,
+      "grad_norm": 0.0388377383351326,
+      "learning_rate": 1.6817593790426907e-06,
+      "loss": 0.0102,
       "step": 13
     },
     {
-      "epoch": 0.12962962962962962,
-      "grad_norm": 0.07107188552618027,
-      "learning_rate": 2.5925925925925925e-05,
-      "loss": 0.0386,
+      "epoch": 0.04525252525252525,
+      "grad_norm": 0.043977439403533936,
+      "learning_rate": 1.8111254851228978e-06,
+      "loss": 0.0099,
       "step": 14
     },
     {
-      "epoch": 0.1388888888888889,
-      "grad_norm": 0.05971238389611244,
-      "learning_rate": 2.777777777777778e-05,
-      "loss": 0.0417,
-      "step": 15
-    },
-    {
-      "epoch": 0.1388888888888889,
-      "eval_loss": 0.04156189784407616,
-      "eval_runtime": 9.1211,
-      "eval_samples_per_second": 5.482,
-      "eval_steps_per_second": 1.425,
+      "epoch": 0.048484848484848485,
+      "grad_norm": 0.036720160394907,
+      "learning_rate": 1.9404915912031045e-06,
+      "loss": 0.0143,
       "step": 15
     },
     {
-      "epoch": 0.14814814814814814,
-      "grad_norm": 0.05262186750769615,
-      "learning_rate": 2.962962962962963e-05,
-      "loss": 0.0384,
+      "epoch": 0.05171717171717172,
+      "grad_norm": 0.03513036668300629,
+      "learning_rate": 2.0698576972833122e-06,
+      "loss": 0.0092,
       "step": 16
     },
     {
-      "epoch": 0.1574074074074074,
-      "grad_norm": 0.05361900106072426,
-      "learning_rate": 3.148148148148148e-05,
-      "loss": 0.0378,
+      "epoch": 0.05494949494949495,
+      "grad_norm": 0.03713906928896904,
+      "learning_rate": 2.199223803363519e-06,
+      "loss": 0.0144,
       "step": 17
     },
     {
-      "epoch": 0.16666666666666666,
-      "grad_norm": 0.05355929210782051,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.0399,
+      "epoch": 0.05818181818181818,
+      "grad_norm": 0.0338338240981102,
+      "learning_rate": 2.328589909443726e-06,
+      "loss": 0.0133,
       "step": 18
     },
     {
-      "epoch": 0.17592592592592593,
-      "grad_norm": 0.04563885182142258,
-      "learning_rate": 3.518518518518519e-05,
-      "loss": 0.0368,
+      "epoch": 0.061414141414141414,
+      "grad_norm": 0.0263836532831192,
+      "learning_rate": 2.457956015523933e-06,
+      "loss": 0.0113,
       "step": 19
     },
     {
-      "epoch": 0.18518518518518517,
-      "grad_norm": 0.060624465346336365,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.0396,
-      "step": 20
-    },
-    {
-      "epoch": 0.18518518518518517,
-      "eval_loss": 0.03584723547101021,
-      "eval_runtime": 9.1162,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.06464646464646465,
+      "grad_norm": 0.01817137934267521,
+      "learning_rate": 2.5873221216041398e-06,
+      "loss": 0.0074,
       "step": 20
     },
     {
-      "epoch": 0.19444444444444445,
-      "grad_norm": 0.0525534488260746,
-      "learning_rate": 3.888888888888889e-05,
-      "loss": 0.0364,
+      "epoch": 0.06787878787878789,
+      "grad_norm": 0.018132060766220093,
+      "learning_rate": 2.716688227684347e-06,
+      "loss": 0.008,
       "step": 21
     },
     {
-      "epoch": 0.2037037037037037,
-      "grad_norm": 0.041657958179712296,
-      "learning_rate": 4.074074074074074e-05,
-      "loss": 0.034,
+      "epoch": 0.07111111111111111,
+      "grad_norm": 0.02117246761918068,
+      "learning_rate": 2.846054333764554e-06,
+      "loss": 0.0091,
       "step": 22
     },
     {
-      "epoch": 0.21296296296296297,
-      "grad_norm": 0.04589791223406792,
-      "learning_rate": 4.259259259259259e-05,
-      "loss": 0.0317,
+      "epoch": 0.07434343434343435,
+      "grad_norm": 0.01761500909924507,
+      "learning_rate": 2.975420439844761e-06,
+      "loss": 0.008,
       "step": 23
     },
     {
-      "epoch": 0.2222222222222222,
-      "grad_norm": 0.04220304638147354,
-      "learning_rate": 4.4444444444444447e-05,
-      "loss": 0.0339,
+      "epoch": 0.07757575757575758,
+      "grad_norm": 0.01908932812511921,
+      "learning_rate": 3.1047865459249677e-06,
+      "loss": 0.008,
       "step": 24
     },
     {
-      "epoch": 0.23148148148148148,
-      "grad_norm": 0.03630352392792702,
-      "learning_rate": 4.62962962962963e-05,
-      "loss": 0.029,
+      "epoch": 0.08080808080808081,
+      "grad_norm": 0.01619753986597061,
+      "learning_rate": 3.234152652005175e-06,
+      "loss": 0.0078,
       "step": 25
     },
     {
-      "epoch": 0.23148148148148148,
-      "eval_loss": 0.03286580368876457,
-      "eval_runtime": 9.1191,
-      "eval_samples_per_second": 5.483,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.08080808080808081,
+      "eval_loss": 0.007939654402434826,
+      "eval_runtime": 20.513,
+      "eval_samples_per_second": 4.875,
+      "eval_steps_per_second": 1.219,
       "step": 25
     },
     {
-      "epoch": 0.24074074074074073,
-      "grad_norm": 0.04235522821545601,
-      "learning_rate": 4.814814814814815e-05,
-      "loss": 0.0326,
+      "epoch": 0.08404040404040404,
+      "grad_norm": 0.01418386958539486,
+      "learning_rate": 3.3635187580853815e-06,
+      "loss": 0.0067,
       "step": 26
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.04675336927175522,
-      "learning_rate": 5e-05,
-      "loss": 0.03,
+      "epoch": 0.08727272727272728,
+      "grad_norm": 0.018947051838040352,
+      "learning_rate": 3.492884864165589e-06,
+      "loss": 0.0086,
       "step": 27
     },
     {
-      "epoch": 0.25925925925925924,
-      "grad_norm": 0.039461418986320496,
-      "learning_rate": 5.185185185185185e-05,
-      "loss": 0.0328,
+      "epoch": 0.0905050505050505,
+      "grad_norm": 0.016921700909733772,
+      "learning_rate": 3.6222509702457957e-06,
+      "loss": 0.0085,
       "step": 28
     },
     {
-      "epoch": 0.26851851851851855,
-      "grad_norm": 0.044042930006980896,
-      "learning_rate": 5.370370370370371e-05,
-      "loss": 0.0294,
+      "epoch": 0.09373737373737374,
+      "grad_norm": 0.01551737543195486,
+      "learning_rate": 3.751617076326003e-06,
+      "loss": 0.0084,
       "step": 29
     },
     {
-      "epoch": 0.2777777777777778,
-      "grad_norm": 0.044502489268779755,
-      "learning_rate": 5.555555555555556e-05,
-      "loss": 0.0311,
-      "step": 30
-    },
-    {
-      "epoch": 0.2777777777777778,
-      "eval_loss": 0.030865700915455818,
-      "eval_runtime": 9.1099,
-      "eval_samples_per_second": 5.489,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.09696969696969697,
+      "grad_norm": 0.014143792912364006,
+      "learning_rate": 3.880983182406209e-06,
+      "loss": 0.0078,
       "step": 30
     },
     {
-      "epoch": 0.28703703703703703,
-      "grad_norm": 0.04979817569255829,
-      "learning_rate": 5.740740740740741e-05,
-      "loss": 0.0292,
+      "epoch": 0.10020202020202021,
+      "grad_norm": 0.011397017166018486,
+      "learning_rate": 4.010349288486417e-06,
+      "loss": 0.0053,
       "step": 31
     },
     {
-      "epoch": 0.2962962962962963,
-      "grad_norm": 0.04573828727006912,
-      "learning_rate": 5.925925925925926e-05,
-      "loss": 0.0346,
+      "epoch": 0.10343434343434343,
+      "grad_norm": 0.015161341056227684,
+      "learning_rate": 4.1397153945666245e-06,
+      "loss": 0.0073,
       "step": 32
     },
     {
-      "epoch": 0.3055555555555556,
-      "grad_norm": 0.0410350002348423,
-      "learning_rate": 6.111111111111112e-05,
-      "loss": 0.0295,
+      "epoch": 0.10666666666666667,
+      "grad_norm": 0.012180107645690441,
+      "learning_rate": 4.2690815006468305e-06,
+      "loss": 0.0063,
       "step": 33
     },
     {
-      "epoch": 0.3148148148148148,
-      "grad_norm": 0.0416686087846756,
-      "learning_rate": 6.296296296296296e-05,
-      "loss": 0.0267,
+      "epoch": 0.1098989898989899,
+      "grad_norm": 0.012770959176123142,
+      "learning_rate": 4.398447606727038e-06,
+      "loss": 0.0079,
       "step": 34
     },
     {
-      "epoch": 0.32407407407407407,
-      "grad_norm": 0.042319901287555695,
-      "learning_rate": 6.481481481481482e-05,
-      "loss": 0.0295,
-      "step": 35
-    },
-    {
-      "epoch": 0.32407407407407407,
-      "eval_loss": 0.028042705729603767,
-      "eval_runtime": 9.1376,
-      "eval_samples_per_second": 5.472,
-      "eval_steps_per_second": 1.423,
+      "epoch": 0.11313131313131314,
+      "grad_norm": 0.014727453701198101,
+      "learning_rate": 4.527813712807244e-06,
+      "loss": 0.0122,
       "step": 35
     },
     {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 0.037845220416784286,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 0.0319,
+      "epoch": 0.11636363636363636,
+      "grad_norm": 0.012778300791978836,
+      "learning_rate": 4.657179818887452e-06,
+      "loss": 0.0071,
       "step": 36
     },
     {
-      "epoch": 0.3425925925925926,
-      "grad_norm": 0.03568718954920769,
-      "learning_rate": 6.851851851851852e-05,
-      "loss": 0.0346,
+      "epoch": 0.1195959595959596,
+      "grad_norm": 0.012685113586485386,
+      "learning_rate": 4.786545924967659e-06,
+      "loss": 0.0065,
       "step": 37
     },
     {
-      "epoch": 0.35185185185185186,
-      "grad_norm": 0.037281136959791183,
-      "learning_rate": 7.037037037037038e-05,
-      "loss": 0.031,
+      "epoch": 0.12282828282828283,
+      "grad_norm": 0.014156874269247055,
+      "learning_rate": 4.915912031047866e-06,
+      "loss": 0.0082,
       "step": 38
     },
     {
-      "epoch": 0.3611111111111111,
-      "grad_norm": 0.03607446327805519,
-      "learning_rate": 7.222222222222222e-05,
-      "loss": 0.0335,
+      "epoch": 0.12606060606060607,
+      "grad_norm": 0.01115910243242979,
+      "learning_rate": 5.045278137128073e-06,
+      "loss": 0.007,
       "step": 39
     },
     {
-      "epoch": 0.37037037037037035,
-      "grad_norm": 0.03654631972312927,
-      "learning_rate": 7.407407407407407e-05,
-      "loss": 0.0262,
-      "step": 40
-    },
-    {
-      "epoch": 0.37037037037037035,
-      "eval_loss": 0.026602942496538162,
-      "eval_runtime": 9.1124,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.1292929292929293,
+      "grad_norm": 0.014089164324104786,
+      "learning_rate": 5.1746442432082795e-06,
+      "loss": 0.0073,
       "step": 40
     },
     {
-      "epoch": 0.37962962962962965,
-      "grad_norm": 0.039490777999162674,
-      "learning_rate": 7.592592592592593e-05,
-      "loss": 0.0252,
+      "epoch": 0.13252525252525252,
+      "grad_norm": 0.013440214097499847,
+      "learning_rate": 5.304010349288486e-06,
+      "loss": 0.0082,
       "step": 41
     },
     {
-      "epoch": 0.3888888888888889,
-      "grad_norm": 0.036680739372968674,
-      "learning_rate": 7.777777777777778e-05,
-      "loss": 0.0242,
+      "epoch": 0.13575757575757577,
+      "grad_norm": 0.01083499938249588,
+      "learning_rate": 5.433376455368694e-06,
+      "loss": 0.006,
       "step": 42
     },
     {
-      "epoch": 0.39814814814814814,
-      "grad_norm": 0.040739599615335464,
-      "learning_rate": 7.962962962962964e-05,
-      "loss": 0.025,
+      "epoch": 0.138989898989899,
+      "grad_norm": 0.011082631535828114,
+      "learning_rate": 5.5627425614489e-06,
+      "loss": 0.007,
       "step": 43
     },
     {
-      "epoch": 0.4074074074074074,
-      "grad_norm": 0.04679260402917862,
-      "learning_rate": 8.148148148148148e-05,
-      "loss": 0.0212,
+      "epoch": 0.14222222222222222,
+      "grad_norm": 0.012490961700677872,
+      "learning_rate": 5.692108667529108e-06,
+      "loss": 0.0062,
       "step": 44
     },
     {
-      "epoch": 0.4166666666666667,
-      "grad_norm": 0.04656214639544487,
-      "learning_rate": 8.333333333333334e-05,
-      "loss": 0.0272,
-      "step": 45
-    },
-    {
-      "epoch": 0.4166666666666667,
-      "eval_loss": 0.02608887106180191,
-      "eval_runtime": 9.1343,
-      "eval_samples_per_second": 5.474,
-      "eval_steps_per_second": 1.423,
+      "epoch": 0.14545454545454545,
+      "grad_norm": 0.012171825394034386,
+      "learning_rate": 5.821474773609315e-06,
+      "loss": 0.0066,
       "step": 45
     },
     {
-      "epoch": 0.42592592592592593,
-      "grad_norm": 0.04525485262274742,
-      "learning_rate": 8.518518518518518e-05,
-      "loss": 0.0274,
+      "epoch": 0.1486868686868687,
+      "grad_norm": 0.010148804634809494,
+      "learning_rate": 5.950840879689522e-06,
+      "loss": 0.0042,
       "step": 46
     },
     {
-      "epoch": 0.4351851851851852,
-      "grad_norm": 0.03210742771625519,
-      "learning_rate": 8.703703703703704e-05,
-      "loss": 0.0283,
+      "epoch": 0.15191919191919193,
+      "grad_norm": 0.013537143357098103,
+      "learning_rate": 6.0802069857697286e-06,
+      "loss": 0.0073,
       "step": 47
     },
     {
-      "epoch": 0.4444444444444444,
-      "grad_norm": 0.03675089031457901,
-      "learning_rate": 8.888888888888889e-05,
-      "loss": 0.0242,
+      "epoch": 0.15515151515151515,
+      "grad_norm": 0.011110929772257805,
+      "learning_rate": 6.2095730918499354e-06,
+      "loss": 0.0072,
       "step": 48
     },
     {
-      "epoch": 0.4537037037037037,
-      "grad_norm": 0.03396710753440857,
-      "learning_rate": 9.074074074074075e-05,
-      "loss": 0.0239,
+      "epoch": 0.15838383838383838,
+      "grad_norm": 0.00937278475612402,
+      "learning_rate": 6.338939197930142e-06,
+      "loss": 0.0054,
       "step": 49
     },
     {
-      "epoch": 0.46296296296296297,
-      "grad_norm": 0.02745971269905567,
-      "learning_rate": 9.25925925925926e-05,
-      "loss": 0.0224,
+      "epoch": 0.16161616161616163,
+      "grad_norm": 0.011824984103441238,
+      "learning_rate": 6.46830530401035e-06,
+      "loss": 0.0119,
       "step": 50
     },
     {
-      "epoch": 0.46296296296296297,
-      "eval_loss": 0.02490057609975338,
-      "eval_runtime": 9.1102,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.16161616161616163,
+      "eval_loss": 0.005149205215275288,
+      "eval_runtime": 18.6062,
+      "eval_samples_per_second": 5.375,
+      "eval_steps_per_second": 1.344,
       "step": 50
     },
     {
-      "epoch": 0.4722222222222222,
-      "grad_norm": 0.04084627702832222,
-      "learning_rate": 9.444444444444444e-05,
-      "loss": 0.0252,
+      "epoch": 0.16484848484848486,
+      "grad_norm": 0.009701536037027836,
+      "learning_rate": 6.597671410090557e-06,
+      "loss": 0.0054,
       "step": 51
     },
     {
-      "epoch": 0.48148148148148145,
-      "grad_norm": 0.033021993935108185,
-      "learning_rate": 9.62962962962963e-05,
-      "loss": 0.0228,
+      "epoch": 0.16808080808080808,
+      "grad_norm": 0.010364706628024578,
+      "learning_rate": 6.727037516170763e-06,
+      "loss": 0.0049,
       "step": 52
     },
     {
-      "epoch": 0.49074074074074076,
-      "grad_norm": 0.034785784780979156,
-      "learning_rate": 9.814814814814815e-05,
-      "loss": 0.0259,
+      "epoch": 0.1713131313131313,
+      "grad_norm": 0.009346776641905308,
+      "learning_rate": 6.856403622250971e-06,
+      "loss": 0.0054,
       "step": 53
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.03407888114452362,
-      "learning_rate": 0.0001,
-      "loss": 0.0239,
+      "epoch": 0.17454545454545456,
+      "grad_norm": 0.008815059438347816,
+      "learning_rate": 6.985769728331178e-06,
+      "loss": 0.0069,
       "step": 54
     },
     {
-      "epoch": 0.5092592592592593,
-      "grad_norm": 0.03268973529338837,
-      "learning_rate": 9.99989553622803e-05,
-      "loss": 0.0229,
-      "step": 55
-    },
-    {
-      "epoch": 0.5092592592592593,
-      "eval_loss": 0.02450372651219368,
-      "eval_runtime": 9.1421,
-      "eval_samples_per_second": 5.469,
-      "eval_steps_per_second": 1.422,
+      "epoch": 0.17777777777777778,
+      "grad_norm": 0.009431697428226471,
+      "learning_rate": 7.115135834411385e-06,
+      "loss": 0.0083,
       "step": 55
     },
     {
-      "epoch": 0.5185185185185185,
-      "grad_norm": 0.032378531992435455,
-      "learning_rate": 9.999582149277187e-05,
-      "loss": 0.0219,
+      "epoch": 0.181010101010101,
+      "grad_norm": 0.012181616388261318,
+      "learning_rate": 7.244501940491591e-06,
+      "loss": 0.0065,
       "step": 56
     },
     {
-      "epoch": 0.5277777777777778,
-      "grad_norm": 0.03997437283396721,
-      "learning_rate": 9.999059852242507e-05,
-      "loss": 0.0248,
+      "epoch": 0.18424242424242424,
+      "grad_norm": 0.009227719157934189,
+      "learning_rate": 7.373868046571798e-06,
+      "loss": 0.0052,
       "step": 57
     },
     {
-      "epoch": 0.5370370370370371,
-      "grad_norm": 0.04024836793541908,
-      "learning_rate": 9.998328666948438e-05,
-      "loss": 0.0194,
+      "epoch": 0.1874747474747475,
+      "grad_norm": 0.009855546988546848,
+      "learning_rate": 7.503234152652006e-06,
+      "loss": 0.0082,
       "step": 58
     },
     {
-      "epoch": 0.5462962962962963,
-      "grad_norm": 0.03850249573588371,
-      "learning_rate": 9.997388623947928e-05,
-      "loss": 0.0251,
+      "epoch": 0.1907070707070707,
+      "grad_norm": 0.008454745635390282,
+      "learning_rate": 7.632600258732213e-06,
+      "loss": 0.0044,
       "step": 59
     },
     {
-      "epoch": 0.5555555555555556,
-      "grad_norm": 0.03326913341879845,
-      "learning_rate": 9.996239762521151e-05,
-      "loss": 0.0233,
-      "step": 60
-    },
-    {
-      "epoch": 0.5555555555555556,
-      "eval_loss": 0.023316912353038788,
-      "eval_runtime": 9.1353,
-      "eval_samples_per_second": 5.473,
-      "eval_steps_per_second": 1.423,
+      "epoch": 0.19393939393939394,
+      "grad_norm": 0.010128876194357872,
+      "learning_rate": 7.761966364812418e-06,
+      "loss": 0.0092,
       "step": 60
     },
     {
-      "epoch": 0.5648148148148148,
-      "grad_norm": 0.034179024398326874,
-      "learning_rate": 9.994882130673868e-05,
-      "loss": 0.0222,
+      "epoch": 0.19717171717171716,
+      "grad_norm": 0.009267722256481647,
+      "learning_rate": 7.891332470892627e-06,
+      "loss": 0.0049,
       "step": 61
     },
     {
-      "epoch": 0.5740740740740741,
-      "grad_norm": 0.031797800213098526,
-      "learning_rate": 9.993315785135416e-05,
-      "loss": 0.0272,
+      "epoch": 0.20040404040404042,
+      "grad_norm": 0.009394655004143715,
+      "learning_rate": 8.020698576972833e-06,
+      "loss": 0.0059,
       "step": 62
     },
     {
-      "epoch": 0.5833333333333334,
-      "grad_norm": 0.03183833882212639,
-      "learning_rate": 9.991540791356342e-05,
-      "loss": 0.0241,
+      "epoch": 0.20363636363636364,
+      "grad_norm": 0.008983040228486061,
+      "learning_rate": 8.15006468305304e-06,
+      "loss": 0.0041,
       "step": 63
     },
     {
-      "epoch": 0.5925925925925926,
-      "grad_norm": 0.025173548609018326,
-      "learning_rate": 9.989557223505661e-05,
-      "loss": 0.0216,
+      "epoch": 0.20686868686868687,
+      "grad_norm": 0.009920783340930939,
+      "learning_rate": 8.279430789133249e-06,
+      "loss": 0.0058,
       "step": 64
     },
     {
-      "epoch": 0.6018518518518519,
-      "grad_norm": 0.04935009032487869,
-      "learning_rate": 9.987365164467767e-05,
-      "loss": 0.0217,
-      "step": 65
-    },
-    {
-      "epoch": 0.6018518518518519,
-      "eval_loss": 0.02255990356206894,
-      "eval_runtime": 9.1207,
-      "eval_samples_per_second": 5.482,
-      "eval_steps_per_second": 1.425,
+      "epoch": 0.2101010101010101,
+      "grad_norm": 0.007694128435105085,
+      "learning_rate": 8.408796895213454e-06,
+      "loss": 0.0033,
       "step": 65
     },
     {
-      "epoch": 0.6111111111111112,
-      "grad_norm": 0.02904060110449791,
-      "learning_rate": 9.98496470583896e-05,
-      "loss": 0.0213,
+      "epoch": 0.21333333333333335,
+      "grad_norm": 0.01226822566241026,
+      "learning_rate": 8.538163001293661e-06,
+      "loss": 0.0099,
       "step": 66
     },
     {
-      "epoch": 0.6203703703703703,
-      "grad_norm": 0.046014755964279175,
-      "learning_rate": 9.982355947923629e-05,
-      "loss": 0.018,
+      "epoch": 0.21656565656565657,
+      "grad_norm": 0.010597337037324905,
+      "learning_rate": 8.66752910737387e-06,
+      "loss": 0.0081,
       "step": 67
     },
     {
-      "epoch": 0.6296296296296297,
-      "grad_norm": 0.0354795977473259,
-      "learning_rate": 9.979538999730047e-05,
-      "loss": 0.0199,
+      "epoch": 0.2197979797979798,
+      "grad_norm": 0.009066218510270119,
+      "learning_rate": 8.796895213454076e-06,
+      "loss": 0.0044,
       "step": 68
     },
     {
-      "epoch": 0.6388888888888888,
-      "grad_norm": 0.03308796137571335,
-      "learning_rate": 9.976513978965829e-05,
-      "loss": 0.0239,
+      "epoch": 0.22303030303030302,
+      "grad_norm": 0.01163114607334137,
+      "learning_rate": 8.926261319534282e-06,
+      "loss": 0.0129,
       "step": 69
     },
     {
-      "epoch": 0.6481481481481481,
-      "grad_norm": 0.03860899433493614,
-      "learning_rate": 9.973281012033007e-05,
-      "loss": 0.0247,
-      "step": 70
-    },
-    {
-      "epoch": 0.6481481481481481,
-      "eval_loss": 0.022898558527231216,
-      "eval_runtime": 9.1074,
-      "eval_samples_per_second": 5.49,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.22626262626262628,
+      "grad_norm": 0.012832598760724068,
+      "learning_rate": 9.055627425614489e-06,
+      "loss": 0.0071,
       "step": 70
     },
     {
-      "epoch": 0.6574074074074074,
-      "grad_norm": 0.028213078156113625,
-      "learning_rate": 9.969840234022749e-05,
-      "loss": 0.0197,
+      "epoch": 0.2294949494949495,
+      "grad_norm": 0.010620299726724625,
+      "learning_rate": 9.184993531694697e-06,
+      "loss": 0.004,
       "step": 71
     },
     {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 0.024581043049693108,
-      "learning_rate": 9.966191788709716e-05,
-      "loss": 0.0207,
+      "epoch": 0.23272727272727273,
+      "grad_norm": 0.00913357175886631,
+      "learning_rate": 9.314359637774904e-06,
+      "loss": 0.0058,
       "step": 72
     },
     {
-      "epoch": 0.6759259259259259,
-      "grad_norm": 0.026658454909920692,
-      "learning_rate": 9.962335828546048e-05,
-      "loss": 0.0214,
+      "epoch": 0.23595959595959595,
+      "grad_norm": 0.011772734113037586,
+      "learning_rate": 9.44372574385511e-06,
+      "loss": 0.0036,
       "step": 73
     },
     {
-      "epoch": 0.6851851851851852,
-      "grad_norm": 0.034941576421260834,
-      "learning_rate": 9.958272514655006e-05,
-      "loss": 0.0205,
+      "epoch": 0.2391919191919192,
+      "grad_norm": 0.011326112784445286,
+      "learning_rate": 9.573091849935318e-06,
+      "loss": 0.0071,
       "step": 74
     },
     {
-      "epoch": 0.6944444444444444,
-      "grad_norm": 0.03060038387775421,
-      "learning_rate": 9.954002016824227e-05,
-      "loss": 0.0193,
+      "epoch": 0.24242424242424243,
+      "grad_norm": 0.010667567141354084,
+      "learning_rate": 9.702457956015525e-06,
+      "loss": 0.0036,
       "step": 75
     },
     {
-      "epoch": 0.6944444444444444,
-      "eval_loss": 0.02283317781984806,
-      "eval_runtime": 9.1512,
-      "eval_samples_per_second": 5.464,
-      "eval_steps_per_second": 1.421,
+      "epoch": 0.24242424242424243,
+      "eval_loss": 0.0031650287564843893,
+      "eval_runtime": 18.6391,
+      "eval_samples_per_second": 5.365,
+      "eval_steps_per_second": 1.341,
       "step": 75
     },
     {
-      "epoch": 0.7037037037037037,
-      "grad_norm": 0.0313015952706337,
-      "learning_rate": 9.949524513498636e-05,
-      "loss": 0.0206,
+      "epoch": 0.24565656565656566,
+      "grad_norm": 0.007108451332896948,
+      "learning_rate": 9.831824062095732e-06,
+      "loss": 0.0023,
       "step": 76
     },
     {
-      "epoch": 0.7129629629629629,
-      "grad_norm": 0.03317766636610031,
-      "learning_rate": 9.944840191772987e-05,
-      "loss": 0.0217,
+      "epoch": 0.24888888888888888,
+      "grad_norm": 0.008278781548142433,
+      "learning_rate": 9.961190168175938e-06,
+      "loss": 0.0026,
       "step": 77
     },
     {
-      "epoch": 0.7222222222222222,
-      "grad_norm": 0.027911782264709473,
-      "learning_rate": 9.939949247384046e-05,
-      "loss": 0.0196,
+      "epoch": 0.25212121212121213,
+      "grad_norm": 0.010979422368109226,
+      "learning_rate": 1.0090556274256145e-05,
+      "loss": 0.0036,
       "step": 78
     },
     {
-      "epoch": 0.7314814814814815,
-      "grad_norm": 0.028807291761040688,
-      "learning_rate": 9.934851884702414e-05,
-      "loss": 0.0223,
+      "epoch": 0.25535353535353533,
+      "grad_norm": 0.007666402496397495,
+      "learning_rate": 1.0219922380336352e-05,
+      "loss": 0.0015,
       "step": 79
     },
     {
-      "epoch": 0.7407407407407407,
-      "grad_norm": 0.03152855485677719,
-      "learning_rate": 9.929548316723982e-05,
-      "loss": 0.0173,
-      "step": 80
-    },
-    {
-      "epoch": 0.7407407407407407,
-      "eval_loss": 0.021335698664188385,
-      "eval_runtime": 9.1689,
-      "eval_samples_per_second": 5.453,
-      "eval_steps_per_second": 1.418,
+      "epoch": 0.2585858585858586,
+      "grad_norm": 0.009133332408964634,
+      "learning_rate": 1.0349288486416559e-05,
+      "loss": 0.0021,
       "step": 80
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.03250882402062416,
-      "learning_rate": 9.924038765061042e-05,
-      "loss": 0.0231,
+      "epoch": 0.26181818181818184,
+      "grad_norm": 0.009566927328705788,
+      "learning_rate": 1.0478654592496766e-05,
+      "loss": 0.0038,
       "step": 81
     },
     {
-      "epoch": 0.7592592592592593,
-      "grad_norm": 0.030853938311338425,
-      "learning_rate": 9.918323459933005e-05,
-      "loss": 0.0224,
+      "epoch": 0.26505050505050504,
+      "grad_norm": 0.011765814386308193,
+      "learning_rate": 1.0608020698576973e-05,
+      "loss": 0.0061,
       "step": 82
     },
     {
-      "epoch": 0.7685185185185185,
-      "grad_norm": 0.03431202098727226,
-      "learning_rate": 9.912402640156811e-05,
-      "loss": 0.0223,
+      "epoch": 0.2682828282828283,
+      "grad_norm": 0.009488740935921669,
+      "learning_rate": 1.073738680465718e-05,
+      "loss": 0.0023,
       "step": 83
     },
     {
-      "epoch": 0.7777777777777778,
-      "grad_norm": 0.027050426229834557,
-      "learning_rate": 9.906276553136923e-05,
-      "loss": 0.0198,
+      "epoch": 0.27151515151515154,
+      "grad_norm": 0.007137450389564037,
+      "learning_rate": 1.0866752910737388e-05,
+      "loss": 0.0021,
       "step": 84
     },
     {
-      "epoch": 0.7870370370370371,
-      "grad_norm": 0.03224191442131996,
-      "learning_rate": 9.899945454855006e-05,
-      "loss": 0.0207,
+      "epoch": 0.27474747474747474,
+      "grad_norm": 0.008984182961285114,
+      "learning_rate": 1.0996119016817593e-05,
+      "loss": 0.0028,
       "step": 85
     },
     {
-      "epoch": 0.7870370370370371,
-      "eval_loss": 0.020375357940793037,
-      "eval_runtime": 9.1362,
-      "eval_samples_per_second": 5.473,
-      "eval_steps_per_second": 1.423,
-      "step": 85
-    },
-    {
-      "epoch": 0.7962962962962963,
-      "grad_norm": 0.028706278651952744,
-      "learning_rate": 9.893409609859222e-05,
-      "loss": 0.0197,
+      "epoch": 0.277979797979798,
+      "grad_norm": 0.010403821244835854,
+      "learning_rate": 1.11254851228978e-05,
+      "loss": 0.0057,
       "step": 86
     },
     {
-      "epoch": 0.8055555555555556,
-      "grad_norm": 0.02814578451216221,
-      "learning_rate": 9.88666929125318e-05,
-      "loss": 0.0199,
+      "epoch": 0.2812121212121212,
+      "grad_norm": 0.0071556540206074715,
+      "learning_rate": 1.1254851228978009e-05,
+      "loss": 0.0017,
       "step": 87
     },
     {
-      "epoch": 0.8148148148148148,
-      "grad_norm": 0.028775395825505257,
-      "learning_rate": 9.879724780684519e-05,
-      "loss": 0.0169,
+      "epoch": 0.28444444444444444,
+      "grad_norm": 0.012338937260210514,
+      "learning_rate": 1.1384217335058216e-05,
+      "loss": 0.0029,
       "step": 88
     },
     {
-      "epoch": 0.8240740740740741,
-      "grad_norm": 0.030078047886490822,
-      "learning_rate": 9.872576368333151e-05,
-      "loss": 0.0209,
+      "epoch": 0.2876767676767677,
+      "grad_norm": 0.010343515314161777,
+      "learning_rate": 1.1513583441138421e-05,
+      "loss": 0.0032,
       "step": 89
     },
     {
-      "epoch": 0.8333333333333334,
-      "grad_norm": 0.031860969960689545,
-      "learning_rate": 9.865224352899119e-05,
-      "loss": 0.0213,
-      "step": 90
-    },
-    {
-      "epoch": 0.8333333333333334,
-      "eval_loss": 0.019939038902521133,
-      "eval_runtime": 9.1287,
-      "eval_samples_per_second": 5.477,
-      "eval_steps_per_second": 1.424,
+      "epoch": 0.2909090909090909,
+      "grad_norm": 0.009991390630602837,
+      "learning_rate": 1.164294954721863e-05,
+      "loss": 0.0035,
       "step": 90
     },
     {
-      "epoch": 0.8425925925925926,
-      "grad_norm": 0.03415157273411751,
-      "learning_rate": 9.857669041590134e-05,
-      "loss": 0.021,
+      "epoch": 0.29414141414141415,
+      "grad_norm": 0.009020160883665085,
+      "learning_rate": 1.1772315653298836e-05,
+      "loss": 0.0033,
       "step": 91
     },
     {
-      "epoch": 0.8518518518518519,
-      "grad_norm": 0.032674115151166916,
-      "learning_rate": 9.849910750108717e-05,
-      "loss": 0.0207,
+      "epoch": 0.2973737373737374,
+      "grad_norm": 0.0067694829776883125,
+      "learning_rate": 1.1901681759379043e-05,
+      "loss": 0.0019,
       "step": 92
     },
     {
-      "epoch": 0.8611111111111112,
-      "grad_norm": 0.02941475249826908,
-      "learning_rate": 9.84194980263903e-05,
-      "loss": 0.0196,
+      "epoch": 0.3006060606060606,
+      "grad_norm": 0.013036763295531273,
+      "learning_rate": 1.203104786545925e-05,
+      "loss": 0.0041,
       "step": 93
     },
     {
-      "epoch": 0.8703703703703703,
-      "grad_norm": 0.036115583032369614,
-      "learning_rate": 9.83378653183331e-05,
-      "loss": 0.0178,
+      "epoch": 0.30383838383838385,
+      "grad_norm": 0.009171461686491966,
+      "learning_rate": 1.2160413971539457e-05,
+      "loss": 0.006,
       "step": 94
     },
     {
-      "epoch": 0.8796296296296297,
-      "grad_norm": 0.03358744457364082,
-      "learning_rate": 9.825421278797983e-05,
-      "loss": 0.0199,
-      "step": 95
-    },
-    {
-      "epoch": 0.8796296296296297,
-      "eval_loss": 0.020193172618746758,
-      "eval_runtime": 9.1141,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.30707070707070705,
+      "grad_norm": 0.007229423616081476,
+      "learning_rate": 1.2289780077619664e-05,
+      "loss": 0.002,
       "step": 95
     },
     {
-      "epoch": 0.8888888888888888,
-      "grad_norm": 0.029014358296990395,
-      "learning_rate": 9.816854393079403e-05,
-      "loss": 0.0219,
+      "epoch": 0.3103030303030303,
+      "grad_norm": 0.009065698832273483,
+      "learning_rate": 1.2419146183699871e-05,
+      "loss": 0.0028,
       "step": 96
     },
     {
-      "epoch": 0.8981481481481481,
-      "grad_norm": 0.042931754142045975,
-      "learning_rate": 9.808086232649246e-05,
-      "loss": 0.0185,
+      "epoch": 0.31353535353535356,
+      "grad_norm": 0.0068460931070148945,
+      "learning_rate": 1.254851228978008e-05,
+      "loss": 0.0024,
       "step": 97
     },
     {
-      "epoch": 0.9074074074074074,
-      "grad_norm": 0.029089825227856636,
-      "learning_rate": 9.799117163889559e-05,
-      "loss": 0.021,
+      "epoch": 0.31676767676767675,
+      "grad_norm": 0.12772586941719055,
+      "learning_rate": 1.2677878395860285e-05,
+      "loss": 0.0046,
       "step": 98
     },
     {
-      "epoch": 0.9166666666666666,
-      "grad_norm": 0.03154176101088524,
-      "learning_rate": 9.789947561577445e-05,
-      "loss": 0.02,
+      "epoch": 0.32,
+      "grad_norm": 0.008569791913032532,
+      "learning_rate": 1.2807244501940493e-05,
+      "loss": 0.0024,
       "step": 99
     },
     {
-      "epoch": 0.9259259259259259,
-      "grad_norm": 0.027786221355199814,
-      "learning_rate": 9.780577808869398e-05,
-      "loss": 0.0188,
+      "epoch": 0.32323232323232326,
+      "grad_norm": 0.010226168669760227,
+      "learning_rate": 1.29366106080207e-05,
+      "loss": 0.004,
       "step": 100
     },
     {
-      "epoch": 0.9259259259259259,
-      "eval_loss": 0.02070247381925583,
-      "eval_runtime": 9.1159,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.32323232323232326,
+      "eval_loss": 0.002474932000041008,
+      "eval_runtime": 18.5931,
+      "eval_samples_per_second": 5.378,
+      "eval_steps_per_second": 1.345,
       "step": 100
     },
     {
-      "epoch": 0.9351851851851852,
-      "grad_norm": 0.030518539249897003,
-      "learning_rate": 9.771008297285307e-05,
-      "loss": 0.0218,
+      "epoch": 0.32646464646464646,
+      "grad_norm": 0.010962710715830326,
+      "learning_rate": 1.3065976714100905e-05,
+      "loss": 0.0018,
       "step": 101
     },
     {
-      "epoch": 0.9444444444444444,
-      "grad_norm": 0.024817178025841713,
-      "learning_rate": 9.761239426692077e-05,
-      "loss": 0.0202,
+      "epoch": 0.3296969696969697,
+      "grad_norm": 0.01257998775690794,
+      "learning_rate": 1.3195342820181114e-05,
+      "loss": 0.0033,
       "step": 102
     },
     {
-      "epoch": 0.9537037037037037,
-      "grad_norm": 0.025192229077219963,
-      "learning_rate": 9.751271605286941e-05,
-      "loss": 0.0197,
+      "epoch": 0.3329292929292929,
+      "grad_norm": 0.013269903138279915,
+      "learning_rate": 1.332470892626132e-05,
+      "loss": 0.0036,
       "step": 103
     },
     {
-      "epoch": 0.9629629629629629,
-      "grad_norm": 0.02538897655904293,
-      "learning_rate": 9.741105249580383e-05,
-      "loss": 0.02,
+      "epoch": 0.33616161616161616,
+      "grad_norm": 0.009509469382464886,
+      "learning_rate": 1.3454075032341526e-05,
+      "loss": 0.0034,
       "step": 104
     },
     {
-      "epoch": 0.9722222222222222,
-      "grad_norm": 0.025440450757741928,
-      "learning_rate": 9.730740784378753e-05,
-      "loss": 0.0193,
-      "step": 105
-    },
-    {
-      "epoch": 0.9722222222222222,
-      "eval_loss": 0.020300446078181267,
-      "eval_runtime": 9.126,
-      "eval_samples_per_second": 5.479,
-      "eval_steps_per_second": 1.425,
+      "epoch": 0.3393939393939394,
+      "grad_norm": 0.00902635883539915,
+      "learning_rate": 1.3583441138421735e-05,
+      "loss": 0.0028,
       "step": 105
     },
     {
-      "epoch": 0.9814814814814815,
-      "grad_norm": 0.02362542785704136,
-      "learning_rate": 9.7201786427665e-05,
-      "loss": 0.0202,
+      "epoch": 0.3426262626262626,
+      "grad_norm": 0.007760872133076191,
+      "learning_rate": 1.3712807244501941e-05,
+      "loss": 0.0024,
       "step": 106
     },
     {
-      "epoch": 0.9907407407407407,
-      "grad_norm": 0.022390421479940414,
-      "learning_rate": 9.709419266088086e-05,
-      "loss": 0.0188,
+      "epoch": 0.34585858585858587,
+      "grad_norm": 0.009263481013476849,
+      "learning_rate": 1.384217335058215e-05,
+      "loss": 0.0031,
       "step": 107
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.026193244382739067,
-      "learning_rate": 9.698463103929542e-05,
-      "loss": 0.022,
+      "epoch": 0.3490909090909091,
+      "grad_norm": 0.00697364890947938,
+      "learning_rate": 1.3971539456662355e-05,
+      "loss": 0.0021,
       "step": 108
     },
     {
-      "epoch": 1.0092592592592593,
-      "grad_norm": 0.028253022581338882,
-      "learning_rate": 9.687310614099675e-05,
-      "loss": 0.0159,
+      "epoch": 0.3523232323232323,
+      "grad_norm": 0.007600918412208557,
+      "learning_rate": 1.4100905562742562e-05,
+      "loss": 0.0018,
       "step": 109
     },
     {
-      "epoch": 1.0185185185185186,
-      "grad_norm": 0.02241157554090023,
-      "learning_rate": 9.67596226261095e-05,
-      "loss": 0.016,
-      "step": 110
-    },
-    {
-      "epoch": 1.0185185185185186,
-      "eval_loss": 0.01969613879919052,
-      "eval_runtime": 9.1053,
-      "eval_samples_per_second": 5.491,
-      "eval_steps_per_second": 1.428,
+      "epoch": 0.35555555555555557,
+      "grad_norm": 0.00856352224946022,
+      "learning_rate": 1.423027166882277e-05,
+      "loss": 0.0031,
       "step": 110
     },
     {
-      "epoch": 1.0277777777777777,
-      "grad_norm": 0.027405373752117157,
-      "learning_rate": 9.664418523660004e-05,
-      "loss": 0.014,
+      "epoch": 0.35878787878787877,
+      "grad_norm": 0.007947358302772045,
+      "learning_rate": 1.4359637774902976e-05,
+      "loss": 0.0025,
       "step": 111
     },
     {
-      "epoch": 1.037037037037037,
-      "grad_norm": 0.032646384090185165,
-      "learning_rate": 9.652679879607843e-05,
-      "loss": 0.0172,
+      "epoch": 0.362020202020202,
+      "grad_norm": 0.007706194184720516,
+      "learning_rate": 1.4489003880983183e-05,
+      "loss": 0.0023,
       "step": 112
     },
     {
-      "epoch": 1.0462962962962963,
-      "grad_norm": 0.02552163228392601,
-      "learning_rate": 9.640746820959684e-05,
-      "loss": 0.014,
+      "epoch": 0.3652525252525253,
+      "grad_norm": 0.008598407730460167,
+      "learning_rate": 1.4618369987063391e-05,
+      "loss": 0.0028,
       "step": 113
     },
     {
-      "epoch": 1.0555555555555556,
-      "grad_norm": 0.022228199988603592,
-      "learning_rate": 9.628619846344454e-05,
-      "loss": 0.0172,
+      "epoch": 0.36848484848484847,
+      "grad_norm": 0.00845644623041153,
+      "learning_rate": 1.4747736093143596e-05,
+      "loss": 0.0029,
       "step": 114
     },
     {
-      "epoch": 1.0648148148148149,
-      "grad_norm": 0.028009962290525436,
-      "learning_rate": 9.616299462493952e-05,
-      "loss": 0.0166,
+      "epoch": 0.3717171717171717,
+      "grad_norm": 0.008533057756721973,
+      "learning_rate": 1.4877102199223805e-05,
+      "loss": 0.0039,
       "step": 115
     },
     {
-      "epoch": 1.0648148148148149,
-      "eval_loss": 0.019864549860358238,
-      "eval_runtime": 9.122,
-      "eval_samples_per_second": 5.481,
-      "eval_steps_per_second": 1.425,
-      "step": 115
-    },
-    {
-      "epoch": 1.074074074074074,
-      "grad_norm": 0.025030331686139107,
-      "learning_rate": 9.603786184221693e-05,
-      "loss": 0.0195,
+      "epoch": 0.374949494949495,
+      "grad_norm": 0.007429471705108881,
+      "learning_rate": 1.5006468305304012e-05,
+      "loss": 0.0024,
       "step": 116
     },
     {
-      "epoch": 1.0833333333333333,
-      "grad_norm": 0.030586065724492073,
-      "learning_rate": 9.591080534401371e-05,
-      "loss": 0.015,
+      "epoch": 0.3781818181818182,
+      "grad_norm": 0.0069627161137759686,
+      "learning_rate": 1.5135834411384217e-05,
+      "loss": 0.0018,
       "step": 117
     },
     {
-      "epoch": 1.0925925925925926,
-      "grad_norm": 0.02425476722419262,
-      "learning_rate": 9.57818304394503e-05,
-      "loss": 0.0183,
+      "epoch": 0.3814141414141414,
+      "grad_norm": 0.007471222430467606,
+      "learning_rate": 1.5265200517464426e-05,
+      "loss": 0.0024,
       "step": 118
     },
     {
-      "epoch": 1.1018518518518519,
-      "grad_norm": 0.03203345090150833,
-      "learning_rate": 9.565094251780871e-05,
-      "loss": 0.0172,
+      "epoch": 0.3846464646464646,
+      "grad_norm": 0.006873416714370251,
+      "learning_rate": 1.5394566623544633e-05,
+      "loss": 0.0019,
       "step": 119
     },
     {
-      "epoch": 1.1111111111111112,
-      "grad_norm": 0.03028124012053013,
-      "learning_rate": 9.551814704830734e-05,
-      "loss": 0.0189,
-      "step": 120
-    },
-    {
-      "epoch": 1.1111111111111112,
-      "eval_loss": 0.019504941999912262,
-      "eval_runtime": 9.1171,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.3878787878787879,
+      "grad_norm": 0.009251467883586884,
+      "learning_rate": 1.5523932729624836e-05,
+      "loss": 0.0027,
       "step": 120
     },
     {
-      "epoch": 1.1203703703703705,
-      "grad_norm": 0.026934562250971794,
-      "learning_rate": 9.538344957987244e-05,
-      "loss": 0.0132,
+      "epoch": 0.39111111111111113,
+      "grad_norm": 0.007302634883671999,
+      "learning_rate": 1.5653298835705046e-05,
+      "loss": 0.0017,
       "step": 121
     },
     {
-      "epoch": 1.1296296296296295,
-      "grad_norm": 0.02392655238509178,
-      "learning_rate": 9.524685574090627e-05,
-      "loss": 0.0184,
+      "epoch": 0.39434343434343433,
+      "grad_norm": 0.006746005266904831,
+      "learning_rate": 1.5782664941785253e-05,
+      "loss": 0.0019,
       "step": 122
     },
     {
-      "epoch": 1.1388888888888888,
-      "grad_norm": 0.02336742728948593,
-      "learning_rate": 9.51083712390519e-05,
-      "loss": 0.0155,
+      "epoch": 0.3975757575757576,
+      "grad_norm": 0.008011849597096443,
+      "learning_rate": 1.591203104786546e-05,
+      "loss": 0.0023,
       "step": 123
     },
     {
-      "epoch": 1.1481481481481481,
-      "grad_norm": 0.025306498631834984,
-      "learning_rate": 9.496800186095466e-05,
-      "loss": 0.0156,
+      "epoch": 0.40080808080808084,
+      "grad_norm": 0.008799983188509941,
+      "learning_rate": 1.6041397153945667e-05,
+      "loss": 0.0023,
       "step": 124
     },
     {
-      "epoch": 1.1574074074074074,
-      "grad_norm": 0.02764940820634365,
-      "learning_rate": 9.482575347202047e-05,
-      "loss": 0.0211,
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.010617449879646301,
+      "learning_rate": 1.6170763260025874e-05,
+      "loss": 0.0019,
       "step": 125
     },
     {
-      "epoch": 1.1574074074074074,
-      "eval_loss": 0.018362991511821747,
-      "eval_runtime": 9.1297,
-      "eval_samples_per_second": 5.477,
-      "eval_steps_per_second": 1.424,
+      "epoch": 0.40404040404040403,
+      "eval_loss": 0.002037045545876026,
+      "eval_runtime": 18.5899,
+      "eval_samples_per_second": 5.379,
+      "eval_steps_per_second": 1.345,
       "step": 125
     },
     {
-      "epoch": 1.1666666666666667,
-      "grad_norm": 0.02213912270963192,
-      "learning_rate": 9.468163201617062e-05,
-      "loss": 0.0178,
+      "epoch": 0.4072727272727273,
+      "grad_norm": 0.006759752053767443,
+      "learning_rate": 1.630012936610608e-05,
+      "loss": 0.0018,
       "step": 126
     },
     {
-      "epoch": 1.175925925925926,
-      "grad_norm": 0.03320689871907234,
-      "learning_rate": 9.453564351559348e-05,
-      "loss": 0.0148,
+      "epoch": 0.4105050505050505,
+      "grad_norm": 0.008995631709694862,
+      "learning_rate": 1.6429495472186288e-05,
+      "loss": 0.0022,
       "step": 127
     },
     {
-      "epoch": 1.1851851851851851,
-      "grad_norm": 0.023370925337076187,
-      "learning_rate": 9.438779407049281e-05,
-      "loss": 0.0174,
+      "epoch": 0.41373737373737374,
+      "grad_norm": 0.00835257675498724,
+      "learning_rate": 1.6558861578266498e-05,
+      "loss": 0.0026,
       "step": 128
     },
     {
-      "epoch": 1.1944444444444444,
-      "grad_norm": 0.02848099358379841,
-      "learning_rate": 9.423808985883289e-05,
-      "loss": 0.0174,
+      "epoch": 0.416969696969697,
+      "grad_norm": 0.009772442281246185,
+      "learning_rate": 1.66882276843467e-05,
+      "loss": 0.0028,
       "step": 129
     },
     {
-      "epoch": 1.2037037037037037,
-      "grad_norm": 0.02608056552708149,
-      "learning_rate": 9.40865371360804e-05,
-      "loss": 0.0171,
-      "step": 130
-    },
-    {
-      "epoch": 1.2037037037037037,
-      "eval_loss": 0.018851976841688156,
-      "eval_runtime": 9.1046,
-      "eval_samples_per_second": 5.492,
-      "eval_steps_per_second": 1.428,
+      "epoch": 0.4202020202020202,
+      "grad_norm": 0.006931178271770477,
+      "learning_rate": 1.6817593790426908e-05,
+      "loss": 0.002,
       "step": 130
     },
     {
-      "epoch": 1.212962962962963,
-      "grad_norm": 0.02152630314230919,
-      "learning_rate": 9.393314223494296e-05,
-      "loss": 0.0172,
+      "epoch": 0.42343434343434344,
+      "grad_norm": 0.012796151451766491,
+      "learning_rate": 1.694695989650712e-05,
+      "loss": 0.0021,
       "step": 131
     },
     {
-      "epoch": 1.2222222222222223,
-      "grad_norm": 0.02550230175256729,
-      "learning_rate": 9.377791156510455e-05,
-      "loss": 0.016,
+      "epoch": 0.4266666666666667,
+      "grad_norm": 0.007226724177598953,
+      "learning_rate": 1.7076326002587322e-05,
+      "loss": 0.0033,
       "step": 132
     },
     {
-      "epoch": 1.2314814814814814,
-      "grad_norm": 0.025004474446177483,
-      "learning_rate": 9.362085161295769e-05,
-      "loss": 0.0163,
+      "epoch": 0.4298989898989899,
+      "grad_norm": 0.0066243987530469894,
+      "learning_rate": 1.720569210866753e-05,
+      "loss": 0.002,
       "step": 133
     },
     {
-      "epoch": 1.2407407407407407,
-      "grad_norm": 0.026416007429361343,
-      "learning_rate": 9.346196894133239e-05,
-      "loss": 0.0165,
+      "epoch": 0.43313131313131314,
+      "grad_norm": 0.007128287572413683,
+      "learning_rate": 1.733505821474774e-05,
+      "loss": 0.0021,
       "step": 134
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.029432326555252075,
-      "learning_rate": 9.330127018922194e-05,
-      "loss": 0.0191,
-      "step": 135
-    },
-    {
-      "epoch": 1.25,
-      "eval_loss": 0.019194327294826508,
-      "eval_runtime": 9.1131,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.43636363636363634,
+      "grad_norm": 0.007028148043900728,
+      "learning_rate": 1.7464424320827943e-05,
+      "loss": 0.0028,
       "step": 135
     },
     {
-      "epoch": 1.2592592592592593,
-      "grad_norm": 0.03440408781170845,
-      "learning_rate": 9.313876207150543e-05,
-      "loss": 0.0165,
+      "epoch": 0.4395959595959596,
+      "grad_norm": 0.00866060983389616,
+      "learning_rate": 1.7593790426908153e-05,
+      "loss": 0.0053,
       "step": 136
     },
     {
-      "epoch": 1.2685185185185186,
-      "grad_norm": 0.025614989921450615,
-      "learning_rate": 9.297445137866727e-05,
-      "loss": 0.0162,
+      "epoch": 0.44282828282828285,
+      "grad_norm": 0.005735491868108511,
+      "learning_rate": 1.7723156532988356e-05,
+      "loss": 0.0023,
       "step": 137
     },
     {
-      "epoch": 1.2777777777777777,
-      "grad_norm": 0.02456337958574295,
-      "learning_rate": 9.280834497651334e-05,
-      "loss": 0.0192,
+      "epoch": 0.44606060606060605,
+      "grad_norm": 0.005858046934008598,
+      "learning_rate": 1.7852522639068563e-05,
+      "loss": 0.0019,
       "step": 138
     },
     {
-      "epoch": 1.287037037037037,
-      "grad_norm": 0.051101330667734146,
-      "learning_rate": 9.264044980588416e-05,
-      "loss": 0.015,
+      "epoch": 0.4492929292929293,
+      "grad_norm": 0.007398343179374933,
+      "learning_rate": 1.7981888745148774e-05,
+      "loss": 0.0024,
       "step": 139
     },
     {
-      "epoch": 1.2962962962962963,
-      "grad_norm": 0.03369716554880142,
-      "learning_rate": 9.247077288236488e-05,
-      "loss": 0.0184,
-      "step": 140
-    },
-    {
-      "epoch": 1.2962962962962963,
-      "eval_loss": 0.018648317083716393,
-      "eval_runtime": 9.1079,
-      "eval_samples_per_second": 5.49,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.45252525252525255,
+      "grad_norm": 0.009320907294750214,
+      "learning_rate": 1.8111254851228977e-05,
+      "loss": 0.0063,
       "step": 140
     },
     {
-      "epoch": 1.3055555555555556,
-      "grad_norm": 0.024168213829398155,
-      "learning_rate": 9.229932129599205e-05,
-      "loss": 0.0166,
+      "epoch": 0.45575757575757575,
+      "grad_norm": 0.008654161356389523,
+      "learning_rate": 1.8240620957309184e-05,
+      "loss": 0.0025,
       "step": 141
     },
     {
-      "epoch": 1.3148148148148149,
-      "grad_norm": 0.027960045263171196,
-      "learning_rate": 9.212610221095748e-05,
-      "loss": 0.0157,
+      "epoch": 0.458989898989899,
+      "grad_norm": 0.004506159573793411,
+      "learning_rate": 1.8369987063389394e-05,
+      "loss": 0.0013,
       "step": 142
     },
     {
-      "epoch": 1.324074074074074,
-      "grad_norm": 0.023985836654901505,
-      "learning_rate": 9.195112286530873e-05,
-      "loss": 0.0178,
+      "epoch": 0.4622222222222222,
+      "grad_norm": 0.009703525342047215,
+      "learning_rate": 1.8499353169469598e-05,
+      "loss": 0.0034,
       "step": 143
     },
     {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 0.026084545999765396,
-      "learning_rate": 9.177439057064683e-05,
-      "loss": 0.0164,
+      "epoch": 0.46545454545454545,
+      "grad_norm": 0.007734385784715414,
+      "learning_rate": 1.8628719275549808e-05,
+      "loss": 0.0029,
       "step": 144
     },
     {
-      "epoch": 1.3425925925925926,
-      "grad_norm": 0.022582337260246277,
-      "learning_rate": 9.159591271182058e-05,
-      "loss": 0.0162,
-      "step": 145
-    },
-    {
-      "epoch": 1.3425925925925926,
-      "eval_loss": 0.018656810745596886,
-      "eval_runtime": 9.1149,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.4686868686868687,
+      "grad_norm": 0.010392666794359684,
+      "learning_rate": 1.8758085381630015e-05,
+      "loss": 0.0028,
       "step": 145
     },
     {
-      "epoch": 1.3518518518518519,
-      "grad_norm": 0.030290907248854637,
-      "learning_rate": 9.141569674661817e-05,
-      "loss": 0.021,
+      "epoch": 0.4719191919191919,
+      "grad_norm": 0.01011224091053009,
+      "learning_rate": 1.888745148771022e-05,
+      "loss": 0.0036,
       "step": 146
     },
     {
-      "epoch": 1.3611111111111112,
-      "grad_norm": 0.026109322905540466,
-      "learning_rate": 9.123375020545535e-05,
-      "loss": 0.0162,
+      "epoch": 0.47515151515151516,
+      "grad_norm": 0.005182855296880007,
+      "learning_rate": 1.901681759379043e-05,
+      "loss": 0.0014,
       "step": 147
     },
     {
-      "epoch": 1.3703703703703702,
-      "grad_norm": 0.02652176469564438,
-      "learning_rate": 9.105008069106093e-05,
-      "loss": 0.0169,
+      "epoch": 0.4783838383838384,
+      "grad_norm": 0.009993299841880798,
+      "learning_rate": 1.9146183699870636e-05,
+      "loss": 0.003,
       "step": 148
     },
     {
-      "epoch": 1.3796296296296298,
-      "grad_norm": 0.024147020652890205,
-      "learning_rate": 9.086469587815904e-05,
-      "loss": 0.0162,
+      "epoch": 0.4816161616161616,
+      "grad_norm": 0.008879208005964756,
+      "learning_rate": 1.927554980595084e-05,
+      "loss": 0.002,
       "step": 149
     },
     {
-      "epoch": 1.3888888888888888,
-      "grad_norm": 0.021294649690389633,
-      "learning_rate": 9.067760351314838e-05,
-      "loss": 0.0165,
+      "epoch": 0.48484848484848486,
+      "grad_norm": 0.006573604419827461,
+      "learning_rate": 1.940491591203105e-05,
+      "loss": 0.0021,
       "step": 150
     },
     {
-      "epoch": 1.3888888888888888,
-      "eval_loss": 0.018213987350463867,
-      "eval_runtime": 9.1247,
-      "eval_samples_per_second": 5.48,
-      "eval_steps_per_second": 1.425,
+      "epoch": 0.48484848484848486,
+      "eval_loss": 0.0017716821748763323,
+      "eval_runtime": 18.6403,
+      "eval_samples_per_second": 5.365,
+      "eval_steps_per_second": 1.341,
       "step": 150
     },
     {
-      "epoch": 1.3981481481481481,
-      "grad_norm": 0.02462903782725334,
-      "learning_rate": 9.048881141377863e-05,
-      "loss": 0.0204,
+      "epoch": 0.48808080808080806,
+      "grad_norm": 0.005542725790292025,
+      "learning_rate": 1.9534282018111256e-05,
+      "loss": 0.0017,
       "step": 151
     },
     {
-      "epoch": 1.4074074074074074,
-      "grad_norm": 0.024652326479554176,
-      "learning_rate": 9.029832746882371e-05,
-      "loss": 0.0164,
+      "epoch": 0.4913131313131313,
+      "grad_norm": 0.008134805597364902,
+      "learning_rate": 1.9663648124191463e-05,
+      "loss": 0.002,
       "step": 152
     },
     {
-      "epoch": 1.4166666666666667,
-      "grad_norm": 0.026834659278392792,
-      "learning_rate": 9.01061596377522e-05,
-      "loss": 0.018,
+      "epoch": 0.49454545454545457,
+      "grad_norm": 0.009408293291926384,
+      "learning_rate": 1.979301423027167e-05,
+      "loss": 0.0031,
       "step": 153
     },
     {
-      "epoch": 1.425925925925926,
-      "grad_norm": 0.02342064492404461,
-      "learning_rate": 8.991231595039465e-05,
-      "loss": 0.0156,
+      "epoch": 0.49777777777777776,
+      "grad_norm": 0.007265687920153141,
+      "learning_rate": 1.9922380336351877e-05,
+      "loss": 0.0021,
       "step": 154
     },
     {
-      "epoch": 1.4351851851851851,
-      "grad_norm": 0.026441222056746483,
-      "learning_rate": 8.97168045066082e-05,
-      "loss": 0.0157,
+      "epoch": 0.501010101010101,
+      "grad_norm": 0.009999910369515419,
+      "learning_rate": 2.0051746442432084e-05,
+      "loss": 0.0014,
       "step": 155
     },
     {
-      "epoch": 1.4351851851851851,
-      "eval_loss": 0.01855114847421646,
-      "eval_runtime": 9.124,
-      "eval_samples_per_second": 5.48,
-      "eval_steps_per_second": 1.425,
-      "step": 155
-    },
-    {
-      "epoch": 1.4444444444444444,
-      "grad_norm": 0.01796615496277809,
-      "learning_rate": 8.951963347593797e-05,
-      "loss": 0.0165,
+      "epoch": 0.5042424242424243,
+      "grad_norm": 0.007147953379899263,
+      "learning_rate": 2.018111254851229e-05,
+      "loss": 0.0022,
       "step": 156
     },
     {
-      "epoch": 1.4537037037037037,
-      "grad_norm": 0.02256671153008938,
-      "learning_rate": 8.932081109727582e-05,
-      "loss": 0.0201,
+      "epoch": 0.5074747474747475,
+      "grad_norm": 0.0072359428741037846,
+      "learning_rate": 2.0310478654592497e-05,
+      "loss": 0.0034,
       "step": 157
     },
     {
-      "epoch": 1.462962962962963,
-      "grad_norm": 0.028528334572911263,
-      "learning_rate": 8.912034567851599e-05,
-      "loss": 0.0182,
+      "epoch": 0.5107070707070707,
+      "grad_norm": 0.005821447819471359,
+      "learning_rate": 2.0439844760672704e-05,
+      "loss": 0.0018,
       "step": 158
     },
     {
-      "epoch": 1.4722222222222223,
-      "grad_norm": 0.029104968532919884,
-      "learning_rate": 8.891824559620801e-05,
-      "loss": 0.0153,
+      "epoch": 0.5139393939393939,
+      "grad_norm": 0.007256666664034128,
+      "learning_rate": 2.056921086675291e-05,
+      "loss": 0.0027,
       "step": 159
     },
     {
-      "epoch": 1.4814814814814814,
-      "grad_norm": 0.02003669925034046,
-      "learning_rate": 8.871451929520663e-05,
-      "loss": 0.0159,
-      "step": 160
-    },
-    {
-      "epoch": 1.4814814814814814,
-      "eval_loss": 0.01888095587491989,
-      "eval_runtime": 9.1172,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.5171717171717172,
+      "grad_norm": 0.00663521233946085,
+      "learning_rate": 2.0698576972833118e-05,
+      "loss": 0.0023,
       "step": 160
     },
     {
-      "epoch": 1.4907407407407407,
-      "grad_norm": 0.019447356462478638,
-      "learning_rate": 8.850917528831899e-05,
-      "loss": 0.0163,
+      "epoch": 0.5204040404040404,
+      "grad_norm": 0.007504627574235201,
+      "learning_rate": 2.0827943078913325e-05,
+      "loss": 0.004,
       "step": 161
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.03438901901245117,
-      "learning_rate": 8.83022221559489e-05,
-      "loss": 0.0125,
+      "epoch": 0.5236363636363637,
+      "grad_norm": 0.005167617462575436,
+      "learning_rate": 2.0957309184993532e-05,
+      "loss": 0.0015,
       "step": 162
     },
     {
-      "epoch": 1.5092592592592593,
-      "grad_norm": 0.026535626500844955,
-      "learning_rate": 8.809366854573831e-05,
-      "loss": 0.0175,
+      "epoch": 0.5268686868686868,
+      "grad_norm": 0.00587807409465313,
+      "learning_rate": 2.108667529107374e-05,
+      "loss": 0.0016,
       "step": 163
     },
     {
-      "epoch": 1.5185185185185186,
-      "grad_norm": 0.029025647789239883,
-      "learning_rate": 8.78835231722059e-05,
-      "loss": 0.0164,
+      "epoch": 0.5301010101010101,
+      "grad_norm": 0.005344127304852009,
+      "learning_rate": 2.1216041397153946e-05,
+      "loss": 0.0011,
       "step": 164
     },
     {
-      "epoch": 1.5277777777777777,
-      "grad_norm": 0.025528129190206528,
-      "learning_rate": 8.767179481638303e-05,
-      "loss": 0.0174,
-      "step": 165
-    },
-    {
-      "epoch": 1.5277777777777777,
-      "eval_loss": 0.018690049648284912,
-      "eval_runtime": 9.1481,
-      "eval_samples_per_second": 5.466,
-      "eval_steps_per_second": 1.421,
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.005351161118596792,
+      "learning_rate": 2.1345407503234156e-05,
+      "loss": 0.0016,
       "step": 165
     },
     {
-      "epoch": 1.5370370370370372,
-      "grad_norm": 0.025675086304545403,
-      "learning_rate": 8.745849232544681e-05,
-      "loss": 0.0179,
+      "epoch": 0.5365656565656566,
+      "grad_norm": 0.006603873800486326,
+      "learning_rate": 2.147477360931436e-05,
+      "loss": 0.0017,
       "step": 166
     },
     {
-      "epoch": 1.5462962962962963,
-      "grad_norm": 0.027451254427433014,
-      "learning_rate": 8.724362461235029e-05,
-      "loss": 0.0169,
+      "epoch": 0.5397979797979798,
+      "grad_norm": 0.006440309341996908,
+      "learning_rate": 2.1604139715394566e-05,
+      "loss": 0.0016,
       "step": 167
     },
     {
-      "epoch": 1.5555555555555556,
-      "grad_norm": 0.026652028784155846,
-      "learning_rate": 8.702720065545024e-05,
-      "loss": 0.0168,
+      "epoch": 0.5430303030303031,
+      "grad_norm": 0.007731405086815357,
+      "learning_rate": 2.1733505821474777e-05,
+      "loss": 0.0024,
       "step": 168
     },
     {
-      "epoch": 1.5648148148148149,
-      "grad_norm": 0.030202018097043037,
-      "learning_rate": 8.680922949813178e-05,
-      "loss": 0.0162,
+      "epoch": 0.5462626262626262,
+      "grad_norm": 0.0070298160426318645,
+      "learning_rate": 2.186287192755498e-05,
+      "loss": 0.0019,
       "step": 169
     },
     {
-      "epoch": 1.574074074074074,
-      "grad_norm": 0.027389824390411377,
-      "learning_rate": 8.658972024843062e-05,
-      "loss": 0.0184,
-      "step": 170
-    },
-    {
-      "epoch": 1.574074074074074,
-      "eval_loss": 0.018272995948791504,
-      "eval_runtime": 9.1448,
-      "eval_samples_per_second": 5.468,
-      "eval_steps_per_second": 1.422,
+      "epoch": 0.5494949494949495,
+      "grad_norm": 0.008382032625377178,
+      "learning_rate": 2.1992238033635187e-05,
+      "loss": 0.0017,
       "step": 170
     },
     {
-      "epoch": 1.5833333333333335,
-      "grad_norm": 0.025648167356848717,
-      "learning_rate": 8.636868207865244e-05,
-      "loss": 0.0152,
+      "epoch": 0.5527272727272727,
+      "grad_norm": 0.007320054341107607,
+      "learning_rate": 2.2121604139715397e-05,
+      "loss": 0.0017,
       "step": 171
     },
     {
-      "epoch": 1.5925925925925926,
-      "grad_norm": 0.02472120150923729,
-      "learning_rate": 8.614612422498964e-05,
-      "loss": 0.0153,
+      "epoch": 0.555959595959596,
+      "grad_norm": 0.008367806673049927,
+      "learning_rate": 2.22509702457956e-05,
+      "loss": 0.0036,
       "step": 172
     },
     {
-      "epoch": 1.6018518518518519,
-      "grad_norm": 0.020042769610881805,
-      "learning_rate": 8.592205598713539e-05,
-      "loss": 0.017,
+      "epoch": 0.5591919191919192,
+      "grad_norm": 0.008995486423373222,
+      "learning_rate": 2.238033635187581e-05,
+      "loss": 0.0019,
       "step": 173
     },
     {
-      "epoch": 1.6111111111111112,
-      "grad_norm": 0.029423648491501808,
-      "learning_rate": 8.569648672789497e-05,
-      "loss": 0.0158,
+      "epoch": 0.5624242424242424,
+      "grad_norm": 0.007300530560314655,
+      "learning_rate": 2.2509702457956018e-05,
+      "loss": 0.0036,
       "step": 174
     },
     {
-      "epoch": 1.6203703703703702,
-      "grad_norm": 0.02159775421023369,
-      "learning_rate": 8.546942587279465e-05,
-      "loss": 0.0165,
+      "epoch": 0.5656565656565656,
+      "grad_norm": 0.005388753954321146,
+      "learning_rate": 2.263906856403622e-05,
+      "loss": 0.0016,
       "step": 175
     },
     {
-      "epoch": 1.6203703703703702,
-      "eval_loss": 0.018273252993822098,
-      "eval_runtime": 9.118,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.5656565656565656,
+      "eval_loss": 0.0016459682956337929,
+      "eval_runtime": 18.6118,
+      "eval_samples_per_second": 5.373,
+      "eval_steps_per_second": 1.343,
       "step": 175
     },
     {
-      "epoch": 1.6296296296296298,
-      "grad_norm": 0.024837305769324303,
-      "learning_rate": 8.524088290968781e-05,
-      "loss": 0.0187,
+      "epoch": 0.5688888888888889,
+      "grad_norm": 0.005545719526708126,
+      "learning_rate": 2.276843467011643e-05,
+      "loss": 0.0016,
       "step": 176
     },
     {
-      "epoch": 1.6388888888888888,
-      "grad_norm": 0.02383432537317276,
-      "learning_rate": 8.501086738835843e-05,
-      "loss": 0.0181,
+      "epoch": 0.5721212121212121,
+      "grad_norm": 0.007394594140350819,
+      "learning_rate": 2.289780077619664e-05,
+      "loss": 0.0021,
       "step": 177
     },
     {
-      "epoch": 1.6481481481481481,
-      "grad_norm": 0.025743911042809486,
-      "learning_rate": 8.47793889201221e-05,
-      "loss": 0.0171,
+      "epoch": 0.5753535353535354,
+      "grad_norm": 0.005089296959340572,
+      "learning_rate": 2.3027166882276842e-05,
+      "loss": 0.0013,
       "step": 178
     },
     {
-      "epoch": 1.6574074074074074,
-      "grad_norm": 0.023100929334759712,
-      "learning_rate": 8.45464571774244e-05,
-      "loss": 0.021,
+      "epoch": 0.5785858585858585,
+      "grad_norm": 0.00883107353001833,
+      "learning_rate": 2.3156532988357052e-05,
+      "loss": 0.0022,
       "step": 179
     },
     {
-      "epoch": 1.6666666666666665,
-      "grad_norm": 0.02667200192809105,
-      "learning_rate": 8.43120818934367e-05,
-      "loss": 0.0173,
+      "epoch": 0.5818181818181818,
+      "grad_norm": 0.012213426642119884,
+      "learning_rate": 2.328589909443726e-05,
+      "loss": 0.005,
       "step": 180
     },
     {
-      "epoch": 1.6666666666666665,
-      "eval_loss": 0.01778573729097843,
-      "eval_runtime": 9.1324,
-      "eval_samples_per_second": 5.475,
-      "eval_steps_per_second": 1.424,
-      "step": 180
-    },
-    {
-      "epoch": 1.675925925925926,
-      "grad_norm": 0.02880384773015976,
-      "learning_rate": 8.407627286164948e-05,
-      "loss": 0.015,
+      "epoch": 0.585050505050505,
+      "grad_norm": 0.007255645003169775,
+      "learning_rate": 2.3415265200517466e-05,
+      "loss": 0.001,
       "step": 181
     },
     {
-      "epoch": 1.6851851851851851,
-      "grad_norm": 0.030301645398139954,
-      "learning_rate": 8.383903993546311e-05,
-      "loss": 0.0157,
+      "epoch": 0.5882828282828283,
+      "grad_norm": 0.006818380672484636,
+      "learning_rate": 2.3544631306597673e-05,
+      "loss": 0.0015,
       "step": 182
     },
     {
-      "epoch": 1.6944444444444444,
-      "grad_norm": 0.021445374935865402,
-      "learning_rate": 8.360039302777612e-05,
-      "loss": 0.0181,
+      "epoch": 0.5915151515151515,
+      "grad_norm": 0.006983195431530476,
+      "learning_rate": 2.367399741267788e-05,
+      "loss": 0.0027,
       "step": 183
     },
     {
-      "epoch": 1.7037037037037037,
-      "grad_norm": 0.023577649146318436,
-      "learning_rate": 8.336034211057098e-05,
-      "loss": 0.0153,
+      "epoch": 0.5947474747474748,
+      "grad_norm": 0.012661872431635857,
+      "learning_rate": 2.3803363518758087e-05,
+      "loss": 0.0053,
       "step": 184
     },
     {
-      "epoch": 1.7129629629629628,
-      "grad_norm": 0.02492811530828476,
-      "learning_rate": 8.31188972144974e-05,
-      "loss": 0.0131,
-      "step": 185
-    },
-    {
-      "epoch": 1.7129629629629628,
-      "eval_loss": 0.017187727615237236,
-      "eval_runtime": 9.1252,
-      "eval_samples_per_second": 5.479,
-      "eval_steps_per_second": 1.425,
+      "epoch": 0.597979797979798,
+      "grad_norm": 0.005046526901423931,
+      "learning_rate": 2.3932729624838294e-05,
+      "loss": 0.0019,
       "step": 185
     },
     {
-      "epoch": 1.7222222222222223,
-      "grad_norm": 0.023155970498919487,
-      "learning_rate": 8.28760684284532e-05,
-      "loss": 0.0162,
+      "epoch": 0.6012121212121212,
+      "grad_norm": 0.004457033704966307,
+      "learning_rate": 2.40620957309185e-05,
+      "loss": 0.0012,
       "step": 186
     },
     {
-      "epoch": 1.7314814814814814,
-      "grad_norm": 0.02491271123290062,
-      "learning_rate": 8.263186589916273e-05,
-      "loss": 0.0137,
+      "epoch": 0.6044444444444445,
+      "grad_norm": 0.00732004176825285,
+      "learning_rate": 2.4191461836998707e-05,
+      "loss": 0.0017,
       "step": 187
     },
     {
-      "epoch": 1.7407407407407407,
-      "grad_norm": 0.02165275253355503,
-      "learning_rate": 8.238629983075294e-05,
-      "loss": 0.0143,
+      "epoch": 0.6076767676767677,
+      "grad_norm": 0.006394708063453436,
+      "learning_rate": 2.4320827943078914e-05,
+      "loss": 0.0019,
       "step": 188
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.024284129962325096,
-      "learning_rate": 8.213938048432697e-05,
-      "loss": 0.0144,
+      "epoch": 0.610909090909091,
+      "grad_norm": 0.007961345836520195,
+      "learning_rate": 2.445019404915912e-05,
+      "loss": 0.0034,
       "step": 189
     },
     {
-      "epoch": 1.7592592592592593,
-      "grad_norm": 0.027395077049732208,
-      "learning_rate": 8.18911181775353e-05,
-      "loss": 0.0132,
-      "step": 190
-    },
-    {
-      "epoch": 1.7592592592592593,
-      "eval_loss": 0.018012873828411102,
-      "eval_runtime": 9.1149,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.6141414141414141,
+      "grad_norm": 0.0068709347397089005,
+      "learning_rate": 2.4579560155239328e-05,
+      "loss": 0.0013,
       "step": 190
     },
     {
-      "epoch": 1.7685185185185186,
-      "grad_norm": 0.02639261819422245,
-      "learning_rate": 8.164152328414476e-05,
-      "loss": 0.0156,
+      "epoch": 0.6173737373737374,
+      "grad_norm": 0.009581703692674637,
+      "learning_rate": 2.4708926261319535e-05,
+      "loss": 0.0026,
       "step": 191
     },
     {
-      "epoch": 1.7777777777777777,
-      "grad_norm": 0.02319464646279812,
-      "learning_rate": 8.139060623360493e-05,
-      "loss": 0.0121,
+      "epoch": 0.6206060606060606,
+      "grad_norm": 0.007400548551231623,
+      "learning_rate": 2.4838292367399742e-05,
+      "loss": 0.0024,
       "step": 192
     },
     {
-      "epoch": 1.7870370370370372,
-      "grad_norm": 0.020444169640541077,
-      "learning_rate": 8.113837751061246e-05,
-      "loss": 0.0156,
+      "epoch": 0.6238383838383839,
+      "grad_norm": 0.00970857497304678,
+      "learning_rate": 2.496765847347995e-05,
+      "loss": 0.0013,
       "step": 193
     },
     {
-      "epoch": 1.7962962962962963,
-      "grad_norm": 0.03843529522418976,
-      "learning_rate": 8.088484765467286e-05,
-      "loss": 0.0202,
+      "epoch": 0.6270707070707071,
+      "grad_norm": 0.007368095684796572,
+      "learning_rate": 2.509702457956016e-05,
+      "loss": 0.0016,
       "step": 194
     },
     {
-      "epoch": 1.8055555555555556,
-      "grad_norm": 0.03014414757490158,
-      "learning_rate": 8.063002725966015e-05,
-      "loss": 0.0157,
-      "step": 195
-    },
-    {
-      "epoch": 1.8055555555555556,
-      "eval_loss": 0.018071575090289116,
-      "eval_runtime": 9.1428,
-      "eval_samples_per_second": 5.469,
-      "eval_steps_per_second": 1.422,
+      "epoch": 0.6303030303030303,
+      "grad_norm": 0.007306639105081558,
+      "learning_rate": 2.5226390685640362e-05,
+      "loss": 0.0015,
       "step": 195
     },
     {
-      "epoch": 1.8148148148148149,
-      "grad_norm": 0.028225911781191826,
-      "learning_rate": 8.037392697337418e-05,
-      "loss": 0.0152,
+      "epoch": 0.6335353535353535,
+      "grad_norm": 0.008257709443569183,
+      "learning_rate": 2.535575679172057e-05,
+      "loss": 0.0017,
       "step": 196
     },
     {
-      "epoch": 1.824074074074074,
-      "grad_norm": 0.022350864484906197,
-      "learning_rate": 8.011655749709575e-05,
-      "loss": 0.0147,
+      "epoch": 0.6367676767676768,
+      "grad_norm": 0.0050062090158462524,
+      "learning_rate": 2.548512289780078e-05,
+      "loss": 0.0012,
       "step": 197
     },
     {
-      "epoch": 1.8333333333333335,
-      "grad_norm": 0.023073699325323105,
-      "learning_rate": 7.985792958513931e-05,
-      "loss": 0.0142,
+      "epoch": 0.64,
+      "grad_norm": 0.007328143808990717,
+      "learning_rate": 2.5614489003880986e-05,
+      "loss": 0.0025,
       "step": 198
     },
     {
-      "epoch": 1.8425925925925926,
-      "grad_norm": 0.027160046622157097,
-      "learning_rate": 7.95980540444038e-05,
-      "loss": 0.0181,
+      "epoch": 0.6432323232323233,
+      "grad_norm": 0.007282217964529991,
+      "learning_rate": 2.574385510996119e-05,
+      "loss": 0.0018,
       "step": 199
     },
     {
-      "epoch": 1.8518518518518519,
-      "grad_norm": 0.02501911297440529,
-      "learning_rate": 7.93369417339209e-05,
-      "loss": 0.0154,
+      "epoch": 0.6464646464646465,
+      "grad_norm": 0.007711055688560009,
+      "learning_rate": 2.58732212160414e-05,
+      "loss": 0.0013,
       "step": 200
     },
     {
-      "epoch": 1.8518518518518519,
-      "eval_loss": 0.01711750030517578,
-      "eval_runtime": 9.1469,
-      "eval_samples_per_second": 5.466,
-      "eval_steps_per_second": 1.421,
+      "epoch": 0.6464646464646465,
+      "eval_loss": 0.0015150802209973335,
+      "eval_runtime": 18.6339,
+      "eval_samples_per_second": 5.367,
+      "eval_steps_per_second": 1.342,
       "step": 200
     },
     {
-      "epoch": 1.8611111111111112,
-      "grad_norm": 0.02209513448178768,
-      "learning_rate": 7.907460356440133e-05,
-      "loss": 0.0156,
+      "epoch": 0.6496969696969697,
+      "grad_norm": 0.006153900176286697,
+      "learning_rate": 2.6002587322121607e-05,
+      "loss": 0.0016,
       "step": 201
     },
     {
-      "epoch": 1.8703703703703702,
-      "grad_norm": 0.022372853010892868,
-      "learning_rate": 7.881105049777901e-05,
-      "loss": 0.0182,
+      "epoch": 0.6529292929292929,
+      "grad_norm": 0.0070861754938960075,
+      "learning_rate": 2.613195342820181e-05,
+      "loss": 0.002,
       "step": 202
     },
     {
-      "epoch": 1.8796296296296298,
-      "grad_norm": 0.02874351665377617,
-      "learning_rate": 7.854629354675291e-05,
-      "loss": 0.0145,
+      "epoch": 0.6561616161616162,
+      "grad_norm": 0.0048012156039476395,
+      "learning_rate": 2.626131953428202e-05,
+      "loss": 0.0012,
       "step": 203
     },
     {
-      "epoch": 1.8888888888888888,
-      "grad_norm": 0.025754928588867188,
-      "learning_rate": 7.828034377432693e-05,
-      "loss": 0.0161,
+      "epoch": 0.6593939393939394,
+      "grad_norm": 0.006167956627905369,
+      "learning_rate": 2.6390685640362228e-05,
+      "loss": 0.0018,
       "step": 204
     },
     {
-      "epoch": 1.8981481481481481,
-      "grad_norm": 0.023868247866630554,
-      "learning_rate": 7.801321229334764e-05,
-      "loss": 0.0139,
+      "epoch": 0.6626262626262627,
+      "grad_norm": 0.0051431735046207905,
+      "learning_rate": 2.652005174644243e-05,
+      "loss": 0.0013,
       "step": 205
     },
     {
-      "epoch": 1.8981481481481481,
-      "eval_loss": 0.01687374897301197,
-      "eval_runtime": 9.1148,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
-      "step": 205
-    },
-    {
-      "epoch": 1.9074074074074074,
-      "grad_norm": 0.02167942002415657,
-      "learning_rate": 7.774491026603985e-05,
-      "loss": 0.0172,
+      "epoch": 0.6658585858585858,
+      "grad_norm": 0.009899413213133812,
+      "learning_rate": 2.664941785252264e-05,
+      "loss": 0.0031,
       "step": 206
     },
     {
-      "epoch": 1.9166666666666665,
-      "grad_norm": 0.028955647721886635,
-      "learning_rate": 7.74754489035403e-05,
-      "loss": 0.0182,
+      "epoch": 0.6690909090909091,
+      "grad_norm": 0.007686274591833353,
+      "learning_rate": 2.677878395860285e-05,
+      "loss": 0.0022,
       "step": 207
     },
     {
-      "epoch": 1.925925925925926,
-      "grad_norm": 0.023490311577916145,
-      "learning_rate": 7.720483946542914e-05,
-      "loss": 0.0176,
+      "epoch": 0.6723232323232323,
+      "grad_norm": 0.004816455766558647,
+      "learning_rate": 2.6908150064683052e-05,
+      "loss": 0.0011,
       "step": 208
     },
     {
-      "epoch": 1.9351851851851851,
-      "grad_norm": 0.02635806053876877,
-      "learning_rate": 7.69330932592594e-05,
-      "loss": 0.0149,
+      "epoch": 0.6755555555555556,
+      "grad_norm": 0.004873145837336779,
+      "learning_rate": 2.7037516170763262e-05,
+      "loss": 0.0011,
       "step": 209
     },
     {
-      "epoch": 1.9444444444444444,
-      "grad_norm": 0.02554040215909481,
-      "learning_rate": 7.666022164008457e-05,
-      "loss": 0.0169,
-      "step": 210
-    },
-    {
-      "epoch": 1.9444444444444444,
-      "eval_loss": 0.016974864527583122,
-      "eval_runtime": 9.1008,
-      "eval_samples_per_second": 5.494,
-      "eval_steps_per_second": 1.428,
+      "epoch": 0.6787878787878788,
+      "grad_norm": 0.006136384792625904,
+      "learning_rate": 2.716688227684347e-05,
+      "loss": 0.0016,
       "step": 210
     },
     {
-      "epoch": 1.9537037037037037,
-      "grad_norm": 0.02924305759370327,
-      "learning_rate": 7.63862360099841e-05,
-      "loss": 0.0148,
+      "epoch": 0.682020202020202,
+      "grad_norm": 0.004508585669100285,
+      "learning_rate": 2.7296248382923673e-05,
+      "loss": 0.0012,
       "step": 211
     },
     {
-      "epoch": 1.9629629629629628,
-      "grad_norm": 0.020948631688952446,
-      "learning_rate": 7.611114781758692e-05,
-      "loss": 0.0158,
+      "epoch": 0.6852525252525252,
+      "grad_norm": 0.005723617039620876,
+      "learning_rate": 2.7425614489003883e-05,
+      "loss": 0.0016,
       "step": 212
     },
     {
-      "epoch": 1.9722222222222223,
-      "grad_norm": 0.021703558042645454,
-      "learning_rate": 7.583496855759316e-05,
-      "loss": 0.0172,
+      "epoch": 0.6884848484848485,
+      "grad_norm": 0.0072562843561172485,
+      "learning_rate": 2.755498059508409e-05,
+      "loss": 0.0015,
       "step": 213
     },
     {
-      "epoch": 1.9814814814814814,
-      "grad_norm": 0.022922605276107788,
-      "learning_rate": 7.555770977029367e-05,
-      "loss": 0.0149,
+      "epoch": 0.6917171717171717,
+      "grad_norm": 0.006079181097447872,
+      "learning_rate": 2.76843467011643e-05,
+      "loss": 0.0017,
       "step": 214
     },
     {
-      "epoch": 1.9907407407407407,
-      "grad_norm": 0.025769095867872238,
-      "learning_rate": 7.527938304108795e-05,
-      "loss": 0.0158,
-      "step": 215
-    },
-    {
-      "epoch": 1.9907407407407407,
-      "eval_loss": 0.017042405903339386,
-      "eval_runtime": 9.1168,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.694949494949495,
+      "grad_norm": 0.011977693997323513,
+      "learning_rate": 2.7813712807244503e-05,
+      "loss": 0.0048,
       "step": 215
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.03371057286858559,
-      "learning_rate": 7.500000000000001e-05,
-      "loss": 0.0126,
+      "epoch": 0.6981818181818182,
+      "grad_norm": 0.007406435441225767,
+      "learning_rate": 2.794307891332471e-05,
+      "loss": 0.0022,
       "step": 216
     },
     {
-      "epoch": 2.009259259259259,
-      "grad_norm": 0.01711084321141243,
-      "learning_rate": 7.471957232119234e-05,
-      "loss": 0.0142,
+      "epoch": 0.7014141414141414,
+      "grad_norm": 0.006909268908202648,
+      "learning_rate": 2.807244501940492e-05,
+      "loss": 0.0022,
       "step": 217
     },
     {
-      "epoch": 2.0185185185185186,
-      "grad_norm": 0.023618614301085472,
-      "learning_rate": 7.443811172247821e-05,
-      "loss": 0.0151,
+      "epoch": 0.7046464646464646,
+      "grad_norm": 0.0051758866757154465,
+      "learning_rate": 2.8201811125485124e-05,
+      "loss": 0.0012,
       "step": 218
     },
     {
-      "epoch": 2.0277777777777777,
-      "grad_norm": 0.02181304432451725,
-      "learning_rate": 7.415562996483192e-05,
-      "loss": 0.0132,
+      "epoch": 0.7078787878787879,
+      "grad_norm": 0.00657995231449604,
+      "learning_rate": 2.833117723156533e-05,
+      "loss": 0.0014,
       "step": 219
     },
     {
-      "epoch": 2.037037037037037,
-      "grad_norm": 0.020521776750683784,
-      "learning_rate": 7.387213885189746e-05,
-      "loss": 0.0139,
-      "step": 220
-    },
-    {
-      "epoch": 2.037037037037037,
-      "eval_loss": 0.01702064275741577,
-      "eval_runtime": 9.1369,
-      "eval_samples_per_second": 5.472,
-      "eval_steps_per_second": 1.423,
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.006561241112649441,
+      "learning_rate": 2.846054333764554e-05,
+      "loss": 0.0018,
       "step": 220
     },
     {
-      "epoch": 2.0462962962962963,
-      "grad_norm": 0.022209780290722847,
-      "learning_rate": 7.358765022949519e-05,
-      "loss": 0.0152,
+      "epoch": 0.7143434343434344,
+      "grad_norm": 0.005906842183321714,
+      "learning_rate": 2.8589909443725745e-05,
+      "loss": 0.0013,
       "step": 221
     },
     {
-      "epoch": 2.0555555555555554,
-      "grad_norm": 0.02240665629506111,
-      "learning_rate": 7.330217598512695e-05,
-      "loss": 0.0136,
+      "epoch": 0.7175757575757575,
+      "grad_norm": 0.004000976216048002,
+      "learning_rate": 2.871927554980595e-05,
+      "loss": 0.0009,
       "step": 222
     },
     {
-      "epoch": 2.064814814814815,
-      "grad_norm": 0.024021176621317863,
-      "learning_rate": 7.30157280474793e-05,
-      "loss": 0.0134,
+      "epoch": 0.7208080808080808,
+      "grad_norm": 0.006313610821962357,
+      "learning_rate": 2.8848641655886162e-05,
+      "loss": 0.0015,
       "step": 223
     },
     {
-      "epoch": 2.074074074074074,
-      "grad_norm": 0.022297382354736328,
-      "learning_rate": 7.272831838592503e-05,
-      "loss": 0.0158,
+      "epoch": 0.724040404040404,
+      "grad_norm": 0.005339731462299824,
+      "learning_rate": 2.8978007761966365e-05,
+      "loss": 0.0014,
       "step": 224
     },
     {
-      "epoch": 2.0833333333333335,
-      "grad_norm": 0.023189576342701912,
-      "learning_rate": 7.243995901002312e-05,
-      "loss": 0.0146,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.00704893097281456,
+      "learning_rate": 2.9107373868046572e-05,
+      "loss": 0.0017,
       "step": 225
     },
     {
-      "epoch": 2.0833333333333335,
-      "eval_loss": 0.017011733725667,
-      "eval_runtime": 9.1385,
-      "eval_samples_per_second": 5.471,
-      "eval_steps_per_second": 1.423,
+      "epoch": 0.7272727272727273,
+      "eval_loss": 0.0014526441227644682,
+      "eval_runtime": 18.5916,
+      "eval_samples_per_second": 5.379,
+      "eval_steps_per_second": 1.345,
       "step": 225
     },
     {
-      "epoch": 2.0925925925925926,
-      "grad_norm": 0.02641259878873825,
-      "learning_rate": 7.215066196901676e-05,
-      "loss": 0.0149,
+      "epoch": 0.7305050505050505,
+      "grad_norm": 0.007863117381930351,
+      "learning_rate": 2.9236739974126783e-05,
+      "loss": 0.002,
       "step": 226
     },
     {
-      "epoch": 2.1018518518518516,
-      "grad_norm": 0.02105395309627056,
-      "learning_rate": 7.186043935133005e-05,
-      "loss": 0.0105,
+      "epoch": 0.7337373737373737,
+      "grad_norm": 0.007929647341370583,
+      "learning_rate": 2.936610608020699e-05,
+      "loss": 0.0014,
       "step": 227
     },
     {
-      "epoch": 2.111111111111111,
-      "grad_norm": 0.020818866789340973,
-      "learning_rate": 7.156930328406268e-05,
-      "loss": 0.0144,
+      "epoch": 0.7369696969696969,
+      "grad_norm": 0.008834928274154663,
+      "learning_rate": 2.9495472186287193e-05,
+      "loss": 0.0033,
       "step": 228
     },
     {
-      "epoch": 2.1203703703703702,
-      "grad_norm": 0.028699271380901337,
-      "learning_rate": 7.127726593248337e-05,
-      "loss": 0.0134,
+      "epoch": 0.7402020202020202,
+      "grad_norm": 0.005479221232235432,
+      "learning_rate": 2.9624838292367403e-05,
+      "loss": 0.0013,
       "step": 229
     },
     {
-      "epoch": 2.1296296296296298,
-      "grad_norm": 0.025844816118478775,
-      "learning_rate": 7.098433949952146e-05,
-      "loss": 0.0115,
-      "step": 230
-    },
-    {
-      "epoch": 2.1296296296296298,
-      "eval_loss": 0.017404422163963318,
-      "eval_runtime": 9.1138,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.7434343434343434,
+      "grad_norm": 0.008690214715898037,
+      "learning_rate": 2.975420439844761e-05,
+      "loss": 0.0031,
       "step": 230
     },
     {
-      "epoch": 2.138888888888889,
-      "grad_norm": 0.02628181129693985,
-      "learning_rate": 7.069053622525696e-05,
-      "loss": 0.0135,
+      "epoch": 0.7466666666666667,
+      "grad_norm": 0.006373909767717123,
+      "learning_rate": 2.9883570504527814e-05,
+      "loss": 0.0013,
       "step": 231
     },
     {
-      "epoch": 2.148148148148148,
-      "grad_norm": 0.03826741501688957,
-      "learning_rate": 7.039586838640919e-05,
-      "loss": 0.013,
+      "epoch": 0.74989898989899,
+      "grad_norm": 0.005924658849835396,
+      "learning_rate": 3.0012936610608024e-05,
+      "loss": 0.0014,
       "step": 232
     },
     {
-      "epoch": 2.1574074074074074,
-      "grad_norm": 0.02549687772989273,
-      "learning_rate": 7.01003482958237e-05,
-      "loss": 0.0112,
+      "epoch": 0.7531313131313131,
+      "grad_norm": 0.007938825525343418,
+      "learning_rate": 3.014230271668823e-05,
+      "loss": 0.0027,
       "step": 233
     },
     {
-      "epoch": 2.1666666666666665,
-      "grad_norm": 0.02850032038986683,
-      "learning_rate": 6.980398830195785e-05,
-      "loss": 0.0114,
+      "epoch": 0.7563636363636363,
+      "grad_norm": 0.007831827737390995,
+      "learning_rate": 3.0271668822768434e-05,
+      "loss": 0.0025,
       "step": 234
     },
     {
-      "epoch": 2.175925925925926,
-      "grad_norm": 0.028789905831217766,
-      "learning_rate": 6.950680078836474e-05,
-      "loss": 0.0138,
-      "step": 235
-    },
-    {
-      "epoch": 2.175925925925926,
-      "eval_loss": 0.016838619485497475,
-      "eval_runtime": 9.1141,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.7595959595959596,
+      "grad_norm": 0.008570835925638676,
+      "learning_rate": 3.0401034928848644e-05,
+      "loss": 0.003,
       "step": 235
     },
     {
-      "epoch": 2.185185185185185,
-      "grad_norm": 0.024276968091726303,
-      "learning_rate": 6.920879817317589e-05,
-      "loss": 0.0156,
+      "epoch": 0.7628282828282829,
+      "grad_norm": 0.005871222820132971,
+      "learning_rate": 3.053040103492885e-05,
+      "loss": 0.0013,
       "step": 236
     },
     {
-      "epoch": 2.1944444444444446,
-      "grad_norm": 0.02652347832918167,
-      "learning_rate": 6.890999290858214e-05,
-      "loss": 0.0111,
+      "epoch": 0.7660606060606061,
+      "grad_norm": 0.007041990291327238,
+      "learning_rate": 3.0659767141009055e-05,
+      "loss": 0.0028,
       "step": 237
     },
     {
-      "epoch": 2.2037037037037037,
-      "grad_norm": 0.03363705053925514,
-      "learning_rate": 6.861039748031351e-05,
-      "loss": 0.0155,
+      "epoch": 0.7692929292929293,
+      "grad_norm": 0.0073891859501600266,
+      "learning_rate": 3.0789133247089265e-05,
+      "loss": 0.0023,
       "step": 238
     },
     {
-      "epoch": 2.212962962962963,
-      "grad_norm": 0.025364842265844345,
-      "learning_rate": 6.83100244071174e-05,
-      "loss": 0.0127,
+      "epoch": 0.7725252525252525,
+      "grad_norm": 0.005527487024664879,
+      "learning_rate": 3.0918499353169475e-05,
+      "loss": 0.0014,
       "step": 239
     },
     {
-      "epoch": 2.2222222222222223,
-      "grad_norm": 0.024912815541028976,
-      "learning_rate": 6.800888624023553e-05,
-      "loss": 0.0138,
-      "step": 240
-    },
-    {
-      "epoch": 2.2222222222222223,
-      "eval_loss": 0.017057882621884346,
-      "eval_runtime": 9.1505,
-      "eval_samples_per_second": 5.464,
-      "eval_steps_per_second": 1.421,
+      "epoch": 0.7757575757575758,
+      "grad_norm": 0.007551091257482767,
+      "learning_rate": 3.104786545924967e-05,
+      "loss": 0.002,
       "step": 240
     },
     {
-      "epoch": 2.2314814814814814,
-      "grad_norm": 0.031296826899051666,
-      "learning_rate": 6.770699556287939e-05,
-      "loss": 0.0138,
+      "epoch": 0.778989898989899,
+      "grad_norm": 0.005599314346909523,
+      "learning_rate": 3.117723156532988e-05,
+      "loss": 0.002,
       "step": 241
     },
     {
-      "epoch": 2.240740740740741,
-      "grad_norm": 0.03207860141992569,
-      "learning_rate": 6.740436498970452e-05,
-      "loss": 0.0128,
+      "epoch": 0.7822222222222223,
+      "grad_norm": 0.007462210953235626,
+      "learning_rate": 3.130659767141009e-05,
+      "loss": 0.0019,
       "step": 242
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 0.027626443654298782,
-      "learning_rate": 6.710100716628344e-05,
-      "loss": 0.0142,
+      "epoch": 0.7854545454545454,
+      "grad_norm": 0.004866201896220446,
+      "learning_rate": 3.14359637774903e-05,
+      "loss": 0.0012,
       "step": 243
     },
     {
-      "epoch": 2.259259259259259,
-      "grad_norm": 0.025963863357901573,
-      "learning_rate": 6.679693476857711e-05,
-      "loss": 0.0137,
+      "epoch": 0.7886868686868687,
+      "grad_norm": 0.005268635228276253,
+      "learning_rate": 3.1565329883570506e-05,
+      "loss": 0.0013,
       "step": 244
     },
     {
-      "epoch": 2.2685185185185186,
-      "grad_norm": 0.022552739828824997,
-      "learning_rate": 6.649216050240539e-05,
-      "loss": 0.0134,
-      "step": 245
-    },
-    {
-      "epoch": 2.2685185185185186,
-      "eval_loss": 0.016679909080266953,
-      "eval_runtime": 9.1095,
-      "eval_samples_per_second": 5.489,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.7919191919191919,
+      "grad_norm": 0.007261104416102171,
+      "learning_rate": 3.169469598965072e-05,
+      "loss": 0.0031,
       "step": 245
     },
     {
-      "epoch": 2.2777777777777777,
-      "grad_norm": 0.0247825738042593,
-      "learning_rate": 6.618669710291606e-05,
-      "loss": 0.0116,
+      "epoch": 0.7951515151515152,
+      "grad_norm": 0.006308354903012514,
+      "learning_rate": 3.182406209573092e-05,
+      "loss": 0.0014,
       "step": 246
     },
     {
-      "epoch": 2.287037037037037,
-      "grad_norm": 0.021808508783578873,
-      "learning_rate": 6.588055733405266e-05,
-      "loss": 0.014,
+      "epoch": 0.7983838383838384,
+      "grad_norm": 0.0074616689234972,
+      "learning_rate": 3.1953428201811124e-05,
+      "loss": 0.0015,
       "step": 247
     },
     {
-      "epoch": 2.2962962962962963,
-      "grad_norm": 0.025087367743253708,
-      "learning_rate": 6.557375398802123e-05,
-      "loss": 0.0167,
+      "epoch": 0.8016161616161617,
+      "grad_norm": 0.007333788555115461,
+      "learning_rate": 3.2082794307891334e-05,
+      "loss": 0.0014,
       "step": 248
     },
     {
-      "epoch": 2.3055555555555554,
-      "grad_norm": 0.022722622379660606,
-      "learning_rate": 6.526629988475567e-05,
-      "loss": 0.013,
+      "epoch": 0.8048484848484848,
+      "grad_norm": 0.006965926848351955,
+      "learning_rate": 3.2212160413971544e-05,
+      "loss": 0.0018,
       "step": 249
     },
     {
-      "epoch": 2.314814814814815,
-      "grad_norm": 0.023495636880397797,
-      "learning_rate": 6.495820787138209e-05,
-      "loss": 0.0167,
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.00535194855183363,
+      "learning_rate": 3.234152652005175e-05,
+      "loss": 0.0015,
       "step": 250
     },
     {
-      "epoch": 2.314814814814815,
-      "eval_loss": 0.016377143561840057,
-      "eval_runtime": 9.1133,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.8080808080808081,
+      "eval_loss": 0.0013973835157230496,
+      "eval_runtime": 18.6113,
+      "eval_samples_per_second": 5.373,
+      "eval_steps_per_second": 1.343,
       "step": 250
     },
     {
-      "epoch": 2.324074074074074,
-      "grad_norm": 0.021211953833699226,
-      "learning_rate": 6.464949082168204e-05,
-      "loss": 0.0125,
+      "epoch": 0.8113131313131313,
+      "grad_norm": 0.006415734998881817,
+      "learning_rate": 3.247089262613196e-05,
+      "loss": 0.0014,
       "step": 251
     },
     {
-      "epoch": 2.3333333333333335,
-      "grad_norm": 0.022748148068785667,
-      "learning_rate": 6.434016163555452e-05,
-      "loss": 0.0121,
+      "epoch": 0.8145454545454546,
+      "grad_norm": 0.00617300346493721,
+      "learning_rate": 3.260025873221216e-05,
+      "loss": 0.0017,
       "step": 252
     },
     {
-      "epoch": 2.3425925925925926,
-      "grad_norm": 0.021960506215691566,
-      "learning_rate": 6.403023323847695e-05,
-      "loss": 0.0159,
+      "epoch": 0.8177777777777778,
+      "grad_norm": 0.007527848239988089,
+      "learning_rate": 3.2729624838292365e-05,
+      "loss": 0.0017,
       "step": 253
     },
     {
-      "epoch": 2.351851851851852,
-      "grad_norm": 0.02572719193994999,
-      "learning_rate": 6.371971858096508e-05,
-      "loss": 0.0137,
+      "epoch": 0.821010101010101,
+      "grad_norm": 0.00747877499088645,
+      "learning_rate": 3.2858990944372575e-05,
+      "loss": 0.0014,
       "step": 254
     },
     {
-      "epoch": 2.361111111111111,
-      "grad_norm": 0.027611717581748962,
-      "learning_rate": 6.340863063803188e-05,
-      "loss": 0.0123,
-      "step": 255
-    },
-    {
-      "epoch": 2.361111111111111,
-      "eval_loss": 0.016414109617471695,
-      "eval_runtime": 9.1093,
-      "eval_samples_per_second": 5.489,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.8242424242424242,
+      "grad_norm": 0.005451089236885309,
+      "learning_rate": 3.2988357050452786e-05,
+      "loss": 0.0012,
       "step": 255
     },
     {
-      "epoch": 2.3703703703703702,
-      "grad_norm": 0.026147907599806786,
-      "learning_rate": 6.30969824086453e-05,
-      "loss": 0.012,
+      "epoch": 0.8274747474747475,
+      "grad_norm": 0.006235205102711916,
+      "learning_rate": 3.3117723156532996e-05,
+      "loss": 0.0014,
       "step": 256
     },
     {
-      "epoch": 2.3796296296296298,
-      "grad_norm": 0.026667073369026184,
-      "learning_rate": 6.27847869151852e-05,
-      "loss": 0.0127,
+      "epoch": 0.8307070707070707,
+      "grad_norm": 0.005764327012002468,
+      "learning_rate": 3.324708926261319e-05,
+      "loss": 0.0013,
       "step": 257
     },
     {
-      "epoch": 2.388888888888889,
-      "grad_norm": 0.023840012028813362,
-      "learning_rate": 6.247205720289907e-05,
-      "loss": 0.0141,
+      "epoch": 0.833939393939394,
+      "grad_norm": 0.007408145349472761,
+      "learning_rate": 3.33764553686934e-05,
+      "loss": 0.0013,
       "step": 258
     },
     {
-      "epoch": 2.398148148148148,
-      "grad_norm": 0.028697500005364418,
-      "learning_rate": 6.215880633935708e-05,
-      "loss": 0.0135,
+      "epoch": 0.8371717171717171,
+      "grad_norm": 0.004855205304920673,
+      "learning_rate": 3.350582147477361e-05,
+      "loss": 0.0009,
       "step": 259
     },
     {
-      "epoch": 2.4074074074074074,
-      "grad_norm": 0.029124466702342033,
-      "learning_rate": 6.184504741390596e-05,
-      "loss": 0.0139,
-      "step": 260
-    },
-    {
-      "epoch": 2.4074074074074074,
-      "eval_loss": 0.016279693692922592,
-      "eval_runtime": 9.1162,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.8404040404040404,
+      "grad_norm": 0.010664681904017925,
+      "learning_rate": 3.3635187580853817e-05,
+      "loss": 0.0016,
       "step": 260
     },
     {
-      "epoch": 2.4166666666666665,
-      "grad_norm": 0.020265506580471992,
-      "learning_rate": 6.153079353712201e-05,
-      "loss": 0.0129,
+      "epoch": 0.8436363636363636,
+      "grad_norm": 0.006055675912648439,
+      "learning_rate": 3.376455368693403e-05,
+      "loss": 0.0016,
       "step": 261
     },
     {
-      "epoch": 2.425925925925926,
-      "grad_norm": 0.020486822351813316,
-      "learning_rate": 6.121605784026339e-05,
-      "loss": 0.0114,
+      "epoch": 0.8468686868686869,
+      "grad_norm": 0.01056759525090456,
+      "learning_rate": 3.389391979301424e-05,
+      "loss": 0.0016,
       "step": 262
     },
     {
-      "epoch": 2.435185185185185,
-      "grad_norm": 0.02432914823293686,
-      "learning_rate": 6.09008534747213e-05,
-      "loss": 0.0138,
+      "epoch": 0.8501010101010101,
+      "grad_norm": 0.007450331002473831,
+      "learning_rate": 3.4023285899094434e-05,
+      "loss": 0.0016,
       "step": 263
     },
     {
-      "epoch": 2.4444444444444446,
-      "grad_norm": 0.027614833787083626,
-      "learning_rate": 6.058519361147055e-05,
-      "loss": 0.0118,
+      "epoch": 0.8533333333333334,
+      "grad_norm": 0.0053154826164245605,
+      "learning_rate": 3.4152652005174644e-05,
+      "loss": 0.0011,
       "step": 264
     },
     {
-      "epoch": 2.4537037037037037,
-      "grad_norm": 0.03493235632777214,
-      "learning_rate": 6.02690914405191e-05,
-      "loss": 0.0125,
-      "step": 265
-    },
-    {
-      "epoch": 2.4537037037037037,
-      "eval_loss": 0.016143780201673508,
-      "eval_runtime": 9.2054,
-      "eval_samples_per_second": 5.432,
-      "eval_steps_per_second": 1.412,
+      "epoch": 0.8565656565656565,
+      "grad_norm": 0.0060371761210262775,
+      "learning_rate": 3.4282018111254854e-05,
+      "loss": 0.0013,
       "step": 265
     },
     {
-      "epoch": 2.462962962962963,
-      "grad_norm": 0.024250265210866928,
-      "learning_rate": 5.995256017035703e-05,
-      "loss": 0.0139,
+      "epoch": 0.8597979797979798,
+      "grad_norm": 0.005466979928314686,
+      "learning_rate": 3.441138421733506e-05,
+      "loss": 0.0012,
       "step": 266
     },
     {
-      "epoch": 2.4722222222222223,
-      "grad_norm": 0.022808292880654335,
-      "learning_rate": 5.963561302740449e-05,
-      "loss": 0.0162,
+      "epoch": 0.863030303030303,
+      "grad_norm": 0.00831522885710001,
+      "learning_rate": 3.454075032341527e-05,
+      "loss": 0.0014,
       "step": 267
     },
     {
-      "epoch": 2.4814814814814814,
-      "grad_norm": 0.03109206259250641,
-      "learning_rate": 5.9318263255459116e-05,
-      "loss": 0.0123,
+      "epoch": 0.8662626262626263,
+      "grad_norm": 0.007963420823216438,
+      "learning_rate": 3.467011642949548e-05,
+      "loss": 0.0025,
       "step": 268
     },
     {
-      "epoch": 2.490740740740741,
-      "grad_norm": 0.02985144406557083,
-      "learning_rate": 5.900052411514257e-05,
-      "loss": 0.015,
+      "epoch": 0.8694949494949495,
+      "grad_norm": 0.004937573801726103,
+      "learning_rate": 3.4799482535575675e-05,
+      "loss": 0.0011,
       "step": 269
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 0.024866314604878426,
-      "learning_rate": 5.868240888334653e-05,
-      "loss": 0.0126,
-      "step": 270
-    },
-    {
-      "epoch": 2.5,
-      "eval_loss": 0.016046511009335518,
-      "eval_runtime": 9.1128,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.8727272727272727,
+      "grad_norm": 0.0058564417995512486,
+      "learning_rate": 3.4928848641655885e-05,
+      "loss": 0.0016,
       "step": 270
     },
     {
-      "epoch": 2.5092592592592595,
-      "grad_norm": 0.0215854924172163,
-      "learning_rate": 5.836393085267776e-05,
-      "loss": 0.0133,
+      "epoch": 0.8759595959595959,
+      "grad_norm": 0.007912451401352882,
+      "learning_rate": 3.5058214747736096e-05,
+      "loss": 0.0026,
       "step": 271
     },
     {
-      "epoch": 2.5185185185185186,
-      "grad_norm": 0.02321489341557026,
-      "learning_rate": 5.804510333090287e-05,
-      "loss": 0.0175,
+      "epoch": 0.8791919191919192,
+      "grad_norm": 0.003929544240236282,
+      "learning_rate": 3.5187580853816306e-05,
+      "loss": 0.0011,
       "step": 272
     },
     {
-      "epoch": 2.5277777777777777,
-      "grad_norm": 0.024908283725380898,
-      "learning_rate": 5.772593964039203e-05,
-      "loss": 0.0116,
+      "epoch": 0.8824242424242424,
+      "grad_norm": 0.008150831796228886,
+      "learning_rate": 3.531694695989651e-05,
+      "loss": 0.0018,
       "step": 273
     },
     {
-      "epoch": 2.537037037037037,
-      "grad_norm": 0.02571980282664299,
-      "learning_rate": 5.740645311756245e-05,
-      "loss": 0.0125,
+      "epoch": 0.8856565656565657,
+      "grad_norm": 0.009232031181454659,
+      "learning_rate": 3.544631306597671e-05,
+      "loss": 0.0042,
       "step": 274
     },
     {
-      "epoch": 2.5462962962962963,
-      "grad_norm": 0.022897284477949142,
-      "learning_rate": 5.708665711232103e-05,
-      "loss": 0.0138,
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.007469587959349155,
+      "learning_rate": 3.557567917205692e-05,
+      "loss": 0.0023,
       "step": 275
     },
     {
-      "epoch": 2.5462962962962963,
-      "eval_loss": 0.016013609245419502,
-      "eval_runtime": 9.1743,
-      "eval_samples_per_second": 5.45,
-      "eval_steps_per_second": 1.417,
+      "epoch": 0.8888888888888888,
+      "eval_loss": 0.0013014678843319416,
+      "eval_runtime": 18.7103,
+      "eval_samples_per_second": 5.345,
+      "eval_steps_per_second": 1.336,
       "step": 275
     },
     {
-      "epoch": 2.5555555555555554,
-      "grad_norm": 0.023732876405119896,
-      "learning_rate": 5.6766564987506566e-05,
-      "loss": 0.0136,
+      "epoch": 0.8921212121212121,
+      "grad_norm": 0.006450500804930925,
+      "learning_rate": 3.570504527813713e-05,
+      "loss": 0.0014,
       "step": 276
     },
     {
-      "epoch": 2.564814814814815,
-      "grad_norm": 0.024980880320072174,
-      "learning_rate": 5.644619011833133e-05,
-      "loss": 0.0131,
+      "epoch": 0.8953535353535353,
+      "grad_norm": 0.004984740167856216,
+      "learning_rate": 3.583441138421734e-05,
+      "loss": 0.0014,
       "step": 277
     },
     {
-      "epoch": 2.574074074074074,
-      "grad_norm": 0.023262949660420418,
-      "learning_rate": 5.6125545891822274e-05,
-      "loss": 0.0143,
+      "epoch": 0.8985858585858586,
+      "grad_norm": 0.006635539699345827,
+      "learning_rate": 3.596377749029755e-05,
+      "loss": 0.0016,
       "step": 278
     },
     {
-      "epoch": 2.5833333333333335,
-      "grad_norm": 0.024468230083584785,
-      "learning_rate": 5.5804645706261514e-05,
-      "loss": 0.0148,
+      "epoch": 0.9018181818181819,
+      "grad_norm": 0.010307137854397297,
+      "learning_rate": 3.609314359637775e-05,
+      "loss": 0.0025,
       "step": 279
     },
     {
-      "epoch": 2.5925925925925926,
-      "grad_norm": 0.020350055769085884,
-      "learning_rate": 5.548350297062659e-05,
-      "loss": 0.0125,
-      "step": 280
-    },
-    {
-      "epoch": 2.5925925925925926,
-      "eval_loss": 0.015153205953538418,
-      "eval_runtime": 9.1126,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.9050505050505051,
+      "grad_norm": 0.008318118751049042,
+      "learning_rate": 3.6222509702457954e-05,
+      "loss": 0.0028,
       "step": 280
     },
     {
-      "epoch": 2.601851851851852,
-      "grad_norm": 0.027165360748767853,
-      "learning_rate": 5.516213110403009e-05,
-      "loss": 0.0093,
+      "epoch": 0.9082828282828282,
+      "grad_norm": 0.007523260544985533,
+      "learning_rate": 3.6351875808538164e-05,
+      "loss": 0.0022,
       "step": 281
     },
     {
-      "epoch": 2.611111111111111,
-      "grad_norm": 0.021070580929517746,
-      "learning_rate": 5.484054353515896e-05,
-      "loss": 0.0138,
+      "epoch": 0.9115151515151515,
+      "grad_norm": 0.00529656745493412,
+      "learning_rate": 3.648124191461837e-05,
+      "loss": 0.001,
       "step": 282
     },
     {
-      "epoch": 2.6203703703703702,
-      "grad_norm": 0.025997430086135864,
-      "learning_rate": 5.451875370171341e-05,
-      "loss": 0.0121,
+      "epoch": 0.9147474747474748,
+      "grad_norm": 0.0050112344324588776,
+      "learning_rate": 3.661060802069858e-05,
+      "loss": 0.0012,
       "step": 283
     },
     {
-      "epoch": 2.6296296296296298,
-      "grad_norm": 0.02517426759004593,
-      "learning_rate": 5.419677504984534e-05,
-      "loss": 0.0126,
+      "epoch": 0.917979797979798,
+      "grad_norm": 0.006318412255495787,
+      "learning_rate": 3.673997412677879e-05,
+      "loss": 0.0017,
       "step": 284
     },
     {
-      "epoch": 2.638888888888889,
-      "grad_norm": 0.025812286883592606,
-      "learning_rate": 5.387462103359655e-05,
-      "loss": 0.0133,
-      "step": 285
-    },
-    {
-      "epoch": 2.638888888888889,
-      "eval_loss": 0.016152961179614067,
-      "eval_runtime": 9.1127,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.427,
+      "epoch": 0.9212121212121213,
+      "grad_norm": 0.006407878361642361,
+      "learning_rate": 3.6869340232859e-05,
+      "loss": 0.0022,
       "step": 285
     },
     {
-      "epoch": 2.648148148148148,
-      "grad_norm": 0.02393972873687744,
-      "learning_rate": 5.355230511433651e-05,
-      "loss": 0.0136,
+      "epoch": 0.9244444444444444,
+      "grad_norm": 0.004356731195002794,
+      "learning_rate": 3.6998706338939195e-05,
+      "loss": 0.0007,
       "step": 286
     },
     {
-      "epoch": 2.6574074074074074,
-      "grad_norm": 0.021706297993659973,
-      "learning_rate": 5.32298407601999e-05,
-      "loss": 0.0133,
+      "epoch": 0.9276767676767677,
+      "grad_norm": 0.004966625012457371,
+      "learning_rate": 3.7128072445019406e-05,
+      "loss": 0.0012,
       "step": 287
     },
     {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 0.026299407705664635,
-      "learning_rate": 5.290724144552379e-05,
-      "loss": 0.0143,
+      "epoch": 0.9309090909090909,
+      "grad_norm": 0.0055012330412864685,
+      "learning_rate": 3.7257438551099616e-05,
+      "loss": 0.0011,
       "step": 288
     },
     {
-      "epoch": 2.675925925925926,
-      "grad_norm": 0.030511364340782166,
-      "learning_rate": 5.258452065028473e-05,
-      "loss": 0.0137,
+      "epoch": 0.9341414141414142,
+      "grad_norm": 0.006766812410205603,
+      "learning_rate": 3.738680465717982e-05,
+      "loss": 0.0017,
       "step": 289
     },
     {
-      "epoch": 2.685185185185185,
-      "grad_norm": 0.024854540824890137,
-      "learning_rate": 5.226169185953532e-05,
-      "loss": 0.0125,
-      "step": 290
-    },
-    {
-      "epoch": 2.685185185185185,
-      "eval_loss": 0.016076602041721344,
-      "eval_runtime": 9.1632,
-      "eval_samples_per_second": 5.457,
-      "eval_steps_per_second": 1.419,
+      "epoch": 0.9373737373737374,
+      "grad_norm": 0.006011553108692169,
+      "learning_rate": 3.751617076326003e-05,
+      "loss": 0.0022,
       "step": 290
     },
     {
-      "epoch": 2.6944444444444446,
-      "grad_norm": 0.022800520062446594,
-      "learning_rate": 5.193876856284085e-05,
-      "loss": 0.012,
+      "epoch": 0.9406060606060606,
+      "grad_norm": 0.0056586177088320255,
+      "learning_rate": 3.764553686934023e-05,
+      "loss": 0.0014,
       "step": 291
     },
     {
-      "epoch": 2.7037037037037037,
-      "grad_norm": 0.021870015189051628,
-      "learning_rate": 5.1615764253715536e-05,
-      "loss": 0.0136,
+      "epoch": 0.9438383838383838,
+      "grad_norm": 0.0064941453747451305,
+      "learning_rate": 3.777490297542044e-05,
+      "loss": 0.0011,
       "step": 292
     },
     {
-      "epoch": 2.712962962962963,
-      "grad_norm": 0.020156167447566986,
-      "learning_rate": 5.129269242905882e-05,
-      "loss": 0.012,
+      "epoch": 0.9470707070707071,
+      "grad_norm": 0.006994554307311773,
+      "learning_rate": 3.790426908150065e-05,
+      "loss": 0.0021,
       "step": 293
     },
     {
-      "epoch": 2.7222222222222223,
-      "grad_norm": 0.019064266234636307,
-      "learning_rate": 5.096956658859122e-05,
-      "loss": 0.0137,
+      "epoch": 0.9503030303030303,
+      "grad_norm": 0.006325817201286554,
+      "learning_rate": 3.803363518758086e-05,
+      "loss": 0.0014,
       "step": 294
     },
     {
-      "epoch": 2.7314814814814814,
-      "grad_norm": 0.027288921177387238,
-      "learning_rate": 5.064640023429043e-05,
-      "loss": 0.0147,
-      "step": 295
-    },
-    {
-      "epoch": 2.7314814814814814,
-      "eval_loss": 0.01584070920944214,
-      "eval_runtime": 9.1151,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 0.9535353535353536,
+      "grad_norm": 0.006501290015876293,
+      "learning_rate": 3.816300129366106e-05,
+      "loss": 0.0017,
       "step": 295
     },
     {
-      "epoch": 2.7407407407407405,
-      "grad_norm": 0.02484748885035515,
-      "learning_rate": 5.0323206869826966e-05,
-      "loss": 0.0111,
+      "epoch": 0.9567676767676768,
+      "grad_norm": 0.006014223676174879,
+      "learning_rate": 3.829236739974127e-05,
+      "loss": 0.0013,
       "step": 296
     },
     {
-      "epoch": 2.75,
-      "grad_norm": 0.02521962858736515,
-      "learning_rate": 5e-05,
-      "loss": 0.0134,
+      "epoch": 0.96,
+      "grad_norm": 0.005024549085646868,
+      "learning_rate": 3.8421733505821475e-05,
+      "loss": 0.0013,
       "step": 297
     },
     {
-      "epoch": 2.7592592592592595,
-      "grad_norm": 0.023346634581685066,
-      "learning_rate": 4.967679313017303e-05,
-      "loss": 0.0124,
+      "epoch": 0.9632323232323232,
+      "grad_norm": 0.00840625911951065,
+      "learning_rate": 3.855109961190168e-05,
+      "loss": 0.0017,
       "step": 298
     },
     {
-      "epoch": 2.7685185185185186,
-      "grad_norm": 0.021654650568962097,
-      "learning_rate": 4.9353599765709584e-05,
-      "loss": 0.0144,
+      "epoch": 0.9664646464646465,
+      "grad_norm": 0.005791939329355955,
+      "learning_rate": 3.868046571798189e-05,
+      "loss": 0.0014,
       "step": 299
     },
     {
-      "epoch": 2.7777777777777777,
-      "grad_norm": 0.021227596327662468,
-      "learning_rate": 4.903043341140879e-05,
-      "loss": 0.0134,
+      "epoch": 0.9696969696969697,
+      "grad_norm": 0.0047224657610058784,
+      "learning_rate": 3.88098318240621e-05,
+      "loss": 0.0012,
       "step": 300
     },
     {
-      "epoch": 2.7777777777777777,
-      "eval_loss": 0.016122175380587578,
-      "eval_runtime": 9.1019,
-      "eval_samples_per_second": 5.493,
-      "eval_steps_per_second": 1.428,
+      "epoch": 0.9696969696969697,
+      "eval_loss": 0.0012624365044757724,
+      "eval_runtime": 18.7089,
+      "eval_samples_per_second": 5.345,
+      "eval_steps_per_second": 1.336,
       "step": 300
     },
     {
-      "epoch": 2.787037037037037,
-      "grad_norm": 0.024656914174556732,
-      "learning_rate": 4.870730757094121e-05,
-      "loss": 0.0123,
+      "epoch": 0.972929292929293,
+      "grad_norm": 0.005727376788854599,
+      "learning_rate": 3.893919793014231e-05,
+      "loss": 0.0013,
       "step": 301
     },
     {
-      "epoch": 2.7962962962962963,
-      "grad_norm": 0.02583468146622181,
-      "learning_rate": 4.8384235746284476e-05,
-      "loss": 0.015,
+      "epoch": 0.9761616161616161,
+      "grad_norm": 0.007024977821856737,
+      "learning_rate": 3.906856403622251e-05,
+      "loss": 0.0018,
       "step": 302
     },
     {
-      "epoch": 2.8055555555555554,
-      "grad_norm": 0.022909915074706078,
-      "learning_rate": 4.806123143715916e-05,
-      "loss": 0.0142,
+      "epoch": 0.9793939393939394,
+      "grad_norm": 0.004745373502373695,
+      "learning_rate": 3.9197930142302716e-05,
+      "loss": 0.001,
       "step": 303
     },
     {
-      "epoch": 2.814814814814815,
-      "grad_norm": 0.02014041878283024,
-      "learning_rate": 4.7738308140464685e-05,
-      "loss": 0.0131,
+      "epoch": 0.9826262626262626,
+      "grad_norm": 0.0050310962833464146,
+      "learning_rate": 3.9327296248382926e-05,
+      "loss": 0.0013,
       "step": 304
     },
     {
-      "epoch": 2.824074074074074,
-      "grad_norm": 0.022683143615722656,
-      "learning_rate": 4.7415479349715275e-05,
-      "loss": 0.0124,
-      "step": 305
-    },
-    {
-      "epoch": 2.824074074074074,
-      "eval_loss": 0.015797268599271774,
-      "eval_runtime": 9.1281,
-      "eval_samples_per_second": 5.478,
-      "eval_steps_per_second": 1.424,
+      "epoch": 0.9858585858585859,
+      "grad_norm": 0.008561142720282078,
+      "learning_rate": 3.945666235446313e-05,
+      "loss": 0.0018,
       "step": 305
     },
     {
-      "epoch": 2.8333333333333335,
-      "grad_norm": 0.025906002148985863,
-      "learning_rate": 4.709275855447621e-05,
-      "loss": 0.0154,
+      "epoch": 0.9890909090909091,
+      "grad_norm": 0.0055971029214560986,
+      "learning_rate": 3.958602846054334e-05,
+      "loss": 0.0013,
       "step": 306
     },
     {
-      "epoch": 2.8425925925925926,
-      "grad_norm": 0.027820315212011337,
-      "learning_rate": 4.677015923980011e-05,
-      "loss": 0.0138,
+      "epoch": 0.9923232323232323,
+      "grad_norm": 0.007602418772876263,
+      "learning_rate": 3.971539456662355e-05,
+      "loss": 0.0014,
       "step": 307
     },
     {
-      "epoch": 2.851851851851852,
-      "grad_norm": 0.023744860664010048,
-      "learning_rate": 4.6447694885663514e-05,
-      "loss": 0.0124,
+      "epoch": 0.9955555555555555,
+      "grad_norm": 0.0054479725658893585,
+      "learning_rate": 3.9844760672703754e-05,
+      "loss": 0.0014,
       "step": 308
     },
     {
-      "epoch": 2.861111111111111,
-      "grad_norm": 0.026518192142248154,
-      "learning_rate": 4.612537896640346e-05,
-      "loss": 0.0155,
+      "epoch": 0.9987878787878788,
+      "grad_norm": 0.006376450881361961,
+      "learning_rate": 3.997412677878396e-05,
+      "loss": 0.0013,
       "step": 309
     },
     {
-      "epoch": 2.8703703703703702,
-      "grad_norm": 0.020426657050848007,
-      "learning_rate": 4.5803224950154656e-05,
-      "loss": 0.0132,
-      "step": 310
-    },
-    {
-      "epoch": 2.8703703703703702,
-      "eval_loss": 0.015400240197777748,
-      "eval_runtime": 9.1185,
-      "eval_samples_per_second": 5.483,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.002020202020202,
+      "grad_norm": 0.00963524729013443,
+      "learning_rate": 4.010349288486417e-05,
+      "loss": 0.0025,
       "step": 310
     },
     {
-      "epoch": 2.8796296296296298,
-      "grad_norm": 0.022766800597310066,
-      "learning_rate": 4.54812462982866e-05,
-      "loss": 0.0139,
+      "epoch": 1.0052525252525253,
+      "grad_norm": 0.004950121510773897,
+      "learning_rate": 4.023285899094437e-05,
+      "loss": 0.0009,
       "step": 311
     },
     {
-      "epoch": 2.888888888888889,
-      "grad_norm": 0.021728193387389183,
-      "learning_rate": 4.515945646484105e-05,
-      "loss": 0.0133,
+      "epoch": 1.0084848484848485,
+      "grad_norm": 0.0047010756097733974,
+      "learning_rate": 4.036222509702458e-05,
+      "loss": 0.0012,
       "step": 312
     },
     {
-      "epoch": 2.898148148148148,
-      "grad_norm": 0.0226016603410244,
-      "learning_rate": 4.4837868895969936e-05,
-      "loss": 0.0126,
+      "epoch": 1.0117171717171718,
+      "grad_norm": 0.004061982501298189,
+      "learning_rate": 4.049159120310479e-05,
+      "loss": 0.0008,
       "step": 313
     },
     {
-      "epoch": 2.9074074074074074,
-      "grad_norm": 0.027723975479602814,
-      "learning_rate": 4.451649702937342e-05,
-      "loss": 0.0106,
+      "epoch": 1.014949494949495,
+      "grad_norm": 0.006061443593353033,
+      "learning_rate": 4.0620957309184995e-05,
+      "loss": 0.0022,
       "step": 314
     },
     {
-      "epoch": 2.9166666666666665,
-      "grad_norm": 0.01856391504406929,
-      "learning_rate": 4.4195354293738484e-05,
-      "loss": 0.0146,
-      "step": 315
-    },
-    {
-      "epoch": 2.9166666666666665,
-      "eval_loss": 0.015166966244578362,
-      "eval_runtime": 9.1172,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.018181818181818,
+      "grad_norm": 0.005154592916369438,
+      "learning_rate": 4.07503234152652e-05,
+      "loss": 0.0023,
       "step": 315
     },
     {
-      "epoch": 2.925925925925926,
-      "grad_norm": 0.019857853651046753,
-      "learning_rate": 4.387445410817774e-05,
-      "loss": 0.0124,
+      "epoch": 1.0214141414141413,
+      "grad_norm": 0.004801798611879349,
+      "learning_rate": 4.087968952134541e-05,
+      "loss": 0.0012,
       "step": 316
     },
     {
-      "epoch": 2.935185185185185,
-      "grad_norm": 0.025410892441868782,
-      "learning_rate": 4.355380988166867e-05,
-      "loss": 0.0119,
+      "epoch": 1.0246464646464646,
+      "grad_norm": 0.004275497514754534,
+      "learning_rate": 4.100905562742562e-05,
+      "loss": 0.0012,
       "step": 317
     },
     {
-      "epoch": 2.9444444444444446,
-      "grad_norm": 0.02312655746936798,
-      "learning_rate": 4.323343501249346e-05,
-      "loss": 0.0144,
+      "epoch": 1.0278787878787878,
+      "grad_norm": 0.006128865294158459,
+      "learning_rate": 4.113842173350582e-05,
+      "loss": 0.0022,
       "step": 318
     },
     {
-      "epoch": 2.9537037037037037,
-      "grad_norm": 0.022076064720749855,
-      "learning_rate": 4.2913342887678985e-05,
-      "loss": 0.0117,
+      "epoch": 1.031111111111111,
+      "grad_norm": 0.004465613514184952,
+      "learning_rate": 4.126778783958603e-05,
+      "loss": 0.0019,
       "step": 319
     },
     {
-      "epoch": 2.962962962962963,
-      "grad_norm": 0.023769903928041458,
-      "learning_rate": 4.259354688243757e-05,
-      "loss": 0.014,
-      "step": 320
-    },
-    {
-      "epoch": 2.962962962962963,
-      "eval_loss": 0.014957955107092857,
-      "eval_runtime": 9.1101,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.0343434343434343,
+      "grad_norm": 0.004612160846590996,
+      "learning_rate": 4.1397153945666236e-05,
+      "loss": 0.0013,
       "step": 320
     },
     {
-      "epoch": 2.9722222222222223,
-      "grad_norm": 0.023904340341687202,
-      "learning_rate": 4.227406035960798e-05,
-      "loss": 0.0121,
+      "epoch": 1.0375757575757576,
+      "grad_norm": 0.00512789748609066,
+      "learning_rate": 4.152652005174644e-05,
+      "loss": 0.0012,
       "step": 321
     },
     {
-      "epoch": 2.9814814814814814,
-      "grad_norm": 0.02383498102426529,
-      "learning_rate": 4.195489666909713e-05,
-      "loss": 0.0119,
+      "epoch": 1.0408080808080808,
+      "grad_norm": 0.006197880953550339,
+      "learning_rate": 4.165588615782665e-05,
+      "loss": 0.0015,
       "step": 322
     },
     {
-      "epoch": 2.9907407407407405,
-      "grad_norm": 0.03048449568450451,
-      "learning_rate": 4.1636069147322246e-05,
-      "loss": 0.0136,
+      "epoch": 1.044040404040404,
+      "grad_norm": 0.005857696291059256,
+      "learning_rate": 4.178525226390686e-05,
+      "loss": 0.0014,
       "step": 323
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 0.023879334330558777,
-      "learning_rate": 4.131759111665349e-05,
-      "loss": 0.0137,
+      "epoch": 1.0472727272727274,
+      "grad_norm": 0.010157118551433086,
+      "learning_rate": 4.1914618369987064e-05,
+      "loss": 0.0015,
       "step": 324
     },
     {
-      "epoch": 3.009259259259259,
-      "grad_norm": 0.025208691135048866,
-      "learning_rate": 4.099947588485744e-05,
-      "loss": 0.0122,
+      "epoch": 1.0505050505050506,
+      "grad_norm": 0.006577750667929649,
+      "learning_rate": 4.2043984476067274e-05,
+      "loss": 0.0011,
       "step": 325
     },
     {
-      "epoch": 3.009259259259259,
-      "eval_loss": 0.015089023858308792,
-      "eval_runtime": 9.116,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.0505050505050506,
+      "eval_loss": 0.0012751913163810968,
+      "eval_runtime": 18.6453,
+      "eval_samples_per_second": 5.363,
+      "eval_steps_per_second": 1.341,
       "step": 325
     },
     {
-      "epoch": 3.0185185185185186,
-      "grad_norm": 0.020718788728117943,
-      "learning_rate": 4.06817367445409e-05,
-      "loss": 0.0095,
+      "epoch": 1.0537373737373736,
+      "grad_norm": 0.004863920155912638,
+      "learning_rate": 4.217335058214748e-05,
+      "loss": 0.0009,
       "step": 326
     },
     {
-      "epoch": 3.0277777777777777,
-      "grad_norm": 0.024810951203107834,
-      "learning_rate": 4.036438697259551e-05,
-      "loss": 0.0134,
+      "epoch": 1.056969696969697,
+      "grad_norm": 0.005144843365997076,
+      "learning_rate": 4.230271668822768e-05,
+      "loss": 0.0013,
       "step": 327
     },
     {
-      "epoch": 3.037037037037037,
-      "grad_norm": 0.019842958077788353,
-      "learning_rate": 4.004743982964298e-05,
-      "loss": 0.0122,
+      "epoch": 1.0602020202020201,
+      "grad_norm": 0.005289267282932997,
+      "learning_rate": 4.243208279430789e-05,
+      "loss": 0.0012,
       "step": 328
     },
     {
-      "epoch": 3.0462962962962963,
-      "grad_norm": 0.01818239875137806,
-      "learning_rate": 3.97309085594809e-05,
-      "loss": 0.0101,
+      "epoch": 1.0634343434343434,
+      "grad_norm": 0.005019436590373516,
+      "learning_rate": 4.25614489003881e-05,
+      "loss": 0.0011,
       "step": 329
     },
     {
-      "epoch": 3.0555555555555554,
-      "grad_norm": 0.022604303434491158,
-      "learning_rate": 3.941480638852948e-05,
-      "loss": 0.0118,
-      "step": 330
-    },
-    {
-      "epoch": 3.0555555555555554,
-      "eval_loss": 0.015503546223044395,
-      "eval_runtime": 9.1063,
-      "eval_samples_per_second": 5.491,
-      "eval_steps_per_second": 1.428,
+      "epoch": 1.0666666666666667,
+      "grad_norm": 0.0038786993827670813,
+      "learning_rate": 4.269081500646831e-05,
+      "loss": 0.0008,
       "step": 330
     },
     {
-      "epoch": 3.064814814814815,
-      "grad_norm": 0.024690452963113785,
-      "learning_rate": 3.909914652527871e-05,
-      "loss": 0.0109,
+      "epoch": 1.06989898989899,
+      "grad_norm": 0.005253227427601814,
+      "learning_rate": 4.2820181112548515e-05,
+      "loss": 0.0012,
       "step": 331
     },
     {
-      "epoch": 3.074074074074074,
-      "grad_norm": 0.02343621291220188,
-      "learning_rate": 3.878394215973663e-05,
-      "loss": 0.0123,
+      "epoch": 1.0731313131313132,
+      "grad_norm": 0.005880107171833515,
+      "learning_rate": 4.294954721862872e-05,
+      "loss": 0.0016,
       "step": 332
     },
     {
-      "epoch": 3.0833333333333335,
-      "grad_norm": 0.026170087978243828,
-      "learning_rate": 3.846920646287799e-05,
-      "loss": 0.0122,
+      "epoch": 1.0763636363636364,
+      "grad_norm": 0.006426130421459675,
+      "learning_rate": 4.307891332470893e-05,
+      "loss": 0.0009,
       "step": 333
     },
     {
-      "epoch": 3.0925925925925926,
-      "grad_norm": 0.024799769744277,
-      "learning_rate": 3.815495258609404e-05,
-      "loss": 0.0125,
+      "epoch": 1.0795959595959597,
+      "grad_norm": 0.0038312741089612246,
+      "learning_rate": 4.320827943078913e-05,
+      "loss": 0.0009,
       "step": 334
     },
     {
-      "epoch": 3.1018518518518516,
-      "grad_norm": 0.02072787657380104,
-      "learning_rate": 3.784119366064293e-05,
-      "loss": 0.0108,
-      "step": 335
-    },
-    {
-      "epoch": 3.1018518518518516,
-      "eval_loss": 0.0155374426394701,
-      "eval_runtime": 9.1152,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.082828282828283,
+      "grad_norm": 0.005139067303389311,
+      "learning_rate": 4.333764553686934e-05,
+      "loss": 0.0011,
       "step": 335
     },
     {
-      "epoch": 3.111111111111111,
-      "grad_norm": 0.021989421918988228,
-      "learning_rate": 3.752794279710094e-05,
-      "loss": 0.0114,
+      "epoch": 1.086060606060606,
+      "grad_norm": 0.006610156502574682,
+      "learning_rate": 4.346701164294955e-05,
+      "loss": 0.0019,
       "step": 336
     },
     {
-      "epoch": 3.1203703703703702,
-      "grad_norm": 0.03829918056726456,
-      "learning_rate": 3.721521308481482e-05,
-      "loss": 0.0101,
+      "epoch": 1.0892929292929292,
+      "grad_norm": 0.003907322883605957,
+      "learning_rate": 4.359637774902976e-05,
+      "loss": 0.001,
       "step": 337
     },
     {
-      "epoch": 3.1296296296296298,
-      "grad_norm": 0.029835987836122513,
-      "learning_rate": 3.6903017591354706e-05,
-      "loss": 0.0107,
+      "epoch": 1.0925252525252525,
+      "grad_norm": 0.006776052061468363,
+      "learning_rate": 4.372574385510996e-05,
+      "loss": 0.0011,
       "step": 338
     },
     {
-      "epoch": 3.138888888888889,
-      "grad_norm": 0.02231847681105137,
-      "learning_rate": 3.6591369361968124e-05,
-      "loss": 0.012,
+      "epoch": 1.0957575757575757,
+      "grad_norm": 0.007625575177371502,
+      "learning_rate": 4.385510996119017e-05,
+      "loss": 0.0016,
       "step": 339
     },
     {
-      "epoch": 3.148148148148148,
-      "grad_norm": 0.02263280376791954,
-      "learning_rate": 3.628028141903493e-05,
-      "loss": 0.0103,
-      "step": 340
-    },
-    {
-      "epoch": 3.148148148148148,
-      "eval_loss": 0.01546421181410551,
-      "eval_runtime": 9.1199,
-      "eval_samples_per_second": 5.483,
-      "eval_steps_per_second": 1.425,
+      "epoch": 1.098989898989899,
+      "grad_norm": 0.012086856178939342,
+      "learning_rate": 4.3984476067270374e-05,
+      "loss": 0.0035,
       "step": 340
     },
     {
-      "epoch": 3.1574074074074074,
-      "grad_norm": 0.023618226870894432,
-      "learning_rate": 3.596976676152306e-05,
-      "loss": 0.0116,
+      "epoch": 1.1022222222222222,
+      "grad_norm": 0.007610958535224199,
+      "learning_rate": 4.4113842173350584e-05,
+      "loss": 0.0014,
       "step": 341
     },
     {
-      "epoch": 3.1666666666666665,
-      "grad_norm": 0.02577986940741539,
-      "learning_rate": 3.5659838364445505e-05,
-      "loss": 0.0108,
+      "epoch": 1.1054545454545455,
+      "grad_norm": 0.006693006958812475,
+      "learning_rate": 4.4243208279430794e-05,
+      "loss": 0.0013,
       "step": 342
     },
     {
-      "epoch": 3.175925925925926,
-      "grad_norm": 0.026071948930621147,
-      "learning_rate": 3.535050917831797e-05,
-      "loss": 0.0108,
+      "epoch": 1.1086868686868687,
+      "grad_norm": 0.0045067970640957355,
+      "learning_rate": 4.4372574385511e-05,
+      "loss": 0.0011,
       "step": 343
     },
     {
-      "epoch": 3.185185185185185,
-      "grad_norm": 0.038238752633333206,
-      "learning_rate": 3.5041792128617927e-05,
-      "loss": 0.0094,
+      "epoch": 1.111919191919192,
+      "grad_norm": 0.006877266336232424,
+      "learning_rate": 4.45019404915912e-05,
+      "loss": 0.001,
       "step": 344
     },
     {
-      "epoch": 3.1944444444444446,
-      "grad_norm": 0.029051663354039192,
-      "learning_rate": 3.473370011524435e-05,
-      "loss": 0.0099,
-      "step": 345
-    },
-    {
-      "epoch": 3.1944444444444446,
-      "eval_loss": 0.015372861176729202,
-      "eval_runtime": 9.1378,
-      "eval_samples_per_second": 5.472,
-      "eval_steps_per_second": 1.423,
+      "epoch": 1.1151515151515152,
+      "grad_norm": 0.0043833632953464985,
+      "learning_rate": 4.463130659767141e-05,
+      "loss": 0.0011,
       "step": 345
     },
     {
-      "epoch": 3.2037037037037037,
-      "grad_norm": 0.022384386509656906,
-      "learning_rate": 3.442624601197877e-05,
-      "loss": 0.0096,
+      "epoch": 1.1183838383838385,
+      "grad_norm": 0.004103951156139374,
+      "learning_rate": 4.476067270375162e-05,
+      "loss": 0.0009,
       "step": 346
     },
     {
-      "epoch": 3.212962962962963,
-      "grad_norm": 0.024341940879821777,
-      "learning_rate": 3.4119442665947344e-05,
-      "loss": 0.0094,
+      "epoch": 1.1216161616161617,
+      "grad_norm": 0.006397966295480728,
+      "learning_rate": 4.4890038809831825e-05,
+      "loss": 0.0008,
       "step": 347
     },
     {
-      "epoch": 3.2222222222222223,
-      "grad_norm": 0.02119499258697033,
-      "learning_rate": 3.381330289708396e-05,
-      "loss": 0.011,
+      "epoch": 1.1248484848484848,
+      "grad_norm": 0.007373814936727285,
+      "learning_rate": 4.5019404915912036e-05,
+      "loss": 0.0016,
       "step": 348
     },
     {
-      "epoch": 3.2314814814814814,
-      "grad_norm": 0.025269504636526108,
-      "learning_rate": 3.350783949759462e-05,
-      "loss": 0.0105,
+      "epoch": 1.128080808080808,
+      "grad_norm": 0.005406413692981005,
+      "learning_rate": 4.514877102199224e-05,
+      "loss": 0.0012,
       "step": 349
     },
     {
-      "epoch": 3.240740740740741,
-      "grad_norm": 0.02428189478814602,
-      "learning_rate": 3.3203065231422904e-05,
-      "loss": 0.0115,
+      "epoch": 1.1313131313131313,
+      "grad_norm": 0.0059000300243496895,
+      "learning_rate": 4.527813712807244e-05,
+      "loss": 0.0011,
       "step": 350
     },
     {
-      "epoch": 3.240740740740741,
-      "eval_loss": 0.015474287793040276,
-      "eval_runtime": 9.1142,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.1313131313131313,
+      "eval_loss": 0.0012826790334656835,
+      "eval_runtime": 18.7277,
+      "eval_samples_per_second": 5.34,
+      "eval_steps_per_second": 1.335,
       "step": 350
     },
     {
-      "epoch": 3.25,
-      "grad_norm": 0.027830710634589195,
-      "learning_rate": 3.289899283371657e-05,
-      "loss": 0.014,
+      "epoch": 1.1345454545454545,
+      "grad_norm": 0.007405865006148815,
+      "learning_rate": 4.540750323415265e-05,
+      "loss": 0.0026,
       "step": 351
     },
     {
-      "epoch": 3.259259259259259,
-      "grad_norm": 0.026644067838788033,
-      "learning_rate": 3.2595635010295475e-05,
-      "loss": 0.0132,
+      "epoch": 1.1377777777777778,
+      "grad_norm": 0.005577345844358206,
+      "learning_rate": 4.553686934023286e-05,
+      "loss": 0.0018,
       "step": 352
     },
     {
-      "epoch": 3.2685185185185186,
-      "grad_norm": 0.028307707980275154,
-      "learning_rate": 3.2293004437120624e-05,
-      "loss": 0.0093,
+      "epoch": 1.141010101010101,
+      "grad_norm": 0.006122430320829153,
+      "learning_rate": 4.566623544631307e-05,
+      "loss": 0.0015,
       "step": 353
     },
     {
-      "epoch": 3.2777777777777777,
-      "grad_norm": 0.03480321913957596,
-      "learning_rate": 3.199111375976449e-05,
-      "loss": 0.0107,
+      "epoch": 1.1442424242424243,
+      "grad_norm": 0.007260697893798351,
+      "learning_rate": 4.579560155239328e-05,
+      "loss": 0.0012,
       "step": 354
     },
     {
-      "epoch": 3.287037037037037,
-      "grad_norm": 0.029546814039349556,
-      "learning_rate": 3.1689975592882603e-05,
-      "loss": 0.0099,
-      "step": 355
-    },
-    {
-      "epoch": 3.287037037037037,
-      "eval_loss": 0.015444349497556686,
-      "eval_runtime": 9.1458,
-      "eval_samples_per_second": 5.467,
-      "eval_steps_per_second": 1.421,
+      "epoch": 1.1474747474747475,
+      "grad_norm": 0.007470736745744944,
+      "learning_rate": 4.592496765847348e-05,
+      "loss": 0.0032,
       "step": 355
     },
     {
-      "epoch": 3.2962962962962963,
-      "grad_norm": 0.02437739446759224,
-      "learning_rate": 3.1389602519686515e-05,
-      "loss": 0.0118,
+      "epoch": 1.1507070707070708,
+      "grad_norm": 0.0061160847544670105,
+      "learning_rate": 4.6054333764553684e-05,
+      "loss": 0.0013,
       "step": 356
     },
     {
-      "epoch": 3.3055555555555554,
-      "grad_norm": 0.029530519619584084,
-      "learning_rate": 3.109000709141788e-05,
-      "loss": 0.0121,
+      "epoch": 1.1539393939393938,
+      "grad_norm": 0.006069364957511425,
+      "learning_rate": 4.6183699870633894e-05,
+      "loss": 0.0011,
       "step": 357
     },
     {
-      "epoch": 3.314814814814815,
-      "grad_norm": 0.029449855908751488,
-      "learning_rate": 3.079120182682412e-05,
-      "loss": 0.0099,
+      "epoch": 1.157171717171717,
+      "grad_norm": 0.004093985538929701,
+      "learning_rate": 4.6313065976714105e-05,
+      "loss": 0.0005,
       "step": 358
     },
     {
-      "epoch": 3.324074074074074,
-      "grad_norm": 0.020589128136634827,
-      "learning_rate": 3.049319921163526e-05,
-      "loss": 0.0119,
+      "epoch": 1.1604040404040403,
+      "grad_norm": 0.0061876364052295685,
+      "learning_rate": 4.6442432082794315e-05,
+      "loss": 0.0017,
       "step": 359
     },
     {
-      "epoch": 3.3333333333333335,
-      "grad_norm": 0.02450876496732235,
-      "learning_rate": 3.019601169804216e-05,
-      "loss": 0.0129,
-      "step": 360
-    },
-    {
-      "epoch": 3.3333333333333335,
-      "eval_loss": 0.0157760102301836,
-      "eval_runtime": 9.1103,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.1636363636363636,
+      "grad_norm": 0.0048703462816774845,
+      "learning_rate": 4.657179818887452e-05,
+      "loss": 0.0009,
       "step": 360
     },
     {
-      "epoch": 3.3425925925925926,
-      "grad_norm": 0.0208604596555233,
-      "learning_rate": 2.9899651704176325e-05,
-      "loss": 0.011,
+      "epoch": 1.1668686868686868,
+      "grad_norm": 0.007618950214236975,
+      "learning_rate": 4.670116429495472e-05,
+      "loss": 0.0019,
       "step": 361
     },
     {
-      "epoch": 3.351851851851852,
-      "grad_norm": 0.025153055787086487,
-      "learning_rate": 2.9604131613590824e-05,
-      "loss": 0.0109,
+      "epoch": 1.17010101010101,
+      "grad_norm": 0.004160857293754816,
+      "learning_rate": 4.683053040103493e-05,
+      "loss": 0.001,
       "step": 362
     },
     {
-      "epoch": 3.361111111111111,
-      "grad_norm": 0.021455859765410423,
-      "learning_rate": 2.9309463774743046e-05,
-      "loss": 0.0122,
+      "epoch": 1.1733333333333333,
+      "grad_norm": 0.007051107473671436,
+      "learning_rate": 4.6959896507115136e-05,
+      "loss": 0.0019,
       "step": 363
     },
     {
-      "epoch": 3.3703703703703702,
-      "grad_norm": 0.01964252069592476,
-      "learning_rate": 2.901566050047855e-05,
-      "loss": 0.0113,
+      "epoch": 1.1765656565656566,
+      "grad_norm": 0.004829261917620897,
+      "learning_rate": 4.7089262613195346e-05,
+      "loss": 0.001,
       "step": 364
     },
     {
-      "epoch": 3.3796296296296298,
-      "grad_norm": 0.020809266716241837,
-      "learning_rate": 2.872273406751664e-05,
-      "loss": 0.0105,
-      "step": 365
-    },
-    {
-      "epoch": 3.3796296296296298,
-      "eval_loss": 0.015391937457025051,
-      "eval_runtime": 9.111,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.1797979797979798,
+      "grad_norm": 0.005092508625239134,
+      "learning_rate": 4.7218628719275556e-05,
+      "loss": 0.0023,
       "step": 365
     },
     {
-      "epoch": 3.388888888888889,
-      "grad_norm": 0.025048566982150078,
-      "learning_rate": 2.8430696715937337e-05,
-      "loss": 0.0107,
+      "epoch": 1.183030303030303,
+      "grad_norm": 0.004417267628014088,
+      "learning_rate": 4.734799482535576e-05,
+      "loss": 0.001,
       "step": 366
     },
     {
-      "epoch": 3.398148148148148,
-      "grad_norm": 0.024674881249666214,
-      "learning_rate": 2.8139560648669962e-05,
-      "loss": 0.0113,
+      "epoch": 1.1862626262626264,
+      "grad_norm": 0.005750073119997978,
+      "learning_rate": 4.747736093143596e-05,
+      "loss": 0.0013,
       "step": 367
     },
     {
-      "epoch": 3.4074074074074074,
-      "grad_norm": 0.025468124076724052,
-      "learning_rate": 2.7849338030983257e-05,
-      "loss": 0.012,
+      "epoch": 1.1894949494949496,
+      "grad_norm": 0.0033691064454615116,
+      "learning_rate": 4.760672703751617e-05,
+      "loss": 0.0008,
       "step": 368
     },
     {
-      "epoch": 3.4166666666666665,
-      "grad_norm": 0.022864418104290962,
-      "learning_rate": 2.7560040989976892e-05,
-      "loss": 0.01,
+      "epoch": 1.1927272727272726,
+      "grad_norm": 0.0044763232581317425,
+      "learning_rate": 4.773609314359638e-05,
+      "loss": 0.0009,
       "step": 369
     },
     {
-      "epoch": 3.425925925925926,
-      "grad_norm": 0.02258789725601673,
-      "learning_rate": 2.7271681614074973e-05,
-      "loss": 0.0121,
-      "step": 370
-    },
-    {
-      "epoch": 3.425925925925926,
-      "eval_loss": 0.015503110364079475,
-      "eval_runtime": 9.1077,
-      "eval_samples_per_second": 5.49,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.195959595959596,
+      "grad_norm": 0.007102371193468571,
+      "learning_rate": 4.786545924967659e-05,
+      "loss": 0.0012,
       "step": 370
     },
     {
-      "epoch": 3.435185185185185,
-      "grad_norm": 0.025097696110606194,
-      "learning_rate": 2.6984271952520722e-05,
-      "loss": 0.0104,
+      "epoch": 1.1991919191919191,
+      "grad_norm": 0.0041382270865142345,
+      "learning_rate": 4.79948253557568e-05,
+      "loss": 0.001,
       "step": 371
     },
     {
-      "epoch": 3.4444444444444446,
-      "grad_norm": 0.028177309781312943,
-      "learning_rate": 2.6697824014873075e-05,
-      "loss": 0.0132,
+      "epoch": 1.2024242424242424,
+      "grad_norm": 0.006266591139137745,
+      "learning_rate": 4.8124191461837e-05,
+      "loss": 0.0015,
       "step": 372
     },
     {
-      "epoch": 3.4537037037037037,
-      "grad_norm": 0.026587417349219322,
-      "learning_rate": 2.641234977050484e-05,
-      "loss": 0.0085,
+      "epoch": 1.2056565656565656,
+      "grad_norm": 0.00487930653616786,
+      "learning_rate": 4.8253557567917204e-05,
+      "loss": 0.0012,
       "step": 373
     },
     {
-      "epoch": 3.462962962962963,
-      "grad_norm": 0.0189076978713274,
-      "learning_rate": 2.612786114810255e-05,
-      "loss": 0.0096,
+      "epoch": 1.208888888888889,
+      "grad_norm": 0.005311232525855303,
+      "learning_rate": 4.8382923673997415e-05,
+      "loss": 0.0014,
       "step": 374
     },
     {
-      "epoch": 3.4722222222222223,
-      "grad_norm": 0.029332995414733887,
-      "learning_rate": 2.5844370035168073e-05,
-      "loss": 0.0096,
+      "epoch": 1.2121212121212122,
+      "grad_norm": 0.004289830103516579,
+      "learning_rate": 4.8512289780077625e-05,
+      "loss": 0.0009,
       "step": 375
     },
     {
-      "epoch": 3.4722222222222223,
-      "eval_loss": 0.015461472794413567,
-      "eval_runtime": 9.1144,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.2121212121212122,
+      "eval_loss": 0.0011982680298388004,
+      "eval_runtime": 18.7329,
+      "eval_samples_per_second": 5.338,
+      "eval_steps_per_second": 1.335,
       "step": 375
     },
     {
-      "epoch": 3.4814814814814814,
-      "grad_norm": 0.02185731939971447,
-      "learning_rate": 2.5561888277521794e-05,
-      "loss": 0.0098,
+      "epoch": 1.2153535353535354,
+      "grad_norm": 0.0062402524054050446,
+      "learning_rate": 4.864165588615783e-05,
+      "loss": 0.0011,
       "step": 376
     },
     {
-      "epoch": 3.490740740740741,
-      "grad_norm": 0.026887575164437294,
-      "learning_rate": 2.528042767880766e-05,
-      "loss": 0.0114,
+      "epoch": 1.2185858585858587,
+      "grad_norm": 0.0034229829907417297,
+      "learning_rate": 4.877102199223804e-05,
+      "loss": 0.0008,
       "step": 377
     },
     {
-      "epoch": 3.5,
-      "grad_norm": 0.023131586611270905,
-      "learning_rate": 2.500000000000001e-05,
-      "loss": 0.0112,
+      "epoch": 1.221818181818182,
+      "grad_norm": 0.0063551426865160465,
+      "learning_rate": 4.890038809831824e-05,
+      "loss": 0.0027,
       "step": 378
     },
     {
-      "epoch": 3.5092592592592595,
-      "grad_norm": 0.028937749564647675,
-      "learning_rate": 2.4720616958912053e-05,
-      "loss": 0.0121,
+      "epoch": 1.225050505050505,
+      "grad_norm": 0.004902714863419533,
+      "learning_rate": 4.9029754204398446e-05,
+      "loss": 0.0015,
       "step": 379
     },
     {
-      "epoch": 3.5185185185185186,
-      "grad_norm": 0.032668791711330414,
-      "learning_rate": 2.4442290229706344e-05,
-      "loss": 0.0112,
-      "step": 380
-    },
-    {
-      "epoch": 3.5185185185185186,
-      "eval_loss": 0.015212837606668472,
-      "eval_runtime": 9.1177,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.2282828282828282,
+      "grad_norm": 0.0044431439600884914,
+      "learning_rate": 4.9159120310478656e-05,
+      "loss": 0.0005,
       "step": 380
     },
     {
-      "epoch": 3.5277777777777777,
-      "grad_norm": 0.02449023723602295,
-      "learning_rate": 2.4165031442406855e-05,
-      "loss": 0.0117,
+      "epoch": 1.2315151515151515,
+      "grad_norm": 0.007968113757669926,
+      "learning_rate": 4.9288486416558866e-05,
+      "loss": 0.0013,
       "step": 381
     },
     {
-      "epoch": 3.537037037037037,
-      "grad_norm": 0.025157004594802856,
-      "learning_rate": 2.3888852182413085e-05,
-      "loss": 0.0091,
+      "epoch": 1.2347474747474747,
+      "grad_norm": 0.007964645512402058,
+      "learning_rate": 4.941785252263907e-05,
+      "loss": 0.0033,
       "step": 382
     },
     {
-      "epoch": 3.5462962962962963,
-      "grad_norm": 0.03108743578195572,
-      "learning_rate": 2.361376399001592e-05,
-      "loss": 0.0108,
+      "epoch": 1.237979797979798,
+      "grad_norm": 0.004984153900295496,
+      "learning_rate": 4.954721862871928e-05,
+      "loss": 0.0008,
       "step": 383
     },
     {
-      "epoch": 3.5555555555555554,
-      "grad_norm": 0.021932488307356834,
-      "learning_rate": 2.333977835991545e-05,
-      "loss": 0.0093,
+      "epoch": 1.2412121212121212,
+      "grad_norm": 0.005040554329752922,
+      "learning_rate": 4.9676584734799483e-05,
+      "loss": 0.0008,
       "step": 384
     },
     {
-      "epoch": 3.564814814814815,
-      "grad_norm": 0.026496881619095802,
-      "learning_rate": 2.3066906740740623e-05,
-      "loss": 0.0118,
-      "step": 385
-    },
-    {
-      "epoch": 3.564814814814815,
-      "eval_loss": 0.01467986311763525,
-      "eval_runtime": 9.1127,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.2444444444444445,
+      "grad_norm": 0.005269059911370277,
+      "learning_rate": 4.980595084087969e-05,
+      "loss": 0.0016,
       "step": 385
     },
     {
-      "epoch": 3.574074074074074,
-      "grad_norm": 0.024211710318922997,
-      "learning_rate": 2.2795160534570864e-05,
-      "loss": 0.0086,
+      "epoch": 1.2476767676767677,
+      "grad_norm": 0.004820747766643763,
+      "learning_rate": 4.99353169469599e-05,
+      "loss": 0.0013,
       "step": 386
     },
     {
-      "epoch": 3.5833333333333335,
-      "grad_norm": 0.023977207019925117,
-      "learning_rate": 2.25245510964597e-05,
-      "loss": 0.0128,
+      "epoch": 1.250909090909091,
+      "grad_norm": 0.003650473430752754,
+      "learning_rate": 5.006468305304011e-05,
+      "loss": 0.001,
       "step": 387
     },
     {
-      "epoch": 3.5925925925925926,
-      "grad_norm": 0.02136526070535183,
-      "learning_rate": 2.225508973396016e-05,
-      "loss": 0.0121,
+      "epoch": 1.2541414141414142,
+      "grad_norm": 0.003410038072615862,
+      "learning_rate": 5.019404915912032e-05,
+      "loss": 0.0007,
       "step": 388
     },
     {
-      "epoch": 3.601851851851852,
-      "grad_norm": 0.026328187435865402,
-      "learning_rate": 2.198678770665238e-05,
-      "loss": 0.0108,
+      "epoch": 1.2573737373737375,
+      "grad_norm": 0.00503958947956562,
+      "learning_rate": 5.032341526520052e-05,
+      "loss": 0.0013,
       "step": 389
     },
     {
-      "epoch": 3.611111111111111,
-      "grad_norm": 0.02159940078854561,
-      "learning_rate": 2.171965622567308e-05,
-      "loss": 0.0082,
-      "step": 390
-    },
-    {
-      "epoch": 3.611111111111111,
-      "eval_loss": 0.014544774778187275,
-      "eval_runtime": 9.1133,
-      "eval_samples_per_second": 5.487,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.2606060606060607,
+      "grad_norm": 0.006551372352987528,
+      "learning_rate": 5.0452781371280725e-05,
+      "loss": 0.0017,
       "step": 390
     },
     {
-      "epoch": 3.6203703703703702,
-      "grad_norm": 0.02303987927734852,
-      "learning_rate": 2.1453706453247087e-05,
-      "loss": 0.0092,
+      "epoch": 1.2638383838383838,
+      "grad_norm": 0.004784473218023777,
+      "learning_rate": 5.058214747736093e-05,
+      "loss": 0.0011,
       "step": 391
     },
     {
-      "epoch": 3.6296296296296298,
-      "grad_norm": 0.027734337374567986,
-      "learning_rate": 2.1188949502220983e-05,
-      "loss": 0.0101,
+      "epoch": 1.267070707070707,
+      "grad_norm": 0.009283591993153095,
+      "learning_rate": 5.071151358344114e-05,
+      "loss": 0.0017,
       "step": 392
     },
     {
-      "epoch": 3.638888888888889,
-      "grad_norm": 0.02069096453487873,
-      "learning_rate": 2.0925396435598664e-05,
-      "loss": 0.0111,
+      "epoch": 1.2703030303030303,
+      "grad_norm": 0.005054951179772615,
+      "learning_rate": 5.084087968952135e-05,
+      "loss": 0.0014,
       "step": 393
     },
     {
-      "epoch": 3.648148148148148,
-      "grad_norm": 0.02777431532740593,
-      "learning_rate": 2.066305826607911e-05,
-      "loss": 0.0091,
+      "epoch": 1.2735353535353535,
+      "grad_norm": 0.007423613220453262,
+      "learning_rate": 5.097024579560156e-05,
+      "loss": 0.0012,
       "step": 394
     },
     {
-      "epoch": 3.6574074074074074,
-      "grad_norm": 0.02333620935678482,
-      "learning_rate": 2.0401945955596206e-05,
-      "loss": 0.0112,
-      "step": 395
-    },
-    {
-      "epoch": 3.6574074074074074,
-      "eval_loss": 0.01460795197635889,
-      "eval_runtime": 9.1059,
-      "eval_samples_per_second": 5.491,
-      "eval_steps_per_second": 1.428,
+      "epoch": 1.2767676767676768,
+      "grad_norm": 0.004986994434148073,
+      "learning_rate": 5.109961190168176e-05,
+      "loss": 0.0017,
       "step": 395
     },
     {
-      "epoch": 3.6666666666666665,
-      "grad_norm": 0.022142188623547554,
-      "learning_rate": 2.0142070414860704e-05,
-      "loss": 0.01,
+      "epoch": 1.28,
+      "grad_norm": 0.004347431473433971,
+      "learning_rate": 5.122897800776197e-05,
+      "loss": 0.0012,
       "step": 396
     },
     {
-      "epoch": 3.675925925925926,
-      "grad_norm": 0.01749616675078869,
-      "learning_rate": 1.9883442502904283e-05,
-      "loss": 0.0095,
+      "epoch": 1.2832323232323233,
+      "grad_norm": 0.006622066255658865,
+      "learning_rate": 5.135834411384217e-05,
+      "loss": 0.0015,
       "step": 397
     },
     {
-      "epoch": 3.685185185185185,
-      "grad_norm": 0.02393367514014244,
-      "learning_rate": 1.9626073026625818e-05,
-      "loss": 0.0095,
+      "epoch": 1.2864646464646465,
+      "grad_norm": 0.006094200070947409,
+      "learning_rate": 5.148771021992238e-05,
+      "loss": 0.0014,
       "step": 398
     },
     {
-      "epoch": 3.6944444444444446,
-      "grad_norm": 0.023465050384402275,
-      "learning_rate": 1.936997274033986e-05,
-      "loss": 0.0108,
+      "epoch": 1.2896969696969696,
+      "grad_norm": 0.004648863337934017,
+      "learning_rate": 5.161707632600259e-05,
+      "loss": 0.0012,
       "step": 399
     },
     {
-      "epoch": 3.7037037037037037,
-      "grad_norm": 0.023157304152846336,
-      "learning_rate": 1.9115152345327152e-05,
-      "loss": 0.0086,
+      "epoch": 1.2929292929292928,
+      "grad_norm": 0.005023527424782515,
+      "learning_rate": 5.17464424320828e-05,
+      "loss": 0.0015,
       "step": 400
     },
     {
-      "epoch": 3.7037037037037037,
-      "eval_loss": 0.014902754686772823,
-      "eval_runtime": 9.1616,
-      "eval_samples_per_second": 5.458,
-      "eval_steps_per_second": 1.419,
+      "epoch": 1.2929292929292928,
+      "eval_loss": 0.001097235712222755,
+      "eval_runtime": 18.6509,
+      "eval_samples_per_second": 5.362,
+      "eval_steps_per_second": 1.34,
       "step": 400
     },
     {
-      "epoch": 3.712962962962963,
-      "grad_norm": 0.021799901500344276,
-      "learning_rate": 1.8861622489387555e-05,
-      "loss": 0.0128,
+      "epoch": 1.296161616161616,
+      "grad_norm": 0.004932883661240339,
+      "learning_rate": 5.1875808538163004e-05,
+      "loss": 0.0012,
       "step": 401
     },
     {
-      "epoch": 3.7222222222222223,
-      "grad_norm": 0.03070679120719433,
-      "learning_rate": 1.8609393766395085e-05,
-      "loss": 0.0123,
+      "epoch": 1.2993939393939393,
+      "grad_norm": 0.004028505180031061,
+      "learning_rate": 5.2005174644243214e-05,
+      "loss": 0.0013,
       "step": 402
     },
     {
-      "epoch": 3.7314814814814814,
-      "grad_norm": 0.02543518878519535,
-      "learning_rate": 1.835847671585526e-05,
-      "loss": 0.0114,
+      "epoch": 1.3026262626262626,
+      "grad_norm": 0.004800683818757534,
+      "learning_rate": 5.213454075032341e-05,
+      "loss": 0.0014,
       "step": 403
     },
     {
-      "epoch": 3.7407407407407405,
-      "grad_norm": 0.027585655450820923,
-      "learning_rate": 1.8108881822464696e-05,
-      "loss": 0.0099,
+      "epoch": 1.3058585858585858,
+      "grad_norm": 0.0042937519028782845,
+      "learning_rate": 5.226390685640362e-05,
+      "loss": 0.001,
       "step": 404
     },
     {
-      "epoch": 3.75,
-      "grad_norm": 0.02352389506995678,
-      "learning_rate": 1.7860619515673033e-05,
-      "loss": 0.0102,
-      "step": 405
-    },
-    {
-      "epoch": 3.75,
-      "eval_loss": 0.014981208369135857,
-      "eval_runtime": 9.1106,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.309090909090909,
+      "grad_norm": 0.004741652403026819,
+      "learning_rate": 5.239327296248383e-05,
+      "loss": 0.0011,
       "step": 405
     },
     {
-      "epoch": 3.7592592592592595,
-      "grad_norm": 0.02560283988714218,
-      "learning_rate": 1.7613700169247056e-05,
-      "loss": 0.012,
+      "epoch": 1.3123232323232323,
+      "grad_norm": 0.004214128013700247,
+      "learning_rate": 5.252263906856404e-05,
+      "loss": 0.0009,
       "step": 406
     },
     {
-      "epoch": 3.7685185185185186,
-      "grad_norm": 0.026089752092957497,
-      "learning_rate": 1.7368134100837287e-05,
-      "loss": 0.0088,
+      "epoch": 1.3155555555555556,
+      "grad_norm": 0.004764182958751917,
+      "learning_rate": 5.2652005174644245e-05,
+      "loss": 0.0011,
       "step": 407
     },
     {
-      "epoch": 3.7777777777777777,
-      "grad_norm": 0.030365899205207825,
-      "learning_rate": 1.7123931571546827e-05,
-      "loss": 0.0119,
+      "epoch": 1.3187878787878788,
+      "grad_norm": 0.005405279342085123,
+      "learning_rate": 5.2781371280724455e-05,
+      "loss": 0.0011,
       "step": 408
     },
     {
-      "epoch": 3.787037037037037,
-      "grad_norm": 0.031558796763420105,
-      "learning_rate": 1.6881102785502616e-05,
-      "loss": 0.011,
+      "epoch": 1.322020202020202,
+      "grad_norm": 0.005355632398277521,
+      "learning_rate": 5.2910737386804666e-05,
+      "loss": 0.0012,
       "step": 409
     },
     {
-      "epoch": 3.7962962962962963,
-      "grad_norm": 0.030366325750947,
-      "learning_rate": 1.6639657889429018e-05,
-      "loss": 0.0116,
-      "step": 410
-    },
-    {
-      "epoch": 3.7962962962962963,
-      "eval_loss": 0.014859426766633987,
-      "eval_runtime": 9.1059,
-      "eval_samples_per_second": 5.491,
-      "eval_steps_per_second": 1.428,
+      "epoch": 1.3252525252525253,
+      "grad_norm": 0.004391840659081936,
+      "learning_rate": 5.304010349288486e-05,
+      "loss": 0.0011,
       "step": 410
     },
     {
-      "epoch": 3.8055555555555554,
-      "grad_norm": 0.025008074939250946,
-      "learning_rate": 1.639960697222388e-05,
-      "loss": 0.0106,
+      "epoch": 1.3284848484848486,
+      "grad_norm": 0.002719137817621231,
+      "learning_rate": 5.316946959896507e-05,
+      "loss": 0.0005,
       "step": 411
     },
     {
-      "epoch": 3.814814814814815,
-      "grad_norm": 0.028196556493639946,
-      "learning_rate": 1.6160960064536908e-05,
-      "loss": 0.0113,
+      "epoch": 1.3317171717171719,
+      "grad_norm": 0.005910648964345455,
+      "learning_rate": 5.329883570504528e-05,
+      "loss": 0.0014,
       "step": 412
     },
     {
-      "epoch": 3.824074074074074,
-      "grad_norm": 0.02165764756500721,
-      "learning_rate": 1.592372713835055e-05,
-      "loss": 0.0115,
+      "epoch": 1.3349494949494949,
+      "grad_norm": 0.004575495608150959,
+      "learning_rate": 5.3428201811125486e-05,
+      "loss": 0.0013,
       "step": 413
     },
     {
-      "epoch": 3.8333333333333335,
-      "grad_norm": 0.020175475627183914,
-      "learning_rate": 1.5687918106563326e-05,
-      "loss": 0.0112,
+      "epoch": 1.3381818181818181,
+      "grad_norm": 0.004968920256942511,
+      "learning_rate": 5.35575679172057e-05,
+      "loss": 0.0012,
       "step": 414
     },
     {
-      "epoch": 3.8425925925925926,
-      "grad_norm": 0.027304671704769135,
-      "learning_rate": 1.545354282257562e-05,
-      "loss": 0.0126,
-      "step": 415
-    },
-    {
-      "epoch": 3.8425925925925926,
-      "eval_loss": 0.014735485427081585,
-      "eval_runtime": 9.198,
-      "eval_samples_per_second": 5.436,
-      "eval_steps_per_second": 1.413,
+      "epoch": 1.3414141414141414,
+      "grad_norm": 0.005680674687027931,
+      "learning_rate": 5.368693402328591e-05,
+      "loss": 0.0015,
       "step": 415
     },
     {
-      "epoch": 3.851851851851852,
-      "grad_norm": 0.026429716497659683,
-      "learning_rate": 1.52206110798779e-05,
-      "loss": 0.0103,
+      "epoch": 1.3446464646464646,
+      "grad_norm": 0.005805244669318199,
+      "learning_rate": 5.3816300129366104e-05,
+      "loss": 0.0017,
       "step": 416
     },
     {
-      "epoch": 3.861111111111111,
-      "grad_norm": 0.02409077063202858,
-      "learning_rate": 1.4989132611641576e-05,
-      "loss": 0.012,
+      "epoch": 1.347878787878788,
+      "grad_norm": 0.004365504253655672,
+      "learning_rate": 5.3945666235446314e-05,
+      "loss": 0.001,
       "step": 417
     },
     {
-      "epoch": 3.8703703703703702,
-      "grad_norm": 0.02310461364686489,
-      "learning_rate": 1.4759117090312197e-05,
-      "loss": 0.0096,
+      "epoch": 1.3511111111111112,
+      "grad_norm": 0.004444981925189495,
+      "learning_rate": 5.4075032341526524e-05,
+      "loss": 0.0011,
       "step": 418
     },
     {
-      "epoch": 3.8796296296296298,
-      "grad_norm": 0.026219584047794342,
-      "learning_rate": 1.453057412720536e-05,
-      "loss": 0.0094,
+      "epoch": 1.3543434343434344,
+      "grad_norm": 0.0048050908371806145,
+      "learning_rate": 5.420439844760673e-05,
+      "loss": 0.0011,
       "step": 419
     },
     {
-      "epoch": 3.888888888888889,
-      "grad_norm": 0.027541201561689377,
-      "learning_rate": 1.4303513272105057e-05,
-      "loss": 0.0112,
-      "step": 420
-    },
-    {
-      "epoch": 3.888888888888889,
-      "eval_loss": 0.014594363048672676,
-      "eval_runtime": 9.1304,
-      "eval_samples_per_second": 5.476,
-      "eval_steps_per_second": 1.424,
+      "epoch": 1.3575757575757577,
+      "grad_norm": 0.0112074613571167,
+      "learning_rate": 5.433376455368694e-05,
+      "loss": 0.0009,
       "step": 420
     },
     {
-      "epoch": 3.898148148148148,
-      "grad_norm": 0.024942217394709587,
-      "learning_rate": 1.4077944012864636e-05,
-      "loss": 0.0093,
+      "epoch": 1.3608080808080807,
+      "grad_norm": 0.005805397406220436,
+      "learning_rate": 5.446313065976715e-05,
+      "loss": 0.001,
       "step": 421
     },
     {
-      "epoch": 3.9074074074074074,
-      "grad_norm": 0.018137283623218536,
-      "learning_rate": 1.3853875775010355e-05,
-      "loss": 0.0102,
+      "epoch": 1.364040404040404,
+      "grad_norm": 0.006793915294110775,
+      "learning_rate": 5.4592496765847345e-05,
+      "loss": 0.0018,
       "step": 422
     },
     {
-      "epoch": 3.9166666666666665,
-      "grad_norm": 0.021817779168486595,
-      "learning_rate": 1.3631317921347563e-05,
-      "loss": 0.0084,
+      "epoch": 1.3672727272727272,
+      "grad_norm": 0.005133500322699547,
+      "learning_rate": 5.4721862871927555e-05,
+      "loss": 0.0015,
       "step": 423
     },
     {
-      "epoch": 3.925925925925926,
-      "grad_norm": 0.023799235001206398,
-      "learning_rate": 1.3410279751569399e-05,
-      "loss": 0.0122,
+      "epoch": 1.3705050505050504,
+      "grad_norm": 0.0034578884951770306,
+      "learning_rate": 5.4851228978007766e-05,
+      "loss": 0.0008,
       "step": 424
     },
     {
-      "epoch": 3.935185185185185,
-      "grad_norm": 0.030764896422624588,
-      "learning_rate": 1.3190770501868243e-05,
-      "loss": 0.0107,
+      "epoch": 1.3737373737373737,
+      "grad_norm": 0.005579716991633177,
+      "learning_rate": 5.498059508408797e-05,
+      "loss": 0.0025,
       "step": 425
     },
     {
-      "epoch": 3.935185185185185,
-      "eval_loss": 0.014631365425884724,
-      "eval_runtime": 9.1149,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.3737373737373737,
+      "eval_loss": 0.0011115286033600569,
+      "eval_runtime": 18.7263,
+      "eval_samples_per_second": 5.34,
+      "eval_steps_per_second": 1.335,
       "step": 425
     },
     {
-      "epoch": 3.9444444444444446,
-      "grad_norm": 0.022886106744408607,
-      "learning_rate": 1.297279934454978e-05,
-      "loss": 0.0096,
+      "epoch": 1.376969696969697,
+      "grad_norm": 0.005262942984700203,
+      "learning_rate": 5.510996119016818e-05,
+      "loss": 0.0011,
       "step": 426
     },
     {
-      "epoch": 3.9537037037037037,
-      "grad_norm": 0.03152737021446228,
-      "learning_rate": 1.2756375387649716e-05,
-      "loss": 0.0124,
+      "epoch": 1.3802020202020202,
+      "grad_norm": 0.004813515581190586,
+      "learning_rate": 5.523932729624839e-05,
+      "loss": 0.0011,
       "step": 427
     },
     {
-      "epoch": 3.962962962962963,
-      "grad_norm": 0.02872036211192608,
-      "learning_rate": 1.25415076745532e-05,
-      "loss": 0.0091,
+      "epoch": 1.3834343434343435,
+      "grad_norm": 0.005556623917073011,
+      "learning_rate": 5.53686934023286e-05,
+      "loss": 0.0014,
       "step": 428
     },
     {
-      "epoch": 3.9722222222222223,
-      "grad_norm": 0.021184636279940605,
-      "learning_rate": 1.2328205183616965e-05,
-      "loss": 0.0105,
+      "epoch": 1.3866666666666667,
+      "grad_norm": 0.0028185530100017786,
+      "learning_rate": 5.5498059508408797e-05,
+      "loss": 0.0007,
       "step": 429
     },
     {
-      "epoch": 3.9814814814814814,
-      "grad_norm": 0.02112959884107113,
-      "learning_rate": 1.2116476827794104e-05,
-      "loss": 0.0113,
-      "step": 430
-    },
-    {
-      "epoch": 3.9814814814814814,
-      "eval_loss": 0.01471536885946989,
-      "eval_runtime": 9.116,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.38989898989899,
+      "grad_norm": 0.004296639934182167,
+      "learning_rate": 5.562742561448901e-05,
+      "loss": 0.001,
       "step": 430
     },
     {
-      "epoch": 3.9907407407407405,
-      "grad_norm": 0.019945990294218063,
-      "learning_rate": 1.1906331454261704e-05,
-      "loss": 0.0093,
+      "epoch": 1.3931313131313132,
+      "grad_norm": 0.004092990420758724,
+      "learning_rate": 5.575679172056921e-05,
+      "loss": 0.001,
       "step": 431
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 0.023910805583000183,
-      "learning_rate": 1.1697777844051105e-05,
-      "loss": 0.011,
+      "epoch": 1.3963636363636365,
+      "grad_norm": 0.005713030230253935,
+      "learning_rate": 5.588615782664942e-05,
+      "loss": 0.0015,
       "step": 432
     },
     {
-      "epoch": 4.0092592592592595,
-      "grad_norm": 0.01957758143544197,
-      "learning_rate": 1.1490824711681025e-05,
-      "loss": 0.0094,
+      "epoch": 1.3995959595959597,
+      "grad_norm": 0.005011020693928003,
+      "learning_rate": 5.601552393272963e-05,
+      "loss": 0.0012,
       "step": 433
     },
     {
-      "epoch": 4.018518518518518,
-      "grad_norm": 0.02563118375837803,
-      "learning_rate": 1.1285480704793377e-05,
-      "loss": 0.0093,
+      "epoch": 1.4028282828282828,
+      "grad_norm": 0.004393043462187052,
+      "learning_rate": 5.614489003880984e-05,
+      "loss": 0.0009,
       "step": 434
     },
     {
-      "epoch": 4.027777777777778,
-      "grad_norm": 0.026251764968037605,
-      "learning_rate": 1.1081754403791999e-05,
-      "loss": 0.0091,
-      "step": 435
-    },
-    {
-      "epoch": 4.027777777777778,
-      "eval_loss": 0.014734329655766487,
-      "eval_runtime": 9.1592,
-      "eval_samples_per_second": 5.459,
-      "eval_steps_per_second": 1.419,
+      "epoch": 1.406060606060606,
+      "grad_norm": 0.0047727166675031185,
+      "learning_rate": 5.627425614489004e-05,
+      "loss": 0.0011,
       "step": 435
     },
     {
-      "epoch": 4.037037037037037,
-      "grad_norm": 0.025834446772933006,
-      "learning_rate": 1.0879654321484012e-05,
-      "loss": 0.0067,
+      "epoch": 1.4092929292929293,
+      "grad_norm": 0.0034502504859119654,
+      "learning_rate": 5.640362225097025e-05,
+      "loss": 0.001,
       "step": 436
     },
     {
-      "epoch": 4.046296296296297,
-      "grad_norm": 0.0185233224183321,
-      "learning_rate": 1.0679188902724191e-05,
-      "loss": 0.0108,
+      "epoch": 1.4125252525252525,
+      "grad_norm": 0.004155455157160759,
+      "learning_rate": 5.653298835705045e-05,
+      "loss": 0.0008,
       "step": 437
     },
     {
-      "epoch": 4.055555555555555,
-      "grad_norm": 0.021918736398220062,
-      "learning_rate": 1.0480366524062042e-05,
-      "loss": 0.0088,
+      "epoch": 1.4157575757575758,
+      "grad_norm": 0.005455946549773216,
+      "learning_rate": 5.666235446313066e-05,
+      "loss": 0.0012,
       "step": 438
     },
     {
-      "epoch": 4.064814814814815,
-      "grad_norm": 0.03142661973834038,
-      "learning_rate": 1.0283195493391823e-05,
-      "loss": 0.0103,
+      "epoch": 1.418989898989899,
+      "grad_norm": 0.005093955434858799,
+      "learning_rate": 5.679172056921087e-05,
+      "loss": 0.0015,
       "step": 439
     },
     {
-      "epoch": 4.074074074074074,
-      "grad_norm": 0.023410873487591743,
-      "learning_rate": 1.008768404960535e-05,
-      "loss": 0.0094,
+      "epoch": 1.4222222222222223,
+      "grad_norm": 0.004076706245541573,
+      "learning_rate": 5.692108667529108e-05,
+      "loss": 0.0012,
       "step": 440
     },
     {
-      "epoch": 4.074074074074074,
-      "eval_loss": 0.014965096488595009,
-      "eval_runtime": 9.1135,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
-      "step": 440
-    },
-    {
-      "epoch": 4.083333333333333,
-      "grad_norm": 0.02943902276456356,
-      "learning_rate": 9.893840362247809e-06,
-      "loss": 0.0056,
+      "epoch": 1.4254545454545455,
+      "grad_norm": 0.004360303282737732,
+      "learning_rate": 5.7050452781371286e-05,
+      "loss": 0.001,
       "step": 441
     },
     {
-      "epoch": 4.092592592592593,
-      "grad_norm": 0.021431270986795425,
-      "learning_rate": 9.701672531176286e-06,
-      "loss": 0.0089,
+      "epoch": 1.4286868686868686,
+      "grad_norm": 0.005434586200863123,
+      "learning_rate": 5.717981888745149e-05,
+      "loss": 0.002,
       "step": 442
     },
     {
-      "epoch": 4.101851851851852,
-      "grad_norm": 0.02797669917345047,
-      "learning_rate": 9.511188586221376e-06,
-      "loss": 0.0092,
+      "epoch": 1.4319191919191918,
+      "grad_norm": 0.006644332781434059,
+      "learning_rate": 5.730918499353169e-05,
+      "loss": 0.0033,
       "step": 443
     },
     {
-      "epoch": 4.111111111111111,
-      "grad_norm": 0.02437691204249859,
-      "learning_rate": 9.322396486851626e-06,
-      "loss": 0.0104,
+      "epoch": 1.435151515151515,
+      "grad_norm": 0.006485591642558575,
+      "learning_rate": 5.74385510996119e-05,
+      "loss": 0.002,
       "step": 444
     },
     {
-      "epoch": 4.12037037037037,
-      "grad_norm": 0.024811841547489166,
-      "learning_rate": 9.135304121840976e-06,
-      "loss": 0.0096,
-      "step": 445
-    },
-    {
-      "epoch": 4.12037037037037,
-      "eval_loss": 0.014996801503002644,
-      "eval_runtime": 9.1094,
-      "eval_samples_per_second": 5.489,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.4383838383838383,
+      "grad_norm": 0.006898663938045502,
+      "learning_rate": 5.7567917205692113e-05,
+      "loss": 0.001,
       "step": 445
     },
     {
-      "epoch": 4.12962962962963,
-      "grad_norm": 0.0309213325381279,
-      "learning_rate": 8.949919308939082e-06,
-      "loss": 0.0109,
+      "epoch": 1.4416161616161616,
+      "grad_norm": 0.006443925201892853,
+      "learning_rate": 5.7697283311772324e-05,
+      "loss": 0.0015,
       "step": 446
     },
     {
-      "epoch": 4.138888888888889,
-      "grad_norm": 0.023763932287693024,
-      "learning_rate": 8.766249794544662e-06,
-      "loss": 0.0073,
+      "epoch": 1.4448484848484848,
+      "grad_norm": 0.003106453223153949,
+      "learning_rate": 5.782664941785253e-05,
+      "loss": 0.0009,
       "step": 447
     },
     {
-      "epoch": 4.148148148148148,
-      "grad_norm": 0.023741643875837326,
-      "learning_rate": 8.584303253381847e-06,
-      "loss": 0.0105,
+      "epoch": 1.448080808080808,
+      "grad_norm": 0.0036977077834308147,
+      "learning_rate": 5.795601552393273e-05,
+      "loss": 0.0009,
       "step": 448
     },
     {
-      "epoch": 4.157407407407407,
-      "grad_norm": 0.02090543322265148,
-      "learning_rate": 8.404087288179424e-06,
-      "loss": 0.0096,
+      "epoch": 1.4513131313131313,
+      "grad_norm": 0.0050973836332559586,
+      "learning_rate": 5.8085381630012934e-05,
+      "loss": 0.0014,
       "step": 449
     },
     {
-      "epoch": 4.166666666666667,
-      "grad_norm": 0.026315612718462944,
-      "learning_rate": 8.225609429353187e-06,
-      "loss": 0.0091,
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.0057764300145208836,
+      "learning_rate": 5.8214747736093145e-05,
+      "loss": 0.0016,
       "step": 450
     },
     {
-      "epoch": 4.166666666666667,
-      "eval_loss": 0.015186839736998081,
-      "eval_runtime": 9.1241,
-      "eval_samples_per_second": 5.48,
-      "eval_steps_per_second": 1.425,
+      "epoch": 1.4545454545454546,
+      "eval_loss": 0.0010863245697692037,
+      "eval_runtime": 18.7321,
+      "eval_samples_per_second": 5.338,
+      "eval_steps_per_second": 1.335,
       "step": 450
     },
     {
-      "epoch": 4.175925925925926,
-      "grad_norm": 0.023099206387996674,
-      "learning_rate": 8.048877134691268e-06,
-      "loss": 0.0091,
+      "epoch": 1.4577777777777778,
+      "grad_norm": 0.004166171886026859,
+      "learning_rate": 5.8344113842173355e-05,
+      "loss": 0.0009,
       "step": 451
     },
     {
-      "epoch": 4.185185185185185,
-      "grad_norm": 0.027901167050004005,
-      "learning_rate": 7.873897789042523e-06,
-      "loss": 0.0092,
+      "epoch": 1.461010101010101,
+      "grad_norm": 0.004900315310806036,
+      "learning_rate": 5.8473479948253565e-05,
+      "loss": 0.0013,
       "step": 452
     },
     {
-      "epoch": 4.194444444444445,
-      "grad_norm": 0.025486482307314873,
-      "learning_rate": 7.700678704007947e-06,
-      "loss": 0.0077,
+      "epoch": 1.4642424242424243,
+      "grad_norm": 0.003387290518730879,
+      "learning_rate": 5.860284605433377e-05,
+      "loss": 0.0007,
       "step": 453
     },
     {
-      "epoch": 4.203703703703703,
-      "grad_norm": 0.0233286302536726,
-      "learning_rate": 7.529227117635135e-06,
-      "loss": 0.0077,
+      "epoch": 1.4674747474747476,
+      "grad_norm": 0.004271840676665306,
+      "learning_rate": 5.873221216041398e-05,
+      "loss": 0.0011,
       "step": 454
     },
     {
-      "epoch": 4.212962962962963,
-      "grad_norm": 0.023314587771892548,
-      "learning_rate": 7.35955019411585e-06,
-      "loss": 0.0089,
+      "epoch": 1.4707070707070706,
+      "grad_norm": 0.0032306541688740253,
+      "learning_rate": 5.8861578266494176e-05,
+      "loss": 0.0007,
       "step": 455
     },
     {
-      "epoch": 4.212962962962963,
-      "eval_loss": 0.015497377142310143,
-      "eval_runtime": 9.1064,
-      "eval_samples_per_second": 5.491,
-      "eval_steps_per_second": 1.428,
-      "step": 455
-    },
-    {
-      "epoch": 4.222222222222222,
-      "grad_norm": 0.021640775725245476,
-      "learning_rate": 7.191655023486682e-06,
-      "loss": 0.01,
+      "epoch": 1.4739393939393939,
+      "grad_norm": 0.0022131800651550293,
+      "learning_rate": 5.8990944372574386e-05,
+      "loss": 0.0006,
       "step": 456
     },
     {
-      "epoch": 4.231481481481482,
-      "grad_norm": 0.027831410989165306,
-      "learning_rate": 7.02554862133275e-06,
-      "loss": 0.0105,
+      "epoch": 1.4771717171717171,
+      "grad_norm": 0.005533823277801275,
+      "learning_rate": 5.9120310478654596e-05,
+      "loss": 0.0017,
       "step": 457
     },
     {
-      "epoch": 4.2407407407407405,
-      "grad_norm": 0.023242153227329254,
-      "learning_rate": 6.861237928494579e-06,
-      "loss": 0.009,
+      "epoch": 1.4804040404040404,
+      "grad_norm": 0.004672409035265446,
+      "learning_rate": 5.9249676584734806e-05,
+      "loss": 0.0014,
       "step": 458
     },
     {
-      "epoch": 4.25,
-      "grad_norm": 0.02775505743920803,
-      "learning_rate": 6.698729810778065e-06,
-      "loss": 0.0102,
+      "epoch": 1.4836363636363636,
+      "grad_norm": 0.006192138884216547,
+      "learning_rate": 5.937904269081501e-05,
+      "loss": 0.0015,
       "step": 459
     },
     {
-      "epoch": 4.2592592592592595,
-      "grad_norm": 0.0267843846231699,
-      "learning_rate": 6.53803105866761e-06,
-      "loss": 0.0063,
-      "step": 460
-    },
-    {
-      "epoch": 4.2592592592592595,
-      "eval_loss": 0.01563325710594654,
-      "eval_runtime": 9.111,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.486868686868687,
+      "grad_norm": 0.011036567389965057,
+      "learning_rate": 5.950840879689522e-05,
+      "loss": 0.0025,
       "step": 460
     },
     {
-      "epoch": 4.268518518518518,
-      "grad_norm": 0.02488654851913452,
-      "learning_rate": 6.379148387042316e-06,
-      "loss": 0.01,
+      "epoch": 1.4901010101010101,
+      "grad_norm": 0.0055129267275333405,
+      "learning_rate": 5.963777490297542e-05,
+      "loss": 0.0015,
       "step": 461
     },
     {
-      "epoch": 4.277777777777778,
-      "grad_norm": 0.024208445101976395,
-      "learning_rate": 6.222088434895462e-06,
-      "loss": 0.0072,
+      "epoch": 1.4933333333333334,
+      "grad_norm": 0.005369866266846657,
+      "learning_rate": 5.976714100905563e-05,
+      "loss": 0.0027,
       "step": 462
     },
     {
-      "epoch": 4.287037037037037,
-      "grad_norm": 0.023147890344262123,
-      "learning_rate": 6.066857765057055e-06,
-      "loss": 0.0088,
+      "epoch": 1.4965656565656564,
+      "grad_norm": 0.005017601884901524,
+      "learning_rate": 5.989650711513584e-05,
+      "loss": 0.0011,
       "step": 463
     },
     {
-      "epoch": 4.296296296296296,
-      "grad_norm": 0.029451172798871994,
-      "learning_rate": 5.9134628639196e-06,
-      "loss": 0.0085,
+      "epoch": 1.4997979797979797,
+      "grad_norm": 0.00623415969312191,
+      "learning_rate": 6.002587322121605e-05,
+      "loss": 0.0028,
       "step": 464
     },
     {
-      "epoch": 4.305555555555555,
-      "grad_norm": 0.02764413133263588,
-      "learning_rate": 5.7619101411671095e-06,
-      "loss": 0.0099,
-      "step": 465
-    },
-    {
-      "epoch": 4.305555555555555,
-      "eval_loss": 0.015693385154008865,
-      "eval_runtime": 9.1176,
-      "eval_samples_per_second": 5.484,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.503030303030303,
+      "grad_norm": 0.005604151636362076,
+      "learning_rate": 6.015523932729625e-05,
+      "loss": 0.0023,
       "step": 465
     },
     {
-      "epoch": 4.314814814814815,
-      "grad_norm": 0.021906448528170586,
-      "learning_rate": 5.6122059295072085e-06,
-      "loss": 0.0096,
+      "epoch": 1.5062626262626262,
+      "grad_norm": 0.004366429056972265,
+      "learning_rate": 6.028460543337646e-05,
+      "loss": 0.001,
       "step": 466
     },
     {
-      "epoch": 4.324074074074074,
-      "grad_norm": 0.02385389618575573,
-      "learning_rate": 5.464356484406535e-06,
-      "loss": 0.0072,
+      "epoch": 1.5094949494949494,
+      "grad_norm": 0.004982110112905502,
+      "learning_rate": 6.041397153945667e-05,
+      "loss": 0.0014,
       "step": 467
     },
     {
-      "epoch": 4.333333333333333,
-      "grad_norm": 0.026357507333159447,
-      "learning_rate": 5.318367983829392e-06,
-      "loss": 0.0079,
+      "epoch": 1.5127272727272727,
+      "grad_norm": 0.005147982854396105,
+      "learning_rate": 6.054333764553687e-05,
+      "loss": 0.0013,
       "step": 468
     },
     {
-      "epoch": 4.342592592592593,
-      "grad_norm": 0.026002187281847,
-      "learning_rate": 5.174246527979531e-06,
-      "loss": 0.0095,
+      "epoch": 1.515959595959596,
+      "grad_norm": 0.004790551029145718,
+      "learning_rate": 6.067270375161708e-05,
+      "loss": 0.0011,
       "step": 469
     },
     {
-      "epoch": 4.351851851851852,
-      "grad_norm": 0.02679777517914772,
-      "learning_rate": 5.031998139045352e-06,
-      "loss": 0.0085,
-      "step": 470
-    },
-    {
-      "epoch": 4.351851851851852,
-      "eval_loss": 0.015615792945027351,
-      "eval_runtime": 9.1365,
-      "eval_samples_per_second": 5.473,
-      "eval_steps_per_second": 1.423,
+      "epoch": 1.5191919191919192,
+      "grad_norm": 0.0038922505918890238,
+      "learning_rate": 6.080206985769729e-05,
+      "loss": 0.0011,
       "step": 470
     },
     {
-      "epoch": 4.361111111111111,
-      "grad_norm": 0.023431269451975822,
-      "learning_rate": 4.891628760948114e-06,
-      "loss": 0.009,
+      "epoch": 1.5224242424242425,
+      "grad_norm": 0.006303661502897739,
+      "learning_rate": 6.093143596377749e-05,
+      "loss": 0.0023,
       "step": 471
     },
     {
-      "epoch": 4.37037037037037,
-      "grad_norm": 0.02848837524652481,
-      "learning_rate": 4.7531442590937335e-06,
-      "loss": 0.0102,
+      "epoch": 1.5256565656565657,
+      "grad_norm": 0.003573813708499074,
+      "learning_rate": 6.10608020698577e-05,
+      "loss": 0.0011,
       "step": 472
     },
     {
-      "epoch": 4.37962962962963,
-      "grad_norm": 0.026586227118968964,
-      "learning_rate": 4.616550420127563e-06,
-      "loss": 0.0078,
+      "epoch": 1.528888888888889,
+      "grad_norm": 0.005556274671107531,
+      "learning_rate": 6.119016817593791e-05,
+      "loss": 0.0033,
       "step": 473
     },
     {
-      "epoch": 4.388888888888889,
-      "grad_norm": 0.025660747662186623,
-      "learning_rate": 4.4818529516926726e-06,
-      "loss": 0.0086,
+      "epoch": 1.5321212121212122,
+      "grad_norm": 0.004455295857042074,
+      "learning_rate": 6.131953428201811e-05,
+      "loss": 0.0014,
       "step": 474
     },
     {
-      "epoch": 4.398148148148148,
-      "grad_norm": 0.02436869405210018,
-      "learning_rate": 4.349057482191299e-06,
-      "loss": 0.011,
+      "epoch": 1.5353535353535355,
+      "grad_norm": 0.003466435242444277,
+      "learning_rate": 6.144890038809832e-05,
+      "loss": 0.001,
       "step": 475
     },
     {
-      "epoch": 4.398148148148148,
-      "eval_loss": 0.015554042533040047,
-      "eval_runtime": 9.1142,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.5353535353535355,
+      "eval_loss": 0.0010681893909350038,
+      "eval_runtime": 18.6082,
+      "eval_samples_per_second": 5.374,
+      "eval_steps_per_second": 1.343,
       "step": 475
     },
     {
-      "epoch": 4.407407407407407,
-      "grad_norm": 0.02513139322400093,
-      "learning_rate": 4.218169560549706e-06,
-      "loss": 0.0108,
+      "epoch": 1.5385858585858587,
+      "grad_norm": 0.004641688894480467,
+      "learning_rate": 6.157826649417853e-05,
+      "loss": 0.0029,
       "step": 476
     },
     {
-      "epoch": 4.416666666666667,
-      "grad_norm": 0.027343349531292915,
-      "learning_rate": 4.089194655986306e-06,
-      "loss": 0.0099,
+      "epoch": 1.541818181818182,
+      "grad_norm": 0.004081512801349163,
+      "learning_rate": 6.170763260025874e-05,
+      "loss": 0.0011,
       "step": 477
     },
     {
-      "epoch": 4.425925925925926,
-      "grad_norm": 0.02374204248189926,
-      "learning_rate": 3.962138157783085e-06,
-      "loss": 0.0095,
+      "epoch": 1.545050505050505,
+      "grad_norm": 0.005640064366161823,
+      "learning_rate": 6.183699870633895e-05,
+      "loss": 0.0024,
       "step": 478
     },
     {
-      "epoch": 4.435185185185185,
-      "grad_norm": 0.04114212468266487,
-      "learning_rate": 3.837005375060482e-06,
-      "loss": 0.0089,
+      "epoch": 1.5482828282828283,
+      "grad_norm": 0.0031765501480549574,
+      "learning_rate": 6.196636481241915e-05,
+      "loss": 0.0009,
       "step": 479
     },
     {
-      "epoch": 4.444444444444445,
-      "grad_norm": 0.024016965180635452,
-      "learning_rate": 3.7138015365554833e-06,
-      "loss": 0.0067,
-      "step": 480
-    },
-    {
-      "epoch": 4.444444444444445,
-      "eval_loss": 0.01539613213390112,
-      "eval_runtime": 9.1246,
-      "eval_samples_per_second": 5.48,
-      "eval_steps_per_second": 1.425,
+      "epoch": 1.5515151515151515,
+      "grad_norm": 0.0049113016575574875,
+      "learning_rate": 6.209573091849934e-05,
+      "loss": 0.0019,
       "step": 480
     },
     {
-      "epoch": 4.453703703703704,
-      "grad_norm": 0.02901994250714779,
-      "learning_rate": 3.5925317904031587e-06,
-      "loss": 0.0087,
+      "epoch": 1.5547474747474748,
+      "grad_norm": 0.003700861008837819,
+      "learning_rate": 6.222509702457955e-05,
+      "loss": 0.0012,
       "step": 481
     },
     {
-      "epoch": 4.462962962962963,
-      "grad_norm": 0.020981522276997566,
-      "learning_rate": 3.4732012039215776e-06,
-      "loss": 0.011,
+      "epoch": 1.557979797979798,
+      "grad_norm": 0.004294991493225098,
+      "learning_rate": 6.235446313065976e-05,
+      "loss": 0.0014,
       "step": 482
     },
     {
-      "epoch": 4.472222222222222,
-      "grad_norm": 0.023783011361956596,
-      "learning_rate": 3.3558147633999728e-06,
-      "loss": 0.0096,
+      "epoch": 1.561212121212121,
+      "grad_norm": 0.003475453006103635,
+      "learning_rate": 6.248382923673998e-05,
+      "loss": 0.0007,
       "step": 483
     },
     {
-      "epoch": 4.481481481481482,
-      "grad_norm": 0.02081628330051899,
-      "learning_rate": 3.2403773738905187e-06,
-      "loss": 0.0087,
+      "epoch": 1.5644444444444443,
+      "grad_norm": 0.007765649352222681,
+      "learning_rate": 6.261319534282019e-05,
+      "loss": 0.001,
       "step": 484
     },
     {
-      "epoch": 4.4907407407407405,
-      "grad_norm": 0.024986054748296738,
-      "learning_rate": 3.126893859003249e-06,
-      "loss": 0.0092,
-      "step": 485
-    },
-    {
-      "epoch": 4.4907407407407405,
-      "eval_loss": 0.015287145972251892,
-      "eval_runtime": 9.1097,
-      "eval_samples_per_second": 5.489,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.5676767676767676,
+      "grad_norm": 0.0032908658031374216,
+      "learning_rate": 6.27425614489004e-05,
+      "loss": 0.0008,
       "step": 485
     },
     {
-      "epoch": 4.5,
-      "grad_norm": 0.032323963940143585,
-      "learning_rate": 3.0153689607045845e-06,
-      "loss": 0.0086,
+      "epoch": 1.5709090909090908,
+      "grad_norm": 0.00453177560120821,
+      "learning_rate": 6.28719275549806e-05,
+      "loss": 0.0013,
       "step": 486
     },
     {
-      "epoch": 4.5092592592592595,
-      "grad_norm": 0.02963520959019661,
-      "learning_rate": 2.9058073391191375e-06,
-      "loss": 0.0068,
+      "epoch": 1.574141414141414,
+      "grad_norm": 0.0038091707974672318,
+      "learning_rate": 6.30012936610608e-05,
+      "loss": 0.0008,
       "step": 487
     },
     {
-      "epoch": 4.518518518518518,
-      "grad_norm": 0.035344675183296204,
-      "learning_rate": 2.798213572335001e-06,
-      "loss": 0.0062,
+      "epoch": 1.5773737373737373,
+      "grad_norm": 0.004123839549720287,
+      "learning_rate": 6.313065976714101e-05,
+      "loss": 0.0011,
       "step": 488
     },
     {
-      "epoch": 4.527777777777778,
-      "grad_norm": 0.026800939813256264,
-      "learning_rate": 2.692592156212487e-06,
-      "loss": 0.0092,
+      "epoch": 1.5806060606060606,
+      "grad_norm": 0.003308449639007449,
+      "learning_rate": 6.326002587322122e-05,
+      "loss": 0.0012,
       "step": 489
     },
     {
-      "epoch": 4.537037037037037,
-      "grad_norm": 0.024116506800055504,
-      "learning_rate": 2.5889475041961765e-06,
-      "loss": 0.0072,
-      "step": 490
-    },
-    {
-      "epoch": 4.537037037037037,
-      "eval_loss": 0.015211592428386211,
-      "eval_runtime": 9.1184,
-      "eval_samples_per_second": 5.483,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.5838383838383838,
+      "grad_norm": 0.005438206251710653,
+      "learning_rate": 6.338939197930143e-05,
+      "loss": 0.0023,
       "step": 490
     },
     {
-      "epoch": 4.546296296296296,
-      "grad_norm": 0.027498748153448105,
-      "learning_rate": 2.4872839471306084e-06,
-      "loss": 0.0082,
+      "epoch": 1.587070707070707,
+      "grad_norm": 0.005823109764605761,
+      "learning_rate": 6.351875808538163e-05,
+      "loss": 0.0014,
       "step": 491
     },
     {
-      "epoch": 4.555555555555555,
-      "grad_norm": 0.026998436078429222,
-      "learning_rate": 2.3876057330792346e-06,
-      "loss": 0.008,
+      "epoch": 1.5903030303030303,
+      "grad_norm": 0.0052726129069924355,
+      "learning_rate": 6.364812419146184e-05,
+      "loss": 0.0012,
       "step": 492
     },
     {
-      "epoch": 4.564814814814815,
-      "grad_norm": 0.023703446611762047,
-      "learning_rate": 2.2899170271469428e-06,
-      "loss": 0.011,
+      "epoch": 1.5935353535353536,
+      "grad_norm": 0.003648497397080064,
+      "learning_rate": 6.377749029754204e-05,
+      "loss": 0.0013,
       "step": 493
     },
     {
-      "epoch": 4.574074074074074,
-      "grad_norm": 0.019968930631875992,
-      "learning_rate": 2.1942219113060212e-06,
-      "loss": 0.0075,
+      "epoch": 1.5967676767676768,
+      "grad_norm": 0.003007176099345088,
+      "learning_rate": 6.390685640362225e-05,
+      "loss": 0.0007,
       "step": 494
     },
     {
-      "epoch": 4.583333333333333,
-      "grad_norm": 0.02214980125427246,
-      "learning_rate": 2.100524384225555e-06,
-      "loss": 0.0078,
-      "step": 495
-    },
-    {
-      "epoch": 4.583333333333333,
-      "eval_loss": 0.015181516297161579,
-      "eval_runtime": 9.1214,
-      "eval_samples_per_second": 5.482,
-      "eval_steps_per_second": 1.425,
+      "epoch": 1.6,
+      "grad_norm": 0.004037541802972555,
+      "learning_rate": 6.403622250970246e-05,
+      "loss": 0.0011,
       "step": 495
     },
     {
-      "epoch": 4.592592592592593,
-      "grad_norm": 0.025330157950520515,
-      "learning_rate": 2.0088283611044036e-06,
-      "loss": 0.0062,
+      "epoch": 1.6032323232323233,
+      "grad_norm": 0.0036643114872276783,
+      "learning_rate": 6.416558861578267e-05,
+      "loss": 0.0009,
       "step": 496
     },
     {
-      "epoch": 4.601851851851852,
-      "grad_norm": 0.019013626500964165,
-      "learning_rate": 1.9191376735075427e-06,
-      "loss": 0.0088,
+      "epoch": 1.6064646464646466,
+      "grad_norm": 0.0038799517787992954,
+      "learning_rate": 6.429495472186288e-05,
+      "loss": 0.001,
       "step": 497
     },
     {
-      "epoch": 4.611111111111111,
-      "grad_norm": 0.022145694121718407,
-      "learning_rate": 1.8314560692059835e-06,
-      "loss": 0.0089,
+      "epoch": 1.6096969696969698,
+      "grad_norm": 0.002598103601485491,
+      "learning_rate": 6.442432082794309e-05,
+      "loss": 0.0005,
       "step": 498
     },
     {
-      "epoch": 4.62037037037037,
-      "grad_norm": 0.023724934086203575,
-      "learning_rate": 1.7457872120201779e-06,
-      "loss": 0.0086,
+      "epoch": 1.6129292929292929,
+      "grad_norm": 0.006756095215678215,
+      "learning_rate": 6.45536869340233e-05,
+      "loss": 0.0014,
       "step": 499
     },
     {
-      "epoch": 4.62962962962963,
-      "grad_norm": 0.020578699186444283,
-      "learning_rate": 1.6621346816668992e-06,
-      "loss": 0.0091,
+      "epoch": 1.6161616161616161,
+      "grad_norm": 0.0038995088543742895,
+      "learning_rate": 6.46830530401035e-05,
+      "loss": 0.0007,
       "step": 500
     },
     {
-      "epoch": 4.62962962962963,
-      "eval_loss": 0.015207822434604168,
-      "eval_runtime": 9.1136,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.6161616161616161,
+      "eval_loss": 0.0010785168269649148,
+      "eval_runtime": 18.596,
+      "eval_samples_per_second": 5.378,
+      "eval_steps_per_second": 1.344,
       "step": 500
     },
     {
-      "epoch": 4.638888888888889,
-      "grad_norm": 0.024306217208504677,
-      "learning_rate": 1.5805019736097104e-06,
-      "loss": 0.009,
+      "epoch": 1.6193939393939394,
+      "grad_norm": 0.004936838988214731,
+      "learning_rate": 6.48124191461837e-05,
+      "loss": 0.0012,
       "step": 501
     },
     {
-      "epoch": 4.648148148148148,
-      "grad_norm": 0.020744021981954575,
-      "learning_rate": 1.5008924989128258e-06,
-      "loss": 0.0089,
+      "epoch": 1.6226262626262626,
+      "grad_norm": 0.006074307020753622,
+      "learning_rate": 6.494178525226392e-05,
+      "loss": 0.0022,
       "step": 502
     },
     {
-      "epoch": 4.657407407407407,
-      "grad_norm": 0.02516799047589302,
-      "learning_rate": 1.4233095840986753e-06,
-      "loss": 0.0093,
+      "epoch": 1.625858585858586,
+      "grad_norm": 0.0044588991440832615,
+      "learning_rate": 6.507115135834411e-05,
+      "loss": 0.0011,
       "step": 503
     },
     {
-      "epoch": 4.666666666666667,
-      "grad_norm": 0.024567998945713043,
-      "learning_rate": 1.3477564710088098e-06,
-      "loss": 0.0094,
+      "epoch": 1.6290909090909091,
+      "grad_norm": 0.004507533740252256,
+      "learning_rate": 6.520051746442432e-05,
+      "loss": 0.0011,
       "step": 504
     },
     {
-      "epoch": 4.675925925925926,
-      "grad_norm": 0.024358859285712242,
-      "learning_rate": 1.2742363166685034e-06,
-      "loss": 0.007,
-      "step": 505
-    },
-    {
-      "epoch": 4.675925925925926,
-      "eval_loss": 0.015200878493487835,
-      "eval_runtime": 9.1155,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.6323232323232322,
+      "grad_norm": 0.0048494781367480755,
+      "learning_rate": 6.532988357050453e-05,
+      "loss": 0.0008,
       "step": 505
     },
     {
-      "epoch": 4.685185185185185,
-      "grad_norm": 0.023163504898548126,
-      "learning_rate": 1.2027521931548214e-06,
-      "loss": 0.0074,
+      "epoch": 1.6355555555555554,
+      "grad_norm": 0.0033034745138138533,
+      "learning_rate": 6.545924967658473e-05,
+      "loss": 0.0007,
       "step": 506
     },
     {
-      "epoch": 4.694444444444445,
-      "grad_norm": 0.023604586720466614,
-      "learning_rate": 1.1333070874682216e-06,
-      "loss": 0.0093,
+      "epoch": 1.6387878787878787,
+      "grad_norm": 0.0053002117201685905,
+      "learning_rate": 6.558861578266494e-05,
+      "loss": 0.0009,
       "step": 507
     },
     {
-      "epoch": 4.703703703703704,
-      "grad_norm": 0.02068418823182583,
-      "learning_rate": 1.0659039014077944e-06,
-      "loss": 0.0084,
+      "epoch": 1.642020202020202,
+      "grad_norm": 0.00404641218483448,
+      "learning_rate": 6.571798188874515e-05,
+      "loss": 0.0008,
       "step": 508
     },
     {
-      "epoch": 4.712962962962963,
-      "grad_norm": 0.02598651312291622,
-      "learning_rate": 1.0005454514499414e-06,
-      "loss": 0.0088,
+      "epoch": 1.6452525252525252,
+      "grad_norm": 0.0036740771029144526,
+      "learning_rate": 6.584734799482536e-05,
+      "loss": 0.0009,
       "step": 509
     },
     {
-      "epoch": 4.722222222222222,
-      "grad_norm": 0.02512424811720848,
-      "learning_rate": 9.372344686307655e-07,
-      "loss": 0.0064,
-      "step": 510
-    },
-    {
-      "epoch": 4.722222222222222,
-      "eval_loss": 0.01521637849509716,
-      "eval_runtime": 9.1143,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.6484848484848484,
+      "grad_norm": 0.005331697873771191,
+      "learning_rate": 6.597671410090557e-05,
+      "loss": 0.0014,
       "step": 510
     },
     {
-      "epoch": 4.731481481481482,
-      "grad_norm": 0.021041063591837883,
-      "learning_rate": 8.759735984318895e-07,
-      "loss": 0.0096,
+      "epoch": 1.6517171717171717,
+      "grad_norm": 0.004965492524206638,
+      "learning_rate": 6.610608020698578e-05,
+      "loss": 0.0019,
       "step": 511
     },
     {
-      "epoch": 4.7407407407407405,
-      "grad_norm": 0.025718161836266518,
-      "learning_rate": 8.167654006699443e-07,
-      "loss": 0.0077,
+      "epoch": 1.654949494949495,
+      "grad_norm": 0.0037727411836385727,
+      "learning_rate": 6.623544631306599e-05,
+      "loss": 0.001,
       "step": 512
     },
     {
-      "epoch": 4.75,
-      "grad_norm": 0.02913082391023636,
-      "learning_rate": 7.596123493895991e-07,
-      "loss": 0.0072,
+      "epoch": 1.6581818181818182,
+      "grad_norm": 0.0048839072696864605,
+      "learning_rate": 6.636481241914619e-05,
+      "loss": 0.0013,
       "step": 513
     },
     {
-      "epoch": 4.7592592592592595,
-      "grad_norm": 0.026588505133986473,
-      "learning_rate": 7.04516832760177e-07,
-      "loss": 0.0094,
+      "epoch": 1.6614141414141415,
+      "grad_norm": 0.004607974551618099,
+      "learning_rate": 6.649417852522638e-05,
+      "loss": 0.0017,
       "step": 514
     },
     {
-      "epoch": 4.768518518518518,
-      "grad_norm": 0.023728126659989357,
-      "learning_rate": 6.514811529758747e-07,
-      "loss": 0.0099,
-      "step": 515
-    },
-    {
-      "epoch": 4.768518518518518,
-      "eval_loss": 0.01521516963839531,
-      "eval_runtime": 9.1511,
-      "eval_samples_per_second": 5.464,
-      "eval_steps_per_second": 1.421,
+      "epoch": 1.6646464646464647,
+      "grad_norm": 0.004647100809961557,
+      "learning_rate": 6.66235446313066e-05,
+      "loss": 0.0024,
       "step": 515
     },
     {
-      "epoch": 4.777777777777778,
-      "grad_norm": 0.03438512608408928,
-      "learning_rate": 6.005075261595494e-07,
-      "loss": 0.0086,
+      "epoch": 1.667878787878788,
+      "grad_norm": 0.004482895601540804,
+      "learning_rate": 6.67529107373868e-05,
+      "loss": 0.0023,
       "step": 516
     },
     {
-      "epoch": 4.787037037037037,
-      "grad_norm": 0.019554298371076584,
-      "learning_rate": 5.515980822701439e-07,
-      "loss": 0.0092,
+      "epoch": 1.6711111111111112,
+      "grad_norm": 0.004681951366364956,
+      "learning_rate": 6.688227684346702e-05,
+      "loss": 0.0031,
       "step": 517
     },
     {
-      "epoch": 4.796296296296296,
-      "grad_norm": 0.0235204566270113,
-      "learning_rate": 5.047548650136513e-07,
-      "loss": 0.009,
+      "epoch": 1.6743434343434345,
+      "grad_norm": 0.004061279818415642,
+      "learning_rate": 6.701164294954723e-05,
+      "loss": 0.0012,
       "step": 518
     },
     {
-      "epoch": 4.805555555555555,
-      "grad_norm": 0.023747643455863,
-      "learning_rate": 4.5997983175773417e-07,
-      "loss": 0.0092,
+      "epoch": 1.6775757575757577,
+      "grad_norm": 0.005936305969953537,
+      "learning_rate": 6.714100905562742e-05,
+      "loss": 0.0018,
       "step": 519
     },
     {
-      "epoch": 4.814814814814815,
-      "grad_norm": 0.02751827985048294,
-      "learning_rate": 4.1727485344994486e-07,
-      "loss": 0.0088,
-      "step": 520
-    },
-    {
-      "epoch": 4.814814814814815,
-      "eval_loss": 0.015235532075166702,
-      "eval_runtime": 9.1256,
-      "eval_samples_per_second": 5.479,
-      "eval_steps_per_second": 1.425,
+      "epoch": 1.680808080808081,
+      "grad_norm": 0.003031841479241848,
+      "learning_rate": 6.727037516170763e-05,
+      "loss": 0.0009,
       "step": 520
     },
     {
-      "epoch": 4.824074074074074,
-      "grad_norm": 0.026621591299772263,
-      "learning_rate": 3.766417145395218e-07,
-      "loss": 0.0086,
+      "epoch": 1.684040404040404,
+      "grad_norm": 0.0044912113808095455,
+      "learning_rate": 6.739974126778784e-05,
+      "loss": 0.0013,
       "step": 521
     },
     {
-      "epoch": 4.833333333333333,
-      "grad_norm": 0.01991841197013855,
-      "learning_rate": 3.380821129028489e-07,
-      "loss": 0.0084,
+      "epoch": 1.6872727272727273,
+      "grad_norm": 0.003882101271301508,
+      "learning_rate": 6.752910737386805e-05,
+      "loss": 0.0014,
       "step": 522
     },
     {
-      "epoch": 4.842592592592593,
-      "grad_norm": 0.023508219048380852,
-      "learning_rate": 3.0159765977250673e-07,
-      "loss": 0.0103,
+      "epoch": 1.6905050505050505,
+      "grad_norm": 0.0031570433638989925,
+      "learning_rate": 6.765847347994826e-05,
+      "loss": 0.0011,
       "step": 523
     },
     {
-      "epoch": 4.851851851851852,
-      "grad_norm": 0.02976732887327671,
-      "learning_rate": 2.671898796699268e-07,
-      "loss": 0.0084,
+      "epoch": 1.6937373737373738,
+      "grad_norm": 0.004184515681117773,
+      "learning_rate": 6.778783958602847e-05,
+      "loss": 0.0015,
       "step": 524
     },
     {
-      "epoch": 4.861111111111111,
-      "grad_norm": 0.02255621738731861,
-      "learning_rate": 2.3486021034170857e-07,
-      "loss": 0.0089,
+      "epoch": 1.696969696969697,
+      "grad_norm": 0.002800683258101344,
+      "learning_rate": 6.791720569210867e-05,
+      "loss": 0.0008,
       "step": 525
     },
     {
-      "epoch": 4.861111111111111,
-      "eval_loss": 0.015216498635709286,
-      "eval_runtime": 9.1106,
-      "eval_samples_per_second": 5.488,
-      "eval_steps_per_second": 1.427,
+      "epoch": 1.696969696969697,
+      "eval_loss": 0.0010542384115979075,
+      "eval_runtime": 18.5989,
+      "eval_samples_per_second": 5.377,
+      "eval_steps_per_second": 1.344,
       "step": 525
     },
     {
-      "epoch": 4.87037037037037,
-      "grad_norm": 0.025215914472937584,
-      "learning_rate": 2.0461000269953456e-07,
-      "loss": 0.0075,
+      "epoch": 1.70020202020202,
+      "grad_norm": 0.003550174878910184,
+      "learning_rate": 6.804657179818887e-05,
+      "loss": 0.0014,
       "step": 526
     },
     {
-      "epoch": 4.87962962962963,
-      "grad_norm": 0.02554066851735115,
-      "learning_rate": 1.7644052076371542e-07,
-      "loss": 0.0083,
+      "epoch": 1.7034343434343433,
+      "grad_norm": 0.0030985972844064236,
+      "learning_rate": 6.817593790426908e-05,
+      "loss": 0.0008,
       "step": 527
     },
     {
-      "epoch": 4.888888888888889,
-      "grad_norm": 0.02162836864590645,
-      "learning_rate": 1.503529416103988e-07,
-      "loss": 0.009,
+      "epoch": 1.7066666666666666,
+      "grad_norm": 0.0048317620530724525,
+      "learning_rate": 6.830530401034929e-05,
+      "loss": 0.0018,
       "step": 528
     },
     {
-      "epoch": 4.898148148148148,
-      "grad_norm": 0.02335723116993904,
-      "learning_rate": 1.2634835532233657e-07,
-      "loss": 0.0093,
+      "epoch": 1.7098989898989898,
+      "grad_norm": 0.005726094823330641,
+      "learning_rate": 6.84346701164295e-05,
+      "loss": 0.0025,
       "step": 529
     },
     {
-      "epoch": 4.907407407407407,
-      "grad_norm": 0.02844967506825924,
-      "learning_rate": 1.044277649433989e-07,
-      "loss": 0.0083,
-      "step": 530
-    },
-    {
-      "epoch": 4.907407407407407,
-      "eval_loss": 0.015229844488203526,
-      "eval_runtime": 9.1406,
-      "eval_samples_per_second": 5.47,
-      "eval_steps_per_second": 1.422,
+      "epoch": 1.713131313131313,
+      "grad_norm": 0.0024808107409626245,
+      "learning_rate": 6.856403622250971e-05,
+      "loss": 0.0007,
       "step": 530
     },
     {
-      "epoch": 4.916666666666667,
-      "grad_norm": 0.02188325859606266,
-      "learning_rate": 8.459208643659122e-08,
-      "loss": 0.0084,
+      "epoch": 1.7163636363636363,
+      "grad_norm": 0.003422652604058385,
+      "learning_rate": 6.869340232858992e-05,
+      "loss": 0.001,
       "step": 531
     },
     {
-      "epoch": 4.925925925925926,
-      "grad_norm": 0.026782654225826263,
-      "learning_rate": 6.684214864584038e-08,
-      "loss": 0.009,
+      "epoch": 1.7195959595959596,
+      "grad_norm": 0.0037957008462399244,
+      "learning_rate": 6.882276843467012e-05,
+      "loss": 0.0009,
       "step": 532
     },
     {
-      "epoch": 4.935185185185185,
-      "grad_norm": 0.024010982364416122,
-      "learning_rate": 5.11786932613223e-08,
-      "loss": 0.0055,
+      "epoch": 1.7228282828282828,
+      "grad_norm": 0.0028711955528706312,
+      "learning_rate": 6.895213454075033e-05,
+      "loss": 0.0008,
       "step": 533
     },
     {
-      "epoch": 4.944444444444445,
-      "grad_norm": 0.02621973119676113,
-      "learning_rate": 3.760237478849793e-08,
-      "loss": 0.0093,
+      "epoch": 1.726060606060606,
+      "grad_norm": 0.002812835620716214,
+      "learning_rate": 6.908150064683054e-05,
+      "loss": 0.0009,
       "step": 534
     },
     {
-      "epoch": 4.953703703703704,
-      "grad_norm": 0.02257387712597847,
-      "learning_rate": 2.6113760520735108e-08,
-      "loss": 0.0103,
-      "step": 535
-    },
-    {
-      "epoch": 4.953703703703704,
-      "eval_loss": 0.015256751328706741,
-      "eval_runtime": 9.1156,
-      "eval_samples_per_second": 5.485,
-      "eval_steps_per_second": 1.426,
+      "epoch": 1.7292929292929293,
+      "grad_norm": 0.003739473642781377,
+      "learning_rate": 6.921086675291075e-05,
+      "loss": 0.0016,
       "step": 535
     },
     {
-      "epoch": 4.962962962962963,
-      "grad_norm": 0.02289225161075592,
-      "learning_rate": 1.6713330515627513e-08,
-      "loss": 0.0106,
+      "epoch": 1.7325252525252526,
+      "grad_norm": 0.004807054530829191,
+      "learning_rate": 6.934023285899096e-05,
+      "loss": 0.0023,
       "step": 536
     },
     {
-      "epoch": 4.972222222222222,
-      "grad_norm": 0.032289694994688034,
-      "learning_rate": 9.401477574932926e-09,
-      "loss": 0.0074,
+      "epoch": 1.7357575757575758,
+      "grad_norm": 0.004576352424919605,
+      "learning_rate": 6.946959896507115e-05,
+      "loss": 0.0024,
       "step": 537
     },
     {
-      "epoch": 4.981481481481482,
-      "grad_norm": 0.0215620007365942,
-      "learning_rate": 4.178507228136397e-09,
-      "loss": 0.0082,
+      "epoch": 1.738989898989899,
+      "grad_norm": 0.0030553669203072786,
+      "learning_rate": 6.959896507115135e-05,
+      "loss": 0.0009,
       "step": 538
     },
     {
-      "epoch": 4.9907407407407405,
-      "grad_norm": 0.02391226962208748,
-      "learning_rate": 1.0446377197104173e-09,
-      "loss": 0.0085,
+      "epoch": 1.7422222222222223,
+      "grad_norm": 0.003879109164699912,
+      "learning_rate": 6.972833117723156e-05,
+      "loss": 0.0012,
       "step": 539
     },
     {
-      "epoch": 5.0,
-      "grad_norm": 0.0241775494068861,
-      "learning_rate": 0.0,
-      "loss": 0.0092,
+      "epoch": 1.7454545454545456,
+      "grad_norm": 0.0031888741068542004,
+      "learning_rate": 6.985769728331177e-05,
+      "loss": 0.0013,
       "step": 540
     },
     {
-      "epoch": 5.0,
-      "eval_loss": 0.01526525616645813,
-      "eval_runtime": 9.1149,
-      "eval_samples_per_second": 5.486,
-      "eval_steps_per_second": 1.426,
-      "step": 540
+      "epoch": 1.7486868686868688,
+      "grad_norm": 0.0037957336753606796,
+      "learning_rate": 6.998706338939198e-05,
+      "loss": 0.0008,
+      "step": 541
     },
     {
-      "epoch": 5.0,
-      "step": 540,
-      "total_flos": 1.2254685925518213e+18,
-      "train_loss": 0.016027936152251506,
-      "train_runtime": 9839.9649,
-      "train_samples_per_second": 1.756,
-      "train_steps_per_second": 0.055
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 540,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
-  "save_steps": 50,
-  "stateful_callbacks": {
+      "epoch": 1.7519191919191919,
+      "grad_norm": 0.002901956904679537,
+      "learning_rate": 7.011642949547219e-05,
+      "loss": 0.0007,
+      "step": 542
+    },
+    {
+      "epoch": 1.7551515151515151,
+      "grad_norm": 0.0029531833715736866,
+      "learning_rate": 7.02457956015524e-05,
+      "loss": 0.0008,
+      "step": 543
+    },
+    {
+      "epoch": 1.7583838383838384,
+      "grad_norm": 0.004063300788402557,
+      "learning_rate": 7.037516170763261e-05,
+      "loss": 0.0012,
+      "step": 544
+    },
+    {
+      "epoch": 1.7616161616161616,
+      "grad_norm": 0.0041914028115570545,
+      "learning_rate": 7.050452781371281e-05,
+      "loss": 0.001,
+      "step": 545
+    },
+    {
+      "epoch": 1.7648484848484849,
+      "grad_norm": 0.004884886089712381,
+      "learning_rate": 7.063389391979302e-05,
+      "loss": 0.0021,
+      "step": 546
+    },
+    {
+      "epoch": 1.768080808080808,
+      "grad_norm": 0.005683641415089369,
+      "learning_rate": 7.076326002587323e-05,
+      "loss": 0.0024,
+      "step": 547
+    },
+    {
+      "epoch": 1.7713131313131312,
+      "grad_norm": 0.0042356885969638824,
+      "learning_rate": 7.089262613195343e-05,
+      "loss": 0.0015,
+      "step": 548
+    },
+    {
+      "epoch": 1.7745454545454544,
+      "grad_norm": 0.005650636274367571,
+      "learning_rate": 7.102199223803364e-05,
+      "loss": 0.0013,
+      "step": 549
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.00460450816899538,
+      "learning_rate": 7.115135834411385e-05,
+      "loss": 0.001,
+      "step": 550
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "eval_loss": 0.0010012522106990218,
+      "eval_runtime": 18.6807,
+      "eval_samples_per_second": 5.353,
+      "eval_steps_per_second": 1.338,
+      "step": 550
+    },
+    {
+      "epoch": 1.781010101010101,
+      "grad_norm": 0.004403329454362392,
+      "learning_rate": 7.128072445019404e-05,
+      "loss": 0.0012,
+      "step": 551
+    },
+    {
+      "epoch": 1.7842424242424242,
+      "grad_norm": 0.0032920974772423506,
+      "learning_rate": 7.141009055627425e-05,
+      "loss": 0.0008,
+      "step": 552
+    },
+    {
+      "epoch": 1.7874747474747474,
+      "grad_norm": 0.0038138548843562603,
+      "learning_rate": 7.153945666235446e-05,
+      "loss": 0.0011,
+      "step": 553
+    },
+    {
+      "epoch": 1.7907070707070707,
+      "grad_norm": 0.00296188285574317,
+      "learning_rate": 7.166882276843467e-05,
+      "loss": 0.0008,
+      "step": 554
+    },
+    {
+      "epoch": 1.793939393939394,
+      "grad_norm": 0.0031317227985709906,
+      "learning_rate": 7.179818887451488e-05,
+      "loss": 0.001,
+      "step": 555
+    },
+    {
+      "epoch": 1.7971717171717172,
+      "grad_norm": 0.003491588868200779,
+      "learning_rate": 7.19275549805951e-05,
+      "loss": 0.0007,
+      "step": 556
+    },
+    {
+      "epoch": 1.8004040404040405,
+      "grad_norm": 0.0031343346927314997,
+      "learning_rate": 7.20569210866753e-05,
+      "loss": 0.0005,
+      "step": 557
+    },
+    {
+      "epoch": 1.8036363636363637,
+      "grad_norm": 0.004438635893166065,
+      "learning_rate": 7.21862871927555e-05,
+      "loss": 0.0011,
+      "step": 558
+    },
+    {
+      "epoch": 1.806868686868687,
+      "grad_norm": 0.00411741528660059,
+      "learning_rate": 7.231565329883571e-05,
+      "loss": 0.0008,
+      "step": 559
+    },
+    {
+      "epoch": 1.8101010101010102,
+      "grad_norm": 0.004590868018567562,
+      "learning_rate": 7.244501940491591e-05,
+      "loss": 0.0009,
+      "step": 560
+    },
+    {
+      "epoch": 1.8133333333333335,
+      "grad_norm": 0.00454489141702652,
+      "learning_rate": 7.257438551099612e-05,
+      "loss": 0.001,
+      "step": 561
+    },
+    {
+      "epoch": 1.8165656565656567,
+      "grad_norm": 0.004015767015516758,
+      "learning_rate": 7.270375161707633e-05,
+      "loss": 0.0007,
+      "step": 562
+    },
+    {
+      "epoch": 1.8197979797979797,
+      "grad_norm": 0.003563391976058483,
+      "learning_rate": 7.283311772315654e-05,
+      "loss": 0.0008,
+      "step": 563
+    },
+    {
+      "epoch": 1.823030303030303,
+      "grad_norm": 0.0037166401743888855,
+      "learning_rate": 7.296248382923674e-05,
+      "loss": 0.0009,
+      "step": 564
+    },
+    {
+      "epoch": 1.8262626262626263,
+      "grad_norm": 0.004116456024348736,
+      "learning_rate": 7.309184993531695e-05,
+      "loss": 0.0009,
+      "step": 565
+    },
+    {
+      "epoch": 1.8294949494949495,
+      "grad_norm": 0.004104538355022669,
+      "learning_rate": 7.322121604139716e-05,
+      "loss": 0.0013,
+      "step": 566
+    },
+    {
+      "epoch": 1.8327272727272728,
+      "grad_norm": 0.005509045906364918,
+      "learning_rate": 7.335058214747737e-05,
+      "loss": 0.0023,
+      "step": 567
+    },
+    {
+      "epoch": 1.835959595959596,
+      "grad_norm": 0.0035526566207408905,
+      "learning_rate": 7.347994825355758e-05,
+      "loss": 0.0008,
+      "step": 568
+    },
+    {
+      "epoch": 1.839191919191919,
+      "grad_norm": 0.003440872300416231,
+      "learning_rate": 7.360931435963779e-05,
+      "loss": 0.0007,
+      "step": 569
+    },
+    {
+      "epoch": 1.8424242424242423,
+      "grad_norm": 0.0029369727708399296,
+      "learning_rate": 7.3738680465718e-05,
+      "loss": 0.0008,
+      "step": 570
+    },
+    {
+      "epoch": 1.8456565656565656,
+      "grad_norm": 0.004200720228254795,
+      "learning_rate": 7.38680465717982e-05,
+      "loss": 0.0011,
+      "step": 571
+    },
+    {
+      "epoch": 1.8488888888888888,
+      "grad_norm": 0.005561929661780596,
+      "learning_rate": 7.399741267787839e-05,
+      "loss": 0.0018,
+      "step": 572
+    },
+    {
+      "epoch": 1.852121212121212,
+      "grad_norm": 0.0033285904210060835,
+      "learning_rate": 7.41267787839586e-05,
+      "loss": 0.0009,
+      "step": 573
+    },
+    {
+      "epoch": 1.8553535353535353,
+      "grad_norm": 0.004428067244589329,
+      "learning_rate": 7.425614489003881e-05,
+      "loss": 0.0015,
+      "step": 574
+    },
+    {
+      "epoch": 1.8585858585858586,
+      "grad_norm": 0.003229588968679309,
+      "learning_rate": 7.438551099611902e-05,
+      "loss": 0.0007,
+      "step": 575
+    },
+    {
+      "epoch": 1.8585858585858586,
+      "eval_loss": 0.0009781663538888097,
+      "eval_runtime": 18.7511,
+      "eval_samples_per_second": 5.333,
+      "eval_steps_per_second": 1.333,
+      "step": 575
+    },
+    {
+      "epoch": 1.8618181818181818,
+      "grad_norm": 0.003516717813909054,
+      "learning_rate": 7.451487710219923e-05,
+      "loss": 0.0011,
+      "step": 576
+    },
+    {
+      "epoch": 1.865050505050505,
+      "grad_norm": 0.002692221663892269,
+      "learning_rate": 7.464424320827943e-05,
+      "loss": 0.0008,
+      "step": 577
+    },
+    {
+      "epoch": 1.8682828282828283,
+      "grad_norm": 0.003864881582558155,
+      "learning_rate": 7.477360931435964e-05,
+      "loss": 0.0013,
+      "step": 578
+    },
+    {
+      "epoch": 1.8715151515151516,
+      "grad_norm": 0.003865283913910389,
+      "learning_rate": 7.490297542043985e-05,
+      "loss": 0.0015,
+      "step": 579
+    },
+    {
+      "epoch": 1.8747474747474748,
+      "grad_norm": 0.0035206254106014967,
+      "learning_rate": 7.503234152652006e-05,
+      "loss": 0.0012,
+      "step": 580
+    },
+    {
+      "epoch": 1.877979797979798,
+      "grad_norm": 0.004602121654897928,
+      "learning_rate": 7.516170763260027e-05,
+      "loss": 0.0018,
+      "step": 581
+    },
+    {
+      "epoch": 1.8812121212121213,
+      "grad_norm": 0.0029543524142354727,
+      "learning_rate": 7.529107373868047e-05,
+      "loss": 0.0009,
+      "step": 582
+    },
+    {
+      "epoch": 1.8844444444444446,
+      "grad_norm": 0.0023640701547265053,
+      "learning_rate": 7.542043984476068e-05,
+      "loss": 0.0007,
+      "step": 583
+    },
+    {
+      "epoch": 1.8876767676767678,
+      "grad_norm": 0.0040098559111356735,
+      "learning_rate": 7.554980595084087e-05,
+      "loss": 0.0012,
+      "step": 584
+    },
+    {
+      "epoch": 1.8909090909090909,
+      "grad_norm": 0.004274715203791857,
+      "learning_rate": 7.567917205692108e-05,
+      "loss": 0.0015,
+      "step": 585
+    },
+    {
+      "epoch": 1.8941414141414141,
+      "grad_norm": 0.0031054418068379164,
+      "learning_rate": 7.58085381630013e-05,
+      "loss": 0.0008,
+      "step": 586
+    },
+    {
+      "epoch": 1.8973737373737374,
+      "grad_norm": 0.003954428713768721,
+      "learning_rate": 7.59379042690815e-05,
+      "loss": 0.0009,
+      "step": 587
+    },
+    {
+      "epoch": 1.9006060606060606,
+      "grad_norm": 0.003786911489441991,
+      "learning_rate": 7.606727037516171e-05,
+      "loss": 0.0006,
+      "step": 588
+    },
+    {
+      "epoch": 1.9038383838383839,
+      "grad_norm": 0.0041464087553322315,
+      "learning_rate": 7.619663648124192e-05,
+      "loss": 0.0011,
+      "step": 589
+    },
+    {
+      "epoch": 1.907070707070707,
+      "grad_norm": 0.004568018950521946,
+      "learning_rate": 7.632600258732212e-05,
+      "loss": 0.0019,
+      "step": 590
+    },
+    {
+      "epoch": 1.9103030303030302,
+      "grad_norm": 0.003288006642833352,
+      "learning_rate": 7.645536869340233e-05,
+      "loss": 0.0009,
+      "step": 591
+    },
+    {
+      "epoch": 1.9135353535353534,
+      "grad_norm": 0.004277890548110008,
+      "learning_rate": 7.658473479948254e-05,
+      "loss": 0.0008,
+      "step": 592
+    },
+    {
+      "epoch": 1.9167676767676767,
+      "grad_norm": 0.004379922058433294,
+      "learning_rate": 7.671410090556275e-05,
+      "loss": 0.001,
+      "step": 593
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.003021504729986191,
+      "learning_rate": 7.684346701164295e-05,
+      "loss": 0.0006,
+      "step": 594
+    },
+    {
+      "epoch": 1.9232323232323232,
+      "grad_norm": 0.0032772128470242023,
+      "learning_rate": 7.697283311772316e-05,
+      "loss": 0.0008,
+      "step": 595
+    },
+    {
+      "epoch": 1.9264646464646464,
+      "grad_norm": 0.004533900413662195,
+      "learning_rate": 7.710219922380336e-05,
+      "loss": 0.0015,
+      "step": 596
+    },
+    {
+      "epoch": 1.9296969696969697,
+      "grad_norm": 0.005284740123897791,
+      "learning_rate": 7.723156532988357e-05,
+      "loss": 0.0011,
+      "step": 597
+    },
+    {
+      "epoch": 1.932929292929293,
+      "grad_norm": 0.005875818431377411,
+      "learning_rate": 7.736093143596378e-05,
+      "loss": 0.0013,
+      "step": 598
+    },
+    {
+      "epoch": 1.9361616161616162,
+      "grad_norm": 0.004483737051486969,
+      "learning_rate": 7.749029754204399e-05,
+      "loss": 0.0009,
+      "step": 599
+    },
+    {
+      "epoch": 1.9393939393939394,
+      "grad_norm": 0.004435641225427389,
+      "learning_rate": 7.76196636481242e-05,
+      "loss": 0.0013,
+      "step": 600
+    },
+    {
+      "epoch": 1.9393939393939394,
+      "eval_loss": 0.0009266917477361858,
+      "eval_runtime": 18.7646,
+      "eval_samples_per_second": 5.329,
+      "eval_steps_per_second": 1.332,
+      "step": 600
+    },
+    {
+      "epoch": 1.9426262626262627,
+      "grad_norm": 0.003876154311001301,
+      "learning_rate": 7.774902975420441e-05,
+      "loss": 0.001,
+      "step": 601
+    },
+    {
+      "epoch": 1.945858585858586,
+      "grad_norm": 0.004851337987929583,
+      "learning_rate": 7.787839586028462e-05,
+      "loss": 0.0014,
+      "step": 602
+    },
+    {
+      "epoch": 1.9490909090909092,
+      "grad_norm": 0.003333737375214696,
+      "learning_rate": 7.800776196636481e-05,
+      "loss": 0.001,
+      "step": 603
+    },
+    {
+      "epoch": 1.9523232323232325,
+      "grad_norm": 0.0032944250851869583,
+      "learning_rate": 7.813712807244502e-05,
+      "loss": 0.0007,
+      "step": 604
+    },
+    {
+      "epoch": 1.9555555555555557,
+      "grad_norm": 0.004762569442391396,
+      "learning_rate": 7.826649417852523e-05,
+      "loss": 0.0027,
+      "step": 605
+    },
+    {
+      "epoch": 1.9587878787878787,
+      "grad_norm": 0.0037507396191358566,
+      "learning_rate": 7.839586028460543e-05,
+      "loss": 0.0014,
+      "step": 606
+    },
+    {
+      "epoch": 1.962020202020202,
+      "grad_norm": 0.005207899492233992,
+      "learning_rate": 7.852522639068564e-05,
+      "loss": 0.0031,
+      "step": 607
+    },
+    {
+      "epoch": 1.9652525252525253,
+      "grad_norm": 0.0035236128605902195,
+      "learning_rate": 7.865459249676585e-05,
+      "loss": 0.001,
+      "step": 608
+    },
+    {
+      "epoch": 1.9684848484848485,
+      "grad_norm": 0.004186858423054218,
+      "learning_rate": 7.878395860284605e-05,
+      "loss": 0.0012,
+      "step": 609
+    },
+    {
+      "epoch": 1.9717171717171718,
+      "grad_norm": 0.0038515774067491293,
+      "learning_rate": 7.891332470892626e-05,
+      "loss": 0.0022,
+      "step": 610
+    },
+    {
+      "epoch": 1.9749494949494948,
+      "grad_norm": 0.0030782243702560663,
+      "learning_rate": 7.904269081500647e-05,
+      "loss": 0.0006,
+      "step": 611
+    },
+    {
+      "epoch": 1.978181818181818,
+      "grad_norm": 0.005162171553820372,
+      "learning_rate": 7.917205692108668e-05,
+      "loss": 0.0013,
+      "step": 612
+    },
+    {
+      "epoch": 1.9814141414141413,
+      "grad_norm": 0.004226456396281719,
+      "learning_rate": 7.930142302716689e-05,
+      "loss": 0.0009,
+      "step": 613
+    },
+    {
+      "epoch": 1.9846464646464645,
+      "grad_norm": 0.003615841967985034,
+      "learning_rate": 7.94307891332471e-05,
+      "loss": 0.0022,
+      "step": 614
+    },
+    {
+      "epoch": 1.9878787878787878,
+      "grad_norm": 0.00408519571647048,
+      "learning_rate": 7.956015523932731e-05,
+      "loss": 0.0015,
+      "step": 615
+    },
+    {
+      "epoch": 1.991111111111111,
+      "grad_norm": 0.0042512728832662106,
+      "learning_rate": 7.968952134540751e-05,
+      "loss": 0.0014,
+      "step": 616
+    },
+    {
+      "epoch": 1.9943434343434343,
+      "grad_norm": 0.003178700339049101,
+      "learning_rate": 7.981888745148772e-05,
+      "loss": 0.0008,
+      "step": 617
+    },
+    {
+      "epoch": 1.9975757575757576,
+      "grad_norm": 0.003822662867605686,
+      "learning_rate": 7.994825355756791e-05,
+      "loss": 0.001,
+      "step": 618
+    },
+    {
+      "epoch": 2.000808080808081,
+      "grad_norm": 0.0046024019829928875,
+      "learning_rate": 8.007761966364812e-05,
+      "loss": 0.0011,
+      "step": 619
+    },
+    {
+      "epoch": 2.004040404040404,
+      "grad_norm": 0.0038011916913092136,
+      "learning_rate": 8.020698576972833e-05,
+      "loss": 0.0008,
+      "step": 620
+    },
+    {
+      "epoch": 2.0072727272727273,
+      "grad_norm": 0.006909184157848358,
+      "learning_rate": 8.033635187580855e-05,
+      "loss": 0.0026,
+      "step": 621
+    },
+    {
+      "epoch": 2.0105050505050506,
+      "grad_norm": 0.005042599979788065,
+      "learning_rate": 8.046571798188874e-05,
+      "loss": 0.0011,
+      "step": 622
+    },
+    {
+      "epoch": 2.013737373737374,
+      "grad_norm": 0.004002322442829609,
+      "learning_rate": 8.059508408796895e-05,
+      "loss": 0.0011,
+      "step": 623
+    },
+    {
+      "epoch": 2.016969696969697,
+      "grad_norm": 0.0031209783628582954,
+      "learning_rate": 8.072445019404916e-05,
+      "loss": 0.0009,
+      "step": 624
+    },
+    {
+      "epoch": 2.0202020202020203,
+      "grad_norm": 0.0026158462278544903,
+      "learning_rate": 8.085381630012937e-05,
+      "loss": 0.0007,
+      "step": 625
+    },
+    {
+      "epoch": 2.0202020202020203,
+      "eval_loss": 0.0009721739334054291,
+      "eval_runtime": 18.7204,
+      "eval_samples_per_second": 5.342,
+      "eval_steps_per_second": 1.335,
+      "step": 625
+    },
+    {
+      "epoch": 2.0234343434343436,
+      "grad_norm": 0.003943873103708029,
+      "learning_rate": 8.098318240620958e-05,
+      "loss": 0.001,
+      "step": 626
+    },
+    {
+      "epoch": 2.026666666666667,
+      "grad_norm": 0.0038578941021114588,
+      "learning_rate": 8.111254851228979e-05,
+      "loss": 0.0017,
+      "step": 627
+    },
+    {
+      "epoch": 2.02989898989899,
+      "grad_norm": 0.002991423010826111,
+      "learning_rate": 8.124191461836999e-05,
+      "loss": 0.0007,
+      "step": 628
+    },
+    {
+      "epoch": 2.0331313131313133,
+      "grad_norm": 0.002815463813021779,
+      "learning_rate": 8.13712807244502e-05,
+      "loss": 0.0009,
+      "step": 629
+    },
+    {
+      "epoch": 2.036363636363636,
+      "grad_norm": 0.0025943731889128685,
+      "learning_rate": 8.15006468305304e-05,
+      "loss": 0.0006,
+      "step": 630
+    },
+    {
+      "epoch": 2.0395959595959594,
+      "grad_norm": 0.0029918155632913113,
+      "learning_rate": 8.163001293661061e-05,
+      "loss": 0.0008,
+      "step": 631
+    },
+    {
+      "epoch": 2.0428282828282827,
+      "grad_norm": 0.00398919777944684,
+      "learning_rate": 8.175937904269082e-05,
+      "loss": 0.0011,
+      "step": 632
+    },
+    {
+      "epoch": 2.046060606060606,
+      "grad_norm": 0.003045483957976103,
+      "learning_rate": 8.188874514877103e-05,
+      "loss": 0.0007,
+      "step": 633
+    },
+    {
+      "epoch": 2.049292929292929,
+      "grad_norm": 0.005803203675895929,
+      "learning_rate": 8.201811125485124e-05,
+      "loss": 0.0008,
+      "step": 634
+    },
+    {
+      "epoch": 2.0525252525252524,
+      "grad_norm": 0.003953828942030668,
+      "learning_rate": 8.214747736093143e-05,
+      "loss": 0.001,
+      "step": 635
+    },
+    {
+      "epoch": 2.0557575757575757,
+      "grad_norm": 0.003973971586674452,
+      "learning_rate": 8.227684346701164e-05,
+      "loss": 0.0006,
+      "step": 636
+    },
+    {
+      "epoch": 2.058989898989899,
+      "grad_norm": 0.0025888276286423206,
+      "learning_rate": 8.240620957309186e-05,
+      "loss": 0.0005,
+      "step": 637
+    },
+    {
+      "epoch": 2.062222222222222,
+      "grad_norm": 0.003989651799201965,
+      "learning_rate": 8.253557567917207e-05,
+      "loss": 0.0008,
+      "step": 638
+    },
+    {
+      "epoch": 2.0654545454545454,
+      "grad_norm": 0.0030330433510243893,
+      "learning_rate": 8.266494178525228e-05,
+      "loss": 0.0009,
+      "step": 639
+    },
+    {
+      "epoch": 2.0686868686868687,
+      "grad_norm": 0.0044325897470116615,
+      "learning_rate": 8.279430789133247e-05,
+      "loss": 0.0016,
+      "step": 640
+    },
+    {
+      "epoch": 2.071919191919192,
+      "grad_norm": 0.002264161128550768,
+      "learning_rate": 8.292367399741268e-05,
+      "loss": 0.0006,
+      "step": 641
+    },
+    {
+      "epoch": 2.075151515151515,
+      "grad_norm": 0.0033110452350229025,
+      "learning_rate": 8.305304010349288e-05,
+      "loss": 0.001,
+      "step": 642
+    },
+    {
+      "epoch": 2.0783838383838384,
+      "grad_norm": 0.0029199772980064154,
+      "learning_rate": 8.318240620957309e-05,
+      "loss": 0.0006,
+      "step": 643
+    },
+    {
+      "epoch": 2.0816161616161617,
+      "grad_norm": 0.0031028217636048794,
+      "learning_rate": 8.33117723156533e-05,
+      "loss": 0.0007,
+      "step": 644
+    },
+    {
+      "epoch": 2.084848484848485,
+      "grad_norm": 0.004011626821011305,
+      "learning_rate": 8.344113842173351e-05,
+      "loss": 0.0009,
+      "step": 645
+    },
+    {
+      "epoch": 2.088080808080808,
+      "grad_norm": 0.003449072130024433,
+      "learning_rate": 8.357050452781372e-05,
+      "loss": 0.0009,
+      "step": 646
+    },
+    {
+      "epoch": 2.0913131313131315,
+      "grad_norm": 0.003367091529071331,
+      "learning_rate": 8.369987063389393e-05,
+      "loss": 0.001,
+      "step": 647
+    },
+    {
+      "epoch": 2.0945454545454547,
+      "grad_norm": 0.003321894910186529,
+      "learning_rate": 8.382923673997413e-05,
+      "loss": 0.0016,
+      "step": 648
+    },
+    {
+      "epoch": 2.097777777777778,
+      "grad_norm": 0.003046546597033739,
+      "learning_rate": 8.395860284605434e-05,
+      "loss": 0.0008,
+      "step": 649
+    },
+    {
+      "epoch": 2.101010101010101,
+      "grad_norm": 0.0023006205447018147,
+      "learning_rate": 8.408796895213455e-05,
+      "loss": 0.0006,
+      "step": 650
+    },
+    {
+      "epoch": 2.101010101010101,
+      "eval_loss": 0.0009433354716747999,
+      "eval_runtime": 18.7374,
+      "eval_samples_per_second": 5.337,
+      "eval_steps_per_second": 1.334,
+      "step": 650
+    },
+    {
+      "epoch": 2.1042424242424245,
+      "grad_norm": 0.004013998433947563,
+      "learning_rate": 8.421733505821476e-05,
+      "loss": 0.001,
+      "step": 651
+    },
+    {
+      "epoch": 2.1074747474747473,
+      "grad_norm": 0.0035437876358628273,
+      "learning_rate": 8.434670116429496e-05,
+      "loss": 0.0007,
+      "step": 652
+    },
+    {
+      "epoch": 2.1107070707070705,
+      "grad_norm": 0.004152386449277401,
+      "learning_rate": 8.447606727037517e-05,
+      "loss": 0.001,
+      "step": 653
+    },
+    {
+      "epoch": 2.113939393939394,
+      "grad_norm": 0.0031009165104478598,
+      "learning_rate": 8.460543337645536e-05,
+      "loss": 0.0007,
+      "step": 654
+    },
+    {
+      "epoch": 2.117171717171717,
+      "grad_norm": 0.002030389616265893,
+      "learning_rate": 8.473479948253557e-05,
+      "loss": 0.0005,
+      "step": 655
+    },
+    {
+      "epoch": 2.1204040404040403,
+      "grad_norm": 0.003986351191997528,
+      "learning_rate": 8.486416558861578e-05,
+      "loss": 0.0011,
+      "step": 656
+    },
+    {
+      "epoch": 2.1236363636363635,
+      "grad_norm": 0.0037633986212313175,
+      "learning_rate": 8.499353169469599e-05,
+      "loss": 0.0017,
+      "step": 657
+    },
+    {
+      "epoch": 2.126868686868687,
+      "grad_norm": 0.003191509749740362,
+      "learning_rate": 8.51228978007762e-05,
+      "loss": 0.0009,
+      "step": 658
+    },
+    {
+      "epoch": 2.13010101010101,
+      "grad_norm": 0.003234416712075472,
+      "learning_rate": 8.525226390685641e-05,
+      "loss": 0.0015,
+      "step": 659
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.00314782140776515,
+      "learning_rate": 8.538163001293662e-05,
+      "loss": 0.0008,
+      "step": 660
+    },
+    {
+      "epoch": 2.1365656565656566,
+      "grad_norm": 0.004626644309610128,
+      "learning_rate": 8.551099611901682e-05,
+      "loss": 0.0017,
+      "step": 661
+    },
+    {
+      "epoch": 2.13979797979798,
+      "grad_norm": 0.0037515638396143913,
+      "learning_rate": 8.564036222509703e-05,
+      "loss": 0.001,
+      "step": 662
+    },
+    {
+      "epoch": 2.143030303030303,
+      "grad_norm": 0.0028777304105460644,
+      "learning_rate": 8.576972833117724e-05,
+      "loss": 0.0009,
+      "step": 663
+    },
+    {
+      "epoch": 2.1462626262626263,
+      "grad_norm": 0.0035881204530596733,
+      "learning_rate": 8.589909443725744e-05,
+      "loss": 0.0008,
+      "step": 664
+    },
+    {
+      "epoch": 2.1494949494949496,
+      "grad_norm": 0.002268304582685232,
+      "learning_rate": 8.602846054333765e-05,
+      "loss": 0.0005,
+      "step": 665
+    },
+    {
+      "epoch": 2.152727272727273,
+      "grad_norm": 0.003100323723629117,
+      "learning_rate": 8.615782664941786e-05,
+      "loss": 0.0008,
+      "step": 666
+    },
+    {
+      "epoch": 2.155959595959596,
+      "grad_norm": 0.003520137397572398,
+      "learning_rate": 8.628719275549805e-05,
+      "loss": 0.0009,
+      "step": 667
+    },
+    {
+      "epoch": 2.1591919191919193,
+      "grad_norm": 0.0042273253202438354,
+      "learning_rate": 8.641655886157827e-05,
+      "loss": 0.0015,
+      "step": 668
+    },
+    {
+      "epoch": 2.1624242424242426,
+      "grad_norm": 0.0030146201606839895,
+      "learning_rate": 8.654592496765848e-05,
+      "loss": 0.0009,
+      "step": 669
+    },
+    {
+      "epoch": 2.165656565656566,
+      "grad_norm": 0.0055503263138234615,
+      "learning_rate": 8.667529107373869e-05,
+      "loss": 0.0025,
+      "step": 670
+    },
+    {
+      "epoch": 2.168888888888889,
+      "grad_norm": 0.0036939766723662615,
+      "learning_rate": 8.68046571798189e-05,
+      "loss": 0.002,
+      "step": 671
+    },
+    {
+      "epoch": 2.172121212121212,
+      "grad_norm": 0.0032185425516217947,
+      "learning_rate": 8.69340232858991e-05,
+      "loss": 0.0009,
+      "step": 672
+    },
+    {
+      "epoch": 2.175353535353535,
+      "grad_norm": 0.0030356363859027624,
+      "learning_rate": 8.706338939197932e-05,
+      "loss": 0.0007,
+      "step": 673
+    },
+    {
+      "epoch": 2.1785858585858584,
+      "grad_norm": 0.006692877039313316,
+      "learning_rate": 8.719275549805951e-05,
+      "loss": 0.0008,
+      "step": 674
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 0.0026474855840206146,
+      "learning_rate": 8.732212160413972e-05,
+      "loss": 0.0007,
+      "step": 675
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "eval_loss": 0.0009410877246409655,
+      "eval_runtime": 18.7748,
+      "eval_samples_per_second": 5.326,
+      "eval_steps_per_second": 1.332,
+      "step": 675
+    },
+    {
+      "epoch": 2.185050505050505,
+      "grad_norm": 0.003672214224934578,
+      "learning_rate": 8.745148771021992e-05,
+      "loss": 0.0018,
+      "step": 676
+    },
+    {
+      "epoch": 2.188282828282828,
+      "grad_norm": 0.0031880387105047703,
+      "learning_rate": 8.758085381630013e-05,
+      "loss": 0.0012,
+      "step": 677
+    },
+    {
+      "epoch": 2.1915151515151514,
+      "grad_norm": 0.0033309224527329206,
+      "learning_rate": 8.771021992238034e-05,
+      "loss": 0.001,
+      "step": 678
+    },
+    {
+      "epoch": 2.1947474747474747,
+      "grad_norm": 0.003027291502803564,
+      "learning_rate": 8.783958602846055e-05,
+      "loss": 0.0008,
+      "step": 679
+    },
+    {
+      "epoch": 2.197979797979798,
+      "grad_norm": 0.0036887172609567642,
+      "learning_rate": 8.796895213454075e-05,
+      "loss": 0.001,
+      "step": 680
+    },
+    {
+      "epoch": 2.201212121212121,
+      "grad_norm": 0.003997828811407089,
+      "learning_rate": 8.809831824062096e-05,
+      "loss": 0.0011,
+      "step": 681
+    },
+    {
+      "epoch": 2.2044444444444444,
+      "grad_norm": 0.003314135130494833,
+      "learning_rate": 8.822768434670117e-05,
+      "loss": 0.0017,
+      "step": 682
+    },
+    {
+      "epoch": 2.2076767676767677,
+      "grad_norm": 0.0036913359072059393,
+      "learning_rate": 8.835705045278138e-05,
+      "loss": 0.0009,
+      "step": 683
+    },
+    {
+      "epoch": 2.210909090909091,
+      "grad_norm": 0.0030470637138932943,
+      "learning_rate": 8.848641655886159e-05,
+      "loss": 0.0008,
+      "step": 684
+    },
+    {
+      "epoch": 2.214141414141414,
+      "grad_norm": 0.004008392803370953,
+      "learning_rate": 8.86157826649418e-05,
+      "loss": 0.0012,
+      "step": 685
+    },
+    {
+      "epoch": 2.2173737373737374,
+      "grad_norm": 0.0055717285722494125,
+      "learning_rate": 8.8745148771022e-05,
+      "loss": 0.0019,
+      "step": 686
+    },
+    {
+      "epoch": 2.2206060606060607,
+      "grad_norm": 0.002990501932799816,
+      "learning_rate": 8.88745148771022e-05,
+      "loss": 0.0009,
+      "step": 687
+    },
+    {
+      "epoch": 2.223838383838384,
+      "grad_norm": 0.002291604643687606,
+      "learning_rate": 8.90038809831824e-05,
+      "loss": 0.0006,
+      "step": 688
+    },
+    {
+      "epoch": 2.227070707070707,
+      "grad_norm": 0.002231605350971222,
+      "learning_rate": 8.913324708926261e-05,
+      "loss": 0.0006,
+      "step": 689
+    },
+    {
+      "epoch": 2.2303030303030305,
+      "grad_norm": 0.002650222275406122,
+      "learning_rate": 8.926261319534282e-05,
+      "loss": 0.0007,
+      "step": 690
+    },
+    {
+      "epoch": 2.2335353535353537,
+      "grad_norm": 0.0019864251371473074,
+      "learning_rate": 8.939197930142303e-05,
+      "loss": 0.0006,
+      "step": 691
+    },
+    {
+      "epoch": 2.236767676767677,
+      "grad_norm": 0.0028250280302017927,
+      "learning_rate": 8.952134540750324e-05,
+      "loss": 0.0006,
+      "step": 692
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0035562533885240555,
+      "learning_rate": 8.965071151358344e-05,
+      "loss": 0.0018,
+      "step": 693
+    },
+    {
+      "epoch": 2.2432323232323235,
+      "grad_norm": 0.0030060771387070417,
+      "learning_rate": 8.978007761966365e-05,
+      "loss": 0.0007,
+      "step": 694
+    },
+    {
+      "epoch": 2.2464646464646463,
+      "grad_norm": 0.002113576978445053,
+      "learning_rate": 8.990944372574386e-05,
+      "loss": 0.0005,
+      "step": 695
+    },
+    {
+      "epoch": 2.2496969696969695,
+      "grad_norm": 0.002114512724801898,
+      "learning_rate": 9.003880983182407e-05,
+      "loss": 0.0006,
+      "step": 696
+    },
+    {
+      "epoch": 2.252929292929293,
+      "grad_norm": 0.0029473064932972193,
+      "learning_rate": 9.016817593790428e-05,
+      "loss": 0.0006,
+      "step": 697
+    },
+    {
+      "epoch": 2.256161616161616,
+      "grad_norm": 0.004192824941128492,
+      "learning_rate": 9.029754204398448e-05,
+      "loss": 0.001,
+      "step": 698
+    },
+    {
+      "epoch": 2.2593939393939393,
+      "grad_norm": 0.003509392263367772,
+      "learning_rate": 9.042690815006469e-05,
+      "loss": 0.0009,
+      "step": 699
+    },
+    {
+      "epoch": 2.2626262626262625,
+      "grad_norm": 0.004787352867424488,
+      "learning_rate": 9.055627425614489e-05,
+      "loss": 0.001,
+      "step": 700
+    },
+    {
+      "epoch": 2.2626262626262625,
+      "eval_loss": 0.000903558568097651,
+      "eval_runtime": 18.694,
+      "eval_samples_per_second": 5.349,
+      "eval_steps_per_second": 1.337,
+      "step": 700
+    },
+    {
+      "epoch": 2.265858585858586,
+      "grad_norm": 0.0027677167672663927,
+      "learning_rate": 9.06856403622251e-05,
+      "loss": 0.0009,
+      "step": 701
+    },
+    {
+      "epoch": 2.269090909090909,
+      "grad_norm": 0.0026491908356547356,
+      "learning_rate": 9.08150064683053e-05,
+      "loss": 0.0007,
+      "step": 702
+    },
+    {
+      "epoch": 2.2723232323232323,
+      "grad_norm": 0.0025004090275615454,
+      "learning_rate": 9.094437257438552e-05,
+      "loss": 0.0007,
+      "step": 703
+    },
+    {
+      "epoch": 2.2755555555555556,
+      "grad_norm": 0.0028262247797101736,
+      "learning_rate": 9.107373868046573e-05,
+      "loss": 0.0008,
+      "step": 704
+    },
+    {
+      "epoch": 2.278787878787879,
+      "grad_norm": 0.003089368110522628,
+      "learning_rate": 9.120310478654594e-05,
+      "loss": 0.0009,
+      "step": 705
+    },
+    {
+      "epoch": 2.282020202020202,
+      "grad_norm": 0.002824244322255254,
+      "learning_rate": 9.133247089262613e-05,
+      "loss": 0.0008,
+      "step": 706
+    },
+    {
+      "epoch": 2.2852525252525253,
+      "grad_norm": 0.0023714362177997828,
+      "learning_rate": 9.146183699870634e-05,
+      "loss": 0.0007,
+      "step": 707
+    },
+    {
+      "epoch": 2.2884848484848486,
+      "grad_norm": 0.0033874395303428173,
+      "learning_rate": 9.159120310478655e-05,
+      "loss": 0.0017,
+      "step": 708
+    },
+    {
+      "epoch": 2.291717171717172,
+      "grad_norm": 0.0033185749780386686,
+      "learning_rate": 9.172056921086676e-05,
+      "loss": 0.0017,
+      "step": 709
+    },
+    {
+      "epoch": 2.294949494949495,
+      "grad_norm": 0.00360478856600821,
+      "learning_rate": 9.184993531694696e-05,
+      "loss": 0.0009,
+      "step": 710
+    },
+    {
+      "epoch": 2.2981818181818183,
+      "grad_norm": 0.0032836326863616705,
+      "learning_rate": 9.197930142302717e-05,
+      "loss": 0.0009,
+      "step": 711
+    },
+    {
+      "epoch": 2.3014141414141416,
+      "grad_norm": 0.0029353760182857513,
+      "learning_rate": 9.210866752910737e-05,
+      "loss": 0.0008,
+      "step": 712
+    },
+    {
+      "epoch": 2.304646464646465,
+      "grad_norm": 0.004789955448359251,
+      "learning_rate": 9.223803363518758e-05,
+      "loss": 0.0015,
+      "step": 713
+    },
+    {
+      "epoch": 2.3078787878787876,
+      "grad_norm": 0.00258410326205194,
+      "learning_rate": 9.236739974126779e-05,
+      "loss": 0.0007,
+      "step": 714
+    },
+    {
+      "epoch": 2.311111111111111,
+      "grad_norm": 0.003211255418136716,
+      "learning_rate": 9.2496765847348e-05,
+      "loss": 0.0009,
+      "step": 715
+    },
+    {
+      "epoch": 2.314343434343434,
+      "grad_norm": 0.003458111546933651,
+      "learning_rate": 9.262613195342821e-05,
+      "loss": 0.0008,
+      "step": 716
+    },
+    {
+      "epoch": 2.3175757575757574,
+      "grad_norm": 0.0020549860782921314,
+      "learning_rate": 9.275549805950842e-05,
+      "loss": 0.0005,
+      "step": 717
+    },
+    {
+      "epoch": 2.3208080808080807,
+      "grad_norm": 0.0034006584901362658,
+      "learning_rate": 9.288486416558863e-05,
+      "loss": 0.0011,
+      "step": 718
+    },
+    {
+      "epoch": 2.324040404040404,
+      "grad_norm": 0.003893098793923855,
+      "learning_rate": 9.301423027166883e-05,
+      "loss": 0.0008,
+      "step": 719
+    },
+    {
+      "epoch": 2.327272727272727,
+      "grad_norm": 0.004374553449451923,
+      "learning_rate": 9.314359637774904e-05,
+      "loss": 0.0014,
+      "step": 720
+    },
+    {
+      "epoch": 2.3305050505050504,
+      "grad_norm": 0.004447360523045063,
+      "learning_rate": 9.327296248382925e-05,
+      "loss": 0.0016,
+      "step": 721
+    },
+    {
+      "epoch": 2.3337373737373737,
+      "grad_norm": 0.003019913798198104,
+      "learning_rate": 9.340232858990944e-05,
+      "loss": 0.0007,
+      "step": 722
+    },
+    {
+      "epoch": 2.336969696969697,
+      "grad_norm": 0.00266702170483768,
+      "learning_rate": 9.353169469598965e-05,
+      "loss": 0.0006,
+      "step": 723
+    },
+    {
+      "epoch": 2.34020202020202,
+      "grad_norm": 0.004037540405988693,
+      "learning_rate": 9.366106080206986e-05,
+      "loss": 0.001,
+      "step": 724
+    },
+    {
+      "epoch": 2.3434343434343434,
+      "grad_norm": 0.003830693429335952,
+      "learning_rate": 9.379042690815006e-05,
+      "loss": 0.0015,
+      "step": 725
+    },
+    {
+      "epoch": 2.3434343434343434,
+      "eval_loss": 0.0009221473592333496,
+      "eval_runtime": 18.682,
+      "eval_samples_per_second": 5.353,
+      "eval_steps_per_second": 1.338,
+      "step": 725
+    },
+    {
+      "epoch": 2.3466666666666667,
+      "grad_norm": 0.002714785747230053,
+      "learning_rate": 9.391979301423027e-05,
+      "loss": 0.0007,
+      "step": 726
+    },
+    {
+      "epoch": 2.34989898989899,
+      "grad_norm": 0.0034484020434319973,
+      "learning_rate": 9.404915912031048e-05,
+      "loss": 0.0021,
+      "step": 727
+    },
+    {
+      "epoch": 2.353131313131313,
+      "grad_norm": 0.0027550137601792812,
+      "learning_rate": 9.417852522639069e-05,
+      "loss": 0.0014,
+      "step": 728
+    },
+    {
+      "epoch": 2.3563636363636364,
+      "grad_norm": 0.004323527216911316,
+      "learning_rate": 9.43078913324709e-05,
+      "loss": 0.0021,
+      "step": 729
+    },
+    {
+      "epoch": 2.3595959595959597,
+      "grad_norm": 0.0036909414920955896,
+      "learning_rate": 9.443725743855111e-05,
+      "loss": 0.0009,
+      "step": 730
+    },
+    {
+      "epoch": 2.362828282828283,
+      "grad_norm": 0.0026773421559482813,
+      "learning_rate": 9.456662354463132e-05,
+      "loss": 0.0007,
+      "step": 731
+    },
+    {
+      "epoch": 2.366060606060606,
+      "grad_norm": 0.0027335931081324816,
+      "learning_rate": 9.469598965071152e-05,
+      "loss": 0.0011,
+      "step": 732
+    },
+    {
+      "epoch": 2.3692929292929295,
+      "grad_norm": 0.0037821868900209665,
+      "learning_rate": 9.482535575679173e-05,
+      "loss": 0.0009,
+      "step": 733
+    },
+    {
+      "epoch": 2.3725252525252527,
+      "grad_norm": 0.0027637695893645287,
+      "learning_rate": 9.495472186287193e-05,
+      "loss": 0.0008,
+      "step": 734
+    },
+    {
+      "epoch": 2.375757575757576,
+      "grad_norm": 0.0028088942635804415,
+      "learning_rate": 9.508408796895214e-05,
+      "loss": 0.0008,
+      "step": 735
+    },
+    {
+      "epoch": 2.378989898989899,
+      "grad_norm": 0.004683246370404959,
+      "learning_rate": 9.521345407503235e-05,
+      "loss": 0.0011,
+      "step": 736
+    },
+    {
+      "epoch": 2.3822222222222225,
+      "grad_norm": 0.0035344541538506746,
+      "learning_rate": 9.534282018111256e-05,
+      "loss": 0.001,
+      "step": 737
+    },
+    {
+      "epoch": 2.3854545454545453,
+      "grad_norm": 0.003326730104163289,
+      "learning_rate": 9.547218628719275e-05,
+      "loss": 0.001,
+      "step": 738
+    },
+    {
+      "epoch": 2.3886868686868685,
+      "grad_norm": 0.0027331975288689137,
+      "learning_rate": 9.560155239327296e-05,
+      "loss": 0.001,
+      "step": 739
+    },
+    {
+      "epoch": 2.391919191919192,
+      "grad_norm": 0.0023180183488875628,
+      "learning_rate": 9.573091849935317e-05,
+      "loss": 0.0005,
+      "step": 740
+    },
+    {
+      "epoch": 2.395151515151515,
+      "grad_norm": 0.00545561034232378,
+      "learning_rate": 9.586028460543338e-05,
+      "loss": 0.0019,
+      "step": 741
+    },
+    {
+      "epoch": 2.3983838383838383,
+      "grad_norm": 0.002278296509757638,
+      "learning_rate": 9.59896507115136e-05,
+      "loss": 0.0008,
+      "step": 742
+    },
+    {
+      "epoch": 2.4016161616161615,
+      "grad_norm": 0.0036029706243425608,
+      "learning_rate": 9.61190168175938e-05,
+      "loss": 0.0011,
+      "step": 743
+    },
+    {
+      "epoch": 2.404848484848485,
+      "grad_norm": 0.0028302932623773813,
+      "learning_rate": 9.6248382923674e-05,
+      "loss": 0.001,
+      "step": 744
+    },
+    {
+      "epoch": 2.408080808080808,
+      "grad_norm": 0.005623773206025362,
+      "learning_rate": 9.63777490297542e-05,
+      "loss": 0.0025,
+      "step": 745
+    },
+    {
+      "epoch": 2.4113131313131313,
+      "grad_norm": 0.003791616763919592,
+      "learning_rate": 9.650711513583441e-05,
+      "loss": 0.001,
+      "step": 746
+    },
+    {
+      "epoch": 2.4145454545454546,
+      "grad_norm": 0.003847538959234953,
+      "learning_rate": 9.663648124191462e-05,
+      "loss": 0.002,
+      "step": 747
+    },
+    {
+      "epoch": 2.417777777777778,
+      "grad_norm": 0.003144781803712249,
+      "learning_rate": 9.676584734799483e-05,
+      "loss": 0.0008,
+      "step": 748
+    },
+    {
+      "epoch": 2.421010101010101,
+      "grad_norm": 0.0030476360116153955,
+      "learning_rate": 9.689521345407504e-05,
+      "loss": 0.002,
+      "step": 749
+    },
+    {
+      "epoch": 2.4242424242424243,
+      "grad_norm": 0.023542512208223343,
+      "learning_rate": 9.702457956015525e-05,
+      "loss": 0.0012,
+      "step": 750
+    },
+    {
+      "epoch": 2.4242424242424243,
+      "eval_loss": 0.0009755383944138885,
+      "eval_runtime": 18.7174,
+      "eval_samples_per_second": 5.343,
+      "eval_steps_per_second": 1.336,
+      "step": 750
+    },
+    {
+      "epoch": 2.4274747474747476,
+      "grad_norm": 0.005616334266960621,
+      "learning_rate": 9.715394566623545e-05,
+      "loss": 0.0009,
+      "step": 751
+    },
+    {
+      "epoch": 2.430707070707071,
+      "grad_norm": 0.007717492058873177,
+      "learning_rate": 9.728331177231566e-05,
+      "loss": 0.0008,
+      "step": 752
+    },
+    {
+      "epoch": 2.433939393939394,
+      "grad_norm": 0.008256220258772373,
+      "learning_rate": 9.741267787839587e-05,
+      "loss": 0.0011,
+      "step": 753
+    },
+    {
+      "epoch": 2.4371717171717173,
+      "grad_norm": 0.005972269922494888,
+      "learning_rate": 9.754204398447608e-05,
+      "loss": 0.0004,
+      "step": 754
+    },
+    {
+      "epoch": 2.4404040404040406,
+      "grad_norm": 0.006538939196616411,
+      "learning_rate": 9.767141009055629e-05,
+      "loss": 0.0009,
+      "step": 755
+    },
+    {
+      "epoch": 2.443636363636364,
+      "grad_norm": 0.003234976204112172,
+      "learning_rate": 9.780077619663648e-05,
+      "loss": 0.0011,
+      "step": 756
+    },
+    {
+      "epoch": 2.4468686868686866,
+      "grad_norm": 0.0033180294558405876,
+      "learning_rate": 9.793014230271668e-05,
+      "loss": 0.0009,
+      "step": 757
+    },
+    {
+      "epoch": 2.45010101010101,
+      "grad_norm": 0.0037402757443487644,
+      "learning_rate": 9.805950840879689e-05,
+      "loss": 0.0007,
+      "step": 758
+    },
+    {
+      "epoch": 2.453333333333333,
+      "grad_norm": 0.0028039240278303623,
+      "learning_rate": 9.81888745148771e-05,
+      "loss": 0.0008,
+      "step": 759
+    },
+    {
+      "epoch": 2.4565656565656564,
+      "grad_norm": 0.0029848841950297356,
+      "learning_rate": 9.831824062095731e-05,
+      "loss": 0.0007,
+      "step": 760
+    },
+    {
+      "epoch": 2.4597979797979797,
+      "grad_norm": 0.002809871220961213,
+      "learning_rate": 9.844760672703752e-05,
+      "loss": 0.0007,
+      "step": 761
+    },
+    {
+      "epoch": 2.463030303030303,
+      "grad_norm": 0.004293316043913364,
+      "learning_rate": 9.857697283311773e-05,
+      "loss": 0.0012,
+      "step": 762
+    },
+    {
+      "epoch": 2.466262626262626,
+      "grad_norm": 0.003171822987496853,
+      "learning_rate": 9.870633893919794e-05,
+      "loss": 0.0008,
+      "step": 763
+    },
+    {
+      "epoch": 2.4694949494949494,
+      "grad_norm": 0.0035177527461200953,
+      "learning_rate": 9.883570504527814e-05,
+      "loss": 0.0008,
+      "step": 764
+    },
+    {
+      "epoch": 2.4727272727272727,
+      "grad_norm": 0.0037470462266355753,
+      "learning_rate": 9.896507115135835e-05,
+      "loss": 0.0011,
+      "step": 765
+    },
+    {
+      "epoch": 2.475959595959596,
+      "grad_norm": 0.004131955560296774,
+      "learning_rate": 9.909443725743856e-05,
+      "loss": 0.0015,
+      "step": 766
+    },
+    {
+      "epoch": 2.479191919191919,
+      "grad_norm": 0.002823730930685997,
+      "learning_rate": 9.922380336351877e-05,
+      "loss": 0.0007,
+      "step": 767
+    },
+    {
+      "epoch": 2.4824242424242424,
+      "grad_norm": 0.0037240665405988693,
+      "learning_rate": 9.935316946959897e-05,
+      "loss": 0.001,
+      "step": 768
+    },
+    {
+      "epoch": 2.4856565656565657,
+      "grad_norm": 0.0038744148332625628,
+      "learning_rate": 9.948253557567918e-05,
+      "loss": 0.0019,
+      "step": 769
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 0.003629435319453478,
+      "learning_rate": 9.961190168175937e-05,
+      "loss": 0.0008,
+      "step": 770
+    },
+    {
+      "epoch": 2.492121212121212,
+      "grad_norm": 0.004831044003367424,
+      "learning_rate": 9.974126778783958e-05,
+      "loss": 0.0011,
+      "step": 771
+    },
+    {
+      "epoch": 2.4953535353535354,
+      "grad_norm": 0.003547330852597952,
+      "learning_rate": 9.98706338939198e-05,
+      "loss": 0.0009,
+      "step": 772
+    },
+    {
+      "epoch": 2.4985858585858587,
+      "grad_norm": 0.003263116115704179,
+      "learning_rate": 0.0001,
+      "loss": 0.0008,
+      "step": 773
+    },
+    {
+      "epoch": 2.501818181818182,
+      "grad_norm": 0.0030386094003915787,
+      "learning_rate": 9.999999489471233e-05,
+      "loss": 0.0007,
+      "step": 774
+    },
+    {
+      "epoch": 2.505050505050505,
+      "grad_norm": 0.0036390256136655807,
+      "learning_rate": 9.99999795788503e-05,
+      "loss": 0.0012,
+      "step": 775
+    },
+    {
+      "epoch": 2.505050505050505,
+      "eval_loss": 0.000891694042365998,
+      "eval_runtime": 18.7365,
+      "eval_samples_per_second": 5.337,
+      "eval_steps_per_second": 1.334,
+      "step": 775
+    },
+    {
+      "epoch": 2.5082828282828284,
+      "grad_norm": 0.0027272645384073257,
+      "learning_rate": 9.99999540524171e-05,
+      "loss": 0.0006,
+      "step": 776
+    },
+    {
+      "epoch": 2.5115151515151517,
+      "grad_norm": 0.004212843254208565,
+      "learning_rate": 9.999991831541789e-05,
+      "loss": 0.0023,
+      "step": 777
+    },
+    {
+      "epoch": 2.514747474747475,
+      "grad_norm": 0.0034173503518104553,
+      "learning_rate": 9.999987236786e-05,
+      "loss": 0.0008,
+      "step": 778
+    },
+    {
+      "epoch": 2.517979797979798,
+      "grad_norm": 0.0028145266696810722,
+      "learning_rate": 9.999981620975281e-05,
+      "loss": 0.0007,
+      "step": 779
+    },
+    {
+      "epoch": 2.5212121212121215,
+      "grad_norm": 0.0030887024477124214,
+      "learning_rate": 9.999974984110779e-05,
+      "loss": 0.0009,
+      "step": 780
+    },
+    {
+      "epoch": 2.5244444444444447,
+      "grad_norm": 0.003968552686274052,
+      "learning_rate": 9.999967326193847e-05,
+      "loss": 0.0009,
+      "step": 781
+    },
+    {
+      "epoch": 2.5276767676767675,
+      "grad_norm": 0.00264795683324337,
+      "learning_rate": 9.999958647226049e-05,
+      "loss": 0.0006,
+      "step": 782
+    },
+    {
+      "epoch": 2.5309090909090908,
+      "grad_norm": 0.004550054203718901,
+      "learning_rate": 9.999948947209162e-05,
+      "loss": 0.002,
+      "step": 783
+    },
+    {
+      "epoch": 2.534141414141414,
+      "grad_norm": 0.0030430385377258062,
+      "learning_rate": 9.999938226145161e-05,
+      "loss": 0.0007,
+      "step": 784
+    },
+    {
+      "epoch": 2.5373737373737373,
+      "grad_norm": 0.003766052657738328,
+      "learning_rate": 9.999926484036237e-05,
+      "loss": 0.001,
+      "step": 785
+    },
+    {
+      "epoch": 2.5406060606060605,
+      "grad_norm": 0.003455114783719182,
+      "learning_rate": 9.999913720884791e-05,
+      "loss": 0.002,
+      "step": 786
+    },
+    {
+      "epoch": 2.543838383838384,
+      "grad_norm": 0.002563537796959281,
+      "learning_rate": 9.999899936693426e-05,
+      "loss": 0.0008,
+      "step": 787
+    },
+    {
+      "epoch": 2.547070707070707,
+      "grad_norm": 0.0025122894439846277,
+      "learning_rate": 9.99988513146496e-05,
+      "loss": 0.0006,
+      "step": 788
+    },
+    {
+      "epoch": 2.5503030303030303,
+      "grad_norm": 0.004589984659105539,
+      "learning_rate": 9.999869305202412e-05,
+      "loss": 0.0015,
+      "step": 789
+    },
+    {
+      "epoch": 2.5535353535353535,
+      "grad_norm": 0.0041996450163424015,
+      "learning_rate": 9.999852457909018e-05,
+      "loss": 0.0019,
+      "step": 790
+    },
+    {
+      "epoch": 2.556767676767677,
+      "grad_norm": 0.0033383311238139868,
+      "learning_rate": 9.999834589588217e-05,
+      "loss": 0.0013,
+      "step": 791
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0018662047805264592,
+      "learning_rate": 9.999815700243656e-05,
+      "loss": 0.0006,
+      "step": 792
+    },
+    {
+      "epoch": 2.5632323232323233,
+      "grad_norm": 0.0032810529228299856,
+      "learning_rate": 9.999795789879196e-05,
+      "loss": 0.0009,
+      "step": 793
+    },
+    {
+      "epoch": 2.5664646464646466,
+      "grad_norm": 0.003441553795710206,
+      "learning_rate": 9.9997748584989e-05,
+      "loss": 0.0013,
+      "step": 794
+    },
+    {
+      "epoch": 2.56969696969697,
+      "grad_norm": 0.0025200124364346266,
+      "learning_rate": 9.999752906107042e-05,
+      "loss": 0.0008,
+      "step": 795
+    },
+    {
+      "epoch": 2.572929292929293,
+      "grad_norm": 0.0030850358307361603,
+      "learning_rate": 9.999729932708109e-05,
+      "loss": 0.0009,
+      "step": 796
+    },
+    {
+      "epoch": 2.5761616161616163,
+      "grad_norm": 0.0029367755632847548,
+      "learning_rate": 9.999705938306789e-05,
+      "loss": 0.0007,
+      "step": 797
+    },
+    {
+      "epoch": 2.579393939393939,
+      "grad_norm": 0.003588201245293021,
+      "learning_rate": 9.999680922907982e-05,
+      "loss": 0.0026,
+      "step": 798
+    },
+    {
+      "epoch": 2.5826262626262624,
+      "grad_norm": 0.0031253646593540907,
+      "learning_rate": 9.999654886516798e-05,
+      "loss": 0.0009,
+      "step": 799
+    },
+    {
+      "epoch": 2.5858585858585856,
+      "grad_norm": 0.0036176196299493313,
+      "learning_rate": 9.999627829138554e-05,
+      "loss": 0.0015,
+      "step": 800
+    },
+    {
+      "epoch": 2.5858585858585856,
+      "eval_loss": 0.0009534513228572905,
+      "eval_runtime": 18.6223,
+      "eval_samples_per_second": 5.37,
+      "eval_steps_per_second": 1.342,
+      "step": 800
+    },
+    {
+      "epoch": 2.589090909090909,
+      "grad_norm": 0.00325636169873178,
+      "learning_rate": 9.999599750778772e-05,
+      "loss": 0.0007,
+      "step": 801
+    },
+    {
+      "epoch": 2.592323232323232,
+      "grad_norm": 0.003968099132180214,
+      "learning_rate": 9.999570651443191e-05,
+      "loss": 0.0011,
+      "step": 802
+    },
+    {
+      "epoch": 2.5955555555555554,
+      "grad_norm": 0.00244711060076952,
+      "learning_rate": 9.99954053113775e-05,
+      "loss": 0.0006,
+      "step": 803
+    },
+    {
+      "epoch": 2.5987878787878786,
+      "grad_norm": 0.003342804964631796,
+      "learning_rate": 9.9995093898686e-05,
+      "loss": 0.0014,
+      "step": 804
+    },
+    {
+      "epoch": 2.602020202020202,
+      "grad_norm": 0.006182527635246515,
+      "learning_rate": 9.999477227642103e-05,
+      "loss": 0.0025,
+      "step": 805
+    },
+    {
+      "epoch": 2.605252525252525,
+      "grad_norm": 0.0032223479356616735,
+      "learning_rate": 9.999444044464823e-05,
+      "loss": 0.0007,
+      "step": 806
+    },
+    {
+      "epoch": 2.6084848484848484,
+      "grad_norm": 0.002208409830927849,
+      "learning_rate": 9.999409840343539e-05,
+      "loss": 0.0007,
+      "step": 807
+    },
+    {
+      "epoch": 2.6117171717171717,
+      "grad_norm": 0.002247196389362216,
+      "learning_rate": 9.999374615285236e-05,
+      "loss": 0.0007,
+      "step": 808
+    },
+    {
+      "epoch": 2.614949494949495,
+      "grad_norm": 0.0024588643573224545,
+      "learning_rate": 9.999338369297106e-05,
+      "loss": 0.0005,
+      "step": 809
+    },
+    {
+      "epoch": 2.618181818181818,
+      "grad_norm": 0.003213444259017706,
+      "learning_rate": 9.999301102386553e-05,
+      "loss": 0.0008,
+      "step": 810
+    },
+    {
+      "epoch": 2.6214141414141414,
+      "grad_norm": 0.0028623330872505903,
+      "learning_rate": 9.999262814561185e-05,
+      "loss": 0.0009,
+      "step": 811
+    },
+    {
+      "epoch": 2.6246464646464647,
+      "grad_norm": 0.0031820102594792843,
+      "learning_rate": 9.999223505828821e-05,
+      "loss": 0.001,
+      "step": 812
+    },
+    {
+      "epoch": 2.627878787878788,
+      "grad_norm": 0.0017483988776803017,
+      "learning_rate": 9.999183176197491e-05,
+      "loss": 0.0006,
+      "step": 813
+    },
+    {
+      "epoch": 2.631111111111111,
+      "grad_norm": 0.0020935633219778538,
+      "learning_rate": 9.999141825675426e-05,
+      "loss": 0.0006,
+      "step": 814
+    },
+    {
+      "epoch": 2.6343434343434344,
+      "grad_norm": 0.003117109416052699,
+      "learning_rate": 9.999099454271074e-05,
+      "loss": 0.0008,
+      "step": 815
+    },
+    {
+      "epoch": 2.6375757575757577,
+      "grad_norm": 0.0037443467881530523,
+      "learning_rate": 9.999056061993089e-05,
+      "loss": 0.001,
+      "step": 816
+    },
+    {
+      "epoch": 2.640808080808081,
+      "grad_norm": 0.0032844438683241606,
+      "learning_rate": 9.999011648850329e-05,
+      "loss": 0.0009,
+      "step": 817
+    },
+    {
+      "epoch": 2.644040404040404,
+      "grad_norm": 0.0027299323119223118,
+      "learning_rate": 9.998966214851864e-05,
+      "loss": 0.001,
+      "step": 818
+    },
+    {
+      "epoch": 2.6472727272727274,
+      "grad_norm": 0.003157002152875066,
+      "learning_rate": 9.998919760006972e-05,
+      "loss": 0.001,
+      "step": 819
+    },
+    {
+      "epoch": 2.6505050505050507,
+      "grad_norm": 0.003021983429789543,
+      "learning_rate": 9.998872284325142e-05,
+      "loss": 0.0011,
+      "step": 820
+    },
+    {
+      "epoch": 2.653737373737374,
+      "grad_norm": 0.0036499700509011745,
+      "learning_rate": 9.998823787816066e-05,
+      "loss": 0.001,
+      "step": 821
+    },
+    {
+      "epoch": 2.656969696969697,
+      "grad_norm": 0.0031615635380148888,
+      "learning_rate": 9.99877427048965e-05,
+      "loss": 0.0009,
+      "step": 822
+    },
+    {
+      "epoch": 2.6602020202020205,
+      "grad_norm": 0.0032300378661602736,
+      "learning_rate": 9.998723732356006e-05,
+      "loss": 0.0009,
+      "step": 823
+    },
+    {
+      "epoch": 2.6634343434343437,
+      "grad_norm": 0.00247188750654459,
+      "learning_rate": 9.998672173425452e-05,
+      "loss": 0.0008,
+      "step": 824
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.004531141836196184,
+      "learning_rate": 9.998619593708518e-05,
+      "loss": 0.0011,
+      "step": 825
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "eval_loss": 0.0008804297540336847,
+      "eval_runtime": 18.6211,
+      "eval_samples_per_second": 5.37,
+      "eval_steps_per_second": 1.343,
+      "step": 825
+    },
+    {
+      "epoch": 2.6698989898989898,
+      "grad_norm": 0.0034359728451818228,
+      "learning_rate": 9.998565993215943e-05,
+      "loss": 0.0009,
+      "step": 826
+    },
+    {
+      "epoch": 2.673131313131313,
+      "grad_norm": 0.002785489894449711,
+      "learning_rate": 9.998511371958672e-05,
+      "loss": 0.0007,
+      "step": 827
+    },
+    {
+      "epoch": 2.6763636363636363,
+      "grad_norm": 0.0025934847071766853,
+      "learning_rate": 9.998455729947858e-05,
+      "loss": 0.0008,
+      "step": 828
+    },
+    {
+      "epoch": 2.6795959595959595,
+      "grad_norm": 0.0026367492973804474,
+      "learning_rate": 9.998399067194864e-05,
+      "loss": 0.0008,
+      "step": 829
+    },
+    {
+      "epoch": 2.682828282828283,
+      "grad_norm": 0.003330858191475272,
+      "learning_rate": 9.998341383711263e-05,
+      "loss": 0.0007,
+      "step": 830
+    },
+    {
+      "epoch": 2.686060606060606,
+      "grad_norm": 0.003155591432005167,
+      "learning_rate": 9.998282679508835e-05,
+      "loss": 0.0017,
+      "step": 831
+    },
+    {
+      "epoch": 2.6892929292929293,
+      "grad_norm": 0.0030347639694809914,
+      "learning_rate": 9.998222954599563e-05,
+      "loss": 0.0007,
+      "step": 832
+    },
+    {
+      "epoch": 2.6925252525252525,
+      "grad_norm": 0.0032036558259278536,
+      "learning_rate": 9.99816220899565e-05,
+      "loss": 0.0009,
+      "step": 833
+    },
+    {
+      "epoch": 2.695757575757576,
+      "grad_norm": 0.003590774955227971,
+      "learning_rate": 9.998100442709497e-05,
+      "loss": 0.0009,
+      "step": 834
+    },
+    {
+      "epoch": 2.698989898989899,
+      "grad_norm": 0.002222040668129921,
+      "learning_rate": 9.998037655753717e-05,
+      "loss": 0.0009,
+      "step": 835
+    },
+    {
+      "epoch": 2.7022222222222223,
+      "grad_norm": 0.0027599751483649015,
+      "learning_rate": 9.997973848141137e-05,
+      "loss": 0.0009,
+      "step": 836
+    },
+    {
+      "epoch": 2.7054545454545456,
+      "grad_norm": 0.00284950970672071,
+      "learning_rate": 9.997909019884781e-05,
+      "loss": 0.0008,
+      "step": 837
+    },
+    {
+      "epoch": 2.708686868686869,
+      "grad_norm": 0.0024073810782283545,
+      "learning_rate": 9.99784317099789e-05,
+      "loss": 0.001,
+      "step": 838
+    },
+    {
+      "epoch": 2.711919191919192,
+      "grad_norm": 0.0033184473868459463,
+      "learning_rate": 9.997776301493914e-05,
+      "loss": 0.0016,
+      "step": 839
+    },
+    {
+      "epoch": 2.7151515151515153,
+      "grad_norm": 0.0023880742955952883,
+      "learning_rate": 9.997708411386501e-05,
+      "loss": 0.0015,
+      "step": 840
+    },
+    {
+      "epoch": 2.718383838383838,
+      "grad_norm": 0.0031732122879475355,
+      "learning_rate": 9.997639500689523e-05,
+      "loss": 0.0007,
+      "step": 841
+    },
+    {
+      "epoch": 2.7216161616161614,
+      "grad_norm": 0.0036892895586788654,
+      "learning_rate": 9.997569569417049e-05,
+      "loss": 0.002,
+      "step": 842
+    },
+    {
+      "epoch": 2.7248484848484846,
+      "grad_norm": 0.002280700486153364,
+      "learning_rate": 9.997498617583358e-05,
+      "loss": 0.0018,
+      "step": 843
+    },
+    {
+      "epoch": 2.728080808080808,
+      "grad_norm": 0.003923532087355852,
+      "learning_rate": 9.997426645202943e-05,
+      "loss": 0.0006,
+      "step": 844
+    },
+    {
+      "epoch": 2.731313131313131,
+      "grad_norm": 0.0025437732692807913,
+      "learning_rate": 9.9973536522905e-05,
+      "loss": 0.0009,
+      "step": 845
+    },
+    {
+      "epoch": 2.7345454545454544,
+      "grad_norm": 0.002573117846623063,
+      "learning_rate": 9.997279638860933e-05,
+      "loss": 0.0005,
+      "step": 846
+    },
+    {
+      "epoch": 2.7377777777777776,
+      "grad_norm": 0.0021492561791092157,
+      "learning_rate": 9.99720460492936e-05,
+      "loss": 0.0008,
+      "step": 847
+    },
+    {
+      "epoch": 2.741010101010101,
+      "grad_norm": 0.0021558962762355804,
+      "learning_rate": 9.997128550511099e-05,
+      "loss": 0.0006,
+      "step": 848
+    },
+    {
+      "epoch": 2.744242424242424,
+      "grad_norm": 0.0031675281934440136,
+      "learning_rate": 9.997051475621687e-05,
+      "loss": 0.001,
+      "step": 849
+    },
+    {
+      "epoch": 2.7474747474747474,
+      "grad_norm": 0.003528911853209138,
+      "learning_rate": 9.996973380276857e-05,
+      "loss": 0.0007,
+      "step": 850
+    },
+    {
+      "epoch": 2.7474747474747474,
+      "eval_loss": 0.000889240182004869,
+      "eval_runtime": 18.6302,
+      "eval_samples_per_second": 5.368,
+      "eval_steps_per_second": 1.342,
+      "step": 850
+    },
+    {
+      "epoch": 2.7507070707070707,
+      "grad_norm": 0.003921836614608765,
+      "learning_rate": 9.996894264492563e-05,
+      "loss": 0.001,
+      "step": 851
+    },
+    {
+      "epoch": 2.753939393939394,
+      "grad_norm": 0.002436129143461585,
+      "learning_rate": 9.99681412828496e-05,
+      "loss": 0.0007,
+      "step": 852
+    },
+    {
+      "epoch": 2.757171717171717,
+      "grad_norm": 0.002964465180411935,
+      "learning_rate": 9.996732971670408e-05,
+      "loss": 0.0008,
+      "step": 853
+    },
+    {
+      "epoch": 2.7604040404040404,
+      "grad_norm": 0.0034572421573102474,
+      "learning_rate": 9.996650794665487e-05,
+      "loss": 0.0007,
+      "step": 854
+    },
+    {
+      "epoch": 2.7636363636363637,
+      "grad_norm": 0.0021866310853511095,
+      "learning_rate": 9.996567597286974e-05,
+      "loss": 0.0005,
+      "step": 855
+    },
+    {
+      "epoch": 2.766868686868687,
+      "grad_norm": 0.0026556740049272776,
+      "learning_rate": 9.996483379551861e-05,
+      "loss": 0.0014,
+      "step": 856
+    },
+    {
+      "epoch": 2.77010101010101,
+      "grad_norm": 0.003023721743375063,
+      "learning_rate": 9.996398141477344e-05,
+      "loss": 0.0007,
+      "step": 857
+    },
+    {
+      "epoch": 2.7733333333333334,
+      "grad_norm": 0.0022941052448004484,
+      "learning_rate": 9.996311883080832e-05,
+      "loss": 0.0006,
+      "step": 858
+    },
+    {
+      "epoch": 2.7765656565656567,
+      "grad_norm": 0.003141100751236081,
+      "learning_rate": 9.996224604379938e-05,
+      "loss": 0.0007,
+      "step": 859
+    },
+    {
+      "epoch": 2.77979797979798,
+      "grad_norm": 0.002024593763053417,
+      "learning_rate": 9.996136305392487e-05,
+      "loss": 0.0006,
+      "step": 860
+    },
+    {
+      "epoch": 2.783030303030303,
+      "grad_norm": 0.004872876685112715,
+      "learning_rate": 9.996046986136509e-05,
+      "loss": 0.0026,
+      "step": 861
+    },
+    {
+      "epoch": 2.7862626262626264,
+      "grad_norm": 0.002601329004392028,
+      "learning_rate": 9.995956646630246e-05,
+      "loss": 0.0009,
+      "step": 862
+    },
+    {
+      "epoch": 2.7894949494949497,
+      "grad_norm": 0.0028132593724876642,
+      "learning_rate": 9.995865286892145e-05,
+      "loss": 0.0009,
+      "step": 863
+    },
+    {
+      "epoch": 2.792727272727273,
+      "grad_norm": 0.002662686165422201,
+      "learning_rate": 9.995772906940864e-05,
+      "loss": 0.0007,
+      "step": 864
+    },
+    {
+      "epoch": 2.795959595959596,
+      "grad_norm": 0.0027195930015295744,
+      "learning_rate": 9.995679506795264e-05,
+      "loss": 0.0007,
+      "step": 865
+    },
+    {
+      "epoch": 2.7991919191919195,
+      "grad_norm": 0.003179864026606083,
+      "learning_rate": 9.995585086474424e-05,
+      "loss": 0.0014,
+      "step": 866
+    },
+    {
+      "epoch": 2.8024242424242423,
+      "grad_norm": 0.0034151843283325434,
+      "learning_rate": 9.995489645997622e-05,
+      "loss": 0.0009,
+      "step": 867
+    },
+    {
+      "epoch": 2.8056565656565655,
+      "grad_norm": 0.0026823675725609064,
+      "learning_rate": 9.99539318538435e-05,
+      "loss": 0.0007,
+      "step": 868
+    },
+    {
+      "epoch": 2.8088888888888888,
+      "grad_norm": 0.0021664374507963657,
+      "learning_rate": 9.995295704654304e-05,
+      "loss": 0.0012,
+      "step": 869
+    },
+    {
+      "epoch": 2.812121212121212,
+      "grad_norm": 0.003440326079726219,
+      "learning_rate": 9.995197203827393e-05,
+      "loss": 0.0009,
+      "step": 870
+    },
+    {
+      "epoch": 2.8153535353535353,
+      "grad_norm": 0.003980646841228008,
+      "learning_rate": 9.995097682923733e-05,
+      "loss": 0.0018,
+      "step": 871
+    },
+    {
+      "epoch": 2.8185858585858585,
+      "grad_norm": 0.003389423480257392,
+      "learning_rate": 9.994997141963644e-05,
+      "loss": 0.001,
+      "step": 872
+    },
+    {
+      "epoch": 2.821818181818182,
+      "grad_norm": 0.0027842579875141382,
+      "learning_rate": 9.994895580967658e-05,
+      "loss": 0.0007,
+      "step": 873
+    },
+    {
+      "epoch": 2.825050505050505,
+      "grad_norm": 0.0033308009151369333,
+      "learning_rate": 9.994792999956518e-05,
+      "loss": 0.0011,
+      "step": 874
+    },
+    {
+      "epoch": 2.8282828282828283,
+      "grad_norm": 0.002514815656468272,
+      "learning_rate": 9.994689398951169e-05,
+      "loss": 0.0009,
+      "step": 875
+    },
+    {
+      "epoch": 2.8282828282828283,
+      "eval_loss": 0.0008539481204934418,
+      "eval_runtime": 18.6059,
+      "eval_samples_per_second": 5.375,
+      "eval_steps_per_second": 1.344,
+      "step": 875
+    },
+    {
+      "epoch": 2.8315151515151515,
+      "grad_norm": 0.0024293591268360615,
+      "learning_rate": 9.994584777972769e-05,
+      "loss": 0.0008,
+      "step": 876
+    },
+    {
+      "epoch": 2.834747474747475,
+      "grad_norm": 0.0035905223339796066,
+      "learning_rate": 9.994479137042683e-05,
+      "loss": 0.002,
+      "step": 877
+    },
+    {
+      "epoch": 2.837979797979798,
+      "grad_norm": 0.002918825950473547,
+      "learning_rate": 9.994372476182484e-05,
+      "loss": 0.0009,
+      "step": 878
+    },
+    {
+      "epoch": 2.8412121212121213,
+      "grad_norm": 0.0031527606770396233,
+      "learning_rate": 9.994264795413953e-05,
+      "loss": 0.0008,
+      "step": 879
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 0.0035410267300903797,
+      "learning_rate": 9.99415609475908e-05,
+      "loss": 0.0021,
+      "step": 880
+    },
+    {
+      "epoch": 2.847676767676768,
+      "grad_norm": 0.0020397889893501997,
+      "learning_rate": 9.994046374240062e-05,
+      "loss": 0.0007,
+      "step": 881
+    },
+    {
+      "epoch": 2.850909090909091,
+      "grad_norm": 0.002342833438888192,
+      "learning_rate": 9.993935633879306e-05,
+      "loss": 0.0007,
+      "step": 882
+    },
+    {
+      "epoch": 2.854141414141414,
+      "grad_norm": 0.003542246064171195,
+      "learning_rate": 9.993823873699426e-05,
+      "loss": 0.0008,
+      "step": 883
+    },
+    {
+      "epoch": 2.857373737373737,
+      "grad_norm": 0.0024878752883523703,
+      "learning_rate": 9.993711093723245e-05,
+      "loss": 0.0008,
+      "step": 884
+    },
+    {
+      "epoch": 2.8606060606060604,
+      "grad_norm": 0.0030815114732831717,
+      "learning_rate": 9.993597293973796e-05,
+      "loss": 0.0018,
+      "step": 885
+    },
+    {
+      "epoch": 2.8638383838383836,
+      "grad_norm": 0.004479612223803997,
+      "learning_rate": 9.993482474474314e-05,
+      "loss": 0.0009,
+      "step": 886
+    },
+    {
+      "epoch": 2.867070707070707,
+      "grad_norm": 0.0033372335601598024,
+      "learning_rate": 9.99336663524825e-05,
+      "loss": 0.0009,
+      "step": 887
+    },
+    {
+      "epoch": 2.87030303030303,
+      "grad_norm": 0.0021939794532954693,
+      "learning_rate": 9.993249776319258e-05,
+      "loss": 0.0006,
+      "step": 888
+    },
+    {
+      "epoch": 2.8735353535353534,
+      "grad_norm": 0.0028632464818656445,
+      "learning_rate": 9.993131897711202e-05,
+      "loss": 0.0009,
+      "step": 889
+    },
+    {
+      "epoch": 2.8767676767676766,
+      "grad_norm": 0.002615751465782523,
+      "learning_rate": 9.993012999448154e-05,
+      "loss": 0.0009,
+      "step": 890
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0017363366205245256,
+      "learning_rate": 9.992893081554397e-05,
+      "loss": 0.0005,
+      "step": 891
+    },
+    {
+      "epoch": 2.883232323232323,
+      "grad_norm": 0.0032067778520286083,
+      "learning_rate": 9.992772144054415e-05,
+      "loss": 0.0009,
+      "step": 892
+    },
+    {
+      "epoch": 2.8864646464646464,
+      "grad_norm": 0.0025459511671215296,
+      "learning_rate": 9.992650186972909e-05,
+      "loss": 0.0008,
+      "step": 893
+    },
+    {
+      "epoch": 2.8896969696969697,
+      "grad_norm": 0.001871303771622479,
+      "learning_rate": 9.99252721033478e-05,
+      "loss": 0.0008,
+      "step": 894
+    },
+    {
+      "epoch": 2.892929292929293,
+      "grad_norm": 0.001998218474909663,
+      "learning_rate": 9.992403214165147e-05,
+      "loss": 0.0005,
+      "step": 895
+    },
+    {
+      "epoch": 2.896161616161616,
+      "grad_norm": 0.0023115125950425863,
+      "learning_rate": 9.992278198489327e-05,
+      "loss": 0.0007,
+      "step": 896
+    },
+    {
+      "epoch": 2.8993939393939394,
+      "grad_norm": 0.0019588919822126627,
+      "learning_rate": 9.99215216333285e-05,
+      "loss": 0.0007,
+      "step": 897
+    },
+    {
+      "epoch": 2.9026262626262627,
+      "grad_norm": 0.001903701457194984,
+      "learning_rate": 9.992025108721454e-05,
+      "loss": 0.0004,
+      "step": 898
+    },
+    {
+      "epoch": 2.905858585858586,
+      "grad_norm": 0.0023024296388030052,
+      "learning_rate": 9.991897034681087e-05,
+      "loss": 0.0006,
+      "step": 899
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.002626319881528616,
+      "learning_rate": 9.9917679412379e-05,
+      "loss": 0.0009,
+      "step": 900
+    },
+    {
+      "epoch": 2.909090909090909,
+      "eval_loss": 0.0008496911614201963,
+      "eval_runtime": 18.6252,
+      "eval_samples_per_second": 5.369,
+      "eval_steps_per_second": 1.342,
+      "step": 900
+    },
+    {
+      "epoch": 2.9123232323232324,
+      "grad_norm": 0.0029552706982940435,
+      "learning_rate": 9.99163782841826e-05,
+      "loss": 0.001,
+      "step": 901
+    },
+    {
+      "epoch": 2.9155555555555557,
+      "grad_norm": 0.002366641303524375,
+      "learning_rate": 9.991506696248731e-05,
+      "loss": 0.0006,
+      "step": 902
+    },
+    {
+      "epoch": 2.918787878787879,
+      "grad_norm": 0.0030002701096236706,
+      "learning_rate": 9.991374544756098e-05,
+      "loss": 0.0009,
+      "step": 903
+    },
+    {
+      "epoch": 2.922020202020202,
+      "grad_norm": 0.002418822841718793,
+      "learning_rate": 9.991241373967344e-05,
+      "loss": 0.0005,
+      "step": 904
+    },
+    {
+      "epoch": 2.9252525252525254,
+      "grad_norm": 0.0027566729113459587,
+      "learning_rate": 9.991107183909664e-05,
+      "loss": 0.002,
+      "step": 905
+    },
+    {
+      "epoch": 2.9284848484848487,
+      "grad_norm": 0.003200074890628457,
+      "learning_rate": 9.990971974610466e-05,
+      "loss": 0.0009,
+      "step": 906
+    },
+    {
+      "epoch": 2.931717171717172,
+      "grad_norm": 0.00368666322901845,
+      "learning_rate": 9.990835746097356e-05,
+      "loss": 0.0008,
+      "step": 907
+    },
+    {
+      "epoch": 2.934949494949495,
+      "grad_norm": 0.0036671683192253113,
+      "learning_rate": 9.990698498398155e-05,
+      "loss": 0.001,
+      "step": 908
+    },
+    {
+      "epoch": 2.9381818181818184,
+      "grad_norm": 0.0037938845343887806,
+      "learning_rate": 9.990560231540889e-05,
+      "loss": 0.0007,
+      "step": 909
+    },
+    {
+      "epoch": 2.9414141414141413,
+      "grad_norm": 0.0027589588426053524,
+      "learning_rate": 9.990420945553797e-05,
+      "loss": 0.0006,
+      "step": 910
+    },
+    {
+      "epoch": 2.9446464646464645,
+      "grad_norm": 0.003152150195091963,
+      "learning_rate": 9.990280640465321e-05,
+      "loss": 0.0008,
+      "step": 911
+    },
+    {
+      "epoch": 2.9478787878787878,
+      "grad_norm": 0.002495568012818694,
+      "learning_rate": 9.990139316304112e-05,
+      "loss": 0.0007,
+      "step": 912
+    },
+    {
+      "epoch": 2.951111111111111,
+      "grad_norm": 0.0025980276986956596,
+      "learning_rate": 9.989996973099032e-05,
+      "loss": 0.0008,
+      "step": 913
+    },
+    {
+      "epoch": 2.9543434343434343,
+      "grad_norm": 0.0013909428380429745,
+      "learning_rate": 9.989853610879147e-05,
+      "loss": 0.0004,
+      "step": 914
+    },
+    {
+      "epoch": 2.9575757575757575,
+      "grad_norm": 0.0025712710339576006,
+      "learning_rate": 9.989709229673736e-05,
+      "loss": 0.0008,
+      "step": 915
+    },
+    {
+      "epoch": 2.9608080808080808,
+      "grad_norm": 0.0029310788959264755,
+      "learning_rate": 9.98956382951228e-05,
+      "loss": 0.0007,
+      "step": 916
+    },
+    {
+      "epoch": 2.964040404040404,
+      "grad_norm": 0.0041606370359659195,
+      "learning_rate": 9.989417410424475e-05,
+      "loss": 0.0027,
+      "step": 917
+    },
+    {
+      "epoch": 2.9672727272727273,
+      "grad_norm": 0.001812944421544671,
+      "learning_rate": 9.98926997244022e-05,
+      "loss": 0.0006,
+      "step": 918
+    },
+    {
+      "epoch": 2.9705050505050505,
+      "grad_norm": 0.0026405269745737314,
+      "learning_rate": 9.989121515589622e-05,
+      "loss": 0.0011,
+      "step": 919
+    },
+    {
+      "epoch": 2.973737373737374,
+      "grad_norm": 0.003111011115834117,
+      "learning_rate": 9.988972039902997e-05,
+      "loss": 0.0009,
+      "step": 920
+    },
+    {
+      "epoch": 2.976969696969697,
+      "grad_norm": 0.0038792812265455723,
+      "learning_rate": 9.988821545410874e-05,
+      "loss": 0.0018,
+      "step": 921
+    },
+    {
+      "epoch": 2.9802020202020203,
+      "grad_norm": 0.0031774002127349377,
+      "learning_rate": 9.988670032143981e-05,
+      "loss": 0.0011,
+      "step": 922
+    },
+    {
+      "epoch": 2.9834343434343435,
+      "grad_norm": 0.002835791325196624,
+      "learning_rate": 9.988517500133262e-05,
+      "loss": 0.0007,
+      "step": 923
+    },
+    {
+      "epoch": 2.986666666666667,
+      "grad_norm": 0.003195343306288123,
+      "learning_rate": 9.988363949409865e-05,
+      "loss": 0.0008,
+      "step": 924
+    },
+    {
+      "epoch": 2.98989898989899,
+      "grad_norm": 0.004220789764076471,
+      "learning_rate": 9.988209380005144e-05,
+      "loss": 0.001,
+      "step": 925
+    },
+    {
+      "epoch": 2.98989898989899,
+      "eval_loss": 0.000867326685693115,
+      "eval_runtime": 18.6533,
+      "eval_samples_per_second": 5.361,
+      "eval_steps_per_second": 1.34,
+      "step": 925
+    },
+    {
+      "epoch": 2.993131313131313,
+      "grad_norm": 0.0027963484171777964,
+      "learning_rate": 9.98805379195067e-05,
+      "loss": 0.001,
+      "step": 926
+    },
+    {
+      "epoch": 2.996363636363636,
+      "grad_norm": 0.002367103472352028,
+      "learning_rate": 9.987897185278208e-05,
+      "loss": 0.0006,
+      "step": 927
+    },
+    {
+      "epoch": 2.9995959595959594,
+      "grad_norm": 0.008955973200500011,
+      "learning_rate": 9.987739560019746e-05,
+      "loss": 0.0042,
+      "step": 928
+    },
+    {
+      "epoch": 3.0028282828282826,
+      "grad_norm": 0.0027045756578445435,
+      "learning_rate": 9.987580916207468e-05,
+      "loss": 0.0019,
+      "step": 929
+    },
+    {
+      "epoch": 3.006060606060606,
+      "grad_norm": 0.002136245835572481,
+      "learning_rate": 9.987421253873775e-05,
+      "loss": 0.0013,
+      "step": 930
+    },
+    {
+      "epoch": 3.009292929292929,
+      "grad_norm": 0.0018668243428692222,
+      "learning_rate": 9.987260573051269e-05,
+      "loss": 0.0006,
+      "step": 931
+    },
+    {
+      "epoch": 3.0125252525252524,
+      "grad_norm": 0.0018371654441580176,
+      "learning_rate": 9.987098873772763e-05,
+      "loss": 0.0004,
+      "step": 932
+    },
+    {
+      "epoch": 3.0157575757575756,
+      "grad_norm": 0.0020080492831766605,
+      "learning_rate": 9.986936156071278e-05,
+      "loss": 0.0007,
+      "step": 933
+    },
+    {
+      "epoch": 3.018989898989899,
+      "grad_norm": 0.0019176624482497573,
+      "learning_rate": 9.986772419980044e-05,
+      "loss": 0.0006,
+      "step": 934
+    },
+    {
+      "epoch": 3.022222222222222,
+      "grad_norm": 0.0024317919742316008,
+      "learning_rate": 9.986607665532497e-05,
+      "loss": 0.0006,
+      "step": 935
+    },
+    {
+      "epoch": 3.0254545454545454,
+      "grad_norm": 0.002591727999970317,
+      "learning_rate": 9.986441892762281e-05,
+      "loss": 0.0006,
+      "step": 936
+    },
+    {
+      "epoch": 3.0286868686868686,
+      "grad_norm": 0.0026875538751482964,
+      "learning_rate": 9.98627510170325e-05,
+      "loss": 0.0008,
+      "step": 937
+    },
+    {
+      "epoch": 3.031919191919192,
+      "grad_norm": 0.003272917354479432,
+      "learning_rate": 9.986107292389464e-05,
+      "loss": 0.0007,
+      "step": 938
+    },
+    {
+      "epoch": 3.035151515151515,
+      "grad_norm": 0.002936862874776125,
+      "learning_rate": 9.985938464855191e-05,
+      "loss": 0.0008,
+      "step": 939
+    },
+    {
+      "epoch": 3.0383838383838384,
+      "grad_norm": 0.0021727578714489937,
+      "learning_rate": 9.985768619134909e-05,
+      "loss": 0.0007,
+      "step": 940
+    },
+    {
+      "epoch": 3.0416161616161617,
+      "grad_norm": 0.002251671627163887,
+      "learning_rate": 9.985597755263302e-05,
+      "loss": 0.0006,
+      "step": 941
+    },
+    {
+      "epoch": 3.044848484848485,
+      "grad_norm": 0.0035938192158937454,
+      "learning_rate": 9.985425873275263e-05,
+      "loss": 0.0011,
+      "step": 942
+    },
+    {
+      "epoch": 3.048080808080808,
+      "grad_norm": 0.0024208389222621918,
+      "learning_rate": 9.98525297320589e-05,
+      "loss": 0.0007,
+      "step": 943
+    },
+    {
+      "epoch": 3.0513131313131314,
+      "grad_norm": 0.0019874447025358677,
+      "learning_rate": 9.985079055090493e-05,
+      "loss": 0.0005,
+      "step": 944
+    },
+    {
+      "epoch": 3.0545454545454547,
+      "grad_norm": 0.0014314191648736596,
+      "learning_rate": 9.984904118964588e-05,
+      "loss": 0.0004,
+      "step": 945
+    },
+    {
+      "epoch": 3.057777777777778,
+      "grad_norm": 0.0019143620738759637,
+      "learning_rate": 9.984728164863898e-05,
+      "loss": 0.0006,
+      "step": 946
+    },
+    {
+      "epoch": 3.061010101010101,
+      "grad_norm": 0.003406926291063428,
+      "learning_rate": 9.984551192824355e-05,
+      "loss": 0.0011,
+      "step": 947
+    },
+    {
+      "epoch": 3.0642424242424244,
+      "grad_norm": 0.0015149825485423207,
+      "learning_rate": 9.9843732028821e-05,
+      "loss": 0.0004,
+      "step": 948
+    },
+    {
+      "epoch": 3.0674747474747477,
+      "grad_norm": 0.0028568189591169357,
+      "learning_rate": 9.98419419507348e-05,
+      "loss": 0.0008,
+      "step": 949
+    },
+    {
+      "epoch": 3.0707070707070705,
+      "grad_norm": 0.00383683480322361,
+      "learning_rate": 9.98401416943505e-05,
+      "loss": 0.0006,
+      "step": 950
+    },
+    {
+      "epoch": 3.0707070707070705,
+      "eval_loss": 0.0009233049931935966,
+      "eval_runtime": 18.6291,
+      "eval_samples_per_second": 5.368,
+      "eval_steps_per_second": 1.342,
+      "step": 950
+    },
+    {
+      "epoch": 3.0739393939393937,
+      "grad_norm": 0.004385147709399462,
+      "learning_rate": 9.983833126003572e-05,
+      "loss": 0.0009,
+      "step": 951
+    },
+    {
+      "epoch": 3.077171717171717,
+      "grad_norm": 0.0024896373506635427,
+      "learning_rate": 9.98365106481602e-05,
+      "loss": 0.0006,
+      "step": 952
+    },
+    {
+      "epoch": 3.0804040404040403,
+      "grad_norm": 0.002700845478102565,
+      "learning_rate": 9.983467985909573e-05,
+      "loss": 0.0006,
+      "step": 953
+    },
+    {
+      "epoch": 3.0836363636363635,
+      "grad_norm": 0.003025912679731846,
+      "learning_rate": 9.983283889321615e-05,
+      "loss": 0.0008,
+      "step": 954
+    },
+    {
+      "epoch": 3.0868686868686868,
+      "grad_norm": 0.003908303566277027,
+      "learning_rate": 9.983098775089742e-05,
+      "loss": 0.0017,
+      "step": 955
+    },
+    {
+      "epoch": 3.09010101010101,
+      "grad_norm": 0.0035844456870108843,
+      "learning_rate": 9.982912643251757e-05,
+      "loss": 0.0024,
+      "step": 956
+    },
+    {
+      "epoch": 3.0933333333333333,
+      "grad_norm": 0.0022256888914853334,
+      "learning_rate": 9.98272549384567e-05,
+      "loss": 0.0005,
+      "step": 957
+    },
+    {
+      "epoch": 3.0965656565656565,
+      "grad_norm": 0.0024845930747687817,
+      "learning_rate": 9.982537326909697e-05,
+      "loss": 0.0007,
+      "step": 958
+    },
+    {
+      "epoch": 3.0997979797979798,
+      "grad_norm": 0.0021472277585417032,
+      "learning_rate": 9.982348142482269e-05,
+      "loss": 0.0006,
+      "step": 959
+    },
+    {
+      "epoch": 3.103030303030303,
+      "grad_norm": 0.0027109517250210047,
+      "learning_rate": 9.982157940602014e-05,
+      "loss": 0.0007,
+      "step": 960
+    },
+    {
+      "epoch": 3.1062626262626263,
+      "grad_norm": 0.0016069613629952073,
+      "learning_rate": 9.981966721307778e-05,
+      "loss": 0.0005,
+      "step": 961
+    },
+    {
+      "epoch": 3.1094949494949495,
+      "grad_norm": 0.002237373497337103,
+      "learning_rate": 9.981774484638606e-05,
+      "loss": 0.0005,
+      "step": 962
+    },
+    {
+      "epoch": 3.112727272727273,
+      "grad_norm": 0.002959214383736253,
+      "learning_rate": 9.981581230633758e-05,
+      "loss": 0.001,
+      "step": 963
+    },
+    {
+      "epoch": 3.115959595959596,
+      "grad_norm": 0.0031158181373029947,
+      "learning_rate": 9.981386959332697e-05,
+      "loss": 0.0008,
+      "step": 964
+    },
+    {
+      "epoch": 3.1191919191919193,
+      "grad_norm": 0.007239287253469229,
+      "learning_rate": 9.981191670775097e-05,
+      "loss": 0.0037,
+      "step": 965
+    },
+    {
+      "epoch": 3.1224242424242425,
+      "grad_norm": 0.0029164832085371017,
+      "learning_rate": 9.980995365000836e-05,
+      "loss": 0.0006,
+      "step": 966
+    },
+    {
+      "epoch": 3.125656565656566,
+      "grad_norm": 0.0018486580811440945,
+      "learning_rate": 9.980798042050004e-05,
+      "loss": 0.0004,
+      "step": 967
+    },
+    {
+      "epoch": 3.128888888888889,
+      "grad_norm": 0.0032664432656019926,
+      "learning_rate": 9.980599701962896e-05,
+      "loss": 0.0009,
+      "step": 968
+    },
+    {
+      "epoch": 3.1321212121212123,
+      "grad_norm": 0.0037746732123196125,
+      "learning_rate": 9.980400344780015e-05,
+      "loss": 0.0015,
+      "step": 969
+    },
+    {
+      "epoch": 3.1353535353535356,
+      "grad_norm": 0.002780639799311757,
+      "learning_rate": 9.98019997054207e-05,
+      "loss": 0.0006,
+      "step": 970
+    },
+    {
+      "epoch": 3.1385858585858584,
+      "grad_norm": 0.0026836313772946596,
+      "learning_rate": 9.979998579289984e-05,
+      "loss": 0.0006,
+      "step": 971
+    },
+    {
+      "epoch": 3.1418181818181816,
+      "grad_norm": 0.002338019199669361,
+      "learning_rate": 9.979796171064881e-05,
+      "loss": 0.0007,
+      "step": 972
+    },
+    {
+      "epoch": 3.145050505050505,
+      "grad_norm": 0.0024902368895709515,
+      "learning_rate": 9.979592745908095e-05,
+      "loss": 0.0011,
+      "step": 973
+    },
+    {
+      "epoch": 3.148282828282828,
+      "grad_norm": 0.0033587226644158363,
+      "learning_rate": 9.979388303861169e-05,
+      "loss": 0.0006,
+      "step": 974
+    },
+    {
+      "epoch": 3.1515151515151514,
+      "grad_norm": 0.002750332234427333,
+      "learning_rate": 9.97918284496585e-05,
+      "loss": 0.0006,
+      "step": 975
+    },
+    {
+      "epoch": 3.1515151515151514,
+      "eval_loss": 0.0009188714902848005,
+      "eval_runtime": 18.6436,
+      "eval_samples_per_second": 5.364,
+      "eval_steps_per_second": 1.341,
+      "step": 975
+    },
+    {
+      "epoch": 3.1547474747474746,
+      "grad_norm": 0.004522048868238926,
+      "learning_rate": 9.978976369264098e-05,
+      "loss": 0.0009,
+      "step": 976
+    },
+    {
+      "epoch": 3.157979797979798,
+      "grad_norm": 0.0020558724645525217,
+      "learning_rate": 9.978768876798075e-05,
+      "loss": 0.0004,
+      "step": 977
+    },
+    {
+      "epoch": 3.161212121212121,
+      "grad_norm": 0.0017784403171390295,
+      "learning_rate": 9.978560367610156e-05,
+      "loss": 0.0004,
+      "step": 978
+    },
+    {
+      "epoch": 3.1644444444444444,
+      "grad_norm": 0.0039355335757136345,
+      "learning_rate": 9.978350841742919e-05,
+      "loss": 0.0017,
+      "step": 979
+    },
+    {
+      "epoch": 3.1676767676767676,
+      "grad_norm": 0.0031672257464379072,
+      "learning_rate": 9.978140299239152e-05,
+      "loss": 0.0017,
+      "step": 980
+    },
+    {
+      "epoch": 3.170909090909091,
+      "grad_norm": 0.0048868078738451,
+      "learning_rate": 9.977928740141851e-05,
+      "loss": 0.001,
+      "step": 981
+    },
+    {
+      "epoch": 3.174141414141414,
+      "grad_norm": 0.002355673350393772,
+      "learning_rate": 9.977716164494217e-05,
+      "loss": 0.0005,
+      "step": 982
+    },
+    {
+      "epoch": 3.1773737373737374,
+      "grad_norm": 0.0035993049386888742,
+      "learning_rate": 9.977502572339664e-05,
+      "loss": 0.001,
+      "step": 983
+    },
+    {
+      "epoch": 3.1806060606060607,
+      "grad_norm": 0.00291255209594965,
+      "learning_rate": 9.977287963721804e-05,
+      "loss": 0.0007,
+      "step": 984
+    },
+    {
+      "epoch": 3.183838383838384,
+      "grad_norm": 0.002027621492743492,
+      "learning_rate": 9.977072338684469e-05,
+      "loss": 0.0006,
+      "step": 985
+    },
+    {
+      "epoch": 3.187070707070707,
+      "grad_norm": 0.0021479360293596983,
+      "learning_rate": 9.976855697271689e-05,
+      "loss": 0.0006,
+      "step": 986
+    },
+    {
+      "epoch": 3.1903030303030304,
+      "grad_norm": 0.002248652745038271,
+      "learning_rate": 9.976638039527704e-05,
+      "loss": 0.0005,
+      "step": 987
+    },
+    {
+      "epoch": 3.1935353535353537,
+      "grad_norm": 0.002659781137481332,
+      "learning_rate": 9.976419365496963e-05,
+      "loss": 0.0008,
+      "step": 988
+    },
+    {
+      "epoch": 3.196767676767677,
+      "grad_norm": 0.0037245461717247963,
+      "learning_rate": 9.976199675224123e-05,
+      "loss": 0.001,
+      "step": 989
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.004409831948578358,
+      "learning_rate": 9.975978968754045e-05,
+      "loss": 0.0022,
+      "step": 990
+    },
+    {
+      "epoch": 3.2032323232323234,
+      "grad_norm": 0.0030354245100170374,
+      "learning_rate": 9.975757246131803e-05,
+      "loss": 0.0006,
+      "step": 991
+    },
+    {
+      "epoch": 3.2064646464646467,
+      "grad_norm": 0.0026450392324477434,
+      "learning_rate": 9.975534507402671e-05,
+      "loss": 0.0008,
+      "step": 992
+    },
+    {
+      "epoch": 3.2096969696969695,
+      "grad_norm": 0.003626378020271659,
+      "learning_rate": 9.975310752612137e-05,
+      "loss": 0.0004,
+      "step": 993
+    },
+    {
+      "epoch": 3.2129292929292927,
+      "grad_norm": 0.0026495913043618202,
+      "learning_rate": 9.975085981805897e-05,
+      "loss": 0.0018,
+      "step": 994
+    },
+    {
+      "epoch": 3.216161616161616,
+      "grad_norm": 0.0032507823780179024,
+      "learning_rate": 9.974860195029847e-05,
+      "loss": 0.0012,
+      "step": 995
+    },
+    {
+      "epoch": 3.2193939393939393,
+      "grad_norm": 0.002977850381284952,
+      "learning_rate": 9.974633392330097e-05,
+      "loss": 0.0006,
+      "step": 996
+    },
+    {
+      "epoch": 3.2226262626262625,
+      "grad_norm": 0.002434935886412859,
+      "learning_rate": 9.974405573752965e-05,
+      "loss": 0.0006,
+      "step": 997
+    },
+    {
+      "epoch": 3.2258585858585858,
+      "grad_norm": 0.002726243808865547,
+      "learning_rate": 9.974176739344971e-05,
+      "loss": 0.0007,
+      "step": 998
+    },
+    {
+      "epoch": 3.229090909090909,
+      "grad_norm": 0.003465299028903246,
+      "learning_rate": 9.973946889152847e-05,
+      "loss": 0.0006,
+      "step": 999
+    },
+    {
+      "epoch": 3.2323232323232323,
+      "grad_norm": 0.002784241922199726,
+      "learning_rate": 9.973716023223531e-05,
+      "loss": 0.0007,
+      "step": 1000
+    },
+    {
+      "epoch": 3.2323232323232323,
+      "eval_loss": 0.0009049187647178769,
+      "eval_runtime": 18.6293,
+      "eval_samples_per_second": 5.368,
+      "eval_steps_per_second": 1.342,
+      "step": 1000
+    },
+    {
+      "epoch": 3.2355555555555555,
+      "grad_norm": 0.0021909528877586126,
+      "learning_rate": 9.97348414160417e-05,
+      "loss": 0.0005,
+      "step": 1001
+    },
+    {
+      "epoch": 3.2387878787878788,
+      "grad_norm": 0.003046189434826374,
+      "learning_rate": 9.973251244342114e-05,
+      "loss": 0.0009,
+      "step": 1002
+    },
+    {
+      "epoch": 3.242020202020202,
+      "grad_norm": 0.004089404363185167,
+      "learning_rate": 9.973017331484926e-05,
+      "loss": 0.0019,
+      "step": 1003
+    },
+    {
+      "epoch": 3.2452525252525253,
+      "grad_norm": 0.002897520549595356,
+      "learning_rate": 9.972782403080372e-05,
+      "loss": 0.0009,
+      "step": 1004
+    },
+    {
+      "epoch": 3.2484848484848485,
+      "grad_norm": 0.0027832165360450745,
+      "learning_rate": 9.972546459176425e-05,
+      "loss": 0.0006,
+      "step": 1005
+    },
+    {
+      "epoch": 3.251717171717172,
+      "grad_norm": 0.0021642649080604315,
+      "learning_rate": 9.972309499821273e-05,
+      "loss": 0.0005,
+      "step": 1006
+    },
+    {
+      "epoch": 3.254949494949495,
+      "grad_norm": 0.0030184059869498014,
+      "learning_rate": 9.972071525063303e-05,
+      "loss": 0.0019,
+      "step": 1007
+    },
+    {
+      "epoch": 3.2581818181818183,
+      "grad_norm": 0.0022592165041714907,
+      "learning_rate": 9.971832534951108e-05,
+      "loss": 0.0005,
+      "step": 1008
+    },
+    {
+      "epoch": 3.2614141414141415,
+      "grad_norm": 0.0013180490350350738,
+      "learning_rate": 9.9715925295335e-05,
+      "loss": 0.0004,
+      "step": 1009
+    },
+    {
+      "epoch": 3.264646464646465,
+      "grad_norm": 0.00403400557115674,
+      "learning_rate": 9.971351508859488e-05,
+      "loss": 0.0007,
+      "step": 1010
+    },
+    {
+      "epoch": 3.267878787878788,
+      "grad_norm": 0.003028259379789233,
+      "learning_rate": 9.971109472978288e-05,
+      "loss": 0.0013,
+      "step": 1011
+    },
+    {
+      "epoch": 3.2711111111111113,
+      "grad_norm": 0.004238371271640062,
+      "learning_rate": 9.97086642193933e-05,
+      "loss": 0.0008,
+      "step": 1012
+    },
+    {
+      "epoch": 3.274343434343434,
+      "grad_norm": 0.003420765744522214,
+      "learning_rate": 9.970622355792247e-05,
+      "loss": 0.002,
+      "step": 1013
+    },
+    {
+      "epoch": 3.2775757575757574,
+      "grad_norm": 0.002874986035749316,
+      "learning_rate": 9.970377274586879e-05,
+      "loss": 0.0007,
+      "step": 1014
+    },
+    {
+      "epoch": 3.2808080808080806,
+      "grad_norm": 0.001124391914345324,
+      "learning_rate": 9.970131178373277e-05,
+      "loss": 0.0003,
+      "step": 1015
+    },
+    {
+      "epoch": 3.284040404040404,
+      "grad_norm": 0.0029891314916312695,
+      "learning_rate": 9.969884067201695e-05,
+      "loss": 0.0007,
+      "step": 1016
+    },
+    {
+      "epoch": 3.287272727272727,
+      "grad_norm": 0.0026401756331324577,
+      "learning_rate": 9.969635941122595e-05,
+      "loss": 0.0007,
+      "step": 1017
+    },
+    {
+      "epoch": 3.2905050505050504,
+      "grad_norm": 0.001897217589430511,
+      "learning_rate": 9.969386800186649e-05,
+      "loss": 0.0004,
+      "step": 1018
+    },
+    {
+      "epoch": 3.2937373737373736,
+      "grad_norm": 0.002726042177528143,
+      "learning_rate": 9.969136644444731e-05,
+      "loss": 0.0008,
+      "step": 1019
+    },
+    {
+      "epoch": 3.296969696969697,
+      "grad_norm": 0.0030468441545963287,
+      "learning_rate": 9.968885473947932e-05,
+      "loss": 0.0008,
+      "step": 1020
+    },
+    {
+      "epoch": 3.30020202020202,
+      "grad_norm": 0.0016252384521067142,
+      "learning_rate": 9.968633288747539e-05,
+      "loss": 0.0004,
+      "step": 1021
+    },
+    {
+      "epoch": 3.3034343434343434,
+      "grad_norm": 0.002642759820446372,
+      "learning_rate": 9.968380088895052e-05,
+      "loss": 0.0007,
+      "step": 1022
+    },
+    {
+      "epoch": 3.3066666666666666,
+      "grad_norm": 0.003615351626649499,
+      "learning_rate": 9.968125874442179e-05,
+      "loss": 0.0008,
+      "step": 1023
+    },
+    {
+      "epoch": 3.30989898989899,
+      "grad_norm": 0.0026971250772476196,
+      "learning_rate": 9.96787064544083e-05,
+      "loss": 0.0007,
+      "step": 1024
+    },
+    {
+      "epoch": 3.313131313131313,
+      "grad_norm": 0.001658923109062016,
+      "learning_rate": 9.96761440194313e-05,
+      "loss": 0.0004,
+      "step": 1025
+    },
+    {
+      "epoch": 3.313131313131313,
+      "eval_loss": 0.0008755004382692277,
+      "eval_runtime": 18.6451,
+      "eval_samples_per_second": 5.363,
+      "eval_steps_per_second": 1.341,
+      "step": 1025
+    },
+    {
+      "epoch": 3.313131313131313,
+      "step": 1025,
+      "total_flos": 2.5838521494131835e+18,
+      "train_loss": 0.001856716921518943,
+      "train_runtime": 20010.9872,
+      "train_samples_per_second": 12.368,
+      "train_steps_per_second": 0.386
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7725,
+  "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": 4
+      }
+    },
     "TrainerControl": {
       "args": {
         "should_epoch_stop": false,
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": true
+        "should_training_stop": false
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.2254685925518213e+18,
+  "total_flos": 2.5838521494131835e+18,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null