diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24444 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1743243,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008604652363439865,
+      "grad_norm": 3.2356650829315186,
+      "learning_rate": 4.99856589127276e-05,
+      "loss": 1.6234,
+      "step": 500
+    },
+    {
+      "epoch": 0.001720930472687973,
+      "grad_norm": 3.1607751846313477,
+      "learning_rate": 4.99713178254552e-05,
+      "loss": 1.4802,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0025813957090319592,
+      "grad_norm": 2.918936252593994,
+      "learning_rate": 4.99569767381828e-05,
+      "loss": 1.439,
+      "step": 1500
+    },
+    {
+      "epoch": 0.003441860945375946,
+      "grad_norm": 2.8961517810821533,
+      "learning_rate": 4.9942635650910404e-05,
+      "loss": 1.4051,
+      "step": 2000
+    },
+    {
+      "epoch": 0.004302326181719932,
+      "grad_norm": 2.8767917156219482,
+      "learning_rate": 4.992829456363801e-05,
+      "loss": 1.37,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0051627914180639185,
+      "grad_norm": 2.6426689624786377,
+      "learning_rate": 4.9913953476365605e-05,
+      "loss": 1.3463,
+      "step": 3000
+    },
+    {
+      "epoch": 0.006023256654407905,
+      "grad_norm": 2.5832014083862305,
+      "learning_rate": 4.98996123890932e-05,
+      "loss": 1.3317,
+      "step": 3500
+    },
+    {
+      "epoch": 0.006883721890751892,
+      "grad_norm": 2.686662435531616,
+      "learning_rate": 4.98852713018208e-05,
+      "loss": 1.3207,
+      "step": 4000
+    },
+    {
+      "epoch": 0.0077441871270958786,
+      "grad_norm": 2.7090184688568115,
+      "learning_rate": 4.98709302145484e-05,
+      "loss": 1.3116,
+      "step": 4500
+    },
+    {
+      "epoch": 0.008604652363439864,
+      "grad_norm": 2.712428569793701,
+      "learning_rate": 4.985658912727601e-05,
+      "loss": 1.2907,
+      "step": 5000
+    },
+    {
+      "epoch": 0.009465117599783851,
+      "grad_norm": 2.6741058826446533,
+      "learning_rate": 4.9842248040003604e-05,
+      "loss": 1.2853,
+      "step": 5500
+    },
+    {
+      "epoch": 0.010325582836127837,
+      "grad_norm": 2.6448793411254883,
+      "learning_rate": 4.982790695273121e-05,
+      "loss": 1.2688,
+      "step": 6000
+    },
+    {
+      "epoch": 0.011186048072471824,
+      "grad_norm": 2.575014352798462,
+      "learning_rate": 4.9813565865458805e-05,
+      "loss": 1.2704,
+      "step": 6500
+    },
+    {
+      "epoch": 0.01204651330881581,
+      "grad_norm": 2.5448801517486572,
+      "learning_rate": 4.97992247781864e-05,
+      "loss": 1.2578,
+      "step": 7000
+    },
+    {
+      "epoch": 0.012906978545159797,
+      "grad_norm": 2.6225686073303223,
+      "learning_rate": 4.9784883690914006e-05,
+      "loss": 1.2471,
+      "step": 7500
+    },
+    {
+      "epoch": 0.013767443781503784,
+      "grad_norm": 2.385652542114258,
+      "learning_rate": 4.977054260364161e-05,
+      "loss": 1.2391,
+      "step": 8000
+    },
+    {
+      "epoch": 0.01462790901784777,
+      "grad_norm": 2.431713819503784,
+      "learning_rate": 4.975620151636921e-05,
+      "loss": 1.2347,
+      "step": 8500
+    },
+    {
+      "epoch": 0.015488374254191757,
+      "grad_norm": 2.5336477756500244,
+      "learning_rate": 4.9741860429096804e-05,
+      "loss": 1.2276,
+      "step": 9000
+    },
+    {
+      "epoch": 0.016348839490535744,
+      "grad_norm": 2.459806203842163,
+      "learning_rate": 4.972751934182441e-05,
+      "loss": 1.2269,
+      "step": 9500
+    },
+    {
+      "epoch": 0.017209304726879728,
+      "grad_norm": 2.6384100914001465,
+      "learning_rate": 4.9713178254552005e-05,
+      "loss": 1.2196,
+      "step": 10000
+    },
+    {
+      "epoch": 0.018069769963223715,
+      "grad_norm": 2.530402898788452,
+      "learning_rate": 4.96988371672796e-05,
+      "loss": 1.2087,
+      "step": 10500
+    },
+    {
+      "epoch": 0.018930235199567703,
+      "grad_norm": 3.5155889987945557,
+      "learning_rate": 4.9684496080007206e-05,
+      "loss": 1.2019,
+      "step": 11000
+    },
+    {
+      "epoch": 0.01979070043591169,
+      "grad_norm": 2.5882487297058105,
+      "learning_rate": 4.967015499273481e-05,
+      "loss": 1.2044,
+      "step": 11500
+    },
+    {
+      "epoch": 0.020651165672255674,
+      "grad_norm": 2.349508047103882,
+      "learning_rate": 4.965581390546241e-05,
+      "loss": 1.1935,
+      "step": 12000
+    },
+    {
+      "epoch": 0.02151163090859966,
+      "grad_norm": 2.4116039276123047,
+      "learning_rate": 4.964147281819001e-05,
+      "loss": 1.1893,
+      "step": 12500
+    },
+    {
+      "epoch": 0.02237209614494365,
+      "grad_norm": 2.585287570953369,
+      "learning_rate": 4.962713173091761e-05,
+      "loss": 1.1874,
+      "step": 13000
+    },
+    {
+      "epoch": 0.023232561381287636,
+      "grad_norm": 2.419912338256836,
+      "learning_rate": 4.9612790643645205e-05,
+      "loss": 1.1769,
+      "step": 13500
+    },
+    {
+      "epoch": 0.02409302661763162,
+      "grad_norm": 2.4008758068084717,
+      "learning_rate": 4.959844955637281e-05,
+      "loss": 1.1772,
+      "step": 14000
+    },
+    {
+      "epoch": 0.024953491853975607,
+      "grad_norm": 2.352156400680542,
+      "learning_rate": 4.958410846910041e-05,
+      "loss": 1.1751,
+      "step": 14500
+    },
+    {
+      "epoch": 0.025813957090319594,
+      "grad_norm": 2.361285448074341,
+      "learning_rate": 4.956976738182801e-05,
+      "loss": 1.1698,
+      "step": 15000
+    },
+    {
+      "epoch": 0.02667442232666358,
+      "grad_norm": 2.274899959564209,
+      "learning_rate": 4.955542629455561e-05,
+      "loss": 1.1721,
+      "step": 15500
+    },
+    {
+      "epoch": 0.02753488756300757,
+      "grad_norm": 2.572073221206665,
+      "learning_rate": 4.954108520728321e-05,
+      "loss": 1.1679,
+      "step": 16000
+    },
+    {
+      "epoch": 0.028395352799351552,
+      "grad_norm": 2.4291021823883057,
+      "learning_rate": 4.952674412001081e-05,
+      "loss": 1.1646,
+      "step": 16500
+    },
+    {
+      "epoch": 0.02925581803569554,
+      "grad_norm": 2.2385716438293457,
+      "learning_rate": 4.9512403032738405e-05,
+      "loss": 1.1583,
+      "step": 17000
+    },
+    {
+      "epoch": 0.030116283272039527,
+      "grad_norm": 2.2407643795013428,
+      "learning_rate": 4.9498061945466015e-05,
+      "loss": 1.1575,
+      "step": 17500
+    },
+    {
+      "epoch": 0.030976748508383514,
+      "grad_norm": 2.469423532485962,
+      "learning_rate": 4.948372085819361e-05,
+      "loss": 1.1584,
+      "step": 18000
+    },
+    {
+      "epoch": 0.0318372137447275,
+      "grad_norm": 2.4228737354278564,
+      "learning_rate": 4.946937977092121e-05,
+      "loss": 1.1456,
+      "step": 18500
+    },
+    {
+      "epoch": 0.03269767898107149,
+      "grad_norm": 2.16351056098938,
+      "learning_rate": 4.9455038683648813e-05,
+      "loss": 1.1475,
+      "step": 19000
+    },
+    {
+      "epoch": 0.03355814421741547,
+      "grad_norm": 2.4227852821350098,
+      "learning_rate": 4.944069759637641e-05,
+      "loss": 1.1538,
+      "step": 19500
+    },
+    {
+      "epoch": 0.034418609453759456,
+      "grad_norm": 2.178745746612549,
+      "learning_rate": 4.942635650910401e-05,
+      "loss": 1.1488,
+      "step": 20000
+    },
+    {
+      "epoch": 0.03527907469010345,
+      "grad_norm": 2.2594761848449707,
+      "learning_rate": 4.941201542183161e-05,
+      "loss": 1.1475,
+      "step": 20500
+    },
+    {
+      "epoch": 0.03613953992644743,
+      "grad_norm": 2.50846791267395,
+      "learning_rate": 4.9397674334559215e-05,
+      "loss": 1.1377,
+      "step": 21000
+    },
+    {
+      "epoch": 0.037000005162791415,
+      "grad_norm": 2.3424227237701416,
+      "learning_rate": 4.938333324728681e-05,
+      "loss": 1.1357,
+      "step": 21500
+    },
+    {
+      "epoch": 0.037860470399135406,
+      "grad_norm": 2.2766966819763184,
+      "learning_rate": 4.936899216001441e-05,
+      "loss": 1.1285,
+      "step": 22000
+    },
+    {
+      "epoch": 0.03872093563547939,
+      "grad_norm": 2.2529232501983643,
+      "learning_rate": 4.9354651072742013e-05,
+      "loss": 1.1261,
+      "step": 22500
+    },
+    {
+      "epoch": 0.03958140087182338,
+      "grad_norm": 2.248143434524536,
+      "learning_rate": 4.934030998546961e-05,
+      "loss": 1.1252,
+      "step": 23000
+    },
+    {
+      "epoch": 0.040441866108167364,
+      "grad_norm": 2.3664751052856445,
+      "learning_rate": 4.9325968898197214e-05,
+      "loss": 1.1214,
+      "step": 23500
+    },
+    {
+      "epoch": 0.04130233134451135,
+      "grad_norm": 2.2858259677886963,
+      "learning_rate": 4.931162781092482e-05,
+      "loss": 1.1162,
+      "step": 24000
+    },
+    {
+      "epoch": 0.04216279658085534,
+      "grad_norm": 2.3754398822784424,
+      "learning_rate": 4.9297286723652415e-05,
+      "loss": 1.1183,
+      "step": 24500
+    },
+    {
+      "epoch": 0.04302326181719932,
+      "grad_norm": 2.450465202331543,
+      "learning_rate": 4.928294563638001e-05,
+      "loss": 1.1198,
+      "step": 25000
+    },
+    {
+      "epoch": 0.04388372705354331,
+      "grad_norm": 2.2866413593292236,
+      "learning_rate": 4.9268604549107616e-05,
+      "loss": 1.1151,
+      "step": 25500
+    },
+    {
+      "epoch": 0.0447441922898873,
+      "grad_norm": 2.2568089962005615,
+      "learning_rate": 4.925426346183521e-05,
+      "loss": 1.1113,
+      "step": 26000
+    },
+    {
+      "epoch": 0.04560465752623128,
+      "grad_norm": 2.668402910232544,
+      "learning_rate": 4.923992237456281e-05,
+      "loss": 1.1109,
+      "step": 26500
+    },
+    {
+      "epoch": 0.04646512276257527,
+      "grad_norm": 2.635807514190674,
+      "learning_rate": 4.9225581287290414e-05,
+      "loss": 1.1176,
+      "step": 27000
+    },
+    {
+      "epoch": 0.047325587998919255,
+      "grad_norm": 2.1348018646240234,
+      "learning_rate": 4.921124020001802e-05,
+      "loss": 1.1063,
+      "step": 27500
+    },
+    {
+      "epoch": 0.04818605323526324,
+      "grad_norm": 2.548943519592285,
+      "learning_rate": 4.9196899112745615e-05,
+      "loss": 1.1102,
+      "step": 28000
+    },
+    {
+      "epoch": 0.04904651847160723,
+      "grad_norm": 2.266880989074707,
+      "learning_rate": 4.918255802547321e-05,
+      "loss": 1.106,
+      "step": 28500
+    },
+    {
+      "epoch": 0.049906983707951214,
+      "grad_norm": 2.3181064128875732,
+      "learning_rate": 4.9168216938200816e-05,
+      "loss": 1.1094,
+      "step": 29000
+    },
+    {
+      "epoch": 0.050767448944295204,
+      "grad_norm": 2.3166842460632324,
+      "learning_rate": 4.915387585092841e-05,
+      "loss": 1.0996,
+      "step": 29500
+    },
+    {
+      "epoch": 0.05162791418063919,
+      "grad_norm": 2.133981704711914,
+      "learning_rate": 4.913953476365602e-05,
+      "loss": 1.1021,
+      "step": 30000
+    },
+    {
+      "epoch": 0.05248837941698317,
+      "grad_norm": 2.3361051082611084,
+      "learning_rate": 4.912519367638362e-05,
+      "loss": 1.0997,
+      "step": 30500
+    },
+    {
+      "epoch": 0.05334884465332716,
+      "grad_norm": 2.381316661834717,
+      "learning_rate": 4.911085258911122e-05,
+      "loss": 1.0956,
+      "step": 31000
+    },
+    {
+      "epoch": 0.054209309889671146,
+      "grad_norm": 2.2757225036621094,
+      "learning_rate": 4.9096511501838815e-05,
+      "loss": 1.0952,
+      "step": 31500
+    },
+    {
+      "epoch": 0.05506977512601514,
+      "grad_norm": 2.4369475841522217,
+      "learning_rate": 4.908217041456641e-05,
+      "loss": 1.0925,
+      "step": 32000
+    },
+    {
+      "epoch": 0.05593024036235912,
+      "grad_norm": 2.6689693927764893,
+      "learning_rate": 4.9067829327294016e-05,
+      "loss": 1.0907,
+      "step": 32500
+    },
+    {
+      "epoch": 0.056790705598703105,
+      "grad_norm": 2.2719929218292236,
+      "learning_rate": 4.905348824002162e-05,
+      "loss": 1.0888,
+      "step": 33000
+    },
+    {
+      "epoch": 0.057651170835047096,
+      "grad_norm": 2.323052406311035,
+      "learning_rate": 4.903914715274922e-05,
+      "loss": 1.0898,
+      "step": 33500
+    },
+    {
+      "epoch": 0.05851163607139108,
+      "grad_norm": 2.3382821083068848,
+      "learning_rate": 4.902480606547682e-05,
+      "loss": 1.0804,
+      "step": 34000
+    },
+    {
+      "epoch": 0.05937210130773506,
+      "grad_norm": 2.1152403354644775,
+      "learning_rate": 4.901046497820442e-05,
+      "loss": 1.0767,
+      "step": 34500
+    },
+    {
+      "epoch": 0.060232566544079054,
+      "grad_norm": 2.0752382278442383,
+      "learning_rate": 4.8996123890932015e-05,
+      "loss": 1.0838,
+      "step": 35000
+    },
+    {
+      "epoch": 0.06109303178042304,
+      "grad_norm": 2.2990305423736572,
+      "learning_rate": 4.898178280365962e-05,
+      "loss": 1.084,
+      "step": 35500
+    },
+    {
+      "epoch": 0.06195349701676703,
+      "grad_norm": 2.2469754219055176,
+      "learning_rate": 4.8967441716387216e-05,
+      "loss": 1.0819,
+      "step": 36000
+    },
+    {
+      "epoch": 0.062813962253111,
+      "grad_norm": 2.147425889968872,
+      "learning_rate": 4.895310062911482e-05,
+      "loss": 1.0785,
+      "step": 36500
+    },
+    {
+      "epoch": 0.063674427489455,
+      "grad_norm": 2.2332513332366943,
+      "learning_rate": 4.8938759541842424e-05,
+      "loss": 1.0752,
+      "step": 37000
+    },
+    {
+      "epoch": 0.06453489272579899,
+      "grad_norm": 2.3052799701690674,
+      "learning_rate": 4.892441845457002e-05,
+      "loss": 1.0749,
+      "step": 37500
+    },
+    {
+      "epoch": 0.06539535796214298,
+      "grad_norm": 2.1933348178863525,
+      "learning_rate": 4.891007736729762e-05,
+      "loss": 1.0737,
+      "step": 38000
+    },
+    {
+      "epoch": 0.06625582319848695,
+      "grad_norm": 2.2870287895202637,
+      "learning_rate": 4.8895736280025215e-05,
+      "loss": 1.0778,
+      "step": 38500
+    },
+    {
+      "epoch": 0.06711628843483095,
+      "grad_norm": 2.422950506210327,
+      "learning_rate": 4.888139519275282e-05,
+      "loss": 1.0725,
+      "step": 39000
+    },
+    {
+      "epoch": 0.06797675367117494,
+      "grad_norm": 2.274256706237793,
+      "learning_rate": 4.886705410548042e-05,
+      "loss": 1.0652,
+      "step": 39500
+    },
+    {
+      "epoch": 0.06883721890751891,
+      "grad_norm": 2.131338596343994,
+      "learning_rate": 4.885271301820802e-05,
+      "loss": 1.0693,
+      "step": 40000
+    },
+    {
+      "epoch": 0.0696976841438629,
+      "grad_norm": 2.249906539916992,
+      "learning_rate": 4.8838371930935624e-05,
+      "loss": 1.0642,
+      "step": 40500
+    },
+    {
+      "epoch": 0.0705581493802069,
+      "grad_norm": 2.419240951538086,
+      "learning_rate": 4.882403084366322e-05,
+      "loss": 1.0743,
+      "step": 41000
+    },
+    {
+      "epoch": 0.07141861461655087,
+      "grad_norm": 2.2712082862854004,
+      "learning_rate": 4.880968975639082e-05,
+      "loss": 1.0711,
+      "step": 41500
+    },
+    {
+      "epoch": 0.07227907985289486,
+      "grad_norm": 2.220015048980713,
+      "learning_rate": 4.879534866911842e-05,
+      "loss": 1.0625,
+      "step": 42000
+    },
+    {
+      "epoch": 0.07313954508923885,
+      "grad_norm": 2.2555344104766846,
+      "learning_rate": 4.8781007581846025e-05,
+      "loss": 1.0638,
+      "step": 42500
+    },
+    {
+      "epoch": 0.07400001032558283,
+      "grad_norm": 2.2342844009399414,
+      "learning_rate": 4.876666649457362e-05,
+      "loss": 1.0577,
+      "step": 43000
+    },
+    {
+      "epoch": 0.07486047556192682,
+      "grad_norm": 2.1202335357666016,
+      "learning_rate": 4.875232540730122e-05,
+      "loss": 1.06,
+      "step": 43500
+    },
+    {
+      "epoch": 0.07572094079827081,
+      "grad_norm": 2.153123140335083,
+      "learning_rate": 4.8737984320028823e-05,
+      "loss": 1.0555,
+      "step": 44000
+    },
+    {
+      "epoch": 0.0765814060346148,
+      "grad_norm": 2.4861624240875244,
+      "learning_rate": 4.872364323275642e-05,
+      "loss": 1.059,
+      "step": 44500
+    },
+    {
+      "epoch": 0.07744187127095878,
+      "grad_norm": 2.0985679626464844,
+      "learning_rate": 4.870930214548402e-05,
+      "loss": 1.0608,
+      "step": 45000
+    },
+    {
+      "epoch": 0.07830233650730277,
+      "grad_norm": 2.197465658187866,
+      "learning_rate": 4.869496105821162e-05,
+      "loss": 1.0531,
+      "step": 45500
+    },
+    {
+      "epoch": 0.07916280174364676,
+      "grad_norm": 2.1348724365234375,
+      "learning_rate": 4.8680619970939225e-05,
+      "loss": 1.0622,
+      "step": 46000
+    },
+    {
+      "epoch": 0.08002326697999074,
+      "grad_norm": 2.1105077266693115,
+      "learning_rate": 4.866627888366682e-05,
+      "loss": 1.0523,
+      "step": 46500
+    },
+    {
+      "epoch": 0.08088373221633473,
+      "grad_norm": 2.1300766468048096,
+      "learning_rate": 4.8651937796394426e-05,
+      "loss": 1.055,
+      "step": 47000
+    },
+    {
+      "epoch": 0.08174419745267872,
+      "grad_norm": 2.1636085510253906,
+      "learning_rate": 4.863759670912202e-05,
+      "loss": 1.0469,
+      "step": 47500
+    },
+    {
+      "epoch": 0.0826046626890227,
+      "grad_norm": 2.230454206466675,
+      "learning_rate": 4.862325562184962e-05,
+      "loss": 1.0504,
+      "step": 48000
+    },
+    {
+      "epoch": 0.08346512792536669,
+      "grad_norm": 2.13574481010437,
+      "learning_rate": 4.8608914534577224e-05,
+      "loss": 1.0484,
+      "step": 48500
+    },
+    {
+      "epoch": 0.08432559316171068,
+      "grad_norm": 2.431797981262207,
+      "learning_rate": 4.859457344730483e-05,
+      "loss": 1.0496,
+      "step": 49000
+    },
+    {
+      "epoch": 0.08518605839805465,
+      "grad_norm": 2.244641065597534,
+      "learning_rate": 4.8580232360032425e-05,
+      "loss": 1.0496,
+      "step": 49500
+    },
+    {
+      "epoch": 0.08604652363439864,
+      "grad_norm": 2.164771556854248,
+      "learning_rate": 4.856589127276002e-05,
+      "loss": 1.0421,
+      "step": 50000
+    },
+    {
+      "epoch": 0.08690698887074264,
+      "grad_norm": 2.0793097019195557,
+      "learning_rate": 4.8551550185487626e-05,
+      "loss": 1.0379,
+      "step": 50500
+    },
+    {
+      "epoch": 0.08776745410708663,
+      "grad_norm": 2.270684242248535,
+      "learning_rate": 4.853720909821522e-05,
+      "loss": 1.0533,
+      "step": 51000
+    },
+    {
+      "epoch": 0.0886279193434306,
+      "grad_norm": 2.2440924644470215,
+      "learning_rate": 4.852286801094282e-05,
+      "loss": 1.0355,
+      "step": 51500
+    },
+    {
+      "epoch": 0.0894883845797746,
+      "grad_norm": 2.041602611541748,
+      "learning_rate": 4.850852692367043e-05,
+      "loss": 1.0345,
+      "step": 52000
+    },
+    {
+      "epoch": 0.09034884981611858,
+      "grad_norm": 2.487091302871704,
+      "learning_rate": 4.849418583639803e-05,
+      "loss": 1.0497,
+      "step": 52500
+    },
+    {
+      "epoch": 0.09120931505246256,
+      "grad_norm": 2.3336119651794434,
+      "learning_rate": 4.8479844749125625e-05,
+      "loss": 1.0399,
+      "step": 53000
+    },
+    {
+      "epoch": 0.09206978028880655,
+      "grad_norm": 2.019739866256714,
+      "learning_rate": 4.846550366185323e-05,
+      "loss": 1.0378,
+      "step": 53500
+    },
+    {
+      "epoch": 0.09293024552515054,
+      "grad_norm": 2.2122652530670166,
+      "learning_rate": 4.8451162574580826e-05,
+      "loss": 1.0397,
+      "step": 54000
+    },
+    {
+      "epoch": 0.09379071076149452,
+      "grad_norm": 2.15346360206604,
+      "learning_rate": 4.843682148730842e-05,
+      "loss": 1.0378,
+      "step": 54500
+    },
+    {
+      "epoch": 0.09465117599783851,
+      "grad_norm": 2.288830518722534,
+      "learning_rate": 4.842248040003603e-05,
+      "loss": 1.0352,
+      "step": 55000
+    },
+    {
+      "epoch": 0.0955116412341825,
+      "grad_norm": 2.551696538925171,
+      "learning_rate": 4.840813931276363e-05,
+      "loss": 1.0329,
+      "step": 55500
+    },
+    {
+      "epoch": 0.09637210647052648,
+      "grad_norm": 2.044560432434082,
+      "learning_rate": 4.839379822549123e-05,
+      "loss": 1.0296,
+      "step": 56000
+    },
+    {
+      "epoch": 0.09723257170687047,
+      "grad_norm": 2.240878105163574,
+      "learning_rate": 4.8379457138218825e-05,
+      "loss": 1.0382,
+      "step": 56500
+    },
+    {
+      "epoch": 0.09809303694321446,
+      "grad_norm": 2.011512517929077,
+      "learning_rate": 4.836511605094643e-05,
+      "loss": 1.0315,
+      "step": 57000
+    },
+    {
+      "epoch": 0.09895350217955845,
+      "grad_norm": 2.134789228439331,
+      "learning_rate": 4.8350774963674026e-05,
+      "loss": 1.0358,
+      "step": 57500
+    },
+    {
+      "epoch": 0.09981396741590243,
+      "grad_norm": 2.1063222885131836,
+      "learning_rate": 4.833643387640163e-05,
+      "loss": 1.0284,
+      "step": 58000
+    },
+    {
+      "epoch": 0.10067443265224642,
+      "grad_norm": 2.1646718978881836,
+      "learning_rate": 4.8322092789129234e-05,
+      "loss": 1.0239,
+      "step": 58500
+    },
+    {
+      "epoch": 0.10153489788859041,
+      "grad_norm": 2.18721866607666,
+      "learning_rate": 4.830775170185683e-05,
+      "loss": 1.0311,
+      "step": 59000
+    },
+    {
+      "epoch": 0.10239536312493439,
+      "grad_norm": 2.0538229942321777,
+      "learning_rate": 4.829341061458443e-05,
+      "loss": 1.0265,
+      "step": 59500
+    },
+    {
+      "epoch": 0.10325582836127838,
+      "grad_norm": 2.197871208190918,
+      "learning_rate": 4.827906952731203e-05,
+      "loss": 1.0317,
+      "step": 60000
+    },
+    {
+      "epoch": 0.10411629359762237,
+      "grad_norm": 2.194244623184204,
+      "learning_rate": 4.826472844003963e-05,
+      "loss": 1.0327,
+      "step": 60500
+    },
+    {
+      "epoch": 0.10497675883396634,
+      "grad_norm": 2.290822744369507,
+      "learning_rate": 4.8250387352767226e-05,
+      "loss": 1.0228,
+      "step": 61000
+    },
+    {
+      "epoch": 0.10583722407031033,
+      "grad_norm": 2.1944305896759033,
+      "learning_rate": 4.823604626549483e-05,
+      "loss": 1.0208,
+      "step": 61500
+    },
+    {
+      "epoch": 0.10669768930665433,
+      "grad_norm": 2.5611112117767334,
+      "learning_rate": 4.8221705178222434e-05,
+      "loss": 1.0299,
+      "step": 62000
+    },
+    {
+      "epoch": 0.1075581545429983,
+      "grad_norm": 2.017531394958496,
+      "learning_rate": 4.820736409095003e-05,
+      "loss": 1.0193,
+      "step": 62500
+    },
+    {
+      "epoch": 0.10841861977934229,
+      "grad_norm": 2.0505359172821045,
+      "learning_rate": 4.819302300367763e-05,
+      "loss": 1.018,
+      "step": 63000
+    },
+    {
+      "epoch": 0.10927908501568628,
+      "grad_norm": 2.133496046066284,
+      "learning_rate": 4.817868191640523e-05,
+      "loss": 1.0268,
+      "step": 63500
+    },
+    {
+      "epoch": 0.11013955025203027,
+      "grad_norm": 2.1716065406799316,
+      "learning_rate": 4.816434082913283e-05,
+      "loss": 1.0231,
+      "step": 64000
+    },
+    {
+      "epoch": 0.11100001548837425,
+      "grad_norm": 2.344346523284912,
+      "learning_rate": 4.814999974186043e-05,
+      "loss": 1.0107,
+      "step": 64500
+    },
+    {
+      "epoch": 0.11186048072471824,
+      "grad_norm": 2.2010936737060547,
+      "learning_rate": 4.8135658654588036e-05,
+      "loss": 1.0268,
+      "step": 65000
+    },
+    {
+      "epoch": 0.11272094596106223,
+      "grad_norm": 2.142312526702881,
+      "learning_rate": 4.8121317567315634e-05,
+      "loss": 1.0243,
+      "step": 65500
+    },
+    {
+      "epoch": 0.11358141119740621,
+      "grad_norm": 2.0880520343780518,
+      "learning_rate": 4.810697648004323e-05,
+      "loss": 1.0151,
+      "step": 66000
+    },
+    {
+      "epoch": 0.1144418764337502,
+      "grad_norm": 1.9939734935760498,
+      "learning_rate": 4.8092635392770834e-05,
+      "loss": 1.0112,
+      "step": 66500
+    },
+    {
+      "epoch": 0.11530234167009419,
+      "grad_norm": 2.421113967895508,
+      "learning_rate": 4.807829430549843e-05,
+      "loss": 1.0165,
+      "step": 67000
+    },
+    {
+      "epoch": 0.11616280690643817,
+      "grad_norm": 2.137781858444214,
+      "learning_rate": 4.8063953218226035e-05,
+      "loss": 1.0141,
+      "step": 67500
+    },
+    {
+      "epoch": 0.11702327214278216,
+      "grad_norm": 2.0419673919677734,
+      "learning_rate": 4.804961213095363e-05,
+      "loss": 1.0135,
+      "step": 68000
+    },
+    {
+      "epoch": 0.11788373737912615,
+      "grad_norm": 2.1220955848693848,
+      "learning_rate": 4.8035271043681236e-05,
+      "loss": 1.0154,
+      "step": 68500
+    },
+    {
+      "epoch": 0.11874420261547013,
+      "grad_norm": 2.096822500228882,
+      "learning_rate": 4.8020929956408833e-05,
+      "loss": 1.013,
+      "step": 69000
+    },
+    {
+      "epoch": 0.11960466785181412,
+      "grad_norm": 2.2863707542419434,
+      "learning_rate": 4.800658886913643e-05,
+      "loss": 1.0105,
+      "step": 69500
+    },
+    {
+      "epoch": 0.12046513308815811,
+      "grad_norm": 2.1425676345825195,
+      "learning_rate": 4.7992247781864034e-05,
+      "loss": 1.0183,
+      "step": 70000
+    },
+    {
+      "epoch": 0.1213255983245021,
+      "grad_norm": 2.125267744064331,
+      "learning_rate": 4.797790669459163e-05,
+      "loss": 1.0169,
+      "step": 70500
+    },
+    {
+      "epoch": 0.12218606356084608,
+      "grad_norm": 2.2705278396606445,
+      "learning_rate": 4.7963565607319235e-05,
+      "loss": 1.0125,
+      "step": 71000
+    },
+    {
+      "epoch": 0.12304652879719007,
+      "grad_norm": 2.3827016353607178,
+      "learning_rate": 4.794922452004684e-05,
+      "loss": 1.0072,
+      "step": 71500
+    },
+    {
+      "epoch": 0.12390699403353406,
+      "grad_norm": 2.2930362224578857,
+      "learning_rate": 4.7934883432774436e-05,
+      "loss": 1.0013,
+      "step": 72000
+    },
+    {
+      "epoch": 0.12476745926987803,
+      "grad_norm": 2.1181716918945312,
+      "learning_rate": 4.792054234550203e-05,
+      "loss": 1.0106,
+      "step": 72500
+    },
+    {
+      "epoch": 0.125627924506222,
+      "grad_norm": 2.098687171936035,
+      "learning_rate": 4.790620125822963e-05,
+      "loss": 1.0028,
+      "step": 73000
+    },
+    {
+      "epoch": 0.12648838974256602,
+      "grad_norm": 2.1100914478302,
+      "learning_rate": 4.7891860170957234e-05,
+      "loss": 1.0112,
+      "step": 73500
+    },
+    {
+      "epoch": 0.12734885497891,
+      "grad_norm": 2.1170222759246826,
+      "learning_rate": 4.787751908368484e-05,
+      "loss": 1.0159,
+      "step": 74000
+    },
+    {
+      "epoch": 0.128209320215254,
+      "grad_norm": 2.1959116458892822,
+      "learning_rate": 4.7863177996412435e-05,
+      "loss": 1.0046,
+      "step": 74500
+    },
+    {
+      "epoch": 0.12906978545159797,
+      "grad_norm": 2.074676990509033,
+      "learning_rate": 4.784883690914004e-05,
+      "loss": 1.0086,
+      "step": 75000
+    },
+    {
+      "epoch": 0.12993025068794195,
+      "grad_norm": 2.09879207611084,
+      "learning_rate": 4.7834495821867636e-05,
+      "loss": 1.0059,
+      "step": 75500
+    },
+    {
+      "epoch": 0.13079071592428596,
+      "grad_norm": 2.1059296131134033,
+      "learning_rate": 4.782015473459523e-05,
+      "loss": 1.0072,
+      "step": 76000
+    },
+    {
+      "epoch": 0.13165118116062993,
+      "grad_norm": 2.043699026107788,
+      "learning_rate": 4.780581364732284e-05,
+      "loss": 1.0002,
+      "step": 76500
+    },
+    {
+      "epoch": 0.1325116463969739,
+      "grad_norm": 2.139575242996216,
+      "learning_rate": 4.779147256005044e-05,
+      "loss": 1.003,
+      "step": 77000
+    },
+    {
+      "epoch": 0.1333721116333179,
+      "grad_norm": 2.240222930908203,
+      "learning_rate": 4.777713147277804e-05,
+      "loss": 1.0075,
+      "step": 77500
+    },
+    {
+      "epoch": 0.1342325768696619,
+      "grad_norm": 1.817133903503418,
+      "learning_rate": 4.776279038550564e-05,
+      "loss": 0.9962,
+      "step": 78000
+    },
+    {
+      "epoch": 0.13509304210600587,
+      "grad_norm": 1.9295564889907837,
+      "learning_rate": 4.774844929823324e-05,
+      "loss": 0.9977,
+      "step": 78500
+    },
+    {
+      "epoch": 0.13595350734234987,
+      "grad_norm": 2.252047538757324,
+      "learning_rate": 4.7734108210960836e-05,
+      "loss": 1.0009,
+      "step": 79000
+    },
+    {
+      "epoch": 0.13681397257869385,
+      "grad_norm": 2.113853931427002,
+      "learning_rate": 4.771976712368843e-05,
+      "loss": 1.0032,
+      "step": 79500
+    },
+    {
+      "epoch": 0.13767443781503783,
+      "grad_norm": 2.0588555335998535,
+      "learning_rate": 4.770542603641604e-05,
+      "loss": 0.999,
+      "step": 80000
+    },
+    {
+      "epoch": 0.13853490305138183,
+      "grad_norm": 2.266942024230957,
+      "learning_rate": 4.769108494914364e-05,
+      "loss": 1.0058,
+      "step": 80500
+    },
+    {
+      "epoch": 0.1393953682877258,
+      "grad_norm": 2.1544504165649414,
+      "learning_rate": 4.767674386187124e-05,
+      "loss": 0.9998,
+      "step": 81000
+    },
+    {
+      "epoch": 0.14025583352406978,
+      "grad_norm": 2.166877508163452,
+      "learning_rate": 4.766240277459884e-05,
+      "loss": 1.0047,
+      "step": 81500
+    },
+    {
+      "epoch": 0.1411162987604138,
+      "grad_norm": 1.9637346267700195,
+      "learning_rate": 4.764806168732644e-05,
+      "loss": 0.9995,
+      "step": 82000
+    },
+    {
+      "epoch": 0.14197676399675777,
+      "grad_norm": 1.9987211227416992,
+      "learning_rate": 4.7633720600054036e-05,
+      "loss": 0.9922,
+      "step": 82500
+    },
+    {
+      "epoch": 0.14283722923310174,
+      "grad_norm": 2.114132881164551,
+      "learning_rate": 4.761937951278164e-05,
+      "loss": 0.9936,
+      "step": 83000
+    },
+    {
+      "epoch": 0.14369769446944575,
+      "grad_norm": 2.1807868480682373,
+      "learning_rate": 4.7605038425509244e-05,
+      "loss": 0.9946,
+      "step": 83500
+    },
+    {
+      "epoch": 0.14455815970578972,
+      "grad_norm": 2.150728940963745,
+      "learning_rate": 4.759069733823684e-05,
+      "loss": 0.9982,
+      "step": 84000
+    },
+    {
+      "epoch": 0.1454186249421337,
+      "grad_norm": 2.000176429748535,
+      "learning_rate": 4.757635625096444e-05,
+      "loss": 0.9922,
+      "step": 84500
+    },
+    {
+      "epoch": 0.1462790901784777,
+      "grad_norm": 2.2587642669677734,
+      "learning_rate": 4.756201516369204e-05,
+      "loss": 0.9921,
+      "step": 85000
+    },
+    {
+      "epoch": 0.14713955541482168,
+      "grad_norm": 2.2462809085845947,
+      "learning_rate": 4.754767407641964e-05,
+      "loss": 0.9948,
+      "step": 85500
+    },
+    {
+      "epoch": 0.14800002065116566,
+      "grad_norm": 2.114885091781616,
+      "learning_rate": 4.7533332989147236e-05,
+      "loss": 0.9977,
+      "step": 86000
+    },
+    {
+      "epoch": 0.14886048588750966,
+      "grad_norm": 2.3015403747558594,
+      "learning_rate": 4.751899190187484e-05,
+      "loss": 0.9924,
+      "step": 86500
+    },
+    {
+      "epoch": 0.14972095112385364,
+      "grad_norm": 2.2971363067626953,
+      "learning_rate": 4.7504650814602444e-05,
+      "loss": 0.988,
+      "step": 87000
+    },
+    {
+      "epoch": 0.15058141636019765,
+      "grad_norm": 2.223186731338501,
+      "learning_rate": 4.749030972733004e-05,
+      "loss": 0.9891,
+      "step": 87500
+    },
+    {
+      "epoch": 0.15144188159654162,
+      "grad_norm": 2.1935551166534424,
+      "learning_rate": 4.7475968640057645e-05,
+      "loss": 0.9885,
+      "step": 88000
+    },
+    {
+      "epoch": 0.1523023468328856,
+      "grad_norm": 2.0309863090515137,
+      "learning_rate": 4.746162755278524e-05,
+      "loss": 0.9989,
+      "step": 88500
+    },
+    {
+      "epoch": 0.1531628120692296,
+      "grad_norm": 2.113100051879883,
+      "learning_rate": 4.744728646551284e-05,
+      "loss": 0.9924,
+      "step": 89000
+    },
+    {
+      "epoch": 0.15402327730557358,
+      "grad_norm": 1.9869849681854248,
+      "learning_rate": 4.743294537824044e-05,
+      "loss": 0.991,
+      "step": 89500
+    },
+    {
+      "epoch": 0.15488374254191756,
+      "grad_norm": 2.1156396865844727,
+      "learning_rate": 4.7418604290968046e-05,
+      "loss": 0.9907,
+      "step": 90000
+    },
+    {
+      "epoch": 0.15574420777826156,
+      "grad_norm": 1.9725556373596191,
+      "learning_rate": 4.7404263203695644e-05,
+      "loss": 0.9938,
+      "step": 90500
+    },
+    {
+      "epoch": 0.15660467301460554,
+      "grad_norm": 2.0329513549804688,
+      "learning_rate": 4.738992211642324e-05,
+      "loss": 0.9935,
+      "step": 91000
+    },
+    {
+      "epoch": 0.15746513825094952,
+      "grad_norm": 2.1494927406311035,
+      "learning_rate": 4.7375581029150844e-05,
+      "loss": 0.9783,
+      "step": 91500
+    },
+    {
+      "epoch": 0.15832560348729352,
+      "grad_norm": 2.001607656478882,
+      "learning_rate": 4.736123994187844e-05,
+      "loss": 0.9936,
+      "step": 92000
+    },
+    {
+      "epoch": 0.1591860687236375,
+      "grad_norm": 1.998681902885437,
+      "learning_rate": 4.7346898854606045e-05,
+      "loss": 0.9894,
+      "step": 92500
+    },
+    {
+      "epoch": 0.16004653395998147,
+      "grad_norm": 2.114757537841797,
+      "learning_rate": 4.733255776733365e-05,
+      "loss": 0.9901,
+      "step": 93000
+    },
+    {
+      "epoch": 0.16090699919632548,
+      "grad_norm": 2.082709789276123,
+      "learning_rate": 4.7318216680061246e-05,
+      "loss": 0.9859,
+      "step": 93500
+    },
+    {
+      "epoch": 0.16176746443266946,
+      "grad_norm": 2.1626060009002686,
+      "learning_rate": 4.7303875592788843e-05,
+      "loss": 0.9848,
+      "step": 94000
+    },
+    {
+      "epoch": 0.16262792966901343,
+      "grad_norm": 2.2629404067993164,
+      "learning_rate": 4.728953450551645e-05,
+      "loss": 0.9872,
+      "step": 94500
+    },
+    {
+      "epoch": 0.16348839490535744,
+      "grad_norm": 2.1497435569763184,
+      "learning_rate": 4.7275193418244044e-05,
+      "loss": 0.984,
+      "step": 95000
+    },
+    {
+      "epoch": 0.16434886014170141,
+      "grad_norm": 2.212721586227417,
+      "learning_rate": 4.726085233097164e-05,
+      "loss": 0.9874,
+      "step": 95500
+    },
+    {
+      "epoch": 0.1652093253780454,
+      "grad_norm": 2.1093273162841797,
+      "learning_rate": 4.7246511243699245e-05,
+      "loss": 0.9752,
+      "step": 96000
+    },
+    {
+      "epoch": 0.1660697906143894,
+      "grad_norm": 2.026782274246216,
+      "learning_rate": 4.723217015642685e-05,
+      "loss": 0.9847,
+      "step": 96500
+    },
+    {
+      "epoch": 0.16693025585073337,
+      "grad_norm": 2.3008055686950684,
+      "learning_rate": 4.7217829069154446e-05,
+      "loss": 0.982,
+      "step": 97000
+    },
+    {
+      "epoch": 0.16779072108707735,
+      "grad_norm": 2.1564080715179443,
+      "learning_rate": 4.720348798188204e-05,
+      "loss": 0.9794,
+      "step": 97500
+    },
+    {
+      "epoch": 0.16865118632342135,
+      "grad_norm": 2.0874173641204834,
+      "learning_rate": 4.718914689460965e-05,
+      "loss": 0.9804,
+      "step": 98000
+    },
+    {
+      "epoch": 0.16951165155976533,
+      "grad_norm": 1.9421424865722656,
+      "learning_rate": 4.7174805807337244e-05,
+      "loss": 0.9862,
+      "step": 98500
+    },
+    {
+      "epoch": 0.1703721167961093,
+      "grad_norm": 2.108234167098999,
+      "learning_rate": 4.716046472006485e-05,
+      "loss": 0.9826,
+      "step": 99000
+    },
+    {
+      "epoch": 0.1712325820324533,
+      "grad_norm": 2.1940762996673584,
+      "learning_rate": 4.714612363279245e-05,
+      "loss": 0.9807,
+      "step": 99500
+    },
+    {
+      "epoch": 0.1720930472687973,
+      "grad_norm": 2.1281466484069824,
+      "learning_rate": 4.713178254552005e-05,
+      "loss": 0.9801,
+      "step": 100000
+    },
+    {
+      "epoch": 0.1729535125051413,
+      "grad_norm": 2.0339930057525635,
+      "learning_rate": 4.7117441458247646e-05,
+      "loss": 0.9703,
+      "step": 100500
+    },
+    {
+      "epoch": 0.17381397774148527,
+      "grad_norm": 2.073927402496338,
+      "learning_rate": 4.710310037097525e-05,
+      "loss": 0.9773,
+      "step": 101000
+    },
+    {
+      "epoch": 0.17467444297782925,
+      "grad_norm": 2.0153615474700928,
+      "learning_rate": 4.708875928370285e-05,
+      "loss": 0.9817,
+      "step": 101500
+    },
+    {
+      "epoch": 0.17553490821417325,
+      "grad_norm": 2.182114839553833,
+      "learning_rate": 4.707441819643045e-05,
+      "loss": 0.9702,
+      "step": 102000
+    },
+    {
+      "epoch": 0.17639537345051723,
+      "grad_norm": 2.2643656730651855,
+      "learning_rate": 4.706007710915805e-05,
+      "loss": 0.9803,
+      "step": 102500
+    },
+    {
+      "epoch": 0.1772558386868612,
+      "grad_norm": 2.1115529537200928,
+      "learning_rate": 4.704573602188565e-05,
+      "loss": 0.9782,
+      "step": 103000
+    },
+    {
+      "epoch": 0.1781163039232052,
+      "grad_norm": 2.089778184890747,
+      "learning_rate": 4.703139493461325e-05,
+      "loss": 0.974,
+      "step": 103500
+    },
+    {
+      "epoch": 0.1789767691595492,
+      "grad_norm": 2.1720387935638428,
+      "learning_rate": 4.7017053847340846e-05,
+      "loss": 0.9726,
+      "step": 104000
+    },
+    {
+      "epoch": 0.17983723439589316,
+      "grad_norm": 2.0718801021575928,
+      "learning_rate": 4.700271276006845e-05,
+      "loss": 0.9693,
+      "step": 104500
+    },
+    {
+      "epoch": 0.18069769963223717,
+      "grad_norm": 1.9460430145263672,
+      "learning_rate": 4.698837167279605e-05,
+      "loss": 0.9661,
+      "step": 105000
+    },
+    {
+      "epoch": 0.18155816486858115,
+      "grad_norm": 2.153810977935791,
+      "learning_rate": 4.697403058552365e-05,
+      "loss": 0.9734,
+      "step": 105500
+    },
+    {
+      "epoch": 0.18241863010492512,
+      "grad_norm": 2.104356050491333,
+      "learning_rate": 4.6959689498251255e-05,
+      "loss": 0.9718,
+      "step": 106000
+    },
+    {
+      "epoch": 0.18327909534126913,
+      "grad_norm": 2.3347885608673096,
+      "learning_rate": 4.694534841097885e-05,
+      "loss": 0.9745,
+      "step": 106500
+    },
+    {
+      "epoch": 0.1841395605776131,
+      "grad_norm": 2.0650954246520996,
+      "learning_rate": 4.693100732370645e-05,
+      "loss": 0.9731,
+      "step": 107000
+    },
+    {
+      "epoch": 0.18500002581395708,
+      "grad_norm": 2.060877561569214,
+      "learning_rate": 4.6916666236434046e-05,
+      "loss": 0.9762,
+      "step": 107500
+    },
+    {
+      "epoch": 0.18586049105030109,
+      "grad_norm": 2.2722277641296387,
+      "learning_rate": 4.690232514916165e-05,
+      "loss": 0.9679,
+      "step": 108000
+    },
+    {
+      "epoch": 0.18672095628664506,
+      "grad_norm": 2.2440106868743896,
+      "learning_rate": 4.6887984061889254e-05,
+      "loss": 0.978,
+      "step": 108500
+    },
+    {
+      "epoch": 0.18758142152298904,
+      "grad_norm": 1.9629226922988892,
+      "learning_rate": 4.687364297461685e-05,
+      "loss": 0.972,
+      "step": 109000
+    },
+    {
+      "epoch": 0.18844188675933304,
+      "grad_norm": 2.1158809661865234,
+      "learning_rate": 4.6859301887344455e-05,
+      "loss": 0.9697,
+      "step": 109500
+    },
+    {
+      "epoch": 0.18930235199567702,
+      "grad_norm": 2.001575231552124,
+      "learning_rate": 4.684496080007205e-05,
+      "loss": 0.9699,
+      "step": 110000
+    },
+    {
+      "epoch": 0.190162817232021,
+      "grad_norm": 2.1017117500305176,
+      "learning_rate": 4.683061971279965e-05,
+      "loss": 0.965,
+      "step": 110500
+    },
+    {
+      "epoch": 0.191023282468365,
+      "grad_norm": 2.198977470397949,
+      "learning_rate": 4.681627862552725e-05,
+      "loss": 0.9682,
+      "step": 111000
+    },
+    {
+      "epoch": 0.19188374770470898,
+      "grad_norm": 2.0062718391418457,
+      "learning_rate": 4.680193753825485e-05,
+      "loss": 0.9709,
+      "step": 111500
+    },
+    {
+      "epoch": 0.19274421294105296,
+      "grad_norm": 2.038653612136841,
+      "learning_rate": 4.6787596450982454e-05,
+      "loss": 0.9668,
+      "step": 112000
+    },
+    {
+      "epoch": 0.19360467817739696,
+      "grad_norm": 2.0191009044647217,
+      "learning_rate": 4.677325536371006e-05,
+      "loss": 0.9669,
+      "step": 112500
+    },
+    {
+      "epoch": 0.19446514341374094,
+      "grad_norm": 2.0882835388183594,
+      "learning_rate": 4.6758914276437655e-05,
+      "loss": 0.9727,
+      "step": 113000
+    },
+    {
+      "epoch": 0.19532560865008491,
+      "grad_norm": 2.0213959217071533,
+      "learning_rate": 4.674457318916525e-05,
+      "loss": 0.9689,
+      "step": 113500
+    },
+    {
+      "epoch": 0.19618607388642892,
+      "grad_norm": 2.1264288425445557,
+      "learning_rate": 4.673023210189285e-05,
+      "loss": 0.9778,
+      "step": 114000
+    },
+    {
+      "epoch": 0.1970465391227729,
+      "grad_norm": 2.0388376712799072,
+      "learning_rate": 4.671589101462045e-05,
+      "loss": 0.9625,
+      "step": 114500
+    },
+    {
+      "epoch": 0.1979070043591169,
+      "grad_norm": 2.0679802894592285,
+      "learning_rate": 4.6701549927348056e-05,
+      "loss": 0.9651,
+      "step": 115000
+    },
+    {
+      "epoch": 0.19876746959546088,
+      "grad_norm": 2.2121481895446777,
+      "learning_rate": 4.6687208840075653e-05,
+      "loss": 0.9602,
+      "step": 115500
+    },
+    {
+      "epoch": 0.19962793483180485,
+      "grad_norm": 1.9560799598693848,
+      "learning_rate": 4.667286775280326e-05,
+      "loss": 0.9625,
+      "step": 116000
+    },
+    {
+      "epoch": 0.20048840006814886,
+      "grad_norm": 2.0703752040863037,
+      "learning_rate": 4.6658526665530854e-05,
+      "loss": 0.965,
+      "step": 116500
+    },
+    {
+      "epoch": 0.20134886530449284,
+      "grad_norm": 2.160184621810913,
+      "learning_rate": 4.664418557825845e-05,
+      "loss": 0.9673,
+      "step": 117000
+    },
+    {
+      "epoch": 0.2022093305408368,
+      "grad_norm": 1.996235728263855,
+      "learning_rate": 4.6629844490986055e-05,
+      "loss": 0.9601,
+      "step": 117500
+    },
+    {
+      "epoch": 0.20306979577718082,
+      "grad_norm": 2.1551196575164795,
+      "learning_rate": 4.661550340371366e-05,
+      "loss": 0.9662,
+      "step": 118000
+    },
+    {
+      "epoch": 0.2039302610135248,
+      "grad_norm": 1.9875051975250244,
+      "learning_rate": 4.6601162316441256e-05,
+      "loss": 0.9648,
+      "step": 118500
+    },
+    {
+      "epoch": 0.20479072624986877,
+      "grad_norm": 2.0819616317749023,
+      "learning_rate": 4.658682122916885e-05,
+      "loss": 0.9612,
+      "step": 119000
+    },
+    {
+      "epoch": 0.20565119148621278,
+      "grad_norm": 2.1169705390930176,
+      "learning_rate": 4.657248014189646e-05,
+      "loss": 0.9575,
+      "step": 119500
+    },
+    {
+      "epoch": 0.20651165672255675,
+      "grad_norm": 2.142817258834839,
+      "learning_rate": 4.6558139054624054e-05,
+      "loss": 0.965,
+      "step": 120000
+    },
+    {
+      "epoch": 0.20737212195890073,
+      "grad_norm": 2.0226452350616455,
+      "learning_rate": 4.654379796735165e-05,
+      "loss": 0.9574,
+      "step": 120500
+    },
+    {
+      "epoch": 0.20823258719524473,
+      "grad_norm": 2.0718817710876465,
+      "learning_rate": 4.6529456880079255e-05,
+      "loss": 0.9624,
+      "step": 121000
+    },
+    {
+      "epoch": 0.2090930524315887,
+      "grad_norm": 2.1692280769348145,
+      "learning_rate": 4.651511579280686e-05,
+      "loss": 0.9608,
+      "step": 121500
+    },
+    {
+      "epoch": 0.2099535176679327,
+      "grad_norm": 2.2331693172454834,
+      "learning_rate": 4.6500774705534456e-05,
+      "loss": 0.9628,
+      "step": 122000
+    },
+    {
+      "epoch": 0.2108139829042767,
+      "grad_norm": 2.127037763595581,
+      "learning_rate": 4.648643361826206e-05,
+      "loss": 0.9657,
+      "step": 122500
+    },
+    {
+      "epoch": 0.21167444814062067,
+      "grad_norm": 2.1657791137695312,
+      "learning_rate": 4.647209253098966e-05,
+      "loss": 0.9615,
+      "step": 123000
+    },
+    {
+      "epoch": 0.21253491337696465,
+      "grad_norm": 2.10907244682312,
+      "learning_rate": 4.6457751443717254e-05,
+      "loss": 0.9575,
+      "step": 123500
+    },
+    {
+      "epoch": 0.21339537861330865,
+      "grad_norm": 2.026007652282715,
+      "learning_rate": 4.644341035644486e-05,
+      "loss": 0.9584,
+      "step": 124000
+    },
+    {
+      "epoch": 0.21425584384965263,
+      "grad_norm": 2.1519062519073486,
+      "learning_rate": 4.642906926917246e-05,
+      "loss": 0.9653,
+      "step": 124500
+    },
+    {
+      "epoch": 0.2151163090859966,
+      "grad_norm": 2.004723310470581,
+      "learning_rate": 4.641472818190006e-05,
+      "loss": 0.9603,
+      "step": 125000
+    },
+    {
+      "epoch": 0.2159767743223406,
+      "grad_norm": 2.032517433166504,
+      "learning_rate": 4.6400387094627656e-05,
+      "loss": 0.9531,
+      "step": 125500
+    },
+    {
+      "epoch": 0.21683723955868459,
+      "grad_norm": 2.0830347537994385,
+      "learning_rate": 4.638604600735526e-05,
+      "loss": 0.9581,
+      "step": 126000
+    },
+    {
+      "epoch": 0.21769770479502856,
+      "grad_norm": 2.0828824043273926,
+      "learning_rate": 4.637170492008286e-05,
+      "loss": 0.95,
+      "step": 126500
+    },
+    {
+      "epoch": 0.21855817003137257,
+      "grad_norm": 1.990696907043457,
+      "learning_rate": 4.635736383281046e-05,
+      "loss": 0.961,
+      "step": 127000
+    },
+    {
+      "epoch": 0.21941863526771654,
+      "grad_norm": 2.3424088954925537,
+      "learning_rate": 4.6343022745538065e-05,
+      "loss": 0.9562,
+      "step": 127500
+    },
+    {
+      "epoch": 0.22027910050406055,
+      "grad_norm": 2.015568494796753,
+      "learning_rate": 4.632868165826566e-05,
+      "loss": 0.9582,
+      "step": 128000
+    },
+    {
+      "epoch": 0.22113956574040453,
+      "grad_norm": 2.360718011856079,
+      "learning_rate": 4.631434057099326e-05,
+      "loss": 0.9551,
+      "step": 128500
+    },
+    {
+      "epoch": 0.2220000309767485,
+      "grad_norm": 2.2896780967712402,
+      "learning_rate": 4.629999948372086e-05,
+      "loss": 0.9588,
+      "step": 129000
+    },
+    {
+      "epoch": 0.2228604962130925,
+      "grad_norm": 2.264202356338501,
+      "learning_rate": 4.628565839644846e-05,
+      "loss": 0.956,
+      "step": 129500
+    },
+    {
+      "epoch": 0.22372096144943648,
+      "grad_norm": 2.0719611644744873,
+      "learning_rate": 4.627131730917606e-05,
+      "loss": 0.9545,
+      "step": 130000
+    },
+    {
+      "epoch": 0.22458142668578046,
+      "grad_norm": 2.0525126457214355,
+      "learning_rate": 4.625697622190366e-05,
+      "loss": 0.9538,
+      "step": 130500
+    },
+    {
+      "epoch": 0.22544189192212447,
+      "grad_norm": 2.126758098602295,
+      "learning_rate": 4.6242635134631265e-05,
+      "loss": 0.9572,
+      "step": 131000
+    },
+    {
+      "epoch": 0.22630235715846844,
+      "grad_norm": 2.1774702072143555,
+      "learning_rate": 4.622829404735886e-05,
+      "loss": 0.9506,
+      "step": 131500
+    },
+    {
+      "epoch": 0.22716282239481242,
+      "grad_norm": 1.9970122575759888,
+      "learning_rate": 4.621395296008646e-05,
+      "loss": 0.9566,
+      "step": 132000
+    },
+    {
+      "epoch": 0.22802328763115642,
+      "grad_norm": 2.1346707344055176,
+      "learning_rate": 4.619961187281406e-05,
+      "loss": 0.9504,
+      "step": 132500
+    },
+    {
+      "epoch": 0.2288837528675004,
+      "grad_norm": 2.0689857006073,
+      "learning_rate": 4.618527078554166e-05,
+      "loss": 0.9561,
+      "step": 133000
+    },
+    {
+      "epoch": 0.22974421810384438,
+      "grad_norm": 1.9728847742080688,
+      "learning_rate": 4.6170929698269264e-05,
+      "loss": 0.9616,
+      "step": 133500
+    },
+    {
+      "epoch": 0.23060468334018838,
+      "grad_norm": 2.1437137126922607,
+      "learning_rate": 4.615658861099687e-05,
+      "loss": 0.9522,
+      "step": 134000
+    },
+    {
+      "epoch": 0.23146514857653236,
+      "grad_norm": 1.8884685039520264,
+      "learning_rate": 4.6142247523724465e-05,
+      "loss": 0.9501,
+      "step": 134500
+    },
+    {
+      "epoch": 0.23232561381287634,
+      "grad_norm": 1.8019119501113892,
+      "learning_rate": 4.612790643645206e-05,
+      "loss": 0.9532,
+      "step": 135000
+    },
+    {
+      "epoch": 0.23318607904922034,
+      "grad_norm": 2.117466688156128,
+      "learning_rate": 4.6113565349179666e-05,
+      "loss": 0.9482,
+      "step": 135500
+    },
+    {
+      "epoch": 0.23404654428556432,
+      "grad_norm": 1.976753830909729,
+      "learning_rate": 4.609922426190726e-05,
+      "loss": 0.9515,
+      "step": 136000
+    },
+    {
+      "epoch": 0.2349070095219083,
+      "grad_norm": 2.071876049041748,
+      "learning_rate": 4.608488317463486e-05,
+      "loss": 0.9518,
+      "step": 136500
+    },
+    {
+      "epoch": 0.2357674747582523,
+      "grad_norm": 2.0765326023101807,
+      "learning_rate": 4.6070542087362464e-05,
+      "loss": 0.947,
+      "step": 137000
+    },
+    {
+      "epoch": 0.23662793999459628,
+      "grad_norm": 1.9267277717590332,
+      "learning_rate": 4.605620100009007e-05,
+      "loss": 0.9525,
+      "step": 137500
+    },
+    {
+      "epoch": 0.23748840523094025,
+      "grad_norm": 2.0131049156188965,
+      "learning_rate": 4.6041859912817664e-05,
+      "loss": 0.9485,
+      "step": 138000
+    },
+    {
+      "epoch": 0.23834887046728426,
+      "grad_norm": 1.991150140762329,
+      "learning_rate": 4.602751882554526e-05,
+      "loss": 0.956,
+      "step": 138500
+    },
+    {
+      "epoch": 0.23920933570362823,
+      "grad_norm": 2.0469107627868652,
+      "learning_rate": 4.6013177738272865e-05,
+      "loss": 0.949,
+      "step": 139000
+    },
+    {
+      "epoch": 0.2400698009399722,
+      "grad_norm": 2.15240216255188,
+      "learning_rate": 4.599883665100046e-05,
+      "loss": 0.9451,
+      "step": 139500
+    },
+    {
+      "epoch": 0.24093026617631622,
+      "grad_norm": 3.1448512077331543,
+      "learning_rate": 4.5984495563728066e-05,
+      "loss": 0.9438,
+      "step": 140000
+    },
+    {
+      "epoch": 0.2417907314126602,
+      "grad_norm": 1.9543808698654175,
+      "learning_rate": 4.597015447645567e-05,
+      "loss": 0.9417,
+      "step": 140500
+    },
+    {
+      "epoch": 0.2426511966490042,
+      "grad_norm": 2.0985629558563232,
+      "learning_rate": 4.595581338918327e-05,
+      "loss": 0.944,
+      "step": 141000
+    },
+    {
+      "epoch": 0.24351166188534817,
+      "grad_norm": 2.0854549407958984,
+      "learning_rate": 4.5941472301910864e-05,
+      "loss": 0.9538,
+      "step": 141500
+    },
+    {
+      "epoch": 0.24437212712169215,
+      "grad_norm": 1.9694116115570068,
+      "learning_rate": 4.592713121463846e-05,
+      "loss": 0.9491,
+      "step": 142000
+    },
+    {
+      "epoch": 0.24523259235803616,
+      "grad_norm": 1.9953951835632324,
+      "learning_rate": 4.5912790127366065e-05,
+      "loss": 0.939,
+      "step": 142500
+    },
+    {
+      "epoch": 0.24609305759438013,
+      "grad_norm": 2.2640116214752197,
+      "learning_rate": 4.589844904009367e-05,
+      "loss": 0.9506,
+      "step": 143000
+    },
+    {
+      "epoch": 0.2469535228307241,
+      "grad_norm": 2.2083499431610107,
+      "learning_rate": 4.5884107952821266e-05,
+      "loss": 0.951,
+      "step": 143500
+    },
+    {
+      "epoch": 0.24781398806706811,
+      "grad_norm": 2.226285457611084,
+      "learning_rate": 4.586976686554887e-05,
+      "loss": 0.9461,
+      "step": 144000
+    },
+    {
+      "epoch": 0.2486744533034121,
+      "grad_norm": 2.1445624828338623,
+      "learning_rate": 4.585542577827647e-05,
+      "loss": 0.9462,
+      "step": 144500
+    },
+    {
+      "epoch": 0.24953491853975607,
+      "grad_norm": 2.054795265197754,
+      "learning_rate": 4.5841084691004064e-05,
+      "loss": 0.9508,
+      "step": 145000
+    },
+    {
+      "epoch": 0.25039538377610004,
+      "grad_norm": 2.2365095615386963,
+      "learning_rate": 4.582674360373167e-05,
+      "loss": 0.944,
+      "step": 145500
+    },
+    {
+      "epoch": 0.251255849012444,
+      "grad_norm": 1.9677467346191406,
+      "learning_rate": 4.5812402516459265e-05,
+      "loss": 0.9462,
+      "step": 146000
+    },
+    {
+      "epoch": 0.25211631424878805,
+      "grad_norm": 2.1352016925811768,
+      "learning_rate": 4.579806142918687e-05,
+      "loss": 0.938,
+      "step": 146500
+    },
+    {
+      "epoch": 0.25297677948513203,
+      "grad_norm": 2.162677526473999,
+      "learning_rate": 4.578372034191447e-05,
+      "loss": 0.9475,
+      "step": 147000
+    },
+    {
+      "epoch": 0.253837244721476,
+      "grad_norm": 2.0151124000549316,
+      "learning_rate": 4.576937925464207e-05,
+      "loss": 0.9448,
+      "step": 147500
+    },
+    {
+      "epoch": 0.25469770995782,
+      "grad_norm": 2.222620964050293,
+      "learning_rate": 4.575503816736967e-05,
+      "loss": 0.9423,
+      "step": 148000
+    },
+    {
+      "epoch": 0.25555817519416396,
+      "grad_norm": 2.1002838611602783,
+      "learning_rate": 4.5740697080097264e-05,
+      "loss": 0.9398,
+      "step": 148500
+    },
+    {
+      "epoch": 0.256418640430508,
+      "grad_norm": 2.1127843856811523,
+      "learning_rate": 4.572635599282487e-05,
+      "loss": 0.9391,
+      "step": 149000
+    },
+    {
+      "epoch": 0.25727910566685197,
+      "grad_norm": 2.1093132495880127,
+      "learning_rate": 4.571201490555247e-05,
+      "loss": 0.9428,
+      "step": 149500
+    },
+    {
+      "epoch": 0.25813957090319595,
+      "grad_norm": 2.0932860374450684,
+      "learning_rate": 4.569767381828007e-05,
+      "loss": 0.9475,
+      "step": 150000
+    },
+    {
+      "epoch": 0.2590000361395399,
+      "grad_norm": 2.1527037620544434,
+      "learning_rate": 4.568333273100767e-05,
+      "loss": 0.9453,
+      "step": 150500
+    },
+    {
+      "epoch": 0.2598605013758839,
+      "grad_norm": 2.1468565464019775,
+      "learning_rate": 4.566899164373527e-05,
+      "loss": 0.9398,
+      "step": 151000
+    },
+    {
+      "epoch": 0.2607209666122279,
+      "grad_norm": 2.050854444503784,
+      "learning_rate": 4.565465055646287e-05,
+      "loss": 0.9412,
+      "step": 151500
+    },
+    {
+      "epoch": 0.2615814318485719,
+      "grad_norm": 2.9939045906066895,
+      "learning_rate": 4.564030946919047e-05,
+      "loss": 0.9459,
+      "step": 152000
+    },
+    {
+      "epoch": 0.2624418970849159,
+      "grad_norm": 2.3177473545074463,
+      "learning_rate": 4.5625968381918075e-05,
+      "loss": 0.9381,
+      "step": 152500
+    },
+    {
+      "epoch": 0.26330236232125986,
+      "grad_norm": 2.2248950004577637,
+      "learning_rate": 4.561162729464567e-05,
+      "loss": 0.9454,
+      "step": 153000
+    },
+    {
+      "epoch": 0.26416282755760384,
+      "grad_norm": 2.0898079872131348,
+      "learning_rate": 4.559728620737327e-05,
+      "loss": 0.9422,
+      "step": 153500
+    },
+    {
+      "epoch": 0.2650232927939478,
+      "grad_norm": 2.272148609161377,
+      "learning_rate": 4.558294512010087e-05,
+      "loss": 0.9464,
+      "step": 154000
+    },
+    {
+      "epoch": 0.2658837580302918,
+      "grad_norm": 1.8870999813079834,
+      "learning_rate": 4.556860403282847e-05,
+      "loss": 0.9379,
+      "step": 154500
+    },
+    {
+      "epoch": 0.2667442232666358,
+      "grad_norm": 2.9211108684539795,
+      "learning_rate": 4.555426294555607e-05,
+      "loss": 0.9381,
+      "step": 155000
+    },
+    {
+      "epoch": 0.2676046885029798,
+      "grad_norm": 1.9935648441314697,
+      "learning_rate": 4.553992185828367e-05,
+      "loss": 0.9415,
+      "step": 155500
+    },
+    {
+      "epoch": 0.2684651537393238,
+      "grad_norm": 2.087963819503784,
+      "learning_rate": 4.5525580771011275e-05,
+      "loss": 0.9431,
+      "step": 156000
+    },
+    {
+      "epoch": 0.26932561897566776,
+      "grad_norm": 2.077944278717041,
+      "learning_rate": 4.551123968373887e-05,
+      "loss": 0.9312,
+      "step": 156500
+    },
+    {
+      "epoch": 0.27018608421201173,
+      "grad_norm": 2.0577354431152344,
+      "learning_rate": 4.5496898596466476e-05,
+      "loss": 0.9431,
+      "step": 157000
+    },
+    {
+      "epoch": 0.2710465494483557,
+      "grad_norm": 1.9151005744934082,
+      "learning_rate": 4.548255750919407e-05,
+      "loss": 0.9417,
+      "step": 157500
+    },
+    {
+      "epoch": 0.27190701468469974,
+      "grad_norm": 2.0716753005981445,
+      "learning_rate": 4.546821642192167e-05,
+      "loss": 0.9411,
+      "step": 158000
+    },
+    {
+      "epoch": 0.2727674799210437,
+      "grad_norm": 2.233751058578491,
+      "learning_rate": 4.5453875334649274e-05,
+      "loss": 0.943,
+      "step": 158500
+    },
+    {
+      "epoch": 0.2736279451573877,
+      "grad_norm": 2.095139265060425,
+      "learning_rate": 4.543953424737688e-05,
+      "loss": 0.9401,
+      "step": 159000
+    },
+    {
+      "epoch": 0.2744884103937317,
+      "grad_norm": 2.020110607147217,
+      "learning_rate": 4.5425193160104475e-05,
+      "loss": 0.9443,
+      "step": 159500
+    },
+    {
+      "epoch": 0.27534887563007565,
+      "grad_norm": 2.152085542678833,
+      "learning_rate": 4.541085207283207e-05,
+      "loss": 0.9357,
+      "step": 160000
+    },
+    {
+      "epoch": 0.27620934086641963,
+      "grad_norm": 2.083021879196167,
+      "learning_rate": 4.5396510985559675e-05,
+      "loss": 0.9367,
+      "step": 160500
+    },
+    {
+      "epoch": 0.27706980610276366,
+      "grad_norm": 2.171924591064453,
+      "learning_rate": 4.538216989828727e-05,
+      "loss": 0.9325,
+      "step": 161000
+    },
+    {
+      "epoch": 0.27793027133910764,
+      "grad_norm": 1.9896093606948853,
+      "learning_rate": 4.5367828811014876e-05,
+      "loss": 0.9379,
+      "step": 161500
+    },
+    {
+      "epoch": 0.2787907365754516,
+      "grad_norm": 2.2307465076446533,
+      "learning_rate": 4.535348772374248e-05,
+      "loss": 0.9376,
+      "step": 162000
+    },
+    {
+      "epoch": 0.2796512018117956,
+      "grad_norm": 2.0789954662323,
+      "learning_rate": 4.533914663647008e-05,
+      "loss": 0.9394,
+      "step": 162500
+    },
+    {
+      "epoch": 0.28051166704813957,
+      "grad_norm": 2.0075278282165527,
+      "learning_rate": 4.5324805549197674e-05,
+      "loss": 0.9353,
+      "step": 163000
+    },
+    {
+      "epoch": 0.2813721322844836,
+      "grad_norm": 2.100450277328491,
+      "learning_rate": 4.531046446192528e-05,
+      "loss": 0.9393,
+      "step": 163500
+    },
+    {
+      "epoch": 0.2822325975208276,
+      "grad_norm": 2.096583604812622,
+      "learning_rate": 4.5296123374652875e-05,
+      "loss": 0.9267,
+      "step": 164000
+    },
+    {
+      "epoch": 0.28309306275717155,
+      "grad_norm": 2.025454044342041,
+      "learning_rate": 4.528178228738047e-05,
+      "loss": 0.934,
+      "step": 164500
+    },
+    {
+      "epoch": 0.28395352799351553,
+      "grad_norm": 2.026967763900757,
+      "learning_rate": 4.5267441200108076e-05,
+      "loss": 0.9324,
+      "step": 165000
+    },
+    {
+      "epoch": 0.2848139932298595,
+      "grad_norm": 1.9689258337020874,
+      "learning_rate": 4.525310011283568e-05,
+      "loss": 0.9298,
+      "step": 165500
+    },
+    {
+      "epoch": 0.2856744584662035,
+      "grad_norm": 2.0960588455200195,
+      "learning_rate": 4.523875902556328e-05,
+      "loss": 0.9262,
+      "step": 166000
+    },
+    {
+      "epoch": 0.2865349237025475,
+      "grad_norm": 2.2745180130004883,
+      "learning_rate": 4.5224417938290874e-05,
+      "loss": 0.93,
+      "step": 166500
+    },
+    {
+      "epoch": 0.2873953889388915,
+      "grad_norm": 2.1345441341400146,
+      "learning_rate": 4.521007685101848e-05,
+      "loss": 0.9297,
+      "step": 167000
+    },
+    {
+      "epoch": 0.28825585417523547,
+      "grad_norm": 1.9947483539581299,
+      "learning_rate": 4.5195735763746075e-05,
+      "loss": 0.9325,
+      "step": 167500
+    },
+    {
+      "epoch": 0.28911631941157945,
+      "grad_norm": 1.983390212059021,
+      "learning_rate": 4.518139467647368e-05,
+      "loss": 0.9317,
+      "step": 168000
+    },
+    {
+      "epoch": 0.2899767846479234,
+      "grad_norm": 2.014618396759033,
+      "learning_rate": 4.516705358920128e-05,
+      "loss": 0.9414,
+      "step": 168500
+    },
+    {
+      "epoch": 0.2908372498842674,
+      "grad_norm": 2.0016281604766846,
+      "learning_rate": 4.515271250192888e-05,
+      "loss": 0.9368,
+      "step": 169000
+    },
+    {
+      "epoch": 0.29169771512061143,
+      "grad_norm": 1.968947410583496,
+      "learning_rate": 4.513837141465648e-05,
+      "loss": 0.9318,
+      "step": 169500
+    },
+    {
+      "epoch": 0.2925581803569554,
+      "grad_norm": 2.0170702934265137,
+      "learning_rate": 4.512403032738408e-05,
+      "loss": 0.9312,
+      "step": 170000
+    },
+    {
+      "epoch": 0.2934186455932994,
+      "grad_norm": 2.0024240016937256,
+      "learning_rate": 4.510968924011168e-05,
+      "loss": 0.9267,
+      "step": 170500
+    },
+    {
+      "epoch": 0.29427911082964336,
+      "grad_norm": 2.0785553455352783,
+      "learning_rate": 4.5095348152839275e-05,
+      "loss": 0.9322,
+      "step": 171000
+    },
+    {
+      "epoch": 0.29513957606598734,
+      "grad_norm": 2.0966298580169678,
+      "learning_rate": 4.508100706556688e-05,
+      "loss": 0.9303,
+      "step": 171500
+    },
+    {
+      "epoch": 0.2960000413023313,
+      "grad_norm": 2.1931986808776855,
+      "learning_rate": 4.506666597829448e-05,
+      "loss": 0.9331,
+      "step": 172000
+    },
+    {
+      "epoch": 0.29686050653867535,
+      "grad_norm": 2.2337117195129395,
+      "learning_rate": 4.505232489102208e-05,
+      "loss": 0.9332,
+      "step": 172500
+    },
+    {
+      "epoch": 0.2977209717750193,
+      "grad_norm": 1.900471568107605,
+      "learning_rate": 4.503798380374968e-05,
+      "loss": 0.926,
+      "step": 173000
+    },
+    {
+      "epoch": 0.2985814370113633,
+      "grad_norm": 2.045421838760376,
+      "learning_rate": 4.502364271647728e-05,
+      "loss": 0.9237,
+      "step": 173500
+    },
+    {
+      "epoch": 0.2994419022477073,
+      "grad_norm": 1.988563060760498,
+      "learning_rate": 4.500930162920488e-05,
+      "loss": 0.9318,
+      "step": 174000
+    },
+    {
+      "epoch": 0.30030236748405126,
+      "grad_norm": 2.1729953289031982,
+      "learning_rate": 4.499496054193248e-05,
+      "loss": 0.9283,
+      "step": 174500
+    },
+    {
+      "epoch": 0.3011628327203953,
+      "grad_norm": 2.043877601623535,
+      "learning_rate": 4.4980619454660086e-05,
+      "loss": 0.9258,
+      "step": 175000
+    },
+    {
+      "epoch": 0.30202329795673927,
+      "grad_norm": 1.941645622253418,
+      "learning_rate": 4.496627836738768e-05,
+      "loss": 0.9342,
+      "step": 175500
+    },
+    {
+      "epoch": 0.30288376319308324,
+      "grad_norm": 2.305682897567749,
+      "learning_rate": 4.495193728011528e-05,
+      "loss": 0.9228,
+      "step": 176000
+    },
+    {
+      "epoch": 0.3037442284294272,
+      "grad_norm": 1.9681456089019775,
+      "learning_rate": 4.493759619284288e-05,
+      "loss": 0.9228,
+      "step": 176500
+    },
+    {
+      "epoch": 0.3046046936657712,
+      "grad_norm": 2.2018189430236816,
+      "learning_rate": 4.492325510557048e-05,
+      "loss": 0.9249,
+      "step": 177000
+    },
+    {
+      "epoch": 0.3054651589021152,
+      "grad_norm": 2.322690963745117,
+      "learning_rate": 4.4908914018298085e-05,
+      "loss": 0.9298,
+      "step": 177500
+    },
+    {
+      "epoch": 0.3063256241384592,
+      "grad_norm": 2.019329786300659,
+      "learning_rate": 4.489457293102568e-05,
+      "loss": 0.9281,
+      "step": 178000
+    },
+    {
+      "epoch": 0.3071860893748032,
+      "grad_norm": 2.0158770084381104,
+      "learning_rate": 4.4880231843753286e-05,
+      "loss": 0.9282,
+      "step": 178500
+    },
+    {
+      "epoch": 0.30804655461114716,
+      "grad_norm": 1.9363014698028564,
+      "learning_rate": 4.486589075648088e-05,
+      "loss": 0.9281,
+      "step": 179000
+    },
+    {
+      "epoch": 0.30890701984749114,
+      "grad_norm": 2.32789945602417,
+      "learning_rate": 4.485154966920848e-05,
+      "loss": 0.9268,
+      "step": 179500
+    },
+    {
+      "epoch": 0.3097674850838351,
+      "grad_norm": 2.1968231201171875,
+      "learning_rate": 4.4837208581936084e-05,
+      "loss": 0.9172,
+      "step": 180000
+    },
+    {
+      "epoch": 0.3106279503201791,
+      "grad_norm": 1.9847798347473145,
+      "learning_rate": 4.482286749466368e-05,
+      "loss": 0.9294,
+      "step": 180500
+    },
+    {
+      "epoch": 0.3114884155565231,
+      "grad_norm": 1.9857689142227173,
+      "learning_rate": 4.4808526407391285e-05,
+      "loss": 0.9336,
+      "step": 181000
+    },
+    {
+      "epoch": 0.3123488807928671,
+      "grad_norm": 2.266566514968872,
+      "learning_rate": 4.479418532011889e-05,
+      "loss": 0.9257,
+      "step": 181500
+    },
+    {
+      "epoch": 0.3132093460292111,
+      "grad_norm": 2.045213222503662,
+      "learning_rate": 4.4779844232846486e-05,
+      "loss": 0.93,
+      "step": 182000
+    },
+    {
+      "epoch": 0.31406981126555505,
+      "grad_norm": 2.0169789791107178,
+      "learning_rate": 4.476550314557408e-05,
+      "loss": 0.9235,
+      "step": 182500
+    },
+    {
+      "epoch": 0.31493027650189903,
+      "grad_norm": 2.1827285289764404,
+      "learning_rate": 4.475116205830168e-05,
+      "loss": 0.9251,
+      "step": 183000
+    },
+    {
+      "epoch": 0.315790741738243,
+      "grad_norm": 2.1066367626190186,
+      "learning_rate": 4.4736820971029284e-05,
+      "loss": 0.9243,
+      "step": 183500
+    },
+    {
+      "epoch": 0.31665120697458704,
+      "grad_norm": 2.0616202354431152,
+      "learning_rate": 4.472247988375689e-05,
+      "loss": 0.9263,
+      "step": 184000
+    },
+    {
+      "epoch": 0.317511672210931,
+      "grad_norm": 2.0447213649749756,
+      "learning_rate": 4.4708138796484485e-05,
+      "loss": 0.9276,
+      "step": 184500
+    },
+    {
+      "epoch": 0.318372137447275,
+      "grad_norm": 2.1739165782928467,
+      "learning_rate": 4.469379770921209e-05,
+      "loss": 0.9239,
+      "step": 185000
+    },
+    {
+      "epoch": 0.31923260268361897,
+      "grad_norm": 3.1863811016082764,
+      "learning_rate": 4.4679456621939685e-05,
+      "loss": 0.9234,
+      "step": 185500
+    },
+    {
+      "epoch": 0.32009306791996295,
+      "grad_norm": 2.1314053535461426,
+      "learning_rate": 4.466511553466728e-05,
+      "loss": 0.925,
+      "step": 186000
+    },
+    {
+      "epoch": 0.3209535331563069,
+      "grad_norm": 2.092536449432373,
+      "learning_rate": 4.4650774447394886e-05,
+      "loss": 0.9256,
+      "step": 186500
+    },
+    {
+      "epoch": 0.32181399839265096,
+      "grad_norm": 2.1120293140411377,
+      "learning_rate": 4.463643336012249e-05,
+      "loss": 0.9252,
+      "step": 187000
+    },
+    {
+      "epoch": 0.32267446362899493,
+      "grad_norm": 1.976959466934204,
+      "learning_rate": 4.462209227285009e-05,
+      "loss": 0.9208,
+      "step": 187500
+    },
+    {
+      "epoch": 0.3235349288653389,
+      "grad_norm": 2.041245937347412,
+      "learning_rate": 4.4607751185577684e-05,
+      "loss": 0.9258,
+      "step": 188000
+    },
+    {
+      "epoch": 0.3243953941016829,
+      "grad_norm": 2.3442962169647217,
+      "learning_rate": 4.459341009830529e-05,
+      "loss": 0.9212,
+      "step": 188500
+    },
+    {
+      "epoch": 0.32525585933802686,
+      "grad_norm": 2.0697786808013916,
+      "learning_rate": 4.4579069011032885e-05,
+      "loss": 0.9139,
+      "step": 189000
+    },
+    {
+      "epoch": 0.3261163245743709,
+      "grad_norm": 2.3078439235687256,
+      "learning_rate": 4.456472792376048e-05,
+      "loss": 0.9223,
+      "step": 189500
+    },
+    {
+      "epoch": 0.3269767898107149,
+      "grad_norm": 2.058361768722534,
+      "learning_rate": 4.4550386836488086e-05,
+      "loss": 0.9224,
+      "step": 190000
+    },
+    {
+      "epoch": 0.32783725504705885,
+      "grad_norm": 2.203118324279785,
+      "learning_rate": 4.453604574921569e-05,
+      "loss": 0.9237,
+      "step": 190500
+    },
+    {
+      "epoch": 0.32869772028340283,
+      "grad_norm": 1.89760422706604,
+      "learning_rate": 4.452170466194329e-05,
+      "loss": 0.9212,
+      "step": 191000
+    },
+    {
+      "epoch": 0.3295581855197468,
+      "grad_norm": 1.964350938796997,
+      "learning_rate": 4.450736357467089e-05,
+      "loss": 0.9237,
+      "step": 191500
+    },
+    {
+      "epoch": 0.3304186507560908,
+      "grad_norm": 2.0023722648620605,
+      "learning_rate": 4.449302248739849e-05,
+      "loss": 0.9227,
+      "step": 192000
+    },
+    {
+      "epoch": 0.3312791159924348,
+      "grad_norm": 2.03126859664917,
+      "learning_rate": 4.4478681400126085e-05,
+      "loss": 0.9167,
+      "step": 192500
+    },
+    {
+      "epoch": 0.3321395812287788,
+      "grad_norm": 2.0440711975097656,
+      "learning_rate": 4.446434031285369e-05,
+      "loss": 0.9125,
+      "step": 193000
+    },
+    {
+      "epoch": 0.33300004646512277,
+      "grad_norm": 2.1581292152404785,
+      "learning_rate": 4.444999922558129e-05,
+      "loss": 0.9168,
+      "step": 193500
+    },
+    {
+      "epoch": 0.33386051170146674,
+      "grad_norm": 2.038708448410034,
+      "learning_rate": 4.443565813830889e-05,
+      "loss": 0.9225,
+      "step": 194000
+    },
+    {
+      "epoch": 0.3347209769378107,
+      "grad_norm": 2.237109422683716,
+      "learning_rate": 4.442131705103649e-05,
+      "loss": 0.9146,
+      "step": 194500
+    },
+    {
+      "epoch": 0.3355814421741547,
+      "grad_norm": 2.081531286239624,
+      "learning_rate": 4.440697596376409e-05,
+      "loss": 0.9151,
+      "step": 195000
+    },
+    {
+      "epoch": 0.33644190741049873,
+      "grad_norm": 1.8253999948501587,
+      "learning_rate": 4.439263487649169e-05,
+      "loss": 0.9221,
+      "step": 195500
+    },
+    {
+      "epoch": 0.3373023726468427,
+      "grad_norm": 2.0977344512939453,
+      "learning_rate": 4.4378293789219285e-05,
+      "loss": 0.9182,
+      "step": 196000
+    },
+    {
+      "epoch": 0.3381628378831867,
+      "grad_norm": 2.108025550842285,
+      "learning_rate": 4.4363952701946896e-05,
+      "loss": 0.9145,
+      "step": 196500
+    },
+    {
+      "epoch": 0.33902330311953066,
+      "grad_norm": 2.023393154144287,
+      "learning_rate": 4.434961161467449e-05,
+      "loss": 0.9179,
+      "step": 197000
+    },
+    {
+      "epoch": 0.33988376835587464,
+      "grad_norm": 2.1165921688079834,
+      "learning_rate": 4.433527052740209e-05,
+      "loss": 0.9206,
+      "step": 197500
+    },
+    {
+      "epoch": 0.3407442335922186,
+      "grad_norm": 1.7804551124572754,
+      "learning_rate": 4.4320929440129694e-05,
+      "loss": 0.9188,
+      "step": 198000
+    },
+    {
+      "epoch": 0.34160469882856265,
+      "grad_norm": 1.9795186519622803,
+      "learning_rate": 4.430658835285729e-05,
+      "loss": 0.9176,
+      "step": 198500
+    },
+    {
+      "epoch": 0.3424651640649066,
+      "grad_norm": 2.0974135398864746,
+      "learning_rate": 4.429224726558489e-05,
+      "loss": 0.921,
+      "step": 199000
+    },
+    {
+      "epoch": 0.3433256293012506,
+      "grad_norm": 2.106494903564453,
+      "learning_rate": 4.427790617831249e-05,
+      "loss": 0.9218,
+      "step": 199500
+    },
+    {
+      "epoch": 0.3441860945375946,
+      "grad_norm": 2.2233493328094482,
+      "learning_rate": 4.4263565091040096e-05,
+      "loss": 0.9184,
+      "step": 200000
+    },
+    {
+      "epoch": 0.34504655977393855,
+      "grad_norm": 1.9548823833465576,
+      "learning_rate": 4.424922400376769e-05,
+      "loss": 0.9166,
+      "step": 200500
+    },
+    {
+      "epoch": 0.3459070250102826,
+      "grad_norm": 1.9309402704238892,
+      "learning_rate": 4.423488291649529e-05,
+      "loss": 0.9133,
+      "step": 201000
+    },
+    {
+      "epoch": 0.34676749024662656,
+      "grad_norm": 2.1494619846343994,
+      "learning_rate": 4.4220541829222894e-05,
+      "loss": 0.9163,
+      "step": 201500
+    },
+    {
+      "epoch": 0.34762795548297054,
+      "grad_norm": 2.0527141094207764,
+      "learning_rate": 4.420620074195049e-05,
+      "loss": 0.9221,
+      "step": 202000
+    },
+    {
+      "epoch": 0.3484884207193145,
+      "grad_norm": 2.224276542663574,
+      "learning_rate": 4.4191859654678095e-05,
+      "loss": 0.9114,
+      "step": 202500
+    },
+    {
+      "epoch": 0.3493488859556585,
+      "grad_norm": 2.2672996520996094,
+      "learning_rate": 4.41775185674057e-05,
+      "loss": 0.9164,
+      "step": 203000
+    },
+    {
+      "epoch": 0.35020935119200247,
+      "grad_norm": 1.9365850687026978,
+      "learning_rate": 4.4163177480133296e-05,
+      "loss": 0.909,
+      "step": 203500
+    },
+    {
+      "epoch": 0.3510698164283465,
+      "grad_norm": 2.1146013736724854,
+      "learning_rate": 4.414883639286089e-05,
+      "loss": 0.9205,
+      "step": 204000
+    },
+    {
+      "epoch": 0.3519302816646905,
+      "grad_norm": 2.0115785598754883,
+      "learning_rate": 4.4134495305588497e-05,
+      "loss": 0.9118,
+      "step": 204500
+    },
+    {
+      "epoch": 0.35279074690103446,
+      "grad_norm": 1.9710958003997803,
+      "learning_rate": 4.4120154218316094e-05,
+      "loss": 0.9147,
+      "step": 205000
+    },
+    {
+      "epoch": 0.35365121213737843,
+      "grad_norm": 2.1901769638061523,
+      "learning_rate": 4.410581313104369e-05,
+      "loss": 0.9157,
+      "step": 205500
+    },
+    {
+      "epoch": 0.3545116773737224,
+      "grad_norm": 1.9956777095794678,
+      "learning_rate": 4.4091472043771295e-05,
+      "loss": 0.9112,
+      "step": 206000
+    },
+    {
+      "epoch": 0.3553721426100664,
+      "grad_norm": 2.0977365970611572,
+      "learning_rate": 4.40771309564989e-05,
+      "loss": 0.9199,
+      "step": 206500
+    },
+    {
+      "epoch": 0.3562326078464104,
+      "grad_norm": 2.092785358428955,
+      "learning_rate": 4.4062789869226496e-05,
+      "loss": 0.9111,
+      "step": 207000
+    },
+    {
+      "epoch": 0.3570930730827544,
+      "grad_norm": 2.122471570968628,
+      "learning_rate": 4.404844878195409e-05,
+      "loss": 0.9114,
+      "step": 207500
+    },
+    {
+      "epoch": 0.3579535383190984,
+      "grad_norm": 1.9477378129959106,
+      "learning_rate": 4.4034107694681696e-05,
+      "loss": 0.9091,
+      "step": 208000
+    },
+    {
+      "epoch": 0.35881400355544235,
+      "grad_norm": 1.773989200592041,
+      "learning_rate": 4.4019766607409294e-05,
+      "loss": 0.9117,
+      "step": 208500
+    },
+    {
+      "epoch": 0.35967446879178633,
+      "grad_norm": 2.234323024749756,
+      "learning_rate": 4.40054255201369e-05,
+      "loss": 0.9144,
+      "step": 209000
+    },
+    {
+      "epoch": 0.3605349340281303,
+      "grad_norm": 2.09895920753479,
+      "learning_rate": 4.39910844328645e-05,
+      "loss": 0.9127,
+      "step": 209500
+    },
+    {
+      "epoch": 0.36139539926447434,
+      "grad_norm": 2.123444080352783,
+      "learning_rate": 4.39767433455921e-05,
+      "loss": 0.9126,
+      "step": 210000
+    },
+    {
+      "epoch": 0.3622558645008183,
+      "grad_norm": 2.052734851837158,
+      "learning_rate": 4.3962402258319695e-05,
+      "loss": 0.9128,
+      "step": 210500
+    },
+    {
+      "epoch": 0.3631163297371623,
+      "grad_norm": 2.132145404815674,
+      "learning_rate": 4.39480611710473e-05,
+      "loss": 0.9117,
+      "step": 211000
+    },
+    {
+      "epoch": 0.36397679497350627,
+      "grad_norm": 2.2205147743225098,
+      "learning_rate": 4.3933720083774896e-05,
+      "loss": 0.9174,
+      "step": 211500
+    },
+    {
+      "epoch": 0.36483726020985024,
+      "grad_norm": 2.234994411468506,
+      "learning_rate": 4.39193789965025e-05,
+      "loss": 0.9121,
+      "step": 212000
+    },
+    {
+      "epoch": 0.3656977254461942,
+      "grad_norm": 1.8479299545288086,
+      "learning_rate": 4.39050379092301e-05,
+      "loss": 0.9083,
+      "step": 212500
+    },
+    {
+      "epoch": 0.36655819068253825,
+      "grad_norm": 2.0153796672821045,
+      "learning_rate": 4.38906968219577e-05,
+      "loss": 0.9124,
+      "step": 213000
+    },
+    {
+      "epoch": 0.36741865591888223,
+      "grad_norm": 2.0594232082366943,
+      "learning_rate": 4.38763557346853e-05,
+      "loss": 0.9075,
+      "step": 213500
+    },
+    {
+      "epoch": 0.3682791211552262,
+      "grad_norm": 2.0284030437469482,
+      "learning_rate": 4.3862014647412895e-05,
+      "loss": 0.9139,
+      "step": 214000
+    },
+    {
+      "epoch": 0.3691395863915702,
+      "grad_norm": 2.0767407417297363,
+      "learning_rate": 4.38476735601405e-05,
+      "loss": 0.9152,
+      "step": 214500
+    },
+    {
+      "epoch": 0.37000005162791416,
+      "grad_norm": 2.1043102741241455,
+      "learning_rate": 4.3833332472868096e-05,
+      "loss": 0.9116,
+      "step": 215000
+    },
+    {
+      "epoch": 0.3708605168642582,
+      "grad_norm": 1.960959792137146,
+      "learning_rate": 4.38189913855957e-05,
+      "loss": 0.9092,
+      "step": 215500
+    },
+    {
+      "epoch": 0.37172098210060217,
+      "grad_norm": 1.9991062879562378,
+      "learning_rate": 4.3804650298323304e-05,
+      "loss": 0.9193,
+      "step": 216000
+    },
+    {
+      "epoch": 0.37258144733694615,
+      "grad_norm": 2.0325255393981934,
+      "learning_rate": 4.37903092110509e-05,
+      "loss": 0.9115,
+      "step": 216500
+    },
+    {
+      "epoch": 0.3734419125732901,
+      "grad_norm": 2.1963534355163574,
+      "learning_rate": 4.37759681237785e-05,
+      "loss": 0.9094,
+      "step": 217000
+    },
+    {
+      "epoch": 0.3743023778096341,
+      "grad_norm": 2.172563314437866,
+      "learning_rate": 4.3761627036506095e-05,
+      "loss": 0.909,
+      "step": 217500
+    },
+    {
+      "epoch": 0.3751628430459781,
+      "grad_norm": 2.0587527751922607,
+      "learning_rate": 4.37472859492337e-05,
+      "loss": 0.9089,
+      "step": 218000
+    },
+    {
+      "epoch": 0.3760233082823221,
+      "grad_norm": 2.1344945430755615,
+      "learning_rate": 4.37329448619613e-05,
+      "loss": 0.9146,
+      "step": 218500
+    },
+    {
+      "epoch": 0.3768837735186661,
+      "grad_norm": 4.116335868835449,
+      "learning_rate": 4.37186037746889e-05,
+      "loss": 0.9093,
+      "step": 219000
+    },
+    {
+      "epoch": 0.37774423875501006,
+      "grad_norm": 1.8812421560287476,
+      "learning_rate": 4.3704262687416504e-05,
+      "loss": 0.9085,
+      "step": 219500
+    },
+    {
+      "epoch": 0.37860470399135404,
+      "grad_norm": 1.841057538986206,
+      "learning_rate": 4.36899216001441e-05,
+      "loss": 0.9108,
+      "step": 220000
+    },
+    {
+      "epoch": 0.379465169227698,
+      "grad_norm": 2.1964564323425293,
+      "learning_rate": 4.36755805128717e-05,
+      "loss": 0.9086,
+      "step": 220500
+    },
+    {
+      "epoch": 0.380325634464042,
+      "grad_norm": 1.981557011604309,
+      "learning_rate": 4.36612394255993e-05,
+      "loss": 0.9109,
+      "step": 221000
+    },
+    {
+      "epoch": 0.381186099700386,
+      "grad_norm": 2.1632473468780518,
+      "learning_rate": 4.3646898338326906e-05,
+      "loss": 0.9012,
+      "step": 221500
+    },
+    {
+      "epoch": 0.38204656493673,
+      "grad_norm": 2.0308001041412354,
+      "learning_rate": 4.36325572510545e-05,
+      "loss": 0.9077,
+      "step": 222000
+    },
+    {
+      "epoch": 0.382907030173074,
+      "grad_norm": 1.95619535446167,
+      "learning_rate": 4.361821616378211e-05,
+      "loss": 0.9083,
+      "step": 222500
+    },
+    {
+      "epoch": 0.38376749540941796,
+      "grad_norm": 2.0333635807037354,
+      "learning_rate": 4.3603875076509704e-05,
+      "loss": 0.9116,
+      "step": 223000
+    },
+    {
+      "epoch": 0.38462796064576193,
+      "grad_norm": 2.3641154766082764,
+      "learning_rate": 4.35895339892373e-05,
+      "loss": 0.9077,
+      "step": 223500
+    },
+    {
+      "epoch": 0.3854884258821059,
+      "grad_norm": 2.0519320964813232,
+      "learning_rate": 4.35751929019649e-05,
+      "loss": 0.9131,
+      "step": 224000
+    },
+    {
+      "epoch": 0.38634889111844994,
+      "grad_norm": 2.1741113662719727,
+      "learning_rate": 4.35608518146925e-05,
+      "loss": 0.9051,
+      "step": 224500
+    },
+    {
+      "epoch": 0.3872093563547939,
+      "grad_norm": 2.2641000747680664,
+      "learning_rate": 4.3546510727420106e-05,
+      "loss": 0.9148,
+      "step": 225000
+    },
+    {
+      "epoch": 0.3880698215911379,
+      "grad_norm": 2.112330675125122,
+      "learning_rate": 4.35321696401477e-05,
+      "loss": 0.9028,
+      "step": 225500
+    },
+    {
+      "epoch": 0.3889302868274819,
+      "grad_norm": 2.238736152648926,
+      "learning_rate": 4.351782855287531e-05,
+      "loss": 0.9075,
+      "step": 226000
+    },
+    {
+      "epoch": 0.38979075206382585,
+      "grad_norm": 2.0583317279815674,
+      "learning_rate": 4.3503487465602904e-05,
+      "loss": 0.9059,
+      "step": 226500
+    },
+    {
+      "epoch": 0.39065121730016983,
+      "grad_norm": 2.021120309829712,
+      "learning_rate": 4.34891463783305e-05,
+      "loss": 0.907,
+      "step": 227000
+    },
+    {
+      "epoch": 0.39151168253651386,
+      "grad_norm": 2.0049400329589844,
+      "learning_rate": 4.3474805291058105e-05,
+      "loss": 0.9028,
+      "step": 227500
+    },
+    {
+      "epoch": 0.39237214777285784,
+      "grad_norm": 2.042653799057007,
+      "learning_rate": 4.346046420378571e-05,
+      "loss": 0.9047,
+      "step": 228000
+    },
+    {
+      "epoch": 0.3932326130092018,
+      "grad_norm": 2.180102586746216,
+      "learning_rate": 4.3446123116513306e-05,
+      "loss": 0.9072,
+      "step": 228500
+    },
+    {
+      "epoch": 0.3940930782455458,
+      "grad_norm": 2.1076409816741943,
+      "learning_rate": 4.34317820292409e-05,
+      "loss": 0.9014,
+      "step": 229000
+    },
+    {
+      "epoch": 0.39495354348188977,
+      "grad_norm": 2.024866819381714,
+      "learning_rate": 4.3417440941968507e-05,
+      "loss": 0.9078,
+      "step": 229500
+    },
+    {
+      "epoch": 0.3958140087182338,
+      "grad_norm": 2.0629196166992188,
+      "learning_rate": 4.3403099854696104e-05,
+      "loss": 0.9021,
+      "step": 230000
+    },
+    {
+      "epoch": 0.3966744739545778,
+      "grad_norm": 2.060455799102783,
+      "learning_rate": 4.33887587674237e-05,
+      "loss": 0.8988,
+      "step": 230500
+    },
+    {
+      "epoch": 0.39753493919092175,
+      "grad_norm": 2.0068538188934326,
+      "learning_rate": 4.337441768015131e-05,
+      "loss": 0.9071,
+      "step": 231000
+    },
+    {
+      "epoch": 0.39839540442726573,
+      "grad_norm": 2.2743167877197266,
+      "learning_rate": 4.336007659287891e-05,
+      "loss": 0.9067,
+      "step": 231500
+    },
+    {
+      "epoch": 0.3992558696636097,
+      "grad_norm": 1.9746699333190918,
+      "learning_rate": 4.3345735505606505e-05,
+      "loss": 0.9028,
+      "step": 232000
+    },
+    {
+      "epoch": 0.4001163348999537,
+      "grad_norm": 2.108346700668335,
+      "learning_rate": 4.333139441833411e-05,
+      "loss": 0.9046,
+      "step": 232500
+    },
+    {
+      "epoch": 0.4009768001362977,
+      "grad_norm": 2.0433006286621094,
+      "learning_rate": 4.3317053331061706e-05,
+      "loss": 0.9079,
+      "step": 233000
+    },
+    {
+      "epoch": 0.4018372653726417,
+      "grad_norm": 2.050401210784912,
+      "learning_rate": 4.3302712243789304e-05,
+      "loss": 0.9032,
+      "step": 233500
+    },
+    {
+      "epoch": 0.40269773060898567,
+      "grad_norm": 2.0607173442840576,
+      "learning_rate": 4.328837115651691e-05,
+      "loss": 0.8957,
+      "step": 234000
+    },
+    {
+      "epoch": 0.40355819584532965,
+      "grad_norm": 2.1468214988708496,
+      "learning_rate": 4.327403006924451e-05,
+      "loss": 0.8996,
+      "step": 234500
+    },
+    {
+      "epoch": 0.4044186610816736,
+      "grad_norm": 2.10554838180542,
+      "learning_rate": 4.325968898197211e-05,
+      "loss": 0.9052,
+      "step": 235000
+    },
+    {
+      "epoch": 0.4052791263180176,
+      "grad_norm": 1.9938172101974487,
+      "learning_rate": 4.3245347894699705e-05,
+      "loss": 0.8994,
+      "step": 235500
+    },
+    {
+      "epoch": 0.40613959155436163,
+      "grad_norm": 2.218541383743286,
+      "learning_rate": 4.323100680742731e-05,
+      "loss": 0.9005,
+      "step": 236000
+    },
+    {
+      "epoch": 0.4070000567907056,
+      "grad_norm": 1.946315050125122,
+      "learning_rate": 4.3216665720154906e-05,
+      "loss": 0.9019,
+      "step": 236500
+    },
+    {
+      "epoch": 0.4078605220270496,
+      "grad_norm": 2.236579179763794,
+      "learning_rate": 4.320232463288251e-05,
+      "loss": 0.9049,
+      "step": 237000
+    },
+    {
+      "epoch": 0.40872098726339356,
+      "grad_norm": 2.0288383960723877,
+      "learning_rate": 4.3187983545610114e-05,
+      "loss": 0.9001,
+      "step": 237500
+    },
+    {
+      "epoch": 0.40958145249973754,
+      "grad_norm": 1.885482907295227,
+      "learning_rate": 4.317364245833771e-05,
+      "loss": 0.9043,
+      "step": 238000
+    },
+    {
+      "epoch": 0.4104419177360815,
+      "grad_norm": 1.9730161428451538,
+      "learning_rate": 4.315930137106531e-05,
+      "loss": 0.9043,
+      "step": 238500
+    },
+    {
+      "epoch": 0.41130238297242555,
+      "grad_norm": 3.811640501022339,
+      "learning_rate": 4.314496028379291e-05,
+      "loss": 0.9007,
+      "step": 239000
+    },
+    {
+      "epoch": 0.41216284820876953,
+      "grad_norm": 1.9219677448272705,
+      "learning_rate": 4.313061919652051e-05,
+      "loss": 0.9059,
+      "step": 239500
+    },
+    {
+      "epoch": 0.4130233134451135,
+      "grad_norm": 2.092369318008423,
+      "learning_rate": 4.3116278109248106e-05,
+      "loss": 0.9017,
+      "step": 240000
+    },
+    {
+      "epoch": 0.4138837786814575,
+      "grad_norm": 4.2498297691345215,
+      "learning_rate": 4.310193702197571e-05,
+      "loss": 0.8922,
+      "step": 240500
+    },
+    {
+      "epoch": 0.41474424391780146,
+      "grad_norm": 2.057892322540283,
+      "learning_rate": 4.3087595934703314e-05,
+      "loss": 0.9016,
+      "step": 241000
+    },
+    {
+      "epoch": 0.4156047091541455,
+      "grad_norm": 1.8440794944763184,
+      "learning_rate": 4.307325484743091e-05,
+      "loss": 0.9037,
+      "step": 241500
+    },
+    {
+      "epoch": 0.41646517439048947,
+      "grad_norm": 2.4426589012145996,
+      "learning_rate": 4.305891376015851e-05,
+      "loss": 0.8994,
+      "step": 242000
+    },
+    {
+      "epoch": 0.41732563962683344,
+      "grad_norm": 2.063756227493286,
+      "learning_rate": 4.304457267288611e-05,
+      "loss": 0.8958,
+      "step": 242500
+    },
+    {
+      "epoch": 0.4181861048631774,
+      "grad_norm": 2.2012550830841064,
+      "learning_rate": 4.303023158561371e-05,
+      "loss": 0.9004,
+      "step": 243000
+    },
+    {
+      "epoch": 0.4190465700995214,
+      "grad_norm": 2.1932969093322754,
+      "learning_rate": 4.301589049834131e-05,
+      "loss": 0.902,
+      "step": 243500
+    },
+    {
+      "epoch": 0.4199070353358654,
+      "grad_norm": 1.9205131530761719,
+      "learning_rate": 4.300154941106892e-05,
+      "loss": 0.9038,
+      "step": 244000
+    },
+    {
+      "epoch": 0.4207675005722094,
+      "grad_norm": 2.281463861465454,
+      "learning_rate": 4.2987208323796514e-05,
+      "loss": 0.901,
+      "step": 244500
+    },
+    {
+      "epoch": 0.4216279658085534,
+      "grad_norm": 1.8984506130218506,
+      "learning_rate": 4.297286723652411e-05,
+      "loss": 0.8934,
+      "step": 245000
+    },
+    {
+      "epoch": 0.42248843104489736,
+      "grad_norm": 2.0649311542510986,
+      "learning_rate": 4.2958526149251715e-05,
+      "loss": 0.8973,
+      "step": 245500
+    },
+    {
+      "epoch": 0.42334889628124134,
+      "grad_norm": 2.0315778255462646,
+      "learning_rate": 4.294418506197931e-05,
+      "loss": 0.9,
+      "step": 246000
+    },
+    {
+      "epoch": 0.4242093615175853,
+      "grad_norm": 2.0501866340637207,
+      "learning_rate": 4.2929843974706916e-05,
+      "loss": 0.9006,
+      "step": 246500
+    },
+    {
+      "epoch": 0.4250698267539293,
+      "grad_norm": 1.8434195518493652,
+      "learning_rate": 4.291550288743451e-05,
+      "loss": 0.8996,
+      "step": 247000
+    },
+    {
+      "epoch": 0.4259302919902733,
+      "grad_norm": 1.9517645835876465,
+      "learning_rate": 4.290116180016212e-05,
+      "loss": 0.8961,
+      "step": 247500
+    },
+    {
+      "epoch": 0.4267907572266173,
+      "grad_norm": 2.049384355545044,
+      "learning_rate": 4.2886820712889714e-05,
+      "loss": 0.8958,
+      "step": 248000
+    },
+    {
+      "epoch": 0.4276512224629613,
+      "grad_norm": 2.0335021018981934,
+      "learning_rate": 4.287247962561731e-05,
+      "loss": 0.9003,
+      "step": 248500
+    },
+    {
+      "epoch": 0.42851168769930525,
+      "grad_norm": 1.9440404176712036,
+      "learning_rate": 4.2858138538344915e-05,
+      "loss": 0.9009,
+      "step": 249000
+    },
+    {
+      "epoch": 0.42937215293564923,
+      "grad_norm": 1.907614827156067,
+      "learning_rate": 4.284379745107251e-05,
+      "loss": 0.8947,
+      "step": 249500
+    },
+    {
+      "epoch": 0.4302326181719932,
+      "grad_norm": 1.9623234272003174,
+      "learning_rate": 4.2829456363800116e-05,
+      "loss": 0.9006,
+      "step": 250000
+    },
+    {
+      "epoch": 0.43109308340833724,
+      "grad_norm": 1.9491050243377686,
+      "learning_rate": 4.281511527652772e-05,
+      "loss": 0.901,
+      "step": 250500
+    },
+    {
+      "epoch": 0.4319535486446812,
+      "grad_norm": 2.1582417488098145,
+      "learning_rate": 4.2800774189255317e-05,
+      "loss": 0.9015,
+      "step": 251000
+    },
+    {
+      "epoch": 0.4328140138810252,
+      "grad_norm": 1.8875306844711304,
+      "learning_rate": 4.2786433101982914e-05,
+      "loss": 0.893,
+      "step": 251500
+    },
+    {
+      "epoch": 0.43367447911736917,
+      "grad_norm": 1.9841439723968506,
+      "learning_rate": 4.277209201471051e-05,
+      "loss": 0.895,
+      "step": 252000
+    },
+    {
+      "epoch": 0.43453494435371315,
+      "grad_norm": 1.9874733686447144,
+      "learning_rate": 4.2757750927438115e-05,
+      "loss": 0.8975,
+      "step": 252500
+    },
+    {
+      "epoch": 0.4353954095900571,
+      "grad_norm": 2.1332311630249023,
+      "learning_rate": 4.274340984016572e-05,
+      "loss": 0.8924,
+      "step": 253000
+    },
+    {
+      "epoch": 0.43625587482640116,
+      "grad_norm": 2.069314956665039,
+      "learning_rate": 4.2729068752893316e-05,
+      "loss": 0.8932,
+      "step": 253500
+    },
+    {
+      "epoch": 0.43711634006274513,
+      "grad_norm": 2.087435245513916,
+      "learning_rate": 4.271472766562092e-05,
+      "loss": 0.8966,
+      "step": 254000
+    },
+    {
+      "epoch": 0.4379768052990891,
+      "grad_norm": 1.9603450298309326,
+      "learning_rate": 4.2700386578348516e-05,
+      "loss": 0.902,
+      "step": 254500
+    },
+    {
+      "epoch": 0.4388372705354331,
+      "grad_norm": 2.302950859069824,
+      "learning_rate": 4.2686045491076114e-05,
+      "loss": 0.8944,
+      "step": 255000
+    },
+    {
+      "epoch": 0.43969773577177707,
+      "grad_norm": 2.033417224884033,
+      "learning_rate": 4.267170440380372e-05,
+      "loss": 0.8938,
+      "step": 255500
+    },
+    {
+      "epoch": 0.4405582010081211,
+      "grad_norm": 2.086822271347046,
+      "learning_rate": 4.265736331653132e-05,
+      "loss": 0.8947,
+      "step": 256000
+    },
+    {
+      "epoch": 0.4414186662444651,
+      "grad_norm": 1.9544868469238281,
+      "learning_rate": 4.264302222925892e-05,
+      "loss": 0.8908,
+      "step": 256500
+    },
+    {
+      "epoch": 0.44227913148080905,
+      "grad_norm": 2.105898380279541,
+      "learning_rate": 4.262868114198652e-05,
+      "loss": 0.8936,
+      "step": 257000
+    },
+    {
+      "epoch": 0.44313959671715303,
+      "grad_norm": 2.320366144180298,
+      "learning_rate": 4.261434005471412e-05,
+      "loss": 0.8963,
+      "step": 257500
+    },
+    {
+      "epoch": 0.444000061953497,
+      "grad_norm": 2.063095808029175,
+      "learning_rate": 4.2599998967441716e-05,
+      "loss": 0.896,
+      "step": 258000
+    },
+    {
+      "epoch": 0.444860527189841,
+      "grad_norm": 1.9787700176239014,
+      "learning_rate": 4.2585657880169313e-05,
+      "loss": 0.893,
+      "step": 258500
+    },
+    {
+      "epoch": 0.445720992426185,
+      "grad_norm": 1.8536509275436401,
+      "learning_rate": 4.257131679289692e-05,
+      "loss": 0.8991,
+      "step": 259000
+    },
+    {
+      "epoch": 0.446581457662529,
+      "grad_norm": 2.1534652709960938,
+      "learning_rate": 4.255697570562452e-05,
+      "loss": 0.901,
+      "step": 259500
+    },
+    {
+      "epoch": 0.44744192289887297,
+      "grad_norm": 2.1015608310699463,
+      "learning_rate": 4.254263461835212e-05,
+      "loss": 0.8984,
+      "step": 260000
+    },
+    {
+      "epoch": 0.44830238813521694,
+      "grad_norm": 2.066599130630493,
+      "learning_rate": 4.252829353107972e-05,
+      "loss": 0.898,
+      "step": 260500
+    },
+    {
+      "epoch": 0.4491628533715609,
+      "grad_norm": 2.0920300483703613,
+      "learning_rate": 4.251395244380732e-05,
+      "loss": 0.8964,
+      "step": 261000
+    },
+    {
+      "epoch": 0.4500233186079049,
+      "grad_norm": 1.9317692518234253,
+      "learning_rate": 4.2499611356534916e-05,
+      "loss": 0.8948,
+      "step": 261500
+    },
+    {
+      "epoch": 0.45088378384424893,
+      "grad_norm": 2.1189675331115723,
+      "learning_rate": 4.248527026926252e-05,
+      "loss": 0.8939,
+      "step": 262000
+    },
+    {
+      "epoch": 0.4517442490805929,
+      "grad_norm": 2.0204405784606934,
+      "learning_rate": 4.2470929181990124e-05,
+      "loss": 0.8919,
+      "step": 262500
+    },
+    {
+      "epoch": 0.4526047143169369,
+      "grad_norm": 2.0056357383728027,
+      "learning_rate": 4.245658809471772e-05,
+      "loss": 0.8933,
+      "step": 263000
+    },
+    {
+      "epoch": 0.45346517955328086,
+      "grad_norm": 2.0812103748321533,
+      "learning_rate": 4.244224700744532e-05,
+      "loss": 0.8939,
+      "step": 263500
+    },
+    {
+      "epoch": 0.45432564478962484,
+      "grad_norm": 2.1630799770355225,
+      "learning_rate": 4.242790592017292e-05,
+      "loss": 0.8919,
+      "step": 264000
+    },
+    {
+      "epoch": 0.4551861100259688,
+      "grad_norm": 2.2746987342834473,
+      "learning_rate": 4.241356483290052e-05,
+      "loss": 0.898,
+      "step": 264500
+    },
+    {
+      "epoch": 0.45604657526231285,
+      "grad_norm": 2.476505756378174,
+      "learning_rate": 4.2399223745628116e-05,
+      "loss": 0.8919,
+      "step": 265000
+    },
+    {
+      "epoch": 0.4569070404986568,
+      "grad_norm": 2.2061052322387695,
+      "learning_rate": 4.238488265835573e-05,
+      "loss": 0.8959,
+      "step": 265500
+    },
+    {
+      "epoch": 0.4577675057350008,
+      "grad_norm": 2.035796642303467,
+      "learning_rate": 4.2370541571083324e-05,
+      "loss": 0.8907,
+      "step": 266000
+    },
+    {
+      "epoch": 0.4586279709713448,
+      "grad_norm": 2.0652384757995605,
+      "learning_rate": 4.235620048381092e-05,
+      "loss": 0.8902,
+      "step": 266500
+    },
+    {
+      "epoch": 0.45948843620768876,
+      "grad_norm": 2.0927200317382812,
+      "learning_rate": 4.2341859396538525e-05,
+      "loss": 0.8942,
+      "step": 267000
+    },
+    {
+      "epoch": 0.46034890144403273,
+      "grad_norm": 1.8390779495239258,
+      "learning_rate": 4.232751830926612e-05,
+      "loss": 0.8916,
+      "step": 267500
+    },
+    {
+      "epoch": 0.46120936668037676,
+      "grad_norm": 2.0620462894439697,
+      "learning_rate": 4.231317722199372e-05,
+      "loss": 0.8914,
+      "step": 268000
+    },
+    {
+      "epoch": 0.46206983191672074,
+      "grad_norm": 1.9553086757659912,
+      "learning_rate": 4.229883613472132e-05,
+      "loss": 0.8897,
+      "step": 268500
+    },
+    {
+      "epoch": 0.4629302971530647,
+      "grad_norm": 2.0123965740203857,
+      "learning_rate": 4.228449504744893e-05,
+      "loss": 0.8935,
+      "step": 269000
+    },
+    {
+      "epoch": 0.4637907623894087,
+      "grad_norm": 1.8187013864517212,
+      "learning_rate": 4.2270153960176524e-05,
+      "loss": 0.8921,
+      "step": 269500
+    },
+    {
+      "epoch": 0.46465122762575267,
+      "grad_norm": 2.2623324394226074,
+      "learning_rate": 4.225581287290412e-05,
+      "loss": 0.8918,
+      "step": 270000
+    },
+    {
+      "epoch": 0.4655116928620967,
+      "grad_norm": 1.965198278427124,
+      "learning_rate": 4.2241471785631725e-05,
+      "loss": 0.8944,
+      "step": 270500
+    },
+    {
+      "epoch": 0.4663721580984407,
+      "grad_norm": 1.9791353940963745,
+      "learning_rate": 4.222713069835932e-05,
+      "loss": 0.8899,
+      "step": 271000
+    },
+    {
+      "epoch": 0.46723262333478466,
+      "grad_norm": 2.0294482707977295,
+      "learning_rate": 4.2212789611086926e-05,
+      "loss": 0.8929,
+      "step": 271500
+    },
+    {
+      "epoch": 0.46809308857112863,
+      "grad_norm": 1.869797706604004,
+      "learning_rate": 4.219844852381453e-05,
+      "loss": 0.8914,
+      "step": 272000
+    },
+    {
+      "epoch": 0.4689535538074726,
+      "grad_norm": 2.0433285236358643,
+      "learning_rate": 4.218410743654213e-05,
+      "loss": 0.8929,
+      "step": 272500
+    },
+    {
+      "epoch": 0.4698140190438166,
+      "grad_norm": 2.2087464332580566,
+      "learning_rate": 4.2169766349269724e-05,
+      "loss": 0.8923,
+      "step": 273000
+    },
+    {
+      "epoch": 0.4706744842801606,
+      "grad_norm": 2.0753061771392822,
+      "learning_rate": 4.215542526199733e-05,
+      "loss": 0.8918,
+      "step": 273500
+    },
+    {
+      "epoch": 0.4715349495165046,
+      "grad_norm": 2.107851266860962,
+      "learning_rate": 4.2141084174724925e-05,
+      "loss": 0.8943,
+      "step": 274000
+    },
+    {
+      "epoch": 0.4723954147528486,
+      "grad_norm": 2.038882255554199,
+      "learning_rate": 4.212674308745252e-05,
+      "loss": 0.8932,
+      "step": 274500
+    },
+    {
+      "epoch": 0.47325587998919255,
+      "grad_norm": 2.12312388420105,
+      "learning_rate": 4.2112402000180126e-05,
+      "loss": 0.8867,
+      "step": 275000
+    },
+    {
+      "epoch": 0.47411634522553653,
+      "grad_norm": 2.0398507118225098,
+      "learning_rate": 4.209806091290773e-05,
+      "loss": 0.8921,
+      "step": 275500
+    },
+    {
+      "epoch": 0.4749768104618805,
+      "grad_norm": 2.09047794342041,
+      "learning_rate": 4.2083719825635327e-05,
+      "loss": 0.8817,
+      "step": 276000
+    },
+    {
+      "epoch": 0.47583727569822454,
+      "grad_norm": 1.9717376232147217,
+      "learning_rate": 4.2069378738362924e-05,
+      "loss": 0.8922,
+      "step": 276500
+    },
+    {
+      "epoch": 0.4766977409345685,
+      "grad_norm": 1.9686635732650757,
+      "learning_rate": 4.205503765109053e-05,
+      "loss": 0.8972,
+      "step": 277000
+    },
+    {
+      "epoch": 0.4775582061709125,
+      "grad_norm": 1.8727903366088867,
+      "learning_rate": 4.2040696563818125e-05,
+      "loss": 0.8892,
+      "step": 277500
+    },
+    {
+      "epoch": 0.47841867140725647,
+      "grad_norm": 2.0806689262390137,
+      "learning_rate": 4.202635547654573e-05,
+      "loss": 0.8869,
+      "step": 278000
+    },
+    {
+      "epoch": 0.47927913664360045,
+      "grad_norm": 1.9705919027328491,
+      "learning_rate": 4.201201438927333e-05,
+      "loss": 0.8844,
+      "step": 278500
+    },
+    {
+      "epoch": 0.4801396018799444,
+      "grad_norm": 2.01405668258667,
+      "learning_rate": 4.199767330200093e-05,
+      "loss": 0.893,
+      "step": 279000
+    },
+    {
+      "epoch": 0.48100006711628845,
+      "grad_norm": 2.08960223197937,
+      "learning_rate": 4.1983332214728526e-05,
+      "loss": 0.8842,
+      "step": 279500
+    },
+    {
+      "epoch": 0.48186053235263243,
+      "grad_norm": 2.0530412197113037,
+      "learning_rate": 4.196899112745613e-05,
+      "loss": 0.8826,
+      "step": 280000
+    },
+    {
+      "epoch": 0.4827209975889764,
+      "grad_norm": 1.8881099224090576,
+      "learning_rate": 4.195465004018373e-05,
+      "loss": 0.8829,
+      "step": 280500
+    },
+    {
+      "epoch": 0.4835814628253204,
+      "grad_norm": 2.0287468433380127,
+      "learning_rate": 4.194030895291133e-05,
+      "loss": 0.8814,
+      "step": 281000
+    },
+    {
+      "epoch": 0.48444192806166436,
+      "grad_norm": 2.0433034896850586,
+      "learning_rate": 4.192596786563893e-05,
+      "loss": 0.8889,
+      "step": 281500
+    },
+    {
+      "epoch": 0.4853023932980084,
+      "grad_norm": 2.0601441860198975,
+      "learning_rate": 4.191162677836653e-05,
+      "loss": 0.8868,
+      "step": 282000
+    },
+    {
+      "epoch": 0.48616285853435237,
+      "grad_norm": 2.3189380168914795,
+      "learning_rate": 4.189728569109413e-05,
+      "loss": 0.8894,
+      "step": 282500
+    },
+    {
+      "epoch": 0.48702332377069635,
+      "grad_norm": 2.302962064743042,
+      "learning_rate": 4.1882944603821726e-05,
+      "loss": 0.8834,
+      "step": 283000
+    },
+    {
+      "epoch": 0.4878837890070403,
+      "grad_norm": 1.9201202392578125,
+      "learning_rate": 4.186860351654933e-05,
+      "loss": 0.8846,
+      "step": 283500
+    },
+    {
+      "epoch": 0.4887442542433843,
+      "grad_norm": 2.085836887359619,
+      "learning_rate": 4.185426242927693e-05,
+      "loss": 0.8919,
+      "step": 284000
+    },
+    {
+      "epoch": 0.4896047194797283,
+      "grad_norm": 2.1023192405700684,
+      "learning_rate": 4.183992134200453e-05,
+      "loss": 0.8881,
+      "step": 284500
+    },
+    {
+      "epoch": 0.4904651847160723,
+      "grad_norm": 2.0571980476379395,
+      "learning_rate": 4.1825580254732135e-05,
+      "loss": 0.8833,
+      "step": 285000
+    },
+    {
+      "epoch": 0.4913256499524163,
+      "grad_norm": 2.038604259490967,
+      "learning_rate": 4.181123916745973e-05,
+      "loss": 0.8816,
+      "step": 285500
+    },
+    {
+      "epoch": 0.49218611518876026,
+      "grad_norm": 2.018510341644287,
+      "learning_rate": 4.179689808018733e-05,
+      "loss": 0.8829,
+      "step": 286000
+    },
+    {
+      "epoch": 0.49304658042510424,
+      "grad_norm": 2.025078535079956,
+      "learning_rate": 4.1782556992914926e-05,
+      "loss": 0.8874,
+      "step": 286500
+    },
+    {
+      "epoch": 0.4939070456614482,
+      "grad_norm": 2.0259318351745605,
+      "learning_rate": 4.176821590564253e-05,
+      "loss": 0.8783,
+      "step": 287000
+    },
+    {
+      "epoch": 0.4947675108977922,
+      "grad_norm": 2.0440592765808105,
+      "learning_rate": 4.1753874818370134e-05,
+      "loss": 0.8852,
+      "step": 287500
+    },
+    {
+      "epoch": 0.49562797613413623,
+      "grad_norm": 2.1548845767974854,
+      "learning_rate": 4.173953373109773e-05,
+      "loss": 0.8843,
+      "step": 288000
+    },
+    {
+      "epoch": 0.4964884413704802,
+      "grad_norm": 2.020847797393799,
+      "learning_rate": 4.1725192643825335e-05,
+      "loss": 0.8881,
+      "step": 288500
+    },
+    {
+      "epoch": 0.4973489066068242,
+      "grad_norm": 2.0484495162963867,
+      "learning_rate": 4.171085155655293e-05,
+      "loss": 0.8902,
+      "step": 289000
+    },
+    {
+      "epoch": 0.49820937184316816,
+      "grad_norm": 1.9658195972442627,
+      "learning_rate": 4.169651046928053e-05,
+      "loss": 0.8805,
+      "step": 289500
+    },
+    {
+      "epoch": 0.49906983707951214,
+      "grad_norm": 2.1585335731506348,
+      "learning_rate": 4.168216938200813e-05,
+      "loss": 0.8885,
+      "step": 290000
+    },
+    {
+      "epoch": 0.4999303023158561,
+      "grad_norm": 2.0190553665161133,
+      "learning_rate": 4.166782829473574e-05,
+      "loss": 0.8833,
+      "step": 290500
+    },
+    {
+      "epoch": 0.5007907675522001,
+      "grad_norm": 2.143428087234497,
+      "learning_rate": 4.1653487207463334e-05,
+      "loss": 0.8837,
+      "step": 291000
+    },
+    {
+      "epoch": 0.5016512327885441,
+      "grad_norm": 2.24603533744812,
+      "learning_rate": 4.163914612019094e-05,
+      "loss": 0.8779,
+      "step": 291500
+    },
+    {
+      "epoch": 0.502511698024888,
+      "grad_norm": 2.109105348587036,
+      "learning_rate": 4.1624805032918535e-05,
+      "loss": 0.8875,
+      "step": 292000
+    },
+    {
+      "epoch": 0.5033721632612321,
+      "grad_norm": 1.9702835083007812,
+      "learning_rate": 4.161046394564613e-05,
+      "loss": 0.8804,
+      "step": 292500
+    },
+    {
+      "epoch": 0.5042326284975761,
+      "grad_norm": 2.784588575363159,
+      "learning_rate": 4.159612285837373e-05,
+      "loss": 0.8879,
+      "step": 293000
+    },
+    {
+      "epoch": 0.5050930937339201,
+      "grad_norm": 1.999629020690918,
+      "learning_rate": 4.158178177110133e-05,
+      "loss": 0.8817,
+      "step": 293500
+    },
+    {
+      "epoch": 0.5059535589702641,
+      "grad_norm": 2.034653663635254,
+      "learning_rate": 4.156744068382894e-05,
+      "loss": 0.8719,
+      "step": 294000
+    },
+    {
+      "epoch": 0.506814024206608,
+      "grad_norm": 2.095053195953369,
+      "learning_rate": 4.1553099596556534e-05,
+      "loss": 0.8823,
+      "step": 294500
+    },
+    {
+      "epoch": 0.507674489442952,
+      "grad_norm": 1.9798400402069092,
+      "learning_rate": 4.153875850928414e-05,
+      "loss": 0.8818,
+      "step": 295000
+    },
+    {
+      "epoch": 0.508534954679296,
+      "grad_norm": 2.06270170211792,
+      "learning_rate": 4.1524417422011735e-05,
+      "loss": 0.8829,
+      "step": 295500
+    },
+    {
+      "epoch": 0.50939541991564,
+      "grad_norm": 2.1127400398254395,
+      "learning_rate": 4.151007633473933e-05,
+      "loss": 0.8815,
+      "step": 296000
+    },
+    {
+      "epoch": 0.510255885151984,
+      "grad_norm": 2.151163101196289,
+      "learning_rate": 4.1495735247466936e-05,
+      "loss": 0.8791,
+      "step": 296500
+    },
+    {
+      "epoch": 0.5111163503883279,
+      "grad_norm": 2.0091350078582764,
+      "learning_rate": 4.148139416019454e-05,
+      "loss": 0.8948,
+      "step": 297000
+    },
+    {
+      "epoch": 0.5119768156246719,
+      "grad_norm": 2.0361227989196777,
+      "learning_rate": 4.146705307292214e-05,
+      "loss": 0.8772,
+      "step": 297500
+    },
+    {
+      "epoch": 0.512837280861016,
+      "grad_norm": 1.9893290996551514,
+      "learning_rate": 4.1452711985649734e-05,
+      "loss": 0.8823,
+      "step": 298000
+    },
+    {
+      "epoch": 0.51369774609736,
+      "grad_norm": 1.988218069076538,
+      "learning_rate": 4.143837089837734e-05,
+      "loss": 0.891,
+      "step": 298500
+    },
+    {
+      "epoch": 0.5145582113337039,
+      "grad_norm": 1.9783134460449219,
+      "learning_rate": 4.1424029811104935e-05,
+      "loss": 0.8815,
+      "step": 299000
+    },
+    {
+      "epoch": 0.5154186765700479,
+      "grad_norm": 2.102525234222412,
+      "learning_rate": 4.140968872383253e-05,
+      "loss": 0.8734,
+      "step": 299500
+    },
+    {
+      "epoch": 0.5162791418063919,
+      "grad_norm": 2.2450239658355713,
+      "learning_rate": 4.1395347636560136e-05,
+      "loss": 0.8838,
+      "step": 300000
+    },
+    {
+      "epoch": 0.5171396070427359,
+      "grad_norm": 2.089251756668091,
+      "learning_rate": 4.138100654928774e-05,
+      "loss": 0.8758,
+      "step": 300500
+    },
+    {
+      "epoch": 0.5180000722790798,
+      "grad_norm": 2.2183520793914795,
+      "learning_rate": 4.1366665462015337e-05,
+      "loss": 0.8814,
+      "step": 301000
+    },
+    {
+      "epoch": 0.5188605375154238,
+      "grad_norm": 1.9809156656265259,
+      "learning_rate": 4.135232437474294e-05,
+      "loss": 0.8751,
+      "step": 301500
+    },
+    {
+      "epoch": 0.5197210027517678,
+      "grad_norm": 2.149174451828003,
+      "learning_rate": 4.133798328747054e-05,
+      "loss": 0.8779,
+      "step": 302000
+    },
+    {
+      "epoch": 0.5205814679881118,
+      "grad_norm": 2.312516927719116,
+      "learning_rate": 4.1323642200198135e-05,
+      "loss": 0.8797,
+      "step": 302500
+    },
+    {
+      "epoch": 0.5214419332244558,
+      "grad_norm": 2.053981065750122,
+      "learning_rate": 4.130930111292574e-05,
+      "loss": 0.8824,
+      "step": 303000
+    },
+    {
+      "epoch": 0.5223023984607997,
+      "grad_norm": 1.9515398740768433,
+      "learning_rate": 4.129496002565334e-05,
+      "loss": 0.8838,
+      "step": 303500
+    },
+    {
+      "epoch": 0.5231628636971438,
+      "grad_norm": 2.1182141304016113,
+      "learning_rate": 4.128061893838094e-05,
+      "loss": 0.8817,
+      "step": 304000
+    },
+    {
+      "epoch": 0.5240233289334878,
+      "grad_norm": 2.140738010406494,
+      "learning_rate": 4.1266277851108536e-05,
+      "loss": 0.8806,
+      "step": 304500
+    },
+    {
+      "epoch": 0.5248837941698318,
+      "grad_norm": 2.0797948837280273,
+      "learning_rate": 4.125193676383614e-05,
+      "loss": 0.88,
+      "step": 305000
+    },
+    {
+      "epoch": 0.5257442594061758,
+      "grad_norm": 2.092442274093628,
+      "learning_rate": 4.123759567656374e-05,
+      "loss": 0.8825,
+      "step": 305500
+    },
+    {
+      "epoch": 0.5266047246425197,
+      "grad_norm": 2.1355602741241455,
+      "learning_rate": 4.122325458929134e-05,
+      "loss": 0.8776,
+      "step": 306000
+    },
+    {
+      "epoch": 0.5274651898788637,
+      "grad_norm": 2.0246939659118652,
+      "learning_rate": 4.1208913502018945e-05,
+      "loss": 0.8827,
+      "step": 306500
+    },
+    {
+      "epoch": 0.5283256551152077,
+      "grad_norm": 1.9206238985061646,
+      "learning_rate": 4.119457241474654e-05,
+      "loss": 0.8833,
+      "step": 307000
+    },
+    {
+      "epoch": 0.5291861203515517,
+      "grad_norm": 2.153930187225342,
+      "learning_rate": 4.118023132747414e-05,
+      "loss": 0.8825,
+      "step": 307500
+    },
+    {
+      "epoch": 0.5300465855878956,
+      "grad_norm": 1.9237273931503296,
+      "learning_rate": 4.116589024020174e-05,
+      "loss": 0.8895,
+      "step": 308000
+    },
+    {
+      "epoch": 0.5309070508242396,
+      "grad_norm": 2.0631232261657715,
+      "learning_rate": 4.115154915292934e-05,
+      "loss": 0.8824,
+      "step": 308500
+    },
+    {
+      "epoch": 0.5317675160605836,
+      "grad_norm": 1.9997292757034302,
+      "learning_rate": 4.113720806565694e-05,
+      "loss": 0.8784,
+      "step": 309000
+    },
+    {
+      "epoch": 0.5326279812969277,
+      "grad_norm": 2.1037609577178955,
+      "learning_rate": 4.112286697838454e-05,
+      "loss": 0.8733,
+      "step": 309500
+    },
+    {
+      "epoch": 0.5334884465332717,
+      "grad_norm": 2.1329736709594727,
+      "learning_rate": 4.1108525891112145e-05,
+      "loss": 0.8767,
+      "step": 310000
+    },
+    {
+      "epoch": 0.5343489117696156,
+      "grad_norm": 2.07114577293396,
+      "learning_rate": 4.109418480383974e-05,
+      "loss": 0.8743,
+      "step": 310500
+    },
+    {
+      "epoch": 0.5352093770059596,
+      "grad_norm": 2.110933780670166,
+      "learning_rate": 4.107984371656734e-05,
+      "loss": 0.8737,
+      "step": 311000
+    },
+    {
+      "epoch": 0.5360698422423036,
+      "grad_norm": 1.9786311388015747,
+      "learning_rate": 4.106550262929494e-05,
+      "loss": 0.877,
+      "step": 311500
+    },
+    {
+      "epoch": 0.5369303074786476,
+      "grad_norm": 1.9082682132720947,
+      "learning_rate": 4.105116154202254e-05,
+      "loss": 0.8783,
+      "step": 312000
+    },
+    {
+      "epoch": 0.5377907727149915,
+      "grad_norm": 2.0248301029205322,
+      "learning_rate": 4.1036820454750144e-05,
+      "loss": 0.8786,
+      "step": 312500
+    },
+    {
+      "epoch": 0.5386512379513355,
+      "grad_norm": 2.0271735191345215,
+      "learning_rate": 4.102247936747775e-05,
+      "loss": 0.8784,
+      "step": 313000
+    },
+    {
+      "epoch": 0.5395117031876795,
+      "grad_norm": 1.8867460489273071,
+      "learning_rate": 4.1008138280205345e-05,
+      "loss": 0.8698,
+      "step": 313500
+    },
+    {
+      "epoch": 0.5403721684240235,
+      "grad_norm": 2.0477957725524902,
+      "learning_rate": 4.099379719293294e-05,
+      "loss": 0.8793,
+      "step": 314000
+    },
+    {
+      "epoch": 0.5412326336603674,
+      "grad_norm": 2.1383612155914307,
+      "learning_rate": 4.0979456105660546e-05,
+      "loss": 0.8828,
+      "step": 314500
+    },
+    {
+      "epoch": 0.5420930988967114,
+      "grad_norm": 1.9967807531356812,
+      "learning_rate": 4.096511501838814e-05,
+      "loss": 0.8816,
+      "step": 315000
+    },
+    {
+      "epoch": 0.5429535641330555,
+      "grad_norm": 2.489108085632324,
+      "learning_rate": 4.095077393111575e-05,
+      "loss": 0.8741,
+      "step": 315500
+    },
+    {
+      "epoch": 0.5438140293693995,
+      "grad_norm": 2.342057466506958,
+      "learning_rate": 4.0936432843843344e-05,
+      "loss": 0.8875,
+      "step": 316000
+    },
+    {
+      "epoch": 0.5446744946057435,
+      "grad_norm": 1.9754986763000488,
+      "learning_rate": 4.092209175657095e-05,
+      "loss": 0.8767,
+      "step": 316500
+    },
+    {
+      "epoch": 0.5455349598420874,
+      "grad_norm": 2.211503267288208,
+      "learning_rate": 4.0907750669298545e-05,
+      "loss": 0.8791,
+      "step": 317000
+    },
+    {
+      "epoch": 0.5463954250784314,
+      "grad_norm": 1.9339312314987183,
+      "learning_rate": 4.089340958202614e-05,
+      "loss": 0.8754,
+      "step": 317500
+    },
+    {
+      "epoch": 0.5472558903147754,
+      "grad_norm": 2.0220510959625244,
+      "learning_rate": 4.0879068494753746e-05,
+      "loss": 0.8705,
+      "step": 318000
+    },
+    {
+      "epoch": 0.5481163555511194,
+      "grad_norm": 1.974448323249817,
+      "learning_rate": 4.086472740748134e-05,
+      "loss": 0.8719,
+      "step": 318500
+    },
+    {
+      "epoch": 0.5489768207874633,
+      "grad_norm": 2.084419012069702,
+      "learning_rate": 4.085038632020895e-05,
+      "loss": 0.8704,
+      "step": 319000
+    },
+    {
+      "epoch": 0.5498372860238073,
+      "grad_norm": 1.9440003633499146,
+      "learning_rate": 4.083604523293655e-05,
+      "loss": 0.8788,
+      "step": 319500
+    },
+    {
+      "epoch": 0.5506977512601513,
+      "grad_norm": 2.0453526973724365,
+      "learning_rate": 4.082170414566415e-05,
+      "loss": 0.8752,
+      "step": 320000
+    },
+    {
+      "epoch": 0.5515582164964953,
+      "grad_norm": 2.0000038146972656,
+      "learning_rate": 4.0807363058391745e-05,
+      "loss": 0.8773,
+      "step": 320500
+    },
+    {
+      "epoch": 0.5524186817328393,
+      "grad_norm": 2.253737688064575,
+      "learning_rate": 4.079302197111935e-05,
+      "loss": 0.8721,
+      "step": 321000
+    },
+    {
+      "epoch": 0.5532791469691833,
+      "grad_norm": 2.0602428913116455,
+      "learning_rate": 4.0778680883846946e-05,
+      "loss": 0.8794,
+      "step": 321500
+    },
+    {
+      "epoch": 0.5541396122055273,
+      "grad_norm": 2.0261833667755127,
+      "learning_rate": 4.076433979657455e-05,
+      "loss": 0.8724,
+      "step": 322000
+    },
+    {
+      "epoch": 0.5550000774418713,
+      "grad_norm": 1.9648793935775757,
+      "learning_rate": 4.0749998709302147e-05,
+      "loss": 0.8724,
+      "step": 322500
+    },
+    {
+      "epoch": 0.5558605426782153,
+      "grad_norm": 1.9868428707122803,
+      "learning_rate": 4.073565762202975e-05,
+      "loss": 0.8721,
+      "step": 323000
+    },
+    {
+      "epoch": 0.5567210079145593,
+      "grad_norm": 1.9594879150390625,
+      "learning_rate": 4.072131653475735e-05,
+      "loss": 0.8717,
+      "step": 323500
+    },
+    {
+      "epoch": 0.5575814731509032,
+      "grad_norm": 2.072599411010742,
+      "learning_rate": 4.0706975447484945e-05,
+      "loss": 0.8758,
+      "step": 324000
+    },
+    {
+      "epoch": 0.5584419383872472,
+      "grad_norm": 2.037163257598877,
+      "learning_rate": 4.069263436021255e-05,
+      "loss": 0.8779,
+      "step": 324500
+    },
+    {
+      "epoch": 0.5593024036235912,
+      "grad_norm": 2.0014655590057373,
+      "learning_rate": 4.0678293272940146e-05,
+      "loss": 0.8753,
+      "step": 325000
+    },
+    {
+      "epoch": 0.5601628688599352,
+      "grad_norm": 1.9225795269012451,
+      "learning_rate": 4.066395218566775e-05,
+      "loss": 0.8713,
+      "step": 325500
+    },
+    {
+      "epoch": 0.5610233340962791,
+      "grad_norm": 1.923448085784912,
+      "learning_rate": 4.064961109839535e-05,
+      "loss": 0.8739,
+      "step": 326000
+    },
+    {
+      "epoch": 0.5618837993326231,
+      "grad_norm": 2.097306251525879,
+      "learning_rate": 4.063527001112295e-05,
+      "loss": 0.8808,
+      "step": 326500
+    },
+    {
+      "epoch": 0.5627442645689672,
+      "grad_norm": 2.081660270690918,
+      "learning_rate": 4.062092892385055e-05,
+      "loss": 0.8812,
+      "step": 327000
+    },
+    {
+      "epoch": 0.5636047298053112,
+      "grad_norm": 2.085153818130493,
+      "learning_rate": 4.0606587836578145e-05,
+      "loss": 0.871,
+      "step": 327500
+    },
+    {
+      "epoch": 0.5644651950416552,
+      "grad_norm": 2.0828776359558105,
+      "learning_rate": 4.059224674930575e-05,
+      "loss": 0.8691,
+      "step": 328000
+    },
+    {
+      "epoch": 0.5653256602779991,
+      "grad_norm": 2.226726531982422,
+      "learning_rate": 4.057790566203335e-05,
+      "loss": 0.879,
+      "step": 328500
+    },
+    {
+      "epoch": 0.5661861255143431,
+      "grad_norm": 1.9523491859436035,
+      "learning_rate": 4.056356457476095e-05,
+      "loss": 0.8706,
+      "step": 329000
+    },
+    {
+      "epoch": 0.5670465907506871,
+      "grad_norm": 2.002776861190796,
+      "learning_rate": 4.054922348748855e-05,
+      "loss": 0.8752,
+      "step": 329500
+    },
+    {
+      "epoch": 0.5679070559870311,
+      "grad_norm": 2.166841745376587,
+      "learning_rate": 4.053488240021615e-05,
+      "loss": 0.8791,
+      "step": 330000
+    },
+    {
+      "epoch": 0.568767521223375,
+      "grad_norm": 2.093510627746582,
+      "learning_rate": 4.052054131294375e-05,
+      "loss": 0.8687,
+      "step": 330500
+    },
+    {
+      "epoch": 0.569627986459719,
+      "grad_norm": 2.178293466567993,
+      "learning_rate": 4.050620022567135e-05,
+      "loss": 0.8676,
+      "step": 331000
+    },
+    {
+      "epoch": 0.570488451696063,
+      "grad_norm": 1.8907274007797241,
+      "learning_rate": 4.0491859138398955e-05,
+      "loss": 0.8666,
+      "step": 331500
+    },
+    {
+      "epoch": 0.571348916932407,
+      "grad_norm": 1.8355967998504639,
+      "learning_rate": 4.047751805112655e-05,
+      "loss": 0.8754,
+      "step": 332000
+    },
+    {
+      "epoch": 0.572209382168751,
+      "grad_norm": 2.001462936401367,
+      "learning_rate": 4.0463176963854156e-05,
+      "loss": 0.8737,
+      "step": 332500
+    },
+    {
+      "epoch": 0.573069847405095,
+      "grad_norm": 2.0783512592315674,
+      "learning_rate": 4.044883587658175e-05,
+      "loss": 0.8642,
+      "step": 333000
+    },
+    {
+      "epoch": 0.573930312641439,
+      "grad_norm": 2.1608786582946777,
+      "learning_rate": 4.043449478930935e-05,
+      "loss": 0.878,
+      "step": 333500
+    },
+    {
+      "epoch": 0.574790777877783,
+      "grad_norm": 2.18611216545105,
+      "learning_rate": 4.042015370203695e-05,
+      "loss": 0.8803,
+      "step": 334000
+    },
+    {
+      "epoch": 0.575651243114127,
+      "grad_norm": 2.0941479206085205,
+      "learning_rate": 4.040581261476455e-05,
+      "loss": 0.874,
+      "step": 334500
+    },
+    {
+      "epoch": 0.5765117083504709,
+      "grad_norm": 2.0728914737701416,
+      "learning_rate": 4.0391471527492155e-05,
+      "loss": 0.8729,
+      "step": 335000
+    },
+    {
+      "epoch": 0.5773721735868149,
+      "grad_norm": 5.831918716430664,
+      "learning_rate": 4.037713044021975e-05,
+      "loss": 0.8658,
+      "step": 335500
+    },
+    {
+      "epoch": 0.5782326388231589,
+      "grad_norm": 2.3052661418914795,
+      "learning_rate": 4.0362789352947356e-05,
+      "loss": 0.8712,
+      "step": 336000
+    },
+    {
+      "epoch": 0.5790931040595029,
+      "grad_norm": 1.9301319122314453,
+      "learning_rate": 4.034844826567495e-05,
+      "loss": 0.8811,
+      "step": 336500
+    },
+    {
+      "epoch": 0.5799535692958468,
+      "grad_norm": 2.0595436096191406,
+      "learning_rate": 4.033410717840255e-05,
+      "loss": 0.8686,
+      "step": 337000
+    },
+    {
+      "epoch": 0.5808140345321908,
+      "grad_norm": 2.0486223697662354,
+      "learning_rate": 4.0319766091130154e-05,
+      "loss": 0.8667,
+      "step": 337500
+    },
+    {
+      "epoch": 0.5816744997685348,
+      "grad_norm": 2.193305492401123,
+      "learning_rate": 4.030542500385776e-05,
+      "loss": 0.8702,
+      "step": 338000
+    },
+    {
+      "epoch": 0.5825349650048789,
+      "grad_norm": 1.9140911102294922,
+      "learning_rate": 4.0291083916585355e-05,
+      "loss": 0.8711,
+      "step": 338500
+    },
+    {
+      "epoch": 0.5833954302412229,
+      "grad_norm": 2.1117045879364014,
+      "learning_rate": 4.027674282931295e-05,
+      "loss": 0.8765,
+      "step": 339000
+    },
+    {
+      "epoch": 0.5842558954775668,
+      "grad_norm": 2.079667806625366,
+      "learning_rate": 4.0262401742040556e-05,
+      "loss": 0.8615,
+      "step": 339500
+    },
+    {
+      "epoch": 0.5851163607139108,
+      "grad_norm": 2.0438807010650635,
+      "learning_rate": 4.024806065476815e-05,
+      "loss": 0.8701,
+      "step": 340000
+    },
+    {
+      "epoch": 0.5859768259502548,
+      "grad_norm": 2.0004425048828125,
+      "learning_rate": 4.023371956749576e-05,
+      "loss": 0.8714,
+      "step": 340500
+    },
+    {
+      "epoch": 0.5868372911865988,
+      "grad_norm": 1.8124200105667114,
+      "learning_rate": 4.021937848022336e-05,
+      "loss": 0.869,
+      "step": 341000
+    },
+    {
+      "epoch": 0.5876977564229428,
+      "grad_norm": 2.239046812057495,
+      "learning_rate": 4.020503739295096e-05,
+      "loss": 0.8687,
+      "step": 341500
+    },
+    {
+      "epoch": 0.5885582216592867,
+      "grad_norm": 1.9475022554397583,
+      "learning_rate": 4.0190696305678555e-05,
+      "loss": 0.8727,
+      "step": 342000
+    },
+    {
+      "epoch": 0.5894186868956307,
+      "grad_norm": 2.0992038249969482,
+      "learning_rate": 4.017635521840616e-05,
+      "loss": 0.8747,
+      "step": 342500
+    },
+    {
+      "epoch": 0.5902791521319747,
+      "grad_norm": 1.9996228218078613,
+      "learning_rate": 4.0162014131133756e-05,
+      "loss": 0.8728,
+      "step": 343000
+    },
+    {
+      "epoch": 0.5911396173683187,
+      "grad_norm": 1.9434428215026855,
+      "learning_rate": 4.014767304386135e-05,
+      "loss": 0.8663,
+      "step": 343500
+    },
+    {
+      "epoch": 0.5920000826046626,
+      "grad_norm": 2.194171667098999,
+      "learning_rate": 4.013333195658896e-05,
+      "loss": 0.8769,
+      "step": 344000
+    },
+    {
+      "epoch": 0.5928605478410067,
+      "grad_norm": 2.1467361450195312,
+      "learning_rate": 4.011899086931656e-05,
+      "loss": 0.8749,
+      "step": 344500
+    },
+    {
+      "epoch": 0.5937210130773507,
+      "grad_norm": 2.1336467266082764,
+      "learning_rate": 4.010464978204416e-05,
+      "loss": 0.8694,
+      "step": 345000
+    },
+    {
+      "epoch": 0.5945814783136947,
+      "grad_norm": 2.085965394973755,
+      "learning_rate": 4.0090308694771755e-05,
+      "loss": 0.8696,
+      "step": 345500
+    },
+    {
+      "epoch": 0.5954419435500387,
+      "grad_norm": 2.0303428173065186,
+      "learning_rate": 4.007596760749936e-05,
+      "loss": 0.8708,
+      "step": 346000
+    },
+    {
+      "epoch": 0.5963024087863826,
+      "grad_norm": 2.0182602405548096,
+      "learning_rate": 4.0061626520226956e-05,
+      "loss": 0.8692,
+      "step": 346500
+    },
+    {
+      "epoch": 0.5971628740227266,
+      "grad_norm": 2.3182501792907715,
+      "learning_rate": 4.004728543295456e-05,
+      "loss": 0.8725,
+      "step": 347000
+    },
+    {
+      "epoch": 0.5980233392590706,
+      "grad_norm": 1.9483355283737183,
+      "learning_rate": 4.003294434568216e-05,
+      "loss": 0.8704,
+      "step": 347500
+    },
+    {
+      "epoch": 0.5988838044954146,
+      "grad_norm": 2.027533531188965,
+      "learning_rate": 4.001860325840976e-05,
+      "loss": 0.8646,
+      "step": 348000
+    },
+    {
+      "epoch": 0.5997442697317585,
+      "grad_norm": 2.079859733581543,
+      "learning_rate": 4.000426217113736e-05,
+      "loss": 0.8679,
+      "step": 348500
+    },
+    {
+      "epoch": 0.6006047349681025,
+      "grad_norm": 2.240821599960327,
+      "learning_rate": 3.998992108386496e-05,
+      "loss": 0.8701,
+      "step": 349000
+    },
+    {
+      "epoch": 0.6014652002044465,
+      "grad_norm": 1.9912835359573364,
+      "learning_rate": 3.997557999659256e-05,
+      "loss": 0.8675,
+      "step": 349500
+    },
+    {
+      "epoch": 0.6023256654407906,
+      "grad_norm": 2.0753517150878906,
+      "learning_rate": 3.9961238909320156e-05,
+      "loss": 0.8649,
+      "step": 350000
+    },
+    {
+      "epoch": 0.6031861306771346,
+      "grad_norm": 1.8598897457122803,
+      "learning_rate": 3.994689782204776e-05,
+      "loss": 0.8681,
+      "step": 350500
+    },
+    {
+      "epoch": 0.6040465959134785,
+      "grad_norm": 1.8939234018325806,
+      "learning_rate": 3.993255673477536e-05,
+      "loss": 0.8702,
+      "step": 351000
+    },
+    {
+      "epoch": 0.6049070611498225,
+      "grad_norm": 1.9680449962615967,
+      "learning_rate": 3.991821564750296e-05,
+      "loss": 0.8656,
+      "step": 351500
+    },
+    {
+      "epoch": 0.6057675263861665,
+      "grad_norm": 1.9677945375442505,
+      "learning_rate": 3.990387456023056e-05,
+      "loss": 0.8692,
+      "step": 352000
+    },
+    {
+      "epoch": 0.6066279916225105,
+      "grad_norm": 2.178060531616211,
+      "learning_rate": 3.988953347295816e-05,
+      "loss": 0.8652,
+      "step": 352500
+    },
+    {
+      "epoch": 0.6074884568588544,
+      "grad_norm": 2.035489082336426,
+      "learning_rate": 3.987519238568576e-05,
+      "loss": 0.8633,
+      "step": 353000
+    },
+    {
+      "epoch": 0.6083489220951984,
+      "grad_norm": 2.296475410461426,
+      "learning_rate": 3.986085129841336e-05,
+      "loss": 0.8695,
+      "step": 353500
+    },
+    {
+      "epoch": 0.6092093873315424,
+      "grad_norm": 2.06257963180542,
+      "learning_rate": 3.9846510211140966e-05,
+      "loss": 0.8701,
+      "step": 354000
+    },
+    {
+      "epoch": 0.6100698525678864,
+      "grad_norm": 1.7742199897766113,
+      "learning_rate": 3.983216912386856e-05,
+      "loss": 0.8653,
+      "step": 354500
+    },
+    {
+      "epoch": 0.6109303178042303,
+      "grad_norm": 2.1947309970855713,
+      "learning_rate": 3.981782803659616e-05,
+      "loss": 0.8722,
+      "step": 355000
+    },
+    {
+      "epoch": 0.6117907830405743,
+      "grad_norm": 1.9710649251937866,
+      "learning_rate": 3.9803486949323764e-05,
+      "loss": 0.8631,
+      "step": 355500
+    },
+    {
+      "epoch": 0.6126512482769184,
+      "grad_norm": 2.3806471824645996,
+      "learning_rate": 3.978914586205136e-05,
+      "loss": 0.8651,
+      "step": 356000
+    },
+    {
+      "epoch": 0.6135117135132624,
+      "grad_norm": 2.1118831634521484,
+      "learning_rate": 3.9774804774778965e-05,
+      "loss": 0.8688,
+      "step": 356500
+    },
+    {
+      "epoch": 0.6143721787496064,
+      "grad_norm": 2.0460760593414307,
+      "learning_rate": 3.976046368750656e-05,
+      "loss": 0.865,
+      "step": 357000
+    },
+    {
+      "epoch": 0.6152326439859503,
+      "grad_norm": 1.9501177072525024,
+      "learning_rate": 3.9746122600234166e-05,
+      "loss": 0.863,
+      "step": 357500
+    },
+    {
+      "epoch": 0.6160931092222943,
+      "grad_norm": 2.000192165374756,
+      "learning_rate": 3.973178151296176e-05,
+      "loss": 0.8625,
+      "step": 358000
+    },
+    {
+      "epoch": 0.6169535744586383,
+      "grad_norm": 2.1251659393310547,
+      "learning_rate": 3.971744042568936e-05,
+      "loss": 0.8684,
+      "step": 358500
+    },
+    {
+      "epoch": 0.6178140396949823,
+      "grad_norm": 2.025874376296997,
+      "learning_rate": 3.9703099338416964e-05,
+      "loss": 0.8685,
+      "step": 359000
+    },
+    {
+      "epoch": 0.6186745049313263,
+      "grad_norm": 1.909056544303894,
+      "learning_rate": 3.968875825114456e-05,
+      "loss": 0.8639,
+      "step": 359500
+    },
+    {
+      "epoch": 0.6195349701676702,
+      "grad_norm": 1.8679569959640503,
+      "learning_rate": 3.9674417163872165e-05,
+      "loss": 0.865,
+      "step": 360000
+    },
+    {
+      "epoch": 0.6203954354040142,
+      "grad_norm": 1.9933589696884155,
+      "learning_rate": 3.966007607659977e-05,
+      "loss": 0.8676,
+      "step": 360500
+    },
+    {
+      "epoch": 0.6212559006403582,
+      "grad_norm": 2.170215129852295,
+      "learning_rate": 3.9645734989327366e-05,
+      "loss": 0.8652,
+      "step": 361000
+    },
+    {
+      "epoch": 0.6221163658767023,
+      "grad_norm": 2.0594518184661865,
+      "learning_rate": 3.963139390205496e-05,
+      "loss": 0.8664,
+      "step": 361500
+    },
+    {
+      "epoch": 0.6229768311130462,
+      "grad_norm": 2.1189212799072266,
+      "learning_rate": 3.961705281478256e-05,
+      "loss": 0.861,
+      "step": 362000
+    },
+    {
+      "epoch": 0.6238372963493902,
+      "grad_norm": 2.0330159664154053,
+      "learning_rate": 3.9602711727510164e-05,
+      "loss": 0.8609,
+      "step": 362500
+    },
+    {
+      "epoch": 0.6246977615857342,
+      "grad_norm": 2.09208345413208,
+      "learning_rate": 3.958837064023777e-05,
+      "loss": 0.8688,
+      "step": 363000
+    },
+    {
+      "epoch": 0.6255582268220782,
+      "grad_norm": 2.081916570663452,
+      "learning_rate": 3.9574029552965365e-05,
+      "loss": 0.8615,
+      "step": 363500
+    },
+    {
+      "epoch": 0.6264186920584222,
+      "grad_norm": 2.043161630630493,
+      "learning_rate": 3.955968846569297e-05,
+      "loss": 0.8625,
+      "step": 364000
+    },
+    {
+      "epoch": 0.6272791572947661,
+      "grad_norm": 2.077566146850586,
+      "learning_rate": 3.9545347378420566e-05,
+      "loss": 0.865,
+      "step": 364500
+    },
+    {
+      "epoch": 0.6281396225311101,
+      "grad_norm": 1.8295567035675049,
+      "learning_rate": 3.953100629114816e-05,
+      "loss": 0.8613,
+      "step": 365000
+    },
+    {
+      "epoch": 0.6290000877674541,
+      "grad_norm": 2.1657068729400635,
+      "learning_rate": 3.951666520387577e-05,
+      "loss": 0.8603,
+      "step": 365500
+    },
+    {
+      "epoch": 0.6298605530037981,
+      "grad_norm": 2.0411875247955322,
+      "learning_rate": 3.950232411660337e-05,
+      "loss": 0.8657,
+      "step": 366000
+    },
+    {
+      "epoch": 0.630721018240142,
+      "grad_norm": 2.146219491958618,
+      "learning_rate": 3.948798302933097e-05,
+      "loss": 0.8648,
+      "step": 366500
+    },
+    {
+      "epoch": 0.631581483476486,
+      "grad_norm": 2.2003700733184814,
+      "learning_rate": 3.947364194205857e-05,
+      "loss": 0.8626,
+      "step": 367000
+    },
+    {
+      "epoch": 0.6324419487128301,
+      "grad_norm": 2.0370073318481445,
+      "learning_rate": 3.945930085478617e-05,
+      "loss": 0.8623,
+      "step": 367500
+    },
+    {
+      "epoch": 0.6333024139491741,
+      "grad_norm": 2.0403857231140137,
+      "learning_rate": 3.9444959767513766e-05,
+      "loss": 0.8634,
+      "step": 368000
+    },
+    {
+      "epoch": 0.6341628791855181,
+      "grad_norm": 2.0806796550750732,
+      "learning_rate": 3.943061868024136e-05,
+      "loss": 0.8657,
+      "step": 368500
+    },
+    {
+      "epoch": 0.635023344421862,
+      "grad_norm": 2.0758426189422607,
+      "learning_rate": 3.941627759296897e-05,
+      "loss": 0.8672,
+      "step": 369000
+    },
+    {
+      "epoch": 0.635883809658206,
+      "grad_norm": 1.9031485319137573,
+      "learning_rate": 3.940193650569657e-05,
+      "loss": 0.862,
+      "step": 369500
+    },
+    {
+      "epoch": 0.63674427489455,
+      "grad_norm": 1.9357956647872925,
+      "learning_rate": 3.938759541842417e-05,
+      "loss": 0.8659,
+      "step": 370000
+    },
+    {
+      "epoch": 0.637604740130894,
+      "grad_norm": 1.8705717325210571,
+      "learning_rate": 3.937325433115177e-05,
+      "loss": 0.8637,
+      "step": 370500
+    },
+    {
+      "epoch": 0.6384652053672379,
+      "grad_norm": 1.963387370109558,
+      "learning_rate": 3.935891324387937e-05,
+      "loss": 0.8609,
+      "step": 371000
+    },
+    {
+      "epoch": 0.6393256706035819,
+      "grad_norm": 2.1126718521118164,
+      "learning_rate": 3.9344572156606966e-05,
+      "loss": 0.867,
+      "step": 371500
+    },
+    {
+      "epoch": 0.6401861358399259,
+      "grad_norm": 1.9715133905410767,
+      "learning_rate": 3.933023106933457e-05,
+      "loss": 0.8633,
+      "step": 372000
+    },
+    {
+      "epoch": 0.6410466010762699,
+      "grad_norm": 2.0479183197021484,
+      "learning_rate": 3.931588998206217e-05,
+      "loss": 0.8678,
+      "step": 372500
+    },
+    {
+      "epoch": 0.6419070663126138,
+      "grad_norm": 2.021559953689575,
+      "learning_rate": 3.930154889478977e-05,
+      "loss": 0.8661,
+      "step": 373000
+    },
+    {
+      "epoch": 0.6427675315489579,
+      "grad_norm": 2.037088394165039,
+      "learning_rate": 3.928720780751737e-05,
+      "loss": 0.8595,
+      "step": 373500
+    },
+    {
+      "epoch": 0.6436279967853019,
+      "grad_norm": 1.8662008047103882,
+      "learning_rate": 3.927286672024497e-05,
+      "loss": 0.859,
+      "step": 374000
+    },
+    {
+      "epoch": 0.6444884620216459,
+      "grad_norm": 2.037175178527832,
+      "learning_rate": 3.925852563297257e-05,
+      "loss": 0.865,
+      "step": 374500
+    },
+    {
+      "epoch": 0.6453489272579899,
+      "grad_norm": 1.90609872341156,
+      "learning_rate": 3.924418454570017e-05,
+      "loss": 0.858,
+      "step": 375000
+    },
+    {
+      "epoch": 0.6462093924943338,
+      "grad_norm": 1.9687128067016602,
+      "learning_rate": 3.9229843458427776e-05,
+      "loss": 0.8596,
+      "step": 375500
+    },
+    {
+      "epoch": 0.6470698577306778,
+      "grad_norm": 1.905707597732544,
+      "learning_rate": 3.921550237115537e-05,
+      "loss": 0.8594,
+      "step": 376000
+    },
+    {
+      "epoch": 0.6479303229670218,
+      "grad_norm": 1.876888394355774,
+      "learning_rate": 3.920116128388297e-05,
+      "loss": 0.8612,
+      "step": 376500
+    },
+    {
+      "epoch": 0.6487907882033658,
+      "grad_norm": 2.0150208473205566,
+      "learning_rate": 3.9186820196610574e-05,
+      "loss": 0.8543,
+      "step": 377000
+    },
+    {
+      "epoch": 0.6496512534397098,
+      "grad_norm": 1.8568073511123657,
+      "learning_rate": 3.917247910933817e-05,
+      "loss": 0.8634,
+      "step": 377500
+    },
+    {
+      "epoch": 0.6505117186760537,
+      "grad_norm": 1.9990817308425903,
+      "learning_rate": 3.915813802206577e-05,
+      "loss": 0.8584,
+      "step": 378000
+    },
+    {
+      "epoch": 0.6513721839123977,
+      "grad_norm": 2.0287017822265625,
+      "learning_rate": 3.914379693479337e-05,
+      "loss": 0.8644,
+      "step": 378500
+    },
+    {
+      "epoch": 0.6522326491487418,
+      "grad_norm": 1.886164665222168,
+      "learning_rate": 3.9129455847520976e-05,
+      "loss": 0.8662,
+      "step": 379000
+    },
+    {
+      "epoch": 0.6530931143850858,
+      "grad_norm": 2.0619380474090576,
+      "learning_rate": 3.911511476024857e-05,
+      "loss": 0.8609,
+      "step": 379500
+    },
+    {
+      "epoch": 0.6539535796214297,
+      "grad_norm": 2.1117541790008545,
+      "learning_rate": 3.910077367297617e-05,
+      "loss": 0.8679,
+      "step": 380000
+    },
+    {
+      "epoch": 0.6548140448577737,
+      "grad_norm": 2.0683281421661377,
+      "learning_rate": 3.9086432585703774e-05,
+      "loss": 0.8558,
+      "step": 380500
+    },
+    {
+      "epoch": 0.6556745100941177,
+      "grad_norm": 2.209404706954956,
+      "learning_rate": 3.907209149843137e-05,
+      "loss": 0.8634,
+      "step": 381000
+    },
+    {
+      "epoch": 0.6565349753304617,
+      "grad_norm": 2.2918388843536377,
+      "learning_rate": 3.9057750411158975e-05,
+      "loss": 0.856,
+      "step": 381500
+    },
+    {
+      "epoch": 0.6573954405668057,
+      "grad_norm": 1.9002126455307007,
+      "learning_rate": 3.904340932388658e-05,
+      "loss": 0.8663,
+      "step": 382000
+    },
+    {
+      "epoch": 0.6582559058031496,
+      "grad_norm": 2.235064744949341,
+      "learning_rate": 3.9029068236614176e-05,
+      "loss": 0.8601,
+      "step": 382500
+    },
+    {
+      "epoch": 0.6591163710394936,
+      "grad_norm": 2.0584187507629395,
+      "learning_rate": 3.901472714934177e-05,
+      "loss": 0.8635,
+      "step": 383000
+    },
+    {
+      "epoch": 0.6599768362758376,
+      "grad_norm": 2.0273942947387695,
+      "learning_rate": 3.900038606206938e-05,
+      "loss": 0.8618,
+      "step": 383500
+    },
+    {
+      "epoch": 0.6608373015121816,
+      "grad_norm": 1.9882392883300781,
+      "learning_rate": 3.8986044974796974e-05,
+      "loss": 0.8552,
+      "step": 384000
+    },
+    {
+      "epoch": 0.6616977667485255,
+      "grad_norm": 1.9432196617126465,
+      "learning_rate": 3.897170388752457e-05,
+      "loss": 0.8597,
+      "step": 384500
+    },
+    {
+      "epoch": 0.6625582319848696,
+      "grad_norm": 2.048971652984619,
+      "learning_rate": 3.8957362800252175e-05,
+      "loss": 0.8623,
+      "step": 385000
+    },
+    {
+      "epoch": 0.6634186972212136,
+      "grad_norm": 2.0021302700042725,
+      "learning_rate": 3.894302171297978e-05,
+      "loss": 0.8611,
+      "step": 385500
+    },
+    {
+      "epoch": 0.6642791624575576,
+      "grad_norm": 2.2216758728027344,
+      "learning_rate": 3.8928680625707376e-05,
+      "loss": 0.8574,
+      "step": 386000
+    },
+    {
+      "epoch": 0.6651396276939016,
+      "grad_norm": 1.8725037574768066,
+      "learning_rate": 3.891433953843497e-05,
+      "loss": 0.8555,
+      "step": 386500
+    },
+    {
+      "epoch": 0.6660000929302455,
+      "grad_norm": 1.8659377098083496,
+      "learning_rate": 3.889999845116258e-05,
+      "loss": 0.8647,
+      "step": 387000
+    },
+    {
+      "epoch": 0.6668605581665895,
+      "grad_norm": 2.034635066986084,
+      "learning_rate": 3.8885657363890174e-05,
+      "loss": 0.8603,
+      "step": 387500
+    },
+    {
+      "epoch": 0.6677210234029335,
+      "grad_norm": 1.9781776666641235,
+      "learning_rate": 3.887131627661778e-05,
+      "loss": 0.8512,
+      "step": 388000
+    },
+    {
+      "epoch": 0.6685814886392775,
+      "grad_norm": 2.0884103775024414,
+      "learning_rate": 3.885697518934538e-05,
+      "loss": 0.8653,
+      "step": 388500
+    },
+    {
+      "epoch": 0.6694419538756214,
+      "grad_norm": 2.025995969772339,
+      "learning_rate": 3.884263410207298e-05,
+      "loss": 0.8617,
+      "step": 389000
+    },
+    {
+      "epoch": 0.6703024191119654,
+      "grad_norm": 2.181551456451416,
+      "learning_rate": 3.8828293014800576e-05,
+      "loss": 0.8572,
+      "step": 389500
+    },
+    {
+      "epoch": 0.6711628843483094,
+      "grad_norm": 1.9846434593200684,
+      "learning_rate": 3.881395192752818e-05,
+      "loss": 0.8558,
+      "step": 390000
+    },
+    {
+      "epoch": 0.6720233495846535,
+      "grad_norm": 2.3479115962982178,
+      "learning_rate": 3.879961084025578e-05,
+      "loss": 0.8532,
+      "step": 390500
+    },
+    {
+      "epoch": 0.6728838148209975,
+      "grad_norm": 1.9880118370056152,
+      "learning_rate": 3.878526975298338e-05,
+      "loss": 0.862,
+      "step": 391000
+    },
+    {
+      "epoch": 0.6737442800573414,
+      "grad_norm": 2.1184558868408203,
+      "learning_rate": 3.877092866571098e-05,
+      "loss": 0.8632,
+      "step": 391500
+    },
+    {
+      "epoch": 0.6746047452936854,
+      "grad_norm": 2.0495119094848633,
+      "learning_rate": 3.875658757843858e-05,
+      "loss": 0.8592,
+      "step": 392000
+    },
+    {
+      "epoch": 0.6754652105300294,
+      "grad_norm": 1.9544061422348022,
+      "learning_rate": 3.874224649116618e-05,
+      "loss": 0.8583,
+      "step": 392500
+    },
+    {
+      "epoch": 0.6763256757663734,
+      "grad_norm": 1.8652530908584595,
+      "learning_rate": 3.8727905403893776e-05,
+      "loss": 0.8523,
+      "step": 393000
+    },
+    {
+      "epoch": 0.6771861410027173,
+      "grad_norm": 1.9426854848861694,
+      "learning_rate": 3.871356431662138e-05,
+      "loss": 0.8517,
+      "step": 393500
+    },
+    {
+      "epoch": 0.6780466062390613,
+      "grad_norm": 1.9407514333724976,
+      "learning_rate": 3.8699223229348977e-05,
+      "loss": 0.8629,
+      "step": 394000
+    },
+    {
+      "epoch": 0.6789070714754053,
+      "grad_norm": 2.0087180137634277,
+      "learning_rate": 3.868488214207658e-05,
+      "loss": 0.8581,
+      "step": 394500
+    },
+    {
+      "epoch": 0.6797675367117493,
+      "grad_norm": 2.013326644897461,
+      "learning_rate": 3.8670541054804184e-05,
+      "loss": 0.859,
+      "step": 395000
+    },
+    {
+      "epoch": 0.6806280019480933,
+      "grad_norm": 2.0862598419189453,
+      "learning_rate": 3.865619996753178e-05,
+      "loss": 0.8652,
+      "step": 395500
+    },
+    {
+      "epoch": 0.6814884671844372,
+      "grad_norm": 2.002089023590088,
+      "learning_rate": 3.864185888025938e-05,
+      "loss": 0.851,
+      "step": 396000
+    },
+    {
+      "epoch": 0.6823489324207813,
+      "grad_norm": 2.0043892860412598,
+      "learning_rate": 3.8627517792986976e-05,
+      "loss": 0.8596,
+      "step": 396500
+    },
+    {
+      "epoch": 0.6832093976571253,
+      "grad_norm": 2.03305721282959,
+      "learning_rate": 3.861317670571458e-05,
+      "loss": 0.8565,
+      "step": 397000
+    },
+    {
+      "epoch": 0.6840698628934693,
+      "grad_norm": 2.194655418395996,
+      "learning_rate": 3.859883561844218e-05,
+      "loss": 0.8539,
+      "step": 397500
+    },
+    {
+      "epoch": 0.6849303281298132,
+      "grad_norm": 2.1509432792663574,
+      "learning_rate": 3.858449453116978e-05,
+      "loss": 0.8588,
+      "step": 398000
+    },
+    {
+      "epoch": 0.6857907933661572,
+      "grad_norm": 2.0242624282836914,
+      "learning_rate": 3.8570153443897384e-05,
+      "loss": 0.8528,
+      "step": 398500
+    },
+    {
+      "epoch": 0.6866512586025012,
+      "grad_norm": 2.357621908187866,
+      "learning_rate": 3.855581235662498e-05,
+      "loss": 0.8556,
+      "step": 399000
+    },
+    {
+      "epoch": 0.6875117238388452,
+      "grad_norm": 4.633720874786377,
+      "learning_rate": 3.854147126935258e-05,
+      "loss": 0.8514,
+      "step": 399500
+    },
+    {
+      "epoch": 0.6883721890751892,
+      "grad_norm": 2.0474085807800293,
+      "learning_rate": 3.852713018208018e-05,
+      "loss": 0.8552,
+      "step": 400000
+    },
+    {
+      "epoch": 0.6892326543115331,
+      "grad_norm": 2.1123034954071045,
+      "learning_rate": 3.8512789094807786e-05,
+      "loss": 0.8535,
+      "step": 400500
+    },
+    {
+      "epoch": 0.6900931195478771,
+      "grad_norm": 2.002837657928467,
+      "learning_rate": 3.849844800753538e-05,
+      "loss": 0.8545,
+      "step": 401000
+    },
+    {
+      "epoch": 0.6909535847842211,
+      "grad_norm": 1.9797070026397705,
+      "learning_rate": 3.848410692026299e-05,
+      "loss": 0.8508,
+      "step": 401500
+    },
+    {
+      "epoch": 0.6918140500205652,
+      "grad_norm": 2.1309895515441895,
+      "learning_rate": 3.8469765832990584e-05,
+      "loss": 0.8537,
+      "step": 402000
+    },
+    {
+      "epoch": 0.6926745152569092,
+      "grad_norm": 1.9080448150634766,
+      "learning_rate": 3.845542474571818e-05,
+      "loss": 0.855,
+      "step": 402500
+    },
+    {
+      "epoch": 0.6935349804932531,
+      "grad_norm": 2.081582546234131,
+      "learning_rate": 3.844108365844578e-05,
+      "loss": 0.8513,
+      "step": 403000
+    },
+    {
+      "epoch": 0.6943954457295971,
+      "grad_norm": 1.9564839601516724,
+      "learning_rate": 3.842674257117338e-05,
+      "loss": 0.851,
+      "step": 403500
+    },
+    {
+      "epoch": 0.6952559109659411,
+      "grad_norm": 1.9080193042755127,
+      "learning_rate": 3.8412401483900986e-05,
+      "loss": 0.8528,
+      "step": 404000
+    },
+    {
+      "epoch": 0.6961163762022851,
+      "grad_norm": 1.936774730682373,
+      "learning_rate": 3.839806039662858e-05,
+      "loss": 0.8536,
+      "step": 404500
+    },
+    {
+      "epoch": 0.696976841438629,
+      "grad_norm": 2.0372822284698486,
+      "learning_rate": 3.838371930935619e-05,
+      "loss": 0.8485,
+      "step": 405000
+    },
+    {
+      "epoch": 0.697837306674973,
+      "grad_norm": 1.938328504562378,
+      "learning_rate": 3.8369378222083784e-05,
+      "loss": 0.8537,
+      "step": 405500
+    },
+    {
+      "epoch": 0.698697771911317,
+      "grad_norm": 2.093485116958618,
+      "learning_rate": 3.835503713481138e-05,
+      "loss": 0.8591,
+      "step": 406000
+    },
+    {
+      "epoch": 0.699558237147661,
+      "grad_norm": 1.9015600681304932,
+      "learning_rate": 3.8340696047538985e-05,
+      "loss": 0.8559,
+      "step": 406500
+    },
+    {
+      "epoch": 0.7004187023840049,
+      "grad_norm": 2.13163685798645,
+      "learning_rate": 3.832635496026659e-05,
+      "loss": 0.853,
+      "step": 407000
+    },
+    {
+      "epoch": 0.7012791676203489,
+      "grad_norm": 2.031601667404175,
+      "learning_rate": 3.8312013872994186e-05,
+      "loss": 0.8601,
+      "step": 407500
+    },
+    {
+      "epoch": 0.702139632856693,
+      "grad_norm": 2.0390450954437256,
+      "learning_rate": 3.829767278572178e-05,
+      "loss": 0.8499,
+      "step": 408000
+    },
+    {
+      "epoch": 0.703000098093037,
+      "grad_norm": 2.225257396697998,
+      "learning_rate": 3.828333169844939e-05,
+      "loss": 0.8594,
+      "step": 408500
+    },
+    {
+      "epoch": 0.703860563329381,
+      "grad_norm": 1.9891399145126343,
+      "learning_rate": 3.8268990611176984e-05,
+      "loss": 0.861,
+      "step": 409000
+    },
+    {
+      "epoch": 0.7047210285657249,
+      "grad_norm": 2.017092227935791,
+      "learning_rate": 3.825464952390458e-05,
+      "loss": 0.8531,
+      "step": 409500
+    },
+    {
+      "epoch": 0.7055814938020689,
+      "grad_norm": 2.091958999633789,
+      "learning_rate": 3.824030843663219e-05,
+      "loss": 0.8517,
+      "step": 410000
+    },
+    {
+      "epoch": 0.7064419590384129,
+      "grad_norm": 1.9318962097167969,
+      "learning_rate": 3.822596734935979e-05,
+      "loss": 0.8515,
+      "step": 410500
+    },
+    {
+      "epoch": 0.7073024242747569,
+      "grad_norm": 2.068328619003296,
+      "learning_rate": 3.8211626262087386e-05,
+      "loss": 0.8513,
+      "step": 411000
+    },
+    {
+      "epoch": 0.7081628895111008,
+      "grad_norm": 2.139561414718628,
+      "learning_rate": 3.819728517481499e-05,
+      "loss": 0.859,
+      "step": 411500
+    },
+    {
+      "epoch": 0.7090233547474448,
+      "grad_norm": 1.9060418605804443,
+      "learning_rate": 3.818294408754259e-05,
+      "loss": 0.8536,
+      "step": 412000
+    },
+    {
+      "epoch": 0.7098838199837888,
+      "grad_norm": 2.385798692703247,
+      "learning_rate": 3.8168603000270184e-05,
+      "loss": 0.8498,
+      "step": 412500
+    },
+    {
+      "epoch": 0.7107442852201328,
+      "grad_norm": 2.1265676021575928,
+      "learning_rate": 3.815426191299779e-05,
+      "loss": 0.8501,
+      "step": 413000
+    },
+    {
+      "epoch": 0.7116047504564768,
+      "grad_norm": 2.1203105449676514,
+      "learning_rate": 3.813992082572539e-05,
+      "loss": 0.8481,
+      "step": 413500
+    },
+    {
+      "epoch": 0.7124652156928208,
+      "grad_norm": 2.1768009662628174,
+      "learning_rate": 3.812557973845299e-05,
+      "loss": 0.8533,
+      "step": 414000
+    },
+    {
+      "epoch": 0.7133256809291648,
+      "grad_norm": 2.061535120010376,
+      "learning_rate": 3.8111238651180586e-05,
+      "loss": 0.8532,
+      "step": 414500
+    },
+    {
+      "epoch": 0.7141861461655088,
+      "grad_norm": 1.9680122137069702,
+      "learning_rate": 3.809689756390819e-05,
+      "loss": 0.8541,
+      "step": 415000
+    },
+    {
+      "epoch": 0.7150466114018528,
+      "grad_norm": 1.9438929557800293,
+      "learning_rate": 3.808255647663579e-05,
+      "loss": 0.8474,
+      "step": 415500
+    },
+    {
+      "epoch": 0.7159070766381967,
+      "grad_norm": 2.017174482345581,
+      "learning_rate": 3.806821538936339e-05,
+      "loss": 0.8526,
+      "step": 416000
+    },
+    {
+      "epoch": 0.7167675418745407,
+      "grad_norm": 2.2368977069854736,
+      "learning_rate": 3.8053874302090994e-05,
+      "loss": 0.8616,
+      "step": 416500
+    },
+    {
+      "epoch": 0.7176280071108847,
+      "grad_norm": 1.999829649925232,
+      "learning_rate": 3.803953321481859e-05,
+      "loss": 0.8525,
+      "step": 417000
+    },
+    {
+      "epoch": 0.7184884723472287,
+      "grad_norm": 2.337951183319092,
+      "learning_rate": 3.802519212754619e-05,
+      "loss": 0.8609,
+      "step": 417500
+    },
+    {
+      "epoch": 0.7193489375835727,
+      "grad_norm": 2.295762300491333,
+      "learning_rate": 3.801085104027379e-05,
+      "loss": 0.852,
+      "step": 418000
+    },
+    {
+      "epoch": 0.7202094028199166,
+      "grad_norm": 2.2162396907806396,
+      "learning_rate": 3.799650995300139e-05,
+      "loss": 0.8486,
+      "step": 418500
+    },
+    {
+      "epoch": 0.7210698680562606,
+      "grad_norm": 2.156071424484253,
+      "learning_rate": 3.7982168865728987e-05,
+      "loss": 0.8534,
+      "step": 419000
+    },
+    {
+      "epoch": 0.7219303332926047,
+      "grad_norm": 2.052499532699585,
+      "learning_rate": 3.796782777845659e-05,
+      "loss": 0.8518,
+      "step": 419500
+    },
+    {
+      "epoch": 0.7227907985289487,
+      "grad_norm": 2.0268807411193848,
+      "learning_rate": 3.7953486691184194e-05,
+      "loss": 0.8492,
+      "step": 420000
+    },
+    {
+      "epoch": 0.7236512637652927,
+      "grad_norm": 2.169954776763916,
+      "learning_rate": 3.793914560391179e-05,
+      "loss": 0.8579,
+      "step": 420500
+    },
+    {
+      "epoch": 0.7245117290016366,
+      "grad_norm": 2.2587361335754395,
+      "learning_rate": 3.792480451663939e-05,
+      "loss": 0.8553,
+      "step": 421000
+    },
+    {
+      "epoch": 0.7253721942379806,
+      "grad_norm": 1.8236783742904663,
+      "learning_rate": 3.791046342936699e-05,
+      "loss": 0.8448,
+      "step": 421500
+    },
+    {
+      "epoch": 0.7262326594743246,
+      "grad_norm": 2.1673099994659424,
+      "learning_rate": 3.789612234209459e-05,
+      "loss": 0.8438,
+      "step": 422000
+    },
+    {
+      "epoch": 0.7270931247106686,
+      "grad_norm": 2.214162588119507,
+      "learning_rate": 3.788178125482219e-05,
+      "loss": 0.8468,
+      "step": 422500
+    },
+    {
+      "epoch": 0.7279535899470125,
+      "grad_norm": 2.025665760040283,
+      "learning_rate": 3.78674401675498e-05,
+      "loss": 0.8503,
+      "step": 423000
+    },
+    {
+      "epoch": 0.7288140551833565,
+      "grad_norm": 1.8908655643463135,
+      "learning_rate": 3.7853099080277394e-05,
+      "loss": 0.8546,
+      "step": 423500
+    },
+    {
+      "epoch": 0.7296745204197005,
+      "grad_norm": 2.1749110221862793,
+      "learning_rate": 3.783875799300499e-05,
+      "loss": 0.8586,
+      "step": 424000
+    },
+    {
+      "epoch": 0.7305349856560445,
+      "grad_norm": 2.0959668159484863,
+      "learning_rate": 3.7824416905732595e-05,
+      "loss": 0.8471,
+      "step": 424500
+    },
+    {
+      "epoch": 0.7313954508923884,
+      "grad_norm": 2.0514888763427734,
+      "learning_rate": 3.781007581846019e-05,
+      "loss": 0.8529,
+      "step": 425000
+    },
+    {
+      "epoch": 0.7322559161287325,
+      "grad_norm": 2.1552107334136963,
+      "learning_rate": 3.7795734731187796e-05,
+      "loss": 0.8497,
+      "step": 425500
+    },
+    {
+      "epoch": 0.7331163813650765,
+      "grad_norm": 2.833772659301758,
+      "learning_rate": 3.778139364391539e-05,
+      "loss": 0.8492,
+      "step": 426000
+    },
+    {
+      "epoch": 0.7339768466014205,
+      "grad_norm": 2.154860019683838,
+      "learning_rate": 3.7767052556643e-05,
+      "loss": 0.8512,
+      "step": 426500
+    },
+    {
+      "epoch": 0.7348373118377645,
+      "grad_norm": 2.1189260482788086,
+      "learning_rate": 3.7752711469370594e-05,
+      "loss": 0.8455,
+      "step": 427000
+    },
+    {
+      "epoch": 0.7356977770741084,
+      "grad_norm": 1.8865413665771484,
+      "learning_rate": 3.773837038209819e-05,
+      "loss": 0.8556,
+      "step": 427500
+    },
+    {
+      "epoch": 0.7365582423104524,
+      "grad_norm": 2.020766258239746,
+      "learning_rate": 3.7724029294825795e-05,
+      "loss": 0.8427,
+      "step": 428000
+    },
+    {
+      "epoch": 0.7374187075467964,
+      "grad_norm": 2.0161490440368652,
+      "learning_rate": 3.770968820755339e-05,
+      "loss": 0.8517,
+      "step": 428500
+    },
+    {
+      "epoch": 0.7382791727831404,
+      "grad_norm": 1.9801963567733765,
+      "learning_rate": 3.7695347120280996e-05,
+      "loss": 0.8501,
+      "step": 429000
+    },
+    {
+      "epoch": 0.7391396380194843,
+      "grad_norm": 2.0916380882263184,
+      "learning_rate": 3.76810060330086e-05,
+      "loss": 0.8462,
+      "step": 429500
+    },
+    {
+      "epoch": 0.7400001032558283,
+      "grad_norm": 2.019502639770508,
+      "learning_rate": 3.76666649457362e-05,
+      "loss": 0.8532,
+      "step": 430000
+    },
+    {
+      "epoch": 0.7408605684921723,
+      "grad_norm": 1.979933500289917,
+      "learning_rate": 3.7652323858463794e-05,
+      "loss": 0.8473,
+      "step": 430500
+    },
+    {
+      "epoch": 0.7417210337285164,
+      "grad_norm": 2.0010876655578613,
+      "learning_rate": 3.763798277119139e-05,
+      "loss": 0.8468,
+      "step": 431000
+    },
+    {
+      "epoch": 0.7425814989648604,
+      "grad_norm": 2.2110819816589355,
+      "learning_rate": 3.7623641683918995e-05,
+      "loss": 0.8453,
+      "step": 431500
+    },
+    {
+      "epoch": 0.7434419642012043,
+      "grad_norm": 2.0687079429626465,
+      "learning_rate": 3.76093005966466e-05,
+      "loss": 0.8522,
+      "step": 432000
+    },
+    {
+      "epoch": 0.7443024294375483,
+      "grad_norm": 1.9159475564956665,
+      "learning_rate": 3.7594959509374196e-05,
+      "loss": 0.8488,
+      "step": 432500
+    },
+    {
+      "epoch": 0.7451628946738923,
+      "grad_norm": 2.194312572479248,
+      "learning_rate": 3.75806184221018e-05,
+      "loss": 0.8513,
+      "step": 433000
+    },
+    {
+      "epoch": 0.7460233599102363,
+      "grad_norm": 1.8975337743759155,
+      "learning_rate": 3.75662773348294e-05,
+      "loss": 0.8454,
+      "step": 433500
+    },
+    {
+      "epoch": 0.7468838251465802,
+      "grad_norm": 2.1459665298461914,
+      "learning_rate": 3.7551936247556994e-05,
+      "loss": 0.8496,
+      "step": 434000
+    },
+    {
+      "epoch": 0.7477442903829242,
+      "grad_norm": 1.8642619848251343,
+      "learning_rate": 3.75375951602846e-05,
+      "loss": 0.8446,
+      "step": 434500
+    },
+    {
+      "epoch": 0.7486047556192682,
+      "grad_norm": 2.064448356628418,
+      "learning_rate": 3.75232540730122e-05,
+      "loss": 0.8488,
+      "step": 435000
+    },
+    {
+      "epoch": 0.7494652208556122,
+      "grad_norm": 2.065342903137207,
+      "learning_rate": 3.75089129857398e-05,
+      "loss": 0.8493,
+      "step": 435500
+    },
+    {
+      "epoch": 0.7503256860919562,
+      "grad_norm": 2.1863017082214355,
+      "learning_rate": 3.74945718984674e-05,
+      "loss": 0.8436,
+      "step": 436000
+    },
+    {
+      "epoch": 0.7511861513283001,
+      "grad_norm": 1.9626061916351318,
+      "learning_rate": 3.7480230811195e-05,
+      "loss": 0.8505,
+      "step": 436500
+    },
+    {
+      "epoch": 0.7520466165646442,
+      "grad_norm": 1.9098788499832153,
+      "learning_rate": 3.74658897239226e-05,
+      "loss": 0.8514,
+      "step": 437000
+    },
+    {
+      "epoch": 0.7529070818009882,
+      "grad_norm": 1.9859448671340942,
+      "learning_rate": 3.7451548636650194e-05,
+      "loss": 0.8524,
+      "step": 437500
+    },
+    {
+      "epoch": 0.7537675470373322,
+      "grad_norm": 1.9206920862197876,
+      "learning_rate": 3.74372075493778e-05,
+      "loss": 0.8515,
+      "step": 438000
+    },
+    {
+      "epoch": 0.7546280122736762,
+      "grad_norm": 1.8871228694915771,
+      "learning_rate": 3.74228664621054e-05,
+      "loss": 0.8504,
+      "step": 438500
+    },
+    {
+      "epoch": 0.7554884775100201,
+      "grad_norm": 1.9660274982452393,
+      "learning_rate": 3.7408525374833e-05,
+      "loss": 0.8548,
+      "step": 439000
+    },
+    {
+      "epoch": 0.7563489427463641,
+      "grad_norm": 2.231797456741333,
+      "learning_rate": 3.73941842875606e-05,
+      "loss": 0.8417,
+      "step": 439500
+    },
+    {
+      "epoch": 0.7572094079827081,
+      "grad_norm": 7.269671440124512,
+      "learning_rate": 3.73798432002882e-05,
+      "loss": 0.8504,
+      "step": 440000
+    },
+    {
+      "epoch": 0.7580698732190521,
+      "grad_norm": 1.7683589458465576,
+      "learning_rate": 3.73655021130158e-05,
+      "loss": 0.8476,
+      "step": 440500
+    },
+    {
+      "epoch": 0.758930338455396,
+      "grad_norm": 2.11978816986084,
+      "learning_rate": 3.73511610257434e-05,
+      "loss": 0.8534,
+      "step": 441000
+    },
+    {
+      "epoch": 0.75979080369174,
+      "grad_norm": 1.9887040853500366,
+      "learning_rate": 3.7336819938471004e-05,
+      "loss": 0.8495,
+      "step": 441500
+    },
+    {
+      "epoch": 0.760651268928084,
+      "grad_norm": 2.1202428340911865,
+      "learning_rate": 3.73224788511986e-05,
+      "loss": 0.8403,
+      "step": 442000
+    },
+    {
+      "epoch": 0.7615117341644281,
+      "grad_norm": 2.0636634826660156,
+      "learning_rate": 3.73081377639262e-05,
+      "loss": 0.8457,
+      "step": 442500
+    },
+    {
+      "epoch": 0.762372199400772,
+      "grad_norm": 1.9846032857894897,
+      "learning_rate": 3.72937966766538e-05,
+      "loss": 0.8473,
+      "step": 443000
+    },
+    {
+      "epoch": 0.763232664637116,
+      "grad_norm": 2.072988748550415,
+      "learning_rate": 3.72794555893814e-05,
+      "loss": 0.847,
+      "step": 443500
+    },
+    {
+      "epoch": 0.76409312987346,
+      "grad_norm": 1.830478310585022,
+      "learning_rate": 3.7265114502108997e-05,
+      "loss": 0.8467,
+      "step": 444000
+    },
+    {
+      "epoch": 0.764953595109804,
+      "grad_norm": 2.1850337982177734,
+      "learning_rate": 3.725077341483661e-05,
+      "loss": 0.8449,
+      "step": 444500
+    },
+    {
+      "epoch": 0.765814060346148,
+      "grad_norm": 1.9432579278945923,
+      "learning_rate": 3.7236432327564204e-05,
+      "loss": 0.8415,
+      "step": 445000
+    },
+    {
+      "epoch": 0.7666745255824919,
+      "grad_norm": 2.1667871475219727,
+      "learning_rate": 3.72220912402918e-05,
+      "loss": 0.8514,
+      "step": 445500
+    },
+    {
+      "epoch": 0.7675349908188359,
+      "grad_norm": 2.2337772846221924,
+      "learning_rate": 3.7207750153019405e-05,
+      "loss": 0.8516,
+      "step": 446000
+    },
+    {
+      "epoch": 0.7683954560551799,
+      "grad_norm": 2.006190776824951,
+      "learning_rate": 3.7193409065747e-05,
+      "loss": 0.844,
+      "step": 446500
+    },
+    {
+      "epoch": 0.7692559212915239,
+      "grad_norm": 1.884323239326477,
+      "learning_rate": 3.71790679784746e-05,
+      "loss": 0.8375,
+      "step": 447000
+    },
+    {
+      "epoch": 0.7701163865278678,
+      "grad_norm": 2.1498095989227295,
+      "learning_rate": 3.71647268912022e-05,
+      "loss": 0.8446,
+      "step": 447500
+    },
+    {
+      "epoch": 0.7709768517642118,
+      "grad_norm": 2.063354253768921,
+      "learning_rate": 3.715038580392981e-05,
+      "loss": 0.8484,
+      "step": 448000
+    },
+    {
+      "epoch": 0.7718373170005559,
+      "grad_norm": 2.038367986679077,
+      "learning_rate": 3.7136044716657404e-05,
+      "loss": 0.8418,
+      "step": 448500
+    },
+    {
+      "epoch": 0.7726977822368999,
+      "grad_norm": 1.953763484954834,
+      "learning_rate": 3.7121703629385e-05,
+      "loss": 0.8511,
+      "step": 449000
+    },
+    {
+      "epoch": 0.7735582474732439,
+      "grad_norm": 1.9874523878097534,
+      "learning_rate": 3.7107362542112605e-05,
+      "loss": 0.8453,
+      "step": 449500
+    },
+    {
+      "epoch": 0.7744187127095878,
+      "grad_norm": 1.9465683698654175,
+      "learning_rate": 3.70930214548402e-05,
+      "loss": 0.8453,
+      "step": 450000
+    },
+    {
+      "epoch": 0.7752791779459318,
+      "grad_norm": 1.8829959630966187,
+      "learning_rate": 3.7078680367567806e-05,
+      "loss": 0.8382,
+      "step": 450500
+    },
+    {
+      "epoch": 0.7761396431822758,
+      "grad_norm": 2.064000368118286,
+      "learning_rate": 3.706433928029541e-05,
+      "loss": 0.8441,
+      "step": 451000
+    },
+    {
+      "epoch": 0.7770001084186198,
+      "grad_norm": 2.108696699142456,
+      "learning_rate": 3.704999819302301e-05,
+      "loss": 0.8486,
+      "step": 451500
+    },
+    {
+      "epoch": 0.7778605736549637,
+      "grad_norm": 2.208709239959717,
+      "learning_rate": 3.7035657105750604e-05,
+      "loss": 0.8443,
+      "step": 452000
+    },
+    {
+      "epoch": 0.7787210388913077,
+      "grad_norm": 1.9275108575820923,
+      "learning_rate": 3.702131601847821e-05,
+      "loss": 0.8413,
+      "step": 452500
+    },
+    {
+      "epoch": 0.7795815041276517,
+      "grad_norm": 2.0423073768615723,
+      "learning_rate": 3.7006974931205805e-05,
+      "loss": 0.8457,
+      "step": 453000
+    },
+    {
+      "epoch": 0.7804419693639957,
+      "grad_norm": 1.9696944952011108,
+      "learning_rate": 3.69926338439334e-05,
+      "loss": 0.8462,
+      "step": 453500
+    },
+    {
+      "epoch": 0.7813024346003397,
+      "grad_norm": 1.974177598953247,
+      "learning_rate": 3.697829275666101e-05,
+      "loss": 0.845,
+      "step": 454000
+    },
+    {
+      "epoch": 0.7821628998366837,
+      "grad_norm": 2.136514663696289,
+      "learning_rate": 3.696395166938861e-05,
+      "loss": 0.8423,
+      "step": 454500
+    },
+    {
+      "epoch": 0.7830233650730277,
+      "grad_norm": 2.2075812816619873,
+      "learning_rate": 3.694961058211621e-05,
+      "loss": 0.8422,
+      "step": 455000
+    },
+    {
+      "epoch": 0.7838838303093717,
+      "grad_norm": 1.9981763362884521,
+      "learning_rate": 3.6935269494843804e-05,
+      "loss": 0.841,
+      "step": 455500
+    },
+    {
+      "epoch": 0.7847442955457157,
+      "grad_norm": 1.9658746719360352,
+      "learning_rate": 3.692092840757141e-05,
+      "loss": 0.843,
+      "step": 456000
+    },
+    {
+      "epoch": 0.7856047607820597,
+      "grad_norm": 1.9719587564468384,
+      "learning_rate": 3.6906587320299005e-05,
+      "loss": 0.8469,
+      "step": 456500
+    },
+    {
+      "epoch": 0.7864652260184036,
+      "grad_norm": 1.9210423231124878,
+      "learning_rate": 3.689224623302661e-05,
+      "loss": 0.8457,
+      "step": 457000
+    },
+    {
+      "epoch": 0.7873256912547476,
+      "grad_norm": 1.938105583190918,
+      "learning_rate": 3.687790514575421e-05,
+      "loss": 0.8455,
+      "step": 457500
+    },
+    {
+      "epoch": 0.7881861564910916,
+      "grad_norm": 1.9795875549316406,
+      "learning_rate": 3.686356405848181e-05,
+      "loss": 0.842,
+      "step": 458000
+    },
+    {
+      "epoch": 0.7890466217274356,
+      "grad_norm": 2.140584945678711,
+      "learning_rate": 3.684922297120941e-05,
+      "loss": 0.8482,
+      "step": 458500
+    },
+    {
+      "epoch": 0.7899070869637795,
+      "grad_norm": 1.879016637802124,
+      "learning_rate": 3.683488188393701e-05,
+      "loss": 0.8426,
+      "step": 459000
+    },
+    {
+      "epoch": 0.7907675522001235,
+      "grad_norm": 1.9474905729293823,
+      "learning_rate": 3.682054079666461e-05,
+      "loss": 0.8496,
+      "step": 459500
+    },
+    {
+      "epoch": 0.7916280174364676,
+      "grad_norm": 2.1440017223358154,
+      "learning_rate": 3.680619970939221e-05,
+      "loss": 0.8457,
+      "step": 460000
+    },
+    {
+      "epoch": 0.7924884826728116,
+      "grad_norm": 2.030857801437378,
+      "learning_rate": 3.679185862211981e-05,
+      "loss": 0.8455,
+      "step": 460500
+    },
+    {
+      "epoch": 0.7933489479091556,
+      "grad_norm": 2.1576220989227295,
+      "learning_rate": 3.677751753484741e-05,
+      "loss": 0.8428,
+      "step": 461000
+    },
+    {
+      "epoch": 0.7942094131454995,
+      "grad_norm": 2.0410971641540527,
+      "learning_rate": 3.676317644757501e-05,
+      "loss": 0.8458,
+      "step": 461500
+    },
+    {
+      "epoch": 0.7950698783818435,
+      "grad_norm": 1.9619431495666504,
+      "learning_rate": 3.674883536030261e-05,
+      "loss": 0.843,
+      "step": 462000
+    },
+    {
+      "epoch": 0.7959303436181875,
+      "grad_norm": 1.9847012758255005,
+      "learning_rate": 3.673449427303021e-05,
+      "loss": 0.8408,
+      "step": 462500
+    },
+    {
+      "epoch": 0.7967908088545315,
+      "grad_norm": 2.00701642036438,
+      "learning_rate": 3.672015318575781e-05,
+      "loss": 0.8382,
+      "step": 463000
+    },
+    {
+      "epoch": 0.7976512740908754,
+      "grad_norm": 1.9219588041305542,
+      "learning_rate": 3.670581209848541e-05,
+      "loss": 0.842,
+      "step": 463500
+    },
+    {
+      "epoch": 0.7985117393272194,
+      "grad_norm": 2.0915541648864746,
+      "learning_rate": 3.6691471011213015e-05,
+      "loss": 0.8383,
+      "step": 464000
+    },
+    {
+      "epoch": 0.7993722045635634,
+      "grad_norm": 2.0045886039733887,
+      "learning_rate": 3.667712992394061e-05,
+      "loss": 0.8454,
+      "step": 464500
+    },
+    {
+      "epoch": 0.8002326697999074,
+      "grad_norm": 2.008676767349243,
+      "learning_rate": 3.666278883666821e-05,
+      "loss": 0.8388,
+      "step": 465000
+    },
+    {
+      "epoch": 0.8010931350362513,
+      "grad_norm": 2.072183609008789,
+      "learning_rate": 3.664844774939581e-05,
+      "loss": 0.8444,
+      "step": 465500
+    },
+    {
+      "epoch": 0.8019536002725954,
+      "grad_norm": 2.069936513900757,
+      "learning_rate": 3.663410666212341e-05,
+      "loss": 0.8409,
+      "step": 466000
+    },
+    {
+      "epoch": 0.8028140655089394,
+      "grad_norm": 2.0726213455200195,
+      "learning_rate": 3.6619765574851014e-05,
+      "loss": 0.8374,
+      "step": 466500
+    },
+    {
+      "epoch": 0.8036745307452834,
+      "grad_norm": 2.0134403705596924,
+      "learning_rate": 3.660542448757861e-05,
+      "loss": 0.8467,
+      "step": 467000
+    },
+    {
+      "epoch": 0.8045349959816274,
+      "grad_norm": 1.969626545906067,
+      "learning_rate": 3.6591083400306215e-05,
+      "loss": 0.8377,
+      "step": 467500
+    },
+    {
+      "epoch": 0.8053954612179713,
+      "grad_norm": 1.9737474918365479,
+      "learning_rate": 3.657674231303381e-05,
+      "loss": 0.8365,
+      "step": 468000
+    },
+    {
+      "epoch": 0.8062559264543153,
+      "grad_norm": 2.0155138969421387,
+      "learning_rate": 3.656240122576141e-05,
+      "loss": 0.8461,
+      "step": 468500
+    },
+    {
+      "epoch": 0.8071163916906593,
+      "grad_norm": 1.9591182470321655,
+      "learning_rate": 3.654806013848901e-05,
+      "loss": 0.847,
+      "step": 469000
+    },
+    {
+      "epoch": 0.8079768569270033,
+      "grad_norm": 2.1773741245269775,
+      "learning_rate": 3.653371905121662e-05,
+      "loss": 0.838,
+      "step": 469500
+    },
+    {
+      "epoch": 0.8088373221633472,
+      "grad_norm": 2.013202428817749,
+      "learning_rate": 3.6519377963944214e-05,
+      "loss": 0.844,
+      "step": 470000
+    },
+    {
+      "epoch": 0.8096977873996912,
+      "grad_norm": 2.2754344940185547,
+      "learning_rate": 3.650503687667182e-05,
+      "loss": 0.8456,
+      "step": 470500
+    },
+    {
+      "epoch": 0.8105582526360352,
+      "grad_norm": 1.956955909729004,
+      "learning_rate": 3.6490695789399415e-05,
+      "loss": 0.8398,
+      "step": 471000
+    },
+    {
+      "epoch": 0.8114187178723793,
+      "grad_norm": 2.2035508155822754,
+      "learning_rate": 3.647635470212701e-05,
+      "loss": 0.8413,
+      "step": 471500
+    },
+    {
+      "epoch": 0.8122791831087233,
+      "grad_norm": 1.993390679359436,
+      "learning_rate": 3.646201361485461e-05,
+      "loss": 0.8417,
+      "step": 472000
+    },
+    {
+      "epoch": 0.8131396483450672,
+      "grad_norm": 2.0903286933898926,
+      "learning_rate": 3.644767252758221e-05,
+      "loss": 0.8392,
+      "step": 472500
+    },
+    {
+      "epoch": 0.8140001135814112,
+      "grad_norm": 1.9314452409744263,
+      "learning_rate": 3.643333144030982e-05,
+      "loss": 0.8336,
+      "step": 473000
+    },
+    {
+      "epoch": 0.8148605788177552,
+      "grad_norm": 2.128455877304077,
+      "learning_rate": 3.6418990353037414e-05,
+      "loss": 0.8382,
+      "step": 473500
+    },
+    {
+      "epoch": 0.8157210440540992,
+      "grad_norm": 2.0214078426361084,
+      "learning_rate": 3.640464926576502e-05,
+      "loss": 0.8408,
+      "step": 474000
+    },
+    {
+      "epoch": 0.8165815092904432,
+      "grad_norm": 1.9782541990280151,
+      "learning_rate": 3.6390308178492615e-05,
+      "loss": 0.8415,
+      "step": 474500
+    },
+    {
+      "epoch": 0.8174419745267871,
+      "grad_norm": 2.0621581077575684,
+      "learning_rate": 3.637596709122021e-05,
+      "loss": 0.8352,
+      "step": 475000
+    },
+    {
+      "epoch": 0.8183024397631311,
+      "grad_norm": 1.955025553703308,
+      "learning_rate": 3.6361626003947816e-05,
+      "loss": 0.844,
+      "step": 475500
+    },
+    {
+      "epoch": 0.8191629049994751,
+      "grad_norm": 1.9758415222167969,
+      "learning_rate": 3.634728491667542e-05,
+      "loss": 0.8393,
+      "step": 476000
+    },
+    {
+      "epoch": 0.8200233702358191,
+      "grad_norm": 1.9560452699661255,
+      "learning_rate": 3.633294382940302e-05,
+      "loss": 0.8439,
+      "step": 476500
+    },
+    {
+      "epoch": 0.820883835472163,
+      "grad_norm": 2.189896583557129,
+      "learning_rate": 3.631860274213062e-05,
+      "loss": 0.8383,
+      "step": 477000
+    },
+    {
+      "epoch": 0.8217443007085071,
+      "grad_norm": 2.1071627140045166,
+      "learning_rate": 3.630426165485822e-05,
+      "loss": 0.8346,
+      "step": 477500
+    },
+    {
+      "epoch": 0.8226047659448511,
+      "grad_norm": 2.060502767562866,
+      "learning_rate": 3.6289920567585815e-05,
+      "loss": 0.8445,
+      "step": 478000
+    },
+    {
+      "epoch": 0.8234652311811951,
+      "grad_norm": 2.4608230590820312,
+      "learning_rate": 3.627557948031341e-05,
+      "loss": 0.8415,
+      "step": 478500
+    },
+    {
+      "epoch": 0.8243256964175391,
+      "grad_norm": 2.0735158920288086,
+      "learning_rate": 3.626123839304102e-05,
+      "loss": 0.8342,
+      "step": 479000
+    },
+    {
+      "epoch": 0.825186161653883,
+      "grad_norm": 2.0641353130340576,
+      "learning_rate": 3.624689730576862e-05,
+      "loss": 0.8389,
+      "step": 479500
+    },
+    {
+      "epoch": 0.826046626890227,
+      "grad_norm": 1.985877513885498,
+      "learning_rate": 3.623255621849622e-05,
+      "loss": 0.8406,
+      "step": 480000
+    },
+    {
+      "epoch": 0.826907092126571,
+      "grad_norm": 2.053553819656372,
+      "learning_rate": 3.621821513122382e-05,
+      "loss": 0.8423,
+      "step": 480500
+    },
+    {
+      "epoch": 0.827767557362915,
+      "grad_norm": 2.022402048110962,
+      "learning_rate": 3.620387404395142e-05,
+      "loss": 0.841,
+      "step": 481000
+    },
+    {
+      "epoch": 0.8286280225992589,
+      "grad_norm": 2.0855982303619385,
+      "learning_rate": 3.6189532956679015e-05,
+      "loss": 0.8371,
+      "step": 481500
+    },
+    {
+      "epoch": 0.8294884878356029,
+      "grad_norm": 2.0884454250335693,
+      "learning_rate": 3.617519186940662e-05,
+      "loss": 0.8377,
+      "step": 482000
+    },
+    {
+      "epoch": 0.8303489530719469,
+      "grad_norm": 2.163231134414673,
+      "learning_rate": 3.616085078213422e-05,
+      "loss": 0.8407,
+      "step": 482500
+    },
+    {
+      "epoch": 0.831209418308291,
+      "grad_norm": 2.1384119987487793,
+      "learning_rate": 3.614650969486182e-05,
+      "loss": 0.8344,
+      "step": 483000
+    },
+    {
+      "epoch": 0.832069883544635,
+      "grad_norm": 2.0148468017578125,
+      "learning_rate": 3.613216860758942e-05,
+      "loss": 0.841,
+      "step": 483500
+    },
+    {
+      "epoch": 0.8329303487809789,
+      "grad_norm": 1.9842921495437622,
+      "learning_rate": 3.611782752031702e-05,
+      "loss": 0.837,
+      "step": 484000
+    },
+    {
+      "epoch": 0.8337908140173229,
+      "grad_norm": 1.9228190183639526,
+      "learning_rate": 3.610348643304462e-05,
+      "loss": 0.8348,
+      "step": 484500
+    },
+    {
+      "epoch": 0.8346512792536669,
+      "grad_norm": 2.2085182666778564,
+      "learning_rate": 3.608914534577222e-05,
+      "loss": 0.8416,
+      "step": 485000
+    },
+    {
+      "epoch": 0.8355117444900109,
+      "grad_norm": 2.118323802947998,
+      "learning_rate": 3.6074804258499825e-05,
+      "loss": 0.8345,
+      "step": 485500
+    },
+    {
+      "epoch": 0.8363722097263548,
+      "grad_norm": 2.1001532077789307,
+      "learning_rate": 3.606046317122742e-05,
+      "loss": 0.8397,
+      "step": 486000
+    },
+    {
+      "epoch": 0.8372326749626988,
+      "grad_norm": 2.194197177886963,
+      "learning_rate": 3.604612208395502e-05,
+      "loss": 0.8437,
+      "step": 486500
+    },
+    {
+      "epoch": 0.8380931401990428,
+      "grad_norm": 1.9958704710006714,
+      "learning_rate": 3.6031780996682623e-05,
+      "loss": 0.8342,
+      "step": 487000
+    },
+    {
+      "epoch": 0.8389536054353868,
+      "grad_norm": 2.2926270961761475,
+      "learning_rate": 3.601743990941022e-05,
+      "loss": 0.8383,
+      "step": 487500
+    },
+    {
+      "epoch": 0.8398140706717308,
+      "grad_norm": 2.2010133266448975,
+      "learning_rate": 3.600309882213782e-05,
+      "loss": 0.8364,
+      "step": 488000
+    },
+    {
+      "epoch": 0.8406745359080747,
+      "grad_norm": 2.061776876449585,
+      "learning_rate": 3.598875773486542e-05,
+      "loss": 0.8385,
+      "step": 488500
+    },
+    {
+      "epoch": 0.8415350011444188,
+      "grad_norm": 2.888331413269043,
+      "learning_rate": 3.5974416647593025e-05,
+      "loss": 0.8407,
+      "step": 489000
+    },
+    {
+      "epoch": 0.8423954663807628,
+      "grad_norm": 2.224823236465454,
+      "learning_rate": 3.596007556032062e-05,
+      "loss": 0.8372,
+      "step": 489500
+    },
+    {
+      "epoch": 0.8432559316171068,
+      "grad_norm": 2.077299118041992,
+      "learning_rate": 3.594573447304822e-05,
+      "loss": 0.8402,
+      "step": 490000
+    },
+    {
+      "epoch": 0.8441163968534507,
+      "grad_norm": 1.9325640201568604,
+      "learning_rate": 3.593139338577582e-05,
+      "loss": 0.8364,
+      "step": 490500
+    },
+    {
+      "epoch": 0.8449768620897947,
+      "grad_norm": 2.32381534576416,
+      "learning_rate": 3.591705229850342e-05,
+      "loss": 0.8398,
+      "step": 491000
+    },
+    {
+      "epoch": 0.8458373273261387,
+      "grad_norm": 1.8531659841537476,
+      "learning_rate": 3.5902711211231024e-05,
+      "loss": 0.8276,
+      "step": 491500
+    },
+    {
+      "epoch": 0.8466977925624827,
+      "grad_norm": 1.9492835998535156,
+      "learning_rate": 3.588837012395863e-05,
+      "loss": 0.8356,
+      "step": 492000
+    },
+    {
+      "epoch": 0.8475582577988267,
+      "grad_norm": 1.9821016788482666,
+      "learning_rate": 3.5874029036686225e-05,
+      "loss": 0.8375,
+      "step": 492500
+    },
+    {
+      "epoch": 0.8484187230351706,
+      "grad_norm": 4.051340103149414,
+      "learning_rate": 3.585968794941382e-05,
+      "loss": 0.8365,
+      "step": 493000
+    },
+    {
+      "epoch": 0.8492791882715146,
+      "grad_norm": 2.0185279846191406,
+      "learning_rate": 3.5845346862141426e-05,
+      "loss": 0.8397,
+      "step": 493500
+    },
+    {
+      "epoch": 0.8501396535078586,
+      "grad_norm": 1.8271918296813965,
+      "learning_rate": 3.583100577486902e-05,
+      "loss": 0.8354,
+      "step": 494000
+    },
+    {
+      "epoch": 0.8510001187442026,
+      "grad_norm": 2.056795597076416,
+      "learning_rate": 3.581666468759663e-05,
+      "loss": 0.8299,
+      "step": 494500
+    },
+    {
+      "epoch": 0.8518605839805466,
+      "grad_norm": 1.9958736896514893,
+      "learning_rate": 3.5802323600324224e-05,
+      "loss": 0.8359,
+      "step": 495000
+    },
+    {
+      "epoch": 0.8527210492168906,
+      "grad_norm": 2.0710606575012207,
+      "learning_rate": 3.578798251305183e-05,
+      "loss": 0.8371,
+      "step": 495500
+    },
+    {
+      "epoch": 0.8535815144532346,
+      "grad_norm": 1.956955909729004,
+      "learning_rate": 3.5773641425779425e-05,
+      "loss": 0.8379,
+      "step": 496000
+    },
+    {
+      "epoch": 0.8544419796895786,
+      "grad_norm": 1.928909420967102,
+      "learning_rate": 3.575930033850702e-05,
+      "loss": 0.837,
+      "step": 496500
+    },
+    {
+      "epoch": 0.8553024449259226,
+      "grad_norm": 2.067176580429077,
+      "learning_rate": 3.5744959251234626e-05,
+      "loss": 0.8416,
+      "step": 497000
+    },
+    {
+      "epoch": 0.8561629101622665,
+      "grad_norm": 2.0131676197052,
+      "learning_rate": 3.573061816396222e-05,
+      "loss": 0.8402,
+      "step": 497500
+    },
+    {
+      "epoch": 0.8570233753986105,
+      "grad_norm": 1.9614657163619995,
+      "learning_rate": 3.571627707668983e-05,
+      "loss": 0.8374,
+      "step": 498000
+    },
+    {
+      "epoch": 0.8578838406349545,
+      "grad_norm": 1.9758917093276978,
+      "learning_rate": 3.570193598941743e-05,
+      "loss": 0.831,
+      "step": 498500
+    },
+    {
+      "epoch": 0.8587443058712985,
+      "grad_norm": 2.080070734024048,
+      "learning_rate": 3.568759490214503e-05,
+      "loss": 0.8359,
+      "step": 499000
+    },
+    {
+      "epoch": 0.8596047711076424,
+      "grad_norm": 2.107738733291626,
+      "learning_rate": 3.5673253814872625e-05,
+      "loss": 0.8351,
+      "step": 499500
+    },
+    {
+      "epoch": 0.8604652363439864,
+      "grad_norm": 2.010946750640869,
+      "learning_rate": 3.565891272760023e-05,
+      "loss": 0.8405,
+      "step": 500000
+    },
+    {
+      "epoch": 0.8613257015803305,
+      "grad_norm": 1.9980274438858032,
+      "learning_rate": 3.5644571640327826e-05,
+      "loss": 0.8326,
+      "step": 500500
+    },
+    {
+      "epoch": 0.8621861668166745,
+      "grad_norm": 1.8315647840499878,
+      "learning_rate": 3.563023055305543e-05,
+      "loss": 0.8287,
+      "step": 501000
+    },
+    {
+      "epoch": 0.8630466320530185,
+      "grad_norm": 1.926206350326538,
+      "learning_rate": 3.561588946578303e-05,
+      "loss": 0.8362,
+      "step": 501500
+    },
+    {
+      "epoch": 0.8639070972893624,
+      "grad_norm": 2.189225196838379,
+      "learning_rate": 3.560154837851063e-05,
+      "loss": 0.8343,
+      "step": 502000
+    },
+    {
+      "epoch": 0.8647675625257064,
+      "grad_norm": 2.096952199935913,
+      "learning_rate": 3.558720729123823e-05,
+      "loss": 0.8339,
+      "step": 502500
+    },
+    {
+      "epoch": 0.8656280277620504,
+      "grad_norm": 2.0683114528656006,
+      "learning_rate": 3.5572866203965825e-05,
+      "loss": 0.8394,
+      "step": 503000
+    },
+    {
+      "epoch": 0.8664884929983944,
+      "grad_norm": 2.128429651260376,
+      "learning_rate": 3.555852511669343e-05,
+      "loss": 0.8332,
+      "step": 503500
+    },
+    {
+      "epoch": 0.8673489582347383,
+      "grad_norm": 2.0516273975372314,
+      "learning_rate": 3.554418402942103e-05,
+      "loss": 0.8383,
+      "step": 504000
+    },
+    {
+      "epoch": 0.8682094234710823,
+      "grad_norm": 1.9619426727294922,
+      "learning_rate": 3.552984294214863e-05,
+      "loss": 0.8247,
+      "step": 504500
+    },
+    {
+      "epoch": 0.8690698887074263,
+      "grad_norm": 2.0234973430633545,
+      "learning_rate": 3.5515501854876234e-05,
+      "loss": 0.8323,
+      "step": 505000
+    },
+    {
+      "epoch": 0.8699303539437703,
+      "grad_norm": 2.0937347412109375,
+      "learning_rate": 3.550116076760383e-05,
+      "loss": 0.8316,
+      "step": 505500
+    },
+    {
+      "epoch": 0.8707908191801143,
+      "grad_norm": 2.096928119659424,
+      "learning_rate": 3.548681968033143e-05,
+      "loss": 0.8333,
+      "step": 506000
+    },
+    {
+      "epoch": 0.8716512844164583,
+      "grad_norm": 2.187704086303711,
+      "learning_rate": 3.5472478593059025e-05,
+      "loss": 0.8403,
+      "step": 506500
+    },
+    {
+      "epoch": 0.8725117496528023,
+      "grad_norm": 1.8772039413452148,
+      "learning_rate": 3.545813750578663e-05,
+      "loss": 0.8332,
+      "step": 507000
+    },
+    {
+      "epoch": 0.8733722148891463,
+      "grad_norm": 2.0619008541107178,
+      "learning_rate": 3.544379641851423e-05,
+      "loss": 0.8351,
+      "step": 507500
+    },
+    {
+      "epoch": 0.8742326801254903,
+      "grad_norm": 1.9941927194595337,
+      "learning_rate": 3.542945533124183e-05,
+      "loss": 0.8345,
+      "step": 508000
+    },
+    {
+      "epoch": 0.8750931453618342,
+      "grad_norm": 1.9549013376235962,
+      "learning_rate": 3.5415114243969434e-05,
+      "loss": 0.8361,
+      "step": 508500
+    },
+    {
+      "epoch": 0.8759536105981782,
+      "grad_norm": 5.23269510269165,
+      "learning_rate": 3.540077315669703e-05,
+      "loss": 0.8344,
+      "step": 509000
+    },
+    {
+      "epoch": 0.8768140758345222,
+      "grad_norm": 2.060516595840454,
+      "learning_rate": 3.538643206942463e-05,
+      "loss": 0.8322,
+      "step": 509500
+    },
+    {
+      "epoch": 0.8776745410708662,
+      "grad_norm": 2.7960243225097656,
+      "learning_rate": 3.537209098215223e-05,
+      "loss": 0.8354,
+      "step": 510000
+    },
+    {
+      "epoch": 0.8785350063072102,
+      "grad_norm": 2.1856071949005127,
+      "learning_rate": 3.5357749894879835e-05,
+      "loss": 0.8338,
+      "step": 510500
+    },
+    {
+      "epoch": 0.8793954715435541,
+      "grad_norm": 2.1425588130950928,
+      "learning_rate": 3.534340880760743e-05,
+      "loss": 0.8303,
+      "step": 511000
+    },
+    {
+      "epoch": 0.8802559367798981,
+      "grad_norm": 2.1811203956604004,
+      "learning_rate": 3.5329067720335036e-05,
+      "loss": 0.8364,
+      "step": 511500
+    },
+    {
+      "epoch": 0.8811164020162422,
+      "grad_norm": 2.2466700077056885,
+      "learning_rate": 3.5314726633062633e-05,
+      "loss": 0.8391,
+      "step": 512000
+    },
+    {
+      "epoch": 0.8819768672525862,
+      "grad_norm": 2.0680465698242188,
+      "learning_rate": 3.530038554579023e-05,
+      "loss": 0.8359,
+      "step": 512500
+    },
+    {
+      "epoch": 0.8828373324889301,
+      "grad_norm": 2.07084321975708,
+      "learning_rate": 3.528604445851783e-05,
+      "loss": 0.8359,
+      "step": 513000
+    },
+    {
+      "epoch": 0.8836977977252741,
+      "grad_norm": 2.203376054763794,
+      "learning_rate": 3.527170337124543e-05,
+      "loss": 0.8351,
+      "step": 513500
+    },
+    {
+      "epoch": 0.8845582629616181,
+      "grad_norm": 2.1163270473480225,
+      "learning_rate": 3.5257362283973035e-05,
+      "loss": 0.8345,
+      "step": 514000
+    },
+    {
+      "epoch": 0.8854187281979621,
+      "grad_norm": 2.2538020610809326,
+      "learning_rate": 3.524302119670063e-05,
+      "loss": 0.8333,
+      "step": 514500
+    },
+    {
+      "epoch": 0.8862791934343061,
+      "grad_norm": 2.0660109519958496,
+      "learning_rate": 3.5228680109428236e-05,
+      "loss": 0.8356,
+      "step": 515000
+    },
+    {
+      "epoch": 0.88713965867065,
+      "grad_norm": 2.139326333999634,
+      "learning_rate": 3.521433902215583e-05,
+      "loss": 0.8319,
+      "step": 515500
+    },
+    {
+      "epoch": 0.888000123906994,
+      "grad_norm": 1.9566582441329956,
+      "learning_rate": 3.519999793488343e-05,
+      "loss": 0.8339,
+      "step": 516000
+    },
+    {
+      "epoch": 0.888860589143338,
+      "grad_norm": 2.0667715072631836,
+      "learning_rate": 3.5185656847611034e-05,
+      "loss": 0.8342,
+      "step": 516500
+    },
+    {
+      "epoch": 0.889721054379682,
+      "grad_norm": 2.1288838386535645,
+      "learning_rate": 3.517131576033864e-05,
+      "loss": 0.8303,
+      "step": 517000
+    },
+    {
+      "epoch": 0.8905815196160259,
+      "grad_norm": 2.0571327209472656,
+      "learning_rate": 3.5156974673066235e-05,
+      "loss": 0.8384,
+      "step": 517500
+    },
+    {
+      "epoch": 0.89144198485237,
+      "grad_norm": 2.157618761062622,
+      "learning_rate": 3.514263358579383e-05,
+      "loss": 0.8314,
+      "step": 518000
+    },
+    {
+      "epoch": 0.892302450088714,
+      "grad_norm": 2.11012864112854,
+      "learning_rate": 3.5128292498521436e-05,
+      "loss": 0.8343,
+      "step": 518500
+    },
+    {
+      "epoch": 0.893162915325058,
+      "grad_norm": 2.0556013584136963,
+      "learning_rate": 3.511395141124903e-05,
+      "loss": 0.8378,
+      "step": 519000
+    },
+    {
+      "epoch": 0.894023380561402,
+      "grad_norm": 1.9106919765472412,
+      "learning_rate": 3.509961032397664e-05,
+      "loss": 0.8331,
+      "step": 519500
+    },
+    {
+      "epoch": 0.8948838457977459,
+      "grad_norm": 2.0491740703582764,
+      "learning_rate": 3.508526923670424e-05,
+      "loss": 0.8359,
+      "step": 520000
+    },
+    {
+      "epoch": 0.8957443110340899,
+      "grad_norm": 1.9946589469909668,
+      "learning_rate": 3.507092814943184e-05,
+      "loss": 0.8289,
+      "step": 520500
+    },
+    {
+      "epoch": 0.8966047762704339,
+      "grad_norm": 2.1237125396728516,
+      "learning_rate": 3.5056587062159435e-05,
+      "loss": 0.8388,
+      "step": 521000
+    },
+    {
+      "epoch": 0.8974652415067779,
+      "grad_norm": 2.3399975299835205,
+      "learning_rate": 3.504224597488704e-05,
+      "loss": 0.828,
+      "step": 521500
+    },
+    {
+      "epoch": 0.8983257067431218,
+      "grad_norm": 1.986943244934082,
+      "learning_rate": 3.5027904887614636e-05,
+      "loss": 0.8326,
+      "step": 522000
+    },
+    {
+      "epoch": 0.8991861719794658,
+      "grad_norm": 2.0902836322784424,
+      "learning_rate": 3.501356380034223e-05,
+      "loss": 0.8319,
+      "step": 522500
+    },
+    {
+      "epoch": 0.9000466372158098,
+      "grad_norm": 2.1451847553253174,
+      "learning_rate": 3.499922271306984e-05,
+      "loss": 0.8367,
+      "step": 523000
+    },
+    {
+      "epoch": 0.9009071024521539,
+      "grad_norm": 1.9373173713684082,
+      "learning_rate": 3.498488162579744e-05,
+      "loss": 0.8332,
+      "step": 523500
+    },
+    {
+      "epoch": 0.9017675676884979,
+      "grad_norm": 2.0312416553497314,
+      "learning_rate": 3.497054053852504e-05,
+      "loss": 0.8292,
+      "step": 524000
+    },
+    {
+      "epoch": 0.9026280329248418,
+      "grad_norm": 2.035531520843506,
+      "learning_rate": 3.4956199451252635e-05,
+      "loss": 0.8249,
+      "step": 524500
+    },
+    {
+      "epoch": 0.9034884981611858,
+      "grad_norm": 2.07161808013916,
+      "learning_rate": 3.494185836398024e-05,
+      "loss": 0.8347,
+      "step": 525000
+    },
+    {
+      "epoch": 0.9043489633975298,
+      "grad_norm": 2.058345317840576,
+      "learning_rate": 3.4927517276707836e-05,
+      "loss": 0.8309,
+      "step": 525500
+    },
+    {
+      "epoch": 0.9052094286338738,
+      "grad_norm": 2.0514707565307617,
+      "learning_rate": 3.491317618943544e-05,
+      "loss": 0.8352,
+      "step": 526000
+    },
+    {
+      "epoch": 0.9060698938702177,
+      "grad_norm": 2.442887544631958,
+      "learning_rate": 3.4898835102163044e-05,
+      "loss": 0.8272,
+      "step": 526500
+    },
+    {
+      "epoch": 0.9069303591065617,
+      "grad_norm": 2.133779764175415,
+      "learning_rate": 3.488449401489064e-05,
+      "loss": 0.8231,
+      "step": 527000
+    },
+    {
+      "epoch": 0.9077908243429057,
+      "grad_norm": 2.0897367000579834,
+      "learning_rate": 3.487015292761824e-05,
+      "loss": 0.8309,
+      "step": 527500
+    },
+    {
+      "epoch": 0.9086512895792497,
+      "grad_norm": 8.865015983581543,
+      "learning_rate": 3.485581184034584e-05,
+      "loss": 0.8311,
+      "step": 528000
+    },
+    {
+      "epoch": 0.9095117548155937,
+      "grad_norm": 1.8066165447235107,
+      "learning_rate": 3.484147075307344e-05,
+      "loss": 0.8274,
+      "step": 528500
+    },
+    {
+      "epoch": 0.9103722200519376,
+      "grad_norm": 1.9701728820800781,
+      "learning_rate": 3.482712966580104e-05,
+      "loss": 0.832,
+      "step": 529000
+    },
+    {
+      "epoch": 0.9112326852882817,
+      "grad_norm": 2.0116164684295654,
+      "learning_rate": 3.481278857852864e-05,
+      "loss": 0.8261,
+      "step": 529500
+    },
+    {
+      "epoch": 0.9120931505246257,
+      "grad_norm": 1.9901094436645508,
+      "learning_rate": 3.4798447491256244e-05,
+      "loss": 0.8324,
+      "step": 530000
+    },
+    {
+      "epoch": 0.9129536157609697,
+      "grad_norm": 1.9889498949050903,
+      "learning_rate": 3.478410640398384e-05,
+      "loss": 0.8182,
+      "step": 530500
+    },
+    {
+      "epoch": 0.9138140809973136,
+      "grad_norm": 2.01224946975708,
+      "learning_rate": 3.476976531671144e-05,
+      "loss": 0.8269,
+      "step": 531000
+    },
+    {
+      "epoch": 0.9146745462336576,
+      "grad_norm": 1.9385786056518555,
+      "learning_rate": 3.475542422943904e-05,
+      "loss": 0.8306,
+      "step": 531500
+    },
+    {
+      "epoch": 0.9155350114700016,
+      "grad_norm": 2.022874593734741,
+      "learning_rate": 3.474108314216664e-05,
+      "loss": 0.8352,
+      "step": 532000
+    },
+    {
+      "epoch": 0.9163954767063456,
+      "grad_norm": 2.1472716331481934,
+      "learning_rate": 3.472674205489424e-05,
+      "loss": 0.8313,
+      "step": 532500
+    },
+    {
+      "epoch": 0.9172559419426896,
+      "grad_norm": 2.008678436279297,
+      "learning_rate": 3.4712400967621846e-05,
+      "loss": 0.83,
+      "step": 533000
+    },
+    {
+      "epoch": 0.9181164071790335,
+      "grad_norm": 1.9580414295196533,
+      "learning_rate": 3.4698059880349443e-05,
+      "loss": 0.8265,
+      "step": 533500
+    },
+    {
+      "epoch": 0.9189768724153775,
+      "grad_norm": 1.7564566135406494,
+      "learning_rate": 3.468371879307704e-05,
+      "loss": 0.8254,
+      "step": 534000
+    },
+    {
+      "epoch": 0.9198373376517215,
+      "grad_norm": 2.0323212146759033,
+      "learning_rate": 3.4669377705804644e-05,
+      "loss": 0.8363,
+      "step": 534500
+    },
+    {
+      "epoch": 0.9206978028880655,
+      "grad_norm": 2.200009822845459,
+      "learning_rate": 3.465503661853224e-05,
+      "loss": 0.8216,
+      "step": 535000
+    },
+    {
+      "epoch": 0.9215582681244096,
+      "grad_norm": 2.065061092376709,
+      "learning_rate": 3.4640695531259845e-05,
+      "loss": 0.83,
+      "step": 535500
+    },
+    {
+      "epoch": 0.9224187333607535,
+      "grad_norm": 2.178332805633545,
+      "learning_rate": 3.462635444398744e-05,
+      "loss": 0.8317,
+      "step": 536000
+    },
+    {
+      "epoch": 0.9232791985970975,
+      "grad_norm": 1.9318639039993286,
+      "learning_rate": 3.4612013356715046e-05,
+      "loss": 0.8322,
+      "step": 536500
+    },
+    {
+      "epoch": 0.9241396638334415,
+      "grad_norm": 2.050119400024414,
+      "learning_rate": 3.459767226944264e-05,
+      "loss": 0.8304,
+      "step": 537000
+    },
+    {
+      "epoch": 0.9250001290697855,
+      "grad_norm": 2.291924238204956,
+      "learning_rate": 3.458333118217024e-05,
+      "loss": 0.8258,
+      "step": 537500
+    },
+    {
+      "epoch": 0.9258605943061294,
+      "grad_norm": 1.9168674945831299,
+      "learning_rate": 3.4568990094897844e-05,
+      "loss": 0.8228,
+      "step": 538000
+    },
+    {
+      "epoch": 0.9267210595424734,
+      "grad_norm": 2.229269027709961,
+      "learning_rate": 3.455464900762544e-05,
+      "loss": 0.8265,
+      "step": 538500
+    },
+    {
+      "epoch": 0.9275815247788174,
+      "grad_norm": 1.979021668434143,
+      "learning_rate": 3.4540307920353045e-05,
+      "loss": 0.8275,
+      "step": 539000
+    },
+    {
+      "epoch": 0.9284419900151614,
+      "grad_norm": 2.030496835708618,
+      "learning_rate": 3.452596683308065e-05,
+      "loss": 0.8302,
+      "step": 539500
+    },
+    {
+      "epoch": 0.9293024552515053,
+      "grad_norm": 2.054203748703003,
+      "learning_rate": 3.4511625745808246e-05,
+      "loss": 0.8313,
+      "step": 540000
+    },
+    {
+      "epoch": 0.9301629204878493,
+      "grad_norm": 2.2082176208496094,
+      "learning_rate": 3.449728465853584e-05,
+      "loss": 0.8259,
+      "step": 540500
+    },
+    {
+      "epoch": 0.9310233857241934,
+      "grad_norm": 1.9945560693740845,
+      "learning_rate": 3.448294357126344e-05,
+      "loss": 0.8277,
+      "step": 541000
+    },
+    {
+      "epoch": 0.9318838509605374,
+      "grad_norm": 2.0513644218444824,
+      "learning_rate": 3.4468602483991044e-05,
+      "loss": 0.8343,
+      "step": 541500
+    },
+    {
+      "epoch": 0.9327443161968814,
+      "grad_norm": 1.8823306560516357,
+      "learning_rate": 3.445426139671865e-05,
+      "loss": 0.8274,
+      "step": 542000
+    },
+    {
+      "epoch": 0.9336047814332253,
+      "grad_norm": 1.9960150718688965,
+      "learning_rate": 3.4439920309446245e-05,
+      "loss": 0.8263,
+      "step": 542500
+    },
+    {
+      "epoch": 0.9344652466695693,
+      "grad_norm": 2.011401653289795,
+      "learning_rate": 3.442557922217385e-05,
+      "loss": 0.8279,
+      "step": 543000
+    },
+    {
+      "epoch": 0.9353257119059133,
+      "grad_norm": 1.9199022054672241,
+      "learning_rate": 3.4411238134901446e-05,
+      "loss": 0.8305,
+      "step": 543500
+    },
+    {
+      "epoch": 0.9361861771422573,
+      "grad_norm": 2.1455698013305664,
+      "learning_rate": 3.439689704762904e-05,
+      "loss": 0.8288,
+      "step": 544000
+    },
+    {
+      "epoch": 0.9370466423786012,
+      "grad_norm": 2.0815937519073486,
+      "learning_rate": 3.438255596035665e-05,
+      "loss": 0.8276,
+      "step": 544500
+    },
+    {
+      "epoch": 0.9379071076149452,
+      "grad_norm": 2.0950100421905518,
+      "learning_rate": 3.436821487308425e-05,
+      "loss": 0.8346,
+      "step": 545000
+    },
+    {
+      "epoch": 0.9387675728512892,
+      "grad_norm": 2.0615501403808594,
+      "learning_rate": 3.435387378581185e-05,
+      "loss": 0.8185,
+      "step": 545500
+    },
+    {
+      "epoch": 0.9396280380876332,
+      "grad_norm": 1.8258684873580933,
+      "learning_rate": 3.433953269853945e-05,
+      "loss": 0.83,
+      "step": 546000
+    },
+    {
+      "epoch": 0.9404885033239772,
+      "grad_norm": 2.053507089614868,
+      "learning_rate": 3.432519161126705e-05,
+      "loss": 0.8301,
+      "step": 546500
+    },
+    {
+      "epoch": 0.9413489685603212,
+      "grad_norm": 2.033759593963623,
+      "learning_rate": 3.4310850523994646e-05,
+      "loss": 0.8303,
+      "step": 547000
+    },
+    {
+      "epoch": 0.9422094337966652,
+      "grad_norm": 1.9754514694213867,
+      "learning_rate": 3.429650943672224e-05,
+      "loss": 0.8307,
+      "step": 547500
+    },
+    {
+      "epoch": 0.9430698990330092,
+      "grad_norm": 2.1358065605163574,
+      "learning_rate": 3.428216834944985e-05,
+      "loss": 0.8237,
+      "step": 548000
+    },
+    {
+      "epoch": 0.9439303642693532,
+      "grad_norm": 2.076956033706665,
+      "learning_rate": 3.426782726217745e-05,
+      "loss": 0.8289,
+      "step": 548500
+    },
+    {
+      "epoch": 0.9447908295056971,
+      "grad_norm": 2.1519205570220947,
+      "learning_rate": 3.425348617490505e-05,
+      "loss": 0.8264,
+      "step": 549000
+    },
+    {
+      "epoch": 0.9456512947420411,
+      "grad_norm": 2.108743906021118,
+      "learning_rate": 3.423914508763265e-05,
+      "loss": 0.8238,
+      "step": 549500
+    },
+    {
+      "epoch": 0.9465117599783851,
+      "grad_norm": 2.0745456218719482,
+      "learning_rate": 3.422480400036025e-05,
+      "loss": 0.827,
+      "step": 550000
+    },
+    {
+      "epoch": 0.9473722252147291,
+      "grad_norm": 2.0193209648132324,
+      "learning_rate": 3.4210462913087846e-05,
+      "loss": 0.8232,
+      "step": 550500
+    },
+    {
+      "epoch": 0.9482326904510731,
+      "grad_norm": 2.215498208999634,
+      "learning_rate": 3.419612182581545e-05,
+      "loss": 0.8314,
+      "step": 551000
+    },
+    {
+      "epoch": 0.949093155687417,
+      "grad_norm": 2.105421304702759,
+      "learning_rate": 3.4181780738543054e-05,
+      "loss": 0.8246,
+      "step": 551500
+    },
+    {
+      "epoch": 0.949953620923761,
+      "grad_norm": 2.102644920349121,
+      "learning_rate": 3.416743965127065e-05,
+      "loss": 0.8283,
+      "step": 552000
+    },
+    {
+      "epoch": 0.9508140861601051,
+      "grad_norm": 1.9746520519256592,
+      "learning_rate": 3.415309856399825e-05,
+      "loss": 0.8286,
+      "step": 552500
+    },
+    {
+      "epoch": 0.9516745513964491,
+      "grad_norm": 2.012468099594116,
+      "learning_rate": 3.413875747672585e-05,
+      "loss": 0.8259,
+      "step": 553000
+    },
+    {
+      "epoch": 0.952535016632793,
+      "grad_norm": 1.770936131477356,
+      "learning_rate": 3.412441638945345e-05,
+      "loss": 0.8236,
+      "step": 553500
+    },
+    {
+      "epoch": 0.953395481869137,
+      "grad_norm": 2.082106351852417,
+      "learning_rate": 3.411007530218105e-05,
+      "loss": 0.8252,
+      "step": 554000
+    },
+    {
+      "epoch": 0.954255947105481,
+      "grad_norm": 2.122305393218994,
+      "learning_rate": 3.4095734214908656e-05,
+      "loss": 0.8295,
+      "step": 554500
+    },
+    {
+      "epoch": 0.955116412341825,
+      "grad_norm": 2.12324595451355,
+      "learning_rate": 3.4081393127636254e-05,
+      "loss": 0.8282,
+      "step": 555000
+    },
+    {
+      "epoch": 0.955976877578169,
+      "grad_norm": 1.997396469116211,
+      "learning_rate": 3.406705204036385e-05,
+      "loss": 0.8297,
+      "step": 555500
+    },
+    {
+      "epoch": 0.9568373428145129,
+      "grad_norm": 2.3350627422332764,
+      "learning_rate": 3.4052710953091454e-05,
+      "loss": 0.8308,
+      "step": 556000
+    },
+    {
+      "epoch": 0.9576978080508569,
+      "grad_norm": 2.1347360610961914,
+      "learning_rate": 3.403836986581905e-05,
+      "loss": 0.8287,
+      "step": 556500
+    },
+    {
+      "epoch": 0.9585582732872009,
+      "grad_norm": 2.196913719177246,
+      "learning_rate": 3.402402877854665e-05,
+      "loss": 0.8236,
+      "step": 557000
+    },
+    {
+      "epoch": 0.9594187385235449,
+      "grad_norm": 2.0057709217071533,
+      "learning_rate": 3.400968769127425e-05,
+      "loss": 0.8312,
+      "step": 557500
+    },
+    {
+      "epoch": 0.9602792037598888,
+      "grad_norm": 2.223766803741455,
+      "learning_rate": 3.3995346604001856e-05,
+      "loss": 0.834,
+      "step": 558000
+    },
+    {
+      "epoch": 0.9611396689962329,
+      "grad_norm": 1.8879119157791138,
+      "learning_rate": 3.3981005516729453e-05,
+      "loss": 0.8227,
+      "step": 558500
+    },
+    {
+      "epoch": 0.9620001342325769,
+      "grad_norm": 2.166381359100342,
+      "learning_rate": 3.396666442945705e-05,
+      "loss": 0.826,
+      "step": 559000
+    },
+    {
+      "epoch": 0.9628605994689209,
+      "grad_norm": 2.1011924743652344,
+      "learning_rate": 3.3952323342184654e-05,
+      "loss": 0.8283,
+      "step": 559500
+    },
+    {
+      "epoch": 0.9637210647052649,
+      "grad_norm": 2.1456775665283203,
+      "learning_rate": 3.393798225491225e-05,
+      "loss": 0.8162,
+      "step": 560000
+    },
+    {
+      "epoch": 0.9645815299416088,
+      "grad_norm": 1.9335885047912598,
+      "learning_rate": 3.3923641167639855e-05,
+      "loss": 0.8227,
+      "step": 560500
+    },
+    {
+      "epoch": 0.9654419951779528,
+      "grad_norm": 3.3997044563293457,
+      "learning_rate": 3.390930008036746e-05,
+      "loss": 0.8252,
+      "step": 561000
+    },
+    {
+      "epoch": 0.9663024604142968,
+      "grad_norm": 2.0262041091918945,
+      "learning_rate": 3.3894958993095056e-05,
+      "loss": 0.8304,
+      "step": 561500
+    },
+    {
+      "epoch": 0.9671629256506408,
+      "grad_norm": 2.062438726425171,
+      "learning_rate": 3.388061790582265e-05,
+      "loss": 0.8305,
+      "step": 562000
+    },
+    {
+      "epoch": 0.9680233908869847,
+      "grad_norm": 1.937840461730957,
+      "learning_rate": 3.386627681855026e-05,
+      "loss": 0.829,
+      "step": 562500
+    },
+    {
+      "epoch": 0.9688838561233287,
+      "grad_norm": 2.0924041271209717,
+      "learning_rate": 3.3851935731277854e-05,
+      "loss": 0.829,
+      "step": 563000
+    },
+    {
+      "epoch": 0.9697443213596727,
+      "grad_norm": 2.0761170387268066,
+      "learning_rate": 3.383759464400546e-05,
+      "loss": 0.8273,
+      "step": 563500
+    },
+    {
+      "epoch": 0.9706047865960168,
+      "grad_norm": 2.1548244953155518,
+      "learning_rate": 3.3823253556733055e-05,
+      "loss": 0.8281,
+      "step": 564000
+    },
+    {
+      "epoch": 0.9714652518323608,
+      "grad_norm": 3.1409645080566406,
+      "learning_rate": 3.380891246946066e-05,
+      "loss": 0.8261,
+      "step": 564500
+    },
+    {
+      "epoch": 0.9723257170687047,
+      "grad_norm": 2.254758358001709,
+      "learning_rate": 3.3794571382188256e-05,
+      "loss": 0.8254,
+      "step": 565000
+    },
+    {
+      "epoch": 0.9731861823050487,
+      "grad_norm": 1.963372826576233,
+      "learning_rate": 3.378023029491585e-05,
+      "loss": 0.8219,
+      "step": 565500
+    },
+    {
+      "epoch": 0.9740466475413927,
+      "grad_norm": 2.0996382236480713,
+      "learning_rate": 3.376588920764346e-05,
+      "loss": 0.8254,
+      "step": 566000
+    },
+    {
+      "epoch": 0.9749071127777367,
+      "grad_norm": 1.9645578861236572,
+      "learning_rate": 3.3751548120371054e-05,
+      "loss": 0.8238,
+      "step": 566500
+    },
+    {
+      "epoch": 0.9757675780140806,
+      "grad_norm": 2.255702495574951,
+      "learning_rate": 3.373720703309866e-05,
+      "loss": 0.8205,
+      "step": 567000
+    },
+    {
+      "epoch": 0.9766280432504246,
+      "grad_norm": 2.045149087905884,
+      "learning_rate": 3.372286594582626e-05,
+      "loss": 0.8257,
+      "step": 567500
+    },
+    {
+      "epoch": 0.9774885084867686,
+      "grad_norm": 1.8754820823669434,
+      "learning_rate": 3.370852485855386e-05,
+      "loss": 0.8262,
+      "step": 568000
+    },
+    {
+      "epoch": 0.9783489737231126,
+      "grad_norm": 2.100236654281616,
+      "learning_rate": 3.3694183771281456e-05,
+      "loss": 0.8206,
+      "step": 568500
+    },
+    {
+      "epoch": 0.9792094389594566,
+      "grad_norm": 2.0266075134277344,
+      "learning_rate": 3.367984268400906e-05,
+      "loss": 0.8205,
+      "step": 569000
+    },
+    {
+      "epoch": 0.9800699041958005,
+      "grad_norm": 2.2990362644195557,
+      "learning_rate": 3.366550159673666e-05,
+      "loss": 0.8178,
+      "step": 569500
+    },
+    {
+      "epoch": 0.9809303694321446,
+      "grad_norm": 2.3041062355041504,
+      "learning_rate": 3.365116050946426e-05,
+      "loss": 0.827,
+      "step": 570000
+    },
+    {
+      "epoch": 0.9817908346684886,
+      "grad_norm": 1.9612354040145874,
+      "learning_rate": 3.363681942219186e-05,
+      "loss": 0.8252,
+      "step": 570500
+    },
+    {
+      "epoch": 0.9826512999048326,
+      "grad_norm": 2.0809218883514404,
+      "learning_rate": 3.362247833491946e-05,
+      "loss": 0.8303,
+      "step": 571000
+    },
+    {
+      "epoch": 0.9835117651411766,
+      "grad_norm": 2.0215840339660645,
+      "learning_rate": 3.360813724764706e-05,
+      "loss": 0.8272,
+      "step": 571500
+    },
+    {
+      "epoch": 0.9843722303775205,
+      "grad_norm": 2.0683231353759766,
+      "learning_rate": 3.3593796160374656e-05,
+      "loss": 0.8242,
+      "step": 572000
+    },
+    {
+      "epoch": 0.9852326956138645,
+      "grad_norm": 2.0237414836883545,
+      "learning_rate": 3.357945507310226e-05,
+      "loss": 0.8251,
+      "step": 572500
+    },
+    {
+      "epoch": 0.9860931608502085,
+      "grad_norm": 2.115190267562866,
+      "learning_rate": 3.356511398582986e-05,
+      "loss": 0.8298,
+      "step": 573000
+    },
+    {
+      "epoch": 0.9869536260865525,
+      "grad_norm": 1.906149983406067,
+      "learning_rate": 3.355077289855746e-05,
+      "loss": 0.8176,
+      "step": 573500
+    },
+    {
+      "epoch": 0.9878140913228964,
+      "grad_norm": 1.9369081258773804,
+      "learning_rate": 3.3536431811285065e-05,
+      "loss": 0.8216,
+      "step": 574000
+    },
+    {
+      "epoch": 0.9886745565592404,
+      "grad_norm": 2.188481092453003,
+      "learning_rate": 3.352209072401266e-05,
+      "loss": 0.8215,
+      "step": 574500
+    },
+    {
+      "epoch": 0.9895350217955844,
+      "grad_norm": 1.9754977226257324,
+      "learning_rate": 3.350774963674026e-05,
+      "loss": 0.8233,
+      "step": 575000
+    },
+    {
+      "epoch": 0.9903954870319284,
+      "grad_norm": 1.8967516422271729,
+      "learning_rate": 3.349340854946786e-05,
+      "loss": 0.8265,
+      "step": 575500
+    },
+    {
+      "epoch": 0.9912559522682725,
+      "grad_norm": 1.9887170791625977,
+      "learning_rate": 3.347906746219546e-05,
+      "loss": 0.8212,
+      "step": 576000
+    },
+    {
+      "epoch": 0.9921164175046164,
+      "grad_norm": 2.0904929637908936,
+      "learning_rate": 3.3464726374923064e-05,
+      "loss": 0.8225,
+      "step": 576500
+    },
+    {
+      "epoch": 0.9929768827409604,
+      "grad_norm": 2.0081520080566406,
+      "learning_rate": 3.345038528765066e-05,
+      "loss": 0.8215,
+      "step": 577000
+    },
+    {
+      "epoch": 0.9938373479773044,
+      "grad_norm": 2.2610244750976562,
+      "learning_rate": 3.3436044200378265e-05,
+      "loss": 0.8178,
+      "step": 577500
+    },
+    {
+      "epoch": 0.9946978132136484,
+      "grad_norm": 1.992607831954956,
+      "learning_rate": 3.342170311310586e-05,
+      "loss": 0.8191,
+      "step": 578000
+    },
+    {
+      "epoch": 0.9955582784499923,
+      "grad_norm": 2.1425204277038574,
+      "learning_rate": 3.340736202583346e-05,
+      "loss": 0.8234,
+      "step": 578500
+    },
+    {
+      "epoch": 0.9964187436863363,
+      "grad_norm": 2.132256269454956,
+      "learning_rate": 3.339302093856106e-05,
+      "loss": 0.8236,
+      "step": 579000
+    },
+    {
+      "epoch": 0.9972792089226803,
+      "grad_norm": 2.027850389480591,
+      "learning_rate": 3.3378679851288666e-05,
+      "loss": 0.8211,
+      "step": 579500
+    },
+    {
+      "epoch": 0.9981396741590243,
+      "grad_norm": 2.0103042125701904,
+      "learning_rate": 3.3364338764016264e-05,
+      "loss": 0.8232,
+      "step": 580000
+    },
+    {
+      "epoch": 0.9990001393953682,
+      "grad_norm": 2.1737022399902344,
+      "learning_rate": 3.334999767674387e-05,
+      "loss": 0.8241,
+      "step": 580500
+    },
+    {
+      "epoch": 0.9998606046317122,
+      "grad_norm": 2.199488401412964,
+      "learning_rate": 3.3335656589471464e-05,
+      "loss": 0.8159,
+      "step": 581000
+    },
+    {
+      "epoch": 1.0007210698680562,
+      "grad_norm": 2.1512362957000732,
+      "learning_rate": 3.332131550219906e-05,
+      "loss": 0.8182,
+      "step": 581500
+    },
+    {
+      "epoch": 1.0015815351044002,
+      "grad_norm": 2.1353979110717773,
+      "learning_rate": 3.330697441492666e-05,
+      "loss": 0.8258,
+      "step": 582000
+    },
+    {
+      "epoch": 1.0024420003407442,
+      "grad_norm": 1.8595000505447388,
+      "learning_rate": 3.329263332765426e-05,
+      "loss": 0.8191,
+      "step": 582500
+    },
+    {
+      "epoch": 1.0033024655770881,
+      "grad_norm": 2.160736322402954,
+      "learning_rate": 3.3278292240381866e-05,
+      "loss": 0.8202,
+      "step": 583000
+    },
+    {
+      "epoch": 1.004162930813432,
+      "grad_norm": 2.159769296646118,
+      "learning_rate": 3.3263951153109463e-05,
+      "loss": 0.8232,
+      "step": 583500
+    },
+    {
+      "epoch": 1.005023396049776,
+      "grad_norm": 1.9886316061019897,
+      "learning_rate": 3.324961006583707e-05,
+      "loss": 0.814,
+      "step": 584000
+    },
+    {
+      "epoch": 1.0058838612861203,
+      "grad_norm": 2.038942337036133,
+      "learning_rate": 3.3235268978564664e-05,
+      "loss": 0.8239,
+      "step": 584500
+    },
+    {
+      "epoch": 1.0067443265224643,
+      "grad_norm": 1.850509524345398,
+      "learning_rate": 3.322092789129226e-05,
+      "loss": 0.8229,
+      "step": 585000
+    },
+    {
+      "epoch": 1.0076047917588082,
+      "grad_norm": 2.1953375339508057,
+      "learning_rate": 3.3206586804019865e-05,
+      "loss": 0.822,
+      "step": 585500
+    },
+    {
+      "epoch": 1.0084652569951522,
+      "grad_norm": 2.1173160076141357,
+      "learning_rate": 3.319224571674747e-05,
+      "loss": 0.8185,
+      "step": 586000
+    },
+    {
+      "epoch": 1.0093257222314962,
+      "grad_norm": 1.989945888519287,
+      "learning_rate": 3.3177904629475066e-05,
+      "loss": 0.8225,
+      "step": 586500
+    },
+    {
+      "epoch": 1.0101861874678402,
+      "grad_norm": 1.9819010496139526,
+      "learning_rate": 3.316356354220267e-05,
+      "loss": 0.8178,
+      "step": 587000
+    },
+    {
+      "epoch": 1.0110466527041841,
+      "grad_norm": 2.016489267349243,
+      "learning_rate": 3.314922245493027e-05,
+      "loss": 0.8168,
+      "step": 587500
+    },
+    {
+      "epoch": 1.0119071179405281,
+      "grad_norm": 2.2139735221862793,
+      "learning_rate": 3.3134881367657864e-05,
+      "loss": 0.8225,
+      "step": 588000
+    },
+    {
+      "epoch": 1.012767583176872,
+      "grad_norm": 2.0623409748077393,
+      "learning_rate": 3.312054028038547e-05,
+      "loss": 0.817,
+      "step": 588500
+    },
+    {
+      "epoch": 1.013628048413216,
+      "grad_norm": 1.8793323040008545,
+      "learning_rate": 3.310619919311307e-05,
+      "loss": 0.8201,
+      "step": 589000
+    },
+    {
+      "epoch": 1.01448851364956,
+      "grad_norm": 1.849631428718567,
+      "learning_rate": 3.309185810584067e-05,
+      "loss": 0.8257,
+      "step": 589500
+    },
+    {
+      "epoch": 1.015348978885904,
+      "grad_norm": 2.2486824989318848,
+      "learning_rate": 3.3077517018568266e-05,
+      "loss": 0.814,
+      "step": 590000
+    },
+    {
+      "epoch": 1.016209444122248,
+      "grad_norm": 1.9355723857879639,
+      "learning_rate": 3.306317593129587e-05,
+      "loss": 0.8148,
+      "step": 590500
+    },
+    {
+      "epoch": 1.017069909358592,
+      "grad_norm": 2.243500232696533,
+      "learning_rate": 3.304883484402347e-05,
+      "loss": 0.8242,
+      "step": 591000
+    },
+    {
+      "epoch": 1.017930374594936,
+      "grad_norm": 2.0630578994750977,
+      "learning_rate": 3.3034493756751064e-05,
+      "loss": 0.8258,
+      "step": 591500
+    },
+    {
+      "epoch": 1.01879083983128,
+      "grad_norm": 2.0572683811187744,
+      "learning_rate": 3.302015266947867e-05,
+      "loss": 0.8181,
+      "step": 592000
+    },
+    {
+      "epoch": 1.019651305067624,
+      "grad_norm": 1.9563729763031006,
+      "learning_rate": 3.300581158220627e-05,
+      "loss": 0.8188,
+      "step": 592500
+    },
+    {
+      "epoch": 1.020511770303968,
+      "grad_norm": 2.0135724544525146,
+      "learning_rate": 3.299147049493387e-05,
+      "loss": 0.828,
+      "step": 593000
+    },
+    {
+      "epoch": 1.0213722355403119,
+      "grad_norm": 2.0690765380859375,
+      "learning_rate": 3.2977129407661466e-05,
+      "loss": 0.8237,
+      "step": 593500
+    },
+    {
+      "epoch": 1.0222327007766558,
+      "grad_norm": 1.8143502473831177,
+      "learning_rate": 3.296278832038907e-05,
+      "loss": 0.8142,
+      "step": 594000
+    },
+    {
+      "epoch": 1.0230931660129998,
+      "grad_norm": 2.014436721801758,
+      "learning_rate": 3.294844723311667e-05,
+      "loss": 0.8287,
+      "step": 594500
+    },
+    {
+      "epoch": 1.0239536312493438,
+      "grad_norm": 2.035250425338745,
+      "learning_rate": 3.293410614584427e-05,
+      "loss": 0.8168,
+      "step": 595000
+    },
+    {
+      "epoch": 1.0248140964856878,
+      "grad_norm": 2.0608718395233154,
+      "learning_rate": 3.2919765058571875e-05,
+      "loss": 0.8156,
+      "step": 595500
+    },
+    {
+      "epoch": 1.025674561722032,
+      "grad_norm": 2.038362741470337,
+      "learning_rate": 3.290542397129947e-05,
+      "loss": 0.8213,
+      "step": 596000
+    },
+    {
+      "epoch": 1.026535026958376,
+      "grad_norm": 1.972121238708496,
+      "learning_rate": 3.289108288402707e-05,
+      "loss": 0.8172,
+      "step": 596500
+    },
+    {
+      "epoch": 1.02739549219472,
+      "grad_norm": 2.067946195602417,
+      "learning_rate": 3.287674179675467e-05,
+      "loss": 0.8188,
+      "step": 597000
+    },
+    {
+      "epoch": 1.028255957431064,
+      "grad_norm": 2.0822267532348633,
+      "learning_rate": 3.286240070948227e-05,
+      "loss": 0.8189,
+      "step": 597500
+    },
+    {
+      "epoch": 1.0291164226674079,
+      "grad_norm": 1.9324904680252075,
+      "learning_rate": 3.284805962220987e-05,
+      "loss": 0.8211,
+      "step": 598000
+    },
+    {
+      "epoch": 1.0299768879037519,
+      "grad_norm": 2.024289608001709,
+      "learning_rate": 3.283371853493748e-05,
+      "loss": 0.8163,
+      "step": 598500
+    },
+    {
+      "epoch": 1.0308373531400958,
+      "grad_norm": 1.9507628679275513,
+      "learning_rate": 3.2819377447665075e-05,
+      "loss": 0.8135,
+      "step": 599000
+    },
+    {
+      "epoch": 1.0316978183764398,
+      "grad_norm": 2.0539042949676514,
+      "learning_rate": 3.280503636039267e-05,
+      "loss": 0.8146,
+      "step": 599500
+    },
+    {
+      "epoch": 1.0325582836127838,
+      "grad_norm": 1.9990108013153076,
+      "learning_rate": 3.279069527312027e-05,
+      "loss": 0.8163,
+      "step": 600000
+    },
+    {
+      "epoch": 1.0334187488491278,
+      "grad_norm": 2.1447951793670654,
+      "learning_rate": 3.277635418584787e-05,
+      "loss": 0.815,
+      "step": 600500
+    },
+    {
+      "epoch": 1.0342792140854717,
+      "grad_norm": 2.0616796016693115,
+      "learning_rate": 3.276201309857547e-05,
+      "loss": 0.8205,
+      "step": 601000
+    },
+    {
+      "epoch": 1.0351396793218157,
+      "grad_norm": 1.999974012374878,
+      "learning_rate": 3.2747672011303074e-05,
+      "loss": 0.8204,
+      "step": 601500
+    },
+    {
+      "epoch": 1.0360001445581597,
+      "grad_norm": 1.998840570449829,
+      "learning_rate": 3.273333092403068e-05,
+      "loss": 0.8181,
+      "step": 602000
+    },
+    {
+      "epoch": 1.0368606097945037,
+      "grad_norm": 2.0231664180755615,
+      "learning_rate": 3.2718989836758275e-05,
+      "loss": 0.8095,
+      "step": 602500
+    },
+    {
+      "epoch": 1.0377210750308477,
+      "grad_norm": 2.1706063747406006,
+      "learning_rate": 3.270464874948587e-05,
+      "loss": 0.812,
+      "step": 603000
+    },
+    {
+      "epoch": 1.0385815402671916,
+      "grad_norm": 2.0731775760650635,
+      "learning_rate": 3.2690307662213475e-05,
+      "loss": 0.824,
+      "step": 603500
+    },
+    {
+      "epoch": 1.0394420055035356,
+      "grad_norm": 2.0796427726745605,
+      "learning_rate": 3.267596657494107e-05,
+      "loss": 0.8137,
+      "step": 604000
+    },
+    {
+      "epoch": 1.0403024707398796,
+      "grad_norm": 2.0877110958099365,
+      "learning_rate": 3.2661625487668676e-05,
+      "loss": 0.8186,
+      "step": 604500
+    },
+    {
+      "epoch": 1.0411629359762236,
+      "grad_norm": 1.938895344734192,
+      "learning_rate": 3.2647284400396273e-05,
+      "loss": 0.8214,
+      "step": 605000
+    },
+    {
+      "epoch": 1.0420234012125675,
+      "grad_norm": 2.034285068511963,
+      "learning_rate": 3.263294331312388e-05,
+      "loss": 0.8204,
+      "step": 605500
+    },
+    {
+      "epoch": 1.0428838664489115,
+      "grad_norm": 1.960330605506897,
+      "learning_rate": 3.2618602225851474e-05,
+      "loss": 0.8137,
+      "step": 606000
+    },
+    {
+      "epoch": 1.0437443316852555,
+      "grad_norm": 2.037123680114746,
+      "learning_rate": 3.260426113857907e-05,
+      "loss": 0.8172,
+      "step": 606500
+    },
+    {
+      "epoch": 1.0446047969215995,
+      "grad_norm": 1.9923410415649414,
+      "learning_rate": 3.2589920051306675e-05,
+      "loss": 0.8164,
+      "step": 607000
+    },
+    {
+      "epoch": 1.0454652621579437,
+      "grad_norm": 2.060210943222046,
+      "learning_rate": 3.257557896403427e-05,
+      "loss": 0.8191,
+      "step": 607500
+    },
+    {
+      "epoch": 1.0463257273942876,
+      "grad_norm": 1.9298114776611328,
+      "learning_rate": 3.2561237876761876e-05,
+      "loss": 0.8167,
+      "step": 608000
+    },
+    {
+      "epoch": 1.0471861926306316,
+      "grad_norm": 2.114866256713867,
+      "learning_rate": 3.254689678948948e-05,
+      "loss": 0.8225,
+      "step": 608500
+    },
+    {
+      "epoch": 1.0480466578669756,
+      "grad_norm": 2.275494337081909,
+      "learning_rate": 3.253255570221708e-05,
+      "loss": 0.8135,
+      "step": 609000
+    },
+    {
+      "epoch": 1.0489071231033196,
+      "grad_norm": 1.9952894449234009,
+      "learning_rate": 3.2518214614944674e-05,
+      "loss": 0.8133,
+      "step": 609500
+    },
+    {
+      "epoch": 1.0497675883396635,
+      "grad_norm": 2.223127841949463,
+      "learning_rate": 3.250387352767228e-05,
+      "loss": 0.8114,
+      "step": 610000
+    },
+    {
+      "epoch": 1.0506280535760075,
+      "grad_norm": 2.092866897583008,
+      "learning_rate": 3.2489532440399875e-05,
+      "loss": 0.817,
+      "step": 610500
+    },
+    {
+      "epoch": 1.0514885188123515,
+      "grad_norm": 1.9106268882751465,
+      "learning_rate": 3.247519135312748e-05,
+      "loss": 0.8218,
+      "step": 611000
+    },
+    {
+      "epoch": 1.0523489840486955,
+      "grad_norm": 2.031660318374634,
+      "learning_rate": 3.2460850265855076e-05,
+      "loss": 0.8157,
+      "step": 611500
+    },
+    {
+      "epoch": 1.0532094492850395,
+      "grad_norm": 2.047598123550415,
+      "learning_rate": 3.244650917858268e-05,
+      "loss": 0.8192,
+      "step": 612000
+    },
+    {
+      "epoch": 1.0540699145213834,
+      "grad_norm": 2.0678765773773193,
+      "learning_rate": 3.243216809131028e-05,
+      "loss": 0.813,
+      "step": 612500
+    },
+    {
+      "epoch": 1.0549303797577274,
+      "grad_norm": 2.12496280670166,
+      "learning_rate": 3.2417827004037874e-05,
+      "loss": 0.8164,
+      "step": 613000
+    },
+    {
+      "epoch": 1.0557908449940714,
+      "grad_norm": 2.0553905963897705,
+      "learning_rate": 3.240348591676548e-05,
+      "loss": 0.8211,
+      "step": 613500
+    },
+    {
+      "epoch": 1.0566513102304154,
+      "grad_norm": 2.044658660888672,
+      "learning_rate": 3.238914482949308e-05,
+      "loss": 0.8096,
+      "step": 614000
+    },
+    {
+      "epoch": 1.0575117754667593,
+      "grad_norm": 2.074327230453491,
+      "learning_rate": 3.237480374222068e-05,
+      "loss": 0.8195,
+      "step": 614500
+    },
+    {
+      "epoch": 1.0583722407031033,
+      "grad_norm": 2.0854201316833496,
+      "learning_rate": 3.236046265494828e-05,
+      "loss": 0.8112,
+      "step": 615000
+    },
+    {
+      "epoch": 1.0592327059394473,
+      "grad_norm": 2.189030408859253,
+      "learning_rate": 3.234612156767588e-05,
+      "loss": 0.8158,
+      "step": 615500
+    },
+    {
+      "epoch": 1.0600931711757913,
+      "grad_norm": 2.0248332023620605,
+      "learning_rate": 3.233178048040348e-05,
+      "loss": 0.8105,
+      "step": 616000
+    },
+    {
+      "epoch": 1.0609536364121352,
+      "grad_norm": 1.9920114278793335,
+      "learning_rate": 3.2317439393131074e-05,
+      "loss": 0.8217,
+      "step": 616500
+    },
+    {
+      "epoch": 1.0618141016484792,
+      "grad_norm": 2.0297579765319824,
+      "learning_rate": 3.230309830585868e-05,
+      "loss": 0.8163,
+      "step": 617000
+    },
+    {
+      "epoch": 1.0626745668848232,
+      "grad_norm": 2.2208950519561768,
+      "learning_rate": 3.228875721858628e-05,
+      "loss": 0.8147,
+      "step": 617500
+    },
+    {
+      "epoch": 1.0635350321211672,
+      "grad_norm": 2.023688316345215,
+      "learning_rate": 3.227441613131388e-05,
+      "loss": 0.8121,
+      "step": 618000
+    },
+    {
+      "epoch": 1.0643954973575112,
+      "grad_norm": 2.1245839595794678,
+      "learning_rate": 3.226007504404148e-05,
+      "loss": 0.8113,
+      "step": 618500
+    },
+    {
+      "epoch": 1.0652559625938554,
+      "grad_norm": 2.0178346633911133,
+      "learning_rate": 3.224573395676908e-05,
+      "loss": 0.8145,
+      "step": 619000
+    },
+    {
+      "epoch": 1.0661164278301993,
+      "grad_norm": 2.1993958950042725,
+      "learning_rate": 3.223139286949668e-05,
+      "loss": 0.8236,
+      "step": 619500
+    },
+    {
+      "epoch": 1.0669768930665433,
+      "grad_norm": 1.9215905666351318,
+      "learning_rate": 3.221705178222428e-05,
+      "loss": 0.8134,
+      "step": 620000
+    },
+    {
+      "epoch": 1.0678373583028873,
+      "grad_norm": 2.226318836212158,
+      "learning_rate": 3.2202710694951885e-05,
+      "loss": 0.8177,
+      "step": 620500
+    },
+    {
+      "epoch": 1.0686978235392313,
+      "grad_norm": 1.9630376100540161,
+      "learning_rate": 3.218836960767948e-05,
+      "loss": 0.8103,
+      "step": 621000
+    },
+    {
+      "epoch": 1.0695582887755752,
+      "grad_norm": 1.9889850616455078,
+      "learning_rate": 3.2174028520407086e-05,
+      "loss": 0.8198,
+      "step": 621500
+    },
+    {
+      "epoch": 1.0704187540119192,
+      "grad_norm": 2.355071544647217,
+      "learning_rate": 3.215968743313468e-05,
+      "loss": 0.8156,
+      "step": 622000
+    },
+    {
+      "epoch": 1.0712792192482632,
+      "grad_norm": 2.0080490112304688,
+      "learning_rate": 3.214534634586228e-05,
+      "loss": 0.8094,
+      "step": 622500
+    },
+    {
+      "epoch": 1.0721396844846072,
+      "grad_norm": 2.099216938018799,
+      "learning_rate": 3.213100525858988e-05,
+      "loss": 0.8163,
+      "step": 623000
+    },
+    {
+      "epoch": 1.0730001497209511,
+      "grad_norm": 2.0304114818573,
+      "learning_rate": 3.211666417131749e-05,
+      "loss": 0.811,
+      "step": 623500
+    },
+    {
+      "epoch": 1.0738606149572951,
+      "grad_norm": 2.1474432945251465,
+      "learning_rate": 3.2102323084045085e-05,
+      "loss": 0.8196,
+      "step": 624000
+    },
+    {
+      "epoch": 1.074721080193639,
+      "grad_norm": 2.216209650039673,
+      "learning_rate": 3.208798199677268e-05,
+      "loss": 0.8149,
+      "step": 624500
+    },
+    {
+      "epoch": 1.075581545429983,
+      "grad_norm": 1.8711814880371094,
+      "learning_rate": 3.2073640909500286e-05,
+      "loss": 0.8258,
+      "step": 625000
+    },
+    {
+      "epoch": 1.076442010666327,
+      "grad_norm": 2.1735618114471436,
+      "learning_rate": 3.205929982222788e-05,
+      "loss": 0.813,
+      "step": 625500
+    },
+    {
+      "epoch": 1.077302475902671,
+      "grad_norm": 2.166551113128662,
+      "learning_rate": 3.204495873495548e-05,
+      "loss": 0.8189,
+      "step": 626000
+    },
+    {
+      "epoch": 1.078162941139015,
+      "grad_norm": 2.1133315563201904,
+      "learning_rate": 3.2030617647683084e-05,
+      "loss": 0.816,
+      "step": 626500
+    },
+    {
+      "epoch": 1.079023406375359,
+      "grad_norm": 2.0182137489318848,
+      "learning_rate": 3.201627656041069e-05,
+      "loss": 0.816,
+      "step": 627000
+    },
+    {
+      "epoch": 1.079883871611703,
+      "grad_norm": 2.062633991241455,
+      "learning_rate": 3.2001935473138284e-05,
+      "loss": 0.8186,
+      "step": 627500
+    },
+    {
+      "epoch": 1.080744336848047,
+      "grad_norm": 2.1116020679473877,
+      "learning_rate": 3.198759438586588e-05,
+      "loss": 0.8134,
+      "step": 628000
+    },
+    {
+      "epoch": 1.081604802084391,
+      "grad_norm": 2.0337700843811035,
+      "learning_rate": 3.1973253298593485e-05,
+      "loss": 0.8133,
+      "step": 628500
+    },
+    {
+      "epoch": 1.082465267320735,
+      "grad_norm": 1.9344936609268188,
+      "learning_rate": 3.195891221132108e-05,
+      "loss": 0.8107,
+      "step": 629000
+    },
+    {
+      "epoch": 1.0833257325570789,
+      "grad_norm": 2.3094825744628906,
+      "learning_rate": 3.1944571124048686e-05,
+      "loss": 0.8086,
+      "step": 629500
+    },
+    {
+      "epoch": 1.0841861977934228,
+      "grad_norm": 2.158808946609497,
+      "learning_rate": 3.193023003677629e-05,
+      "loss": 0.811,
+      "step": 630000
+    },
+    {
+      "epoch": 1.0850466630297668,
+      "grad_norm": 2.1025898456573486,
+      "learning_rate": 3.191588894950389e-05,
+      "loss": 0.8131,
+      "step": 630500
+    },
+    {
+      "epoch": 1.085907128266111,
+      "grad_norm": 2.234091281890869,
+      "learning_rate": 3.1901547862231484e-05,
+      "loss": 0.8186,
+      "step": 631000
+    },
+    {
+      "epoch": 1.086767593502455,
+      "grad_norm": 2.186819553375244,
+      "learning_rate": 3.188720677495909e-05,
+      "loss": 0.8114,
+      "step": 631500
+    },
+    {
+      "epoch": 1.087628058738799,
+      "grad_norm": 2.036576271057129,
+      "learning_rate": 3.1872865687686685e-05,
+      "loss": 0.8144,
+      "step": 632000
+    },
+    {
+      "epoch": 1.088488523975143,
+      "grad_norm": 2.1673030853271484,
+      "learning_rate": 3.185852460041428e-05,
+      "loss": 0.815,
+      "step": 632500
+    },
+    {
+      "epoch": 1.089348989211487,
+      "grad_norm": 2.0698742866516113,
+      "learning_rate": 3.184418351314189e-05,
+      "loss": 0.818,
+      "step": 633000
+    },
+    {
+      "epoch": 1.090209454447831,
+      "grad_norm": 2.974994659423828,
+      "learning_rate": 3.182984242586949e-05,
+      "loss": 0.8106,
+      "step": 633500
+    },
+    {
+      "epoch": 1.0910699196841749,
+      "grad_norm": 2.020418643951416,
+      "learning_rate": 3.181550133859709e-05,
+      "loss": 0.809,
+      "step": 634000
+    },
+    {
+      "epoch": 1.0919303849205189,
+      "grad_norm": 1.9625979661941528,
+      "learning_rate": 3.1801160251324684e-05,
+      "loss": 0.8067,
+      "step": 634500
+    },
+    {
+      "epoch": 1.0927908501568628,
+      "grad_norm": 2.004453420639038,
+      "learning_rate": 3.178681916405229e-05,
+      "loss": 0.81,
+      "step": 635000
+    },
+    {
+      "epoch": 1.0936513153932068,
+      "grad_norm": 2.179042339324951,
+      "learning_rate": 3.1772478076779885e-05,
+      "loss": 0.8178,
+      "step": 635500
+    },
+    {
+      "epoch": 1.0945117806295508,
+      "grad_norm": 2.153991937637329,
+      "learning_rate": 3.175813698950749e-05,
+      "loss": 0.8144,
+      "step": 636000
+    },
+    {
+      "epoch": 1.0953722458658948,
+      "grad_norm": 2.238384962081909,
+      "learning_rate": 3.174379590223509e-05,
+      "loss": 0.81,
+      "step": 636500
+    },
+    {
+      "epoch": 1.0962327111022387,
+      "grad_norm": 1.7696861028671265,
+      "learning_rate": 3.172945481496269e-05,
+      "loss": 0.8139,
+      "step": 637000
+    },
+    {
+      "epoch": 1.0970931763385827,
+      "grad_norm": 1.9494590759277344,
+      "learning_rate": 3.171511372769029e-05,
+      "loss": 0.8122,
+      "step": 637500
+    },
+    {
+      "epoch": 1.0979536415749267,
+      "grad_norm": 1.9551820755004883,
+      "learning_rate": 3.170077264041789e-05,
+      "loss": 0.8104,
+      "step": 638000
+    },
+    {
+      "epoch": 1.0988141068112707,
+      "grad_norm": 2.033318042755127,
+      "learning_rate": 3.168643155314549e-05,
+      "loss": 0.8071,
+      "step": 638500
+    },
+    {
+      "epoch": 1.0996745720476147,
+      "grad_norm": 2.080103874206543,
+      "learning_rate": 3.167209046587309e-05,
+      "loss": 0.8099,
+      "step": 639000
+    },
+    {
+      "epoch": 1.1005350372839586,
+      "grad_norm": 1.9861481189727783,
+      "learning_rate": 3.165774937860069e-05,
+      "loss": 0.8104,
+      "step": 639500
+    },
+    {
+      "epoch": 1.1013955025203026,
+      "grad_norm": 2.102166175842285,
+      "learning_rate": 3.164340829132829e-05,
+      "loss": 0.8092,
+      "step": 640000
+    },
+    {
+      "epoch": 1.1022559677566466,
+      "grad_norm": 2.0474326610565186,
+      "learning_rate": 3.162906720405589e-05,
+      "loss": 0.8123,
+      "step": 640500
+    },
+    {
+      "epoch": 1.1031164329929906,
+      "grad_norm": 1.8780696392059326,
+      "learning_rate": 3.161472611678349e-05,
+      "loss": 0.8145,
+      "step": 641000
+    },
+    {
+      "epoch": 1.1039768982293345,
+      "grad_norm": 2.027871608734131,
+      "learning_rate": 3.160038502951109e-05,
+      "loss": 0.8119,
+      "step": 641500
+    },
+    {
+      "epoch": 1.1048373634656787,
+      "grad_norm": 2.076012372970581,
+      "learning_rate": 3.158604394223869e-05,
+      "loss": 0.8084,
+      "step": 642000
+    },
+    {
+      "epoch": 1.1056978287020227,
+      "grad_norm": 2.0570199489593506,
+      "learning_rate": 3.157170285496629e-05,
+      "loss": 0.809,
+      "step": 642500
+    },
+    {
+      "epoch": 1.1065582939383667,
+      "grad_norm": 2.052778959274292,
+      "learning_rate": 3.1557361767693896e-05,
+      "loss": 0.8123,
+      "step": 643000
+    },
+    {
+      "epoch": 1.1074187591747107,
+      "grad_norm": 2.0170114040374756,
+      "learning_rate": 3.154302068042149e-05,
+      "loss": 0.8137,
+      "step": 643500
+    },
+    {
+      "epoch": 1.1082792244110546,
+      "grad_norm": 2.030034065246582,
+      "learning_rate": 3.152867959314909e-05,
+      "loss": 0.8127,
+      "step": 644000
+    },
+    {
+      "epoch": 1.1091396896473986,
+      "grad_norm": 2.0542829036712646,
+      "learning_rate": 3.1514338505876694e-05,
+      "loss": 0.8167,
+      "step": 644500
+    },
+    {
+      "epoch": 1.1100001548837426,
+      "grad_norm": 2.1526877880096436,
+      "learning_rate": 3.149999741860429e-05,
+      "loss": 0.8128,
+      "step": 645000
+    },
+    {
+      "epoch": 1.1108606201200866,
+      "grad_norm": 2.040294647216797,
+      "learning_rate": 3.1485656331331895e-05,
+      "loss": 0.815,
+      "step": 645500
+    },
+    {
+      "epoch": 1.1117210853564305,
+      "grad_norm": 2.10483717918396,
+      "learning_rate": 3.147131524405949e-05,
+      "loss": 0.8155,
+      "step": 646000
+    },
+    {
+      "epoch": 1.1125815505927745,
+      "grad_norm": 2.1790919303894043,
+      "learning_rate": 3.1456974156787096e-05,
+      "loss": 0.8048,
+      "step": 646500
+    },
+    {
+      "epoch": 1.1134420158291185,
+      "grad_norm": 2.011568784713745,
+      "learning_rate": 3.144263306951469e-05,
+      "loss": 0.8075,
+      "step": 647000
+    },
+    {
+      "epoch": 1.1143024810654625,
+      "grad_norm": 2.0351784229278564,
+      "learning_rate": 3.142829198224229e-05,
+      "loss": 0.8207,
+      "step": 647500
+    },
+    {
+      "epoch": 1.1151629463018065,
+      "grad_norm": 1.9416171312332153,
+      "learning_rate": 3.1413950894969894e-05,
+      "loss": 0.8096,
+      "step": 648000
+    },
+    {
+      "epoch": 1.1160234115381504,
+      "grad_norm": 1.9289367198944092,
+      "learning_rate": 3.13996098076975e-05,
+      "loss": 0.8144,
+      "step": 648500
+    },
+    {
+      "epoch": 1.1168838767744944,
+      "grad_norm": 2.048586845397949,
+      "learning_rate": 3.1385268720425095e-05,
+      "loss": 0.8122,
+      "step": 649000
+    },
+    {
+      "epoch": 1.1177443420108384,
+      "grad_norm": 2.136021852493286,
+      "learning_rate": 3.13709276331527e-05,
+      "loss": 0.8064,
+      "step": 649500
+    },
+    {
+      "epoch": 1.1186048072471824,
+      "grad_norm": 2.0434165000915527,
+      "learning_rate": 3.1356586545880295e-05,
+      "loss": 0.807,
+      "step": 650000
+    },
+    {
+      "epoch": 1.1194652724835263,
+      "grad_norm": 2.0496816635131836,
+      "learning_rate": 3.134224545860789e-05,
+      "loss": 0.8124,
+      "step": 650500
+    },
+    {
+      "epoch": 1.1203257377198703,
+      "grad_norm": 2.304429292678833,
+      "learning_rate": 3.132790437133549e-05,
+      "loss": 0.809,
+      "step": 651000
+    },
+    {
+      "epoch": 1.1211862029562143,
+      "grad_norm": 2.278761148452759,
+      "learning_rate": 3.1313563284063094e-05,
+      "loss": 0.8135,
+      "step": 651500
+    },
+    {
+      "epoch": 1.1220466681925583,
+      "grad_norm": 2.150486707687378,
+      "learning_rate": 3.12992221967907e-05,
+      "loss": 0.813,
+      "step": 652000
+    },
+    {
+      "epoch": 1.1229071334289022,
+      "grad_norm": 2.233739137649536,
+      "learning_rate": 3.1284881109518294e-05,
+      "loss": 0.811,
+      "step": 652500
+    },
+    {
+      "epoch": 1.1237675986652462,
+      "grad_norm": 2.1455652713775635,
+      "learning_rate": 3.12705400222459e-05,
+      "loss": 0.8113,
+      "step": 653000
+    },
+    {
+      "epoch": 1.1246280639015902,
+      "grad_norm": 2.0067386627197266,
+      "learning_rate": 3.1256198934973495e-05,
+      "loss": 0.8094,
+      "step": 653500
+    },
+    {
+      "epoch": 1.1254885291379342,
+      "grad_norm": 1.9603629112243652,
+      "learning_rate": 3.124185784770109e-05,
+      "loss": 0.8063,
+      "step": 654000
+    },
+    {
+      "epoch": 1.1263489943742784,
+      "grad_norm": 2.033787727355957,
+      "learning_rate": 3.1227516760428696e-05,
+      "loss": 0.812,
+      "step": 654500
+    },
+    {
+      "epoch": 1.1272094596106224,
+      "grad_norm": 2.0316808223724365,
+      "learning_rate": 3.12131756731563e-05,
+      "loss": 0.8169,
+      "step": 655000
+    },
+    {
+      "epoch": 1.1280699248469663,
+      "grad_norm": 1.9483590126037598,
+      "learning_rate": 3.11988345858839e-05,
+      "loss": 0.8063,
+      "step": 655500
+    },
+    {
+      "epoch": 1.1289303900833103,
+      "grad_norm": 1.9957387447357178,
+      "learning_rate": 3.11844934986115e-05,
+      "loss": 0.8052,
+      "step": 656000
+    },
+    {
+      "epoch": 1.1297908553196543,
+      "grad_norm": 2.2134573459625244,
+      "learning_rate": 3.11701524113391e-05,
+      "loss": 0.8047,
+      "step": 656500
+    },
+    {
+      "epoch": 1.1306513205559983,
+      "grad_norm": 2.1203839778900146,
+      "learning_rate": 3.1155811324066695e-05,
+      "loss": 0.8128,
+      "step": 657000
+    },
+    {
+      "epoch": 1.1315117857923422,
+      "grad_norm": 1.9745606184005737,
+      "learning_rate": 3.114147023679429e-05,
+      "loss": 0.8106,
+      "step": 657500
+    },
+    {
+      "epoch": 1.1323722510286862,
+      "grad_norm": 1.8720873594284058,
+      "learning_rate": 3.11271291495219e-05,
+      "loss": 0.803,
+      "step": 658000
+    },
+    {
+      "epoch": 1.1332327162650302,
+      "grad_norm": 2.2003602981567383,
+      "learning_rate": 3.11127880622495e-05,
+      "loss": 0.8057,
+      "step": 658500
+    },
+    {
+      "epoch": 1.1340931815013742,
+      "grad_norm": 1.8851550817489624,
+      "learning_rate": 3.10984469749771e-05,
+      "loss": 0.8054,
+      "step": 659000
+    },
+    {
+      "epoch": 1.1349536467377181,
+      "grad_norm": 2.158602714538574,
+      "learning_rate": 3.10841058877047e-05,
+      "loss": 0.8099,
+      "step": 659500
+    },
+    {
+      "epoch": 1.1358141119740621,
+      "grad_norm": 2.1378893852233887,
+      "learning_rate": 3.10697648004323e-05,
+      "loss": 0.8043,
+      "step": 660000
+    },
+    {
+      "epoch": 1.136674577210406,
+      "grad_norm": 2.137439489364624,
+      "learning_rate": 3.1055423713159895e-05,
+      "loss": 0.8173,
+      "step": 660500
+    },
+    {
+      "epoch": 1.13753504244675,
+      "grad_norm": 2.1251001358032227,
+      "learning_rate": 3.10410826258875e-05,
+      "loss": 0.8084,
+      "step": 661000
+    },
+    {
+      "epoch": 1.138395507683094,
+      "grad_norm": 1.9800820350646973,
+      "learning_rate": 3.10267415386151e-05,
+      "loss": 0.8085,
+      "step": 661500
+    },
+    {
+      "epoch": 1.139255972919438,
+      "grad_norm": 2.1576268672943115,
+      "learning_rate": 3.10124004513427e-05,
+      "loss": 0.8082,
+      "step": 662000
+    },
+    {
+      "epoch": 1.140116438155782,
+      "grad_norm": 2.003831624984741,
+      "learning_rate": 3.09980593640703e-05,
+      "loss": 0.816,
+      "step": 662500
+    },
+    {
+      "epoch": 1.140976903392126,
+      "grad_norm": 1.969831943511963,
+      "learning_rate": 3.09837182767979e-05,
+      "loss": 0.8124,
+      "step": 663000
+    },
+    {
+      "epoch": 1.14183736862847,
+      "grad_norm": 2.072331190109253,
+      "learning_rate": 3.09693771895255e-05,
+      "loss": 0.8093,
+      "step": 663500
+    },
+    {
+      "epoch": 1.142697833864814,
+      "grad_norm": 1.948601484298706,
+      "learning_rate": 3.09550361022531e-05,
+      "loss": 0.809,
+      "step": 664000
+    },
+    {
+      "epoch": 1.1435582991011581,
+      "grad_norm": 2.0233547687530518,
+      "learning_rate": 3.0940695014980706e-05,
+      "loss": 0.8039,
+      "step": 664500
+    },
+    {
+      "epoch": 1.1444187643375021,
+      "grad_norm": 2.019453287124634,
+      "learning_rate": 3.09263539277083e-05,
+      "loss": 0.8007,
+      "step": 665000
+    },
+    {
+      "epoch": 1.145279229573846,
+      "grad_norm": 1.9295461177825928,
+      "learning_rate": 3.09120128404359e-05,
+      "loss": 0.8081,
+      "step": 665500
+    },
+    {
+      "epoch": 1.14613969481019,
+      "grad_norm": 1.9981800317764282,
+      "learning_rate": 3.0897671753163504e-05,
+      "loss": 0.808,
+      "step": 666000
+    },
+    {
+      "epoch": 1.147000160046534,
+      "grad_norm": 2.1135873794555664,
+      "learning_rate": 3.08833306658911e-05,
+      "loss": 0.8186,
+      "step": 666500
+    },
+    {
+      "epoch": 1.147860625282878,
+      "grad_norm": 2.176304817199707,
+      "learning_rate": 3.08689895786187e-05,
+      "loss": 0.8106,
+      "step": 667000
+    },
+    {
+      "epoch": 1.148721090519222,
+      "grad_norm": 2.132354974746704,
+      "learning_rate": 3.085464849134631e-05,
+      "loss": 0.8062,
+      "step": 667500
+    },
+    {
+      "epoch": 1.149581555755566,
+      "grad_norm": 2.330860137939453,
+      "learning_rate": 3.0840307404073906e-05,
+      "loss": 0.8081,
+      "step": 668000
+    },
+    {
+      "epoch": 1.15044202099191,
+      "grad_norm": 2.1148693561553955,
+      "learning_rate": 3.08259663168015e-05,
+      "loss": 0.8119,
+      "step": 668500
+    },
+    {
+      "epoch": 1.151302486228254,
+      "grad_norm": 2.028310537338257,
+      "learning_rate": 3.08116252295291e-05,
+      "loss": 0.8051,
+      "step": 669000
+    },
+    {
+      "epoch": 1.152162951464598,
+      "grad_norm": 2.107956647872925,
+      "learning_rate": 3.0797284142256704e-05,
+      "loss": 0.811,
+      "step": 669500
+    },
+    {
+      "epoch": 1.1530234167009419,
+      "grad_norm": 2.11625599861145,
+      "learning_rate": 3.07829430549843e-05,
+      "loss": 0.8095,
+      "step": 670000
+    },
+    {
+      "epoch": 1.1538838819372859,
+      "grad_norm": 2.1002602577209473,
+      "learning_rate": 3.0768601967711905e-05,
+      "loss": 0.8072,
+      "step": 670500
+    },
+    {
+      "epoch": 1.1547443471736298,
+      "grad_norm": 2.5475449562072754,
+      "learning_rate": 3.075426088043951e-05,
+      "loss": 0.8082,
+      "step": 671000
+    },
+    {
+      "epoch": 1.1556048124099738,
+      "grad_norm": 1.9919356107711792,
+      "learning_rate": 3.0739919793167106e-05,
+      "loss": 0.813,
+      "step": 671500
+    },
+    {
+      "epoch": 1.1564652776463178,
+      "grad_norm": 1.9276983737945557,
+      "learning_rate": 3.07255787058947e-05,
+      "loss": 0.8069,
+      "step": 672000
+    },
+    {
+      "epoch": 1.1573257428826618,
+      "grad_norm": 2.14060115814209,
+      "learning_rate": 3.0711237618622306e-05,
+      "loss": 0.8045,
+      "step": 672500
+    },
+    {
+      "epoch": 1.1581862081190057,
+      "grad_norm": 2.1278932094573975,
+      "learning_rate": 3.0696896531349904e-05,
+      "loss": 0.7996,
+      "step": 673000
+    },
+    {
+      "epoch": 1.1590466733553497,
+      "grad_norm": 2.2851386070251465,
+      "learning_rate": 3.068255544407751e-05,
+      "loss": 0.8064,
+      "step": 673500
+    },
+    {
+      "epoch": 1.1599071385916937,
+      "grad_norm": 1.997027039527893,
+      "learning_rate": 3.0668214356805105e-05,
+      "loss": 0.8101,
+      "step": 674000
+    },
+    {
+      "epoch": 1.1607676038280377,
+      "grad_norm": 1.8951047658920288,
+      "learning_rate": 3.065387326953271e-05,
+      "loss": 0.8115,
+      "step": 674500
+    },
+    {
+      "epoch": 1.1616280690643817,
+      "grad_norm": 2.016145706176758,
+      "learning_rate": 3.0639532182260305e-05,
+      "loss": 0.8105,
+      "step": 675000
+    },
+    {
+      "epoch": 1.1624885343007256,
+      "grad_norm": 2.0284934043884277,
+      "learning_rate": 3.06251910949879e-05,
+      "loss": 0.8099,
+      "step": 675500
+    },
+    {
+      "epoch": 1.1633489995370696,
+      "grad_norm": 2.1791024208068848,
+      "learning_rate": 3.0610850007715506e-05,
+      "loss": 0.8073,
+      "step": 676000
+    },
+    {
+      "epoch": 1.1642094647734136,
+      "grad_norm": 2.1079273223876953,
+      "learning_rate": 3.0596508920443103e-05,
+      "loss": 0.8077,
+      "step": 676500
+    },
+    {
+      "epoch": 1.1650699300097576,
+      "grad_norm": 1.8703237771987915,
+      "learning_rate": 3.058216783317071e-05,
+      "loss": 0.8091,
+      "step": 677000
+    },
+    {
+      "epoch": 1.1659303952461018,
+      "grad_norm": 2.1863420009613037,
+      "learning_rate": 3.056782674589831e-05,
+      "loss": 0.804,
+      "step": 677500
+    },
+    {
+      "epoch": 1.1667908604824457,
+      "grad_norm": 2.140691041946411,
+      "learning_rate": 3.055348565862591e-05,
+      "loss": 0.806,
+      "step": 678000
+    },
+    {
+      "epoch": 1.1676513257187897,
+      "grad_norm": 2.0398032665252686,
+      "learning_rate": 3.0539144571353505e-05,
+      "loss": 0.8098,
+      "step": 678500
+    },
+    {
+      "epoch": 1.1685117909551337,
+      "grad_norm": 2.1374642848968506,
+      "learning_rate": 3.052480348408111e-05,
+      "loss": 0.8047,
+      "step": 679000
+    },
+    {
+      "epoch": 1.1693722561914777,
+      "grad_norm": 2.0204110145568848,
+      "learning_rate": 3.051046239680871e-05,
+      "loss": 0.8092,
+      "step": 679500
+    },
+    {
+      "epoch": 1.1702327214278216,
+      "grad_norm": 2.0928399562835693,
+      "learning_rate": 3.0496121309536307e-05,
+      "loss": 0.802,
+      "step": 680000
+    },
+    {
+      "epoch": 1.1710931866641656,
+      "grad_norm": 2.136254072189331,
+      "learning_rate": 3.0481780222263907e-05,
+      "loss": 0.8055,
+      "step": 680500
+    },
+    {
+      "epoch": 1.1719536519005096,
+      "grad_norm": 1.9934322834014893,
+      "learning_rate": 3.046743913499151e-05,
+      "loss": 0.8011,
+      "step": 681000
+    },
+    {
+      "epoch": 1.1728141171368536,
+      "grad_norm": 2.214151382446289,
+      "learning_rate": 3.0453098047719108e-05,
+      "loss": 0.8094,
+      "step": 681500
+    },
+    {
+      "epoch": 1.1736745823731976,
+      "grad_norm": 2.0825462341308594,
+      "learning_rate": 3.043875696044671e-05,
+      "loss": 0.8137,
+      "step": 682000
+    },
+    {
+      "epoch": 1.1745350476095415,
+      "grad_norm": 1.945638656616211,
+      "learning_rate": 3.0424415873174313e-05,
+      "loss": 0.8055,
+      "step": 682500
+    },
+    {
+      "epoch": 1.1753955128458855,
+      "grad_norm": 1.9353575706481934,
+      "learning_rate": 3.041007478590191e-05,
+      "loss": 0.8028,
+      "step": 683000
+    },
+    {
+      "epoch": 1.1762559780822295,
+      "grad_norm": 2.180316686630249,
+      "learning_rate": 3.0395733698629507e-05,
+      "loss": 0.8095,
+      "step": 683500
+    },
+    {
+      "epoch": 1.1771164433185735,
+      "grad_norm": 2.010213851928711,
+      "learning_rate": 3.0381392611357114e-05,
+      "loss": 0.806,
+      "step": 684000
+    },
+    {
+      "epoch": 1.1779769085549174,
+      "grad_norm": 2.019655227661133,
+      "learning_rate": 3.036705152408471e-05,
+      "loss": 0.8023,
+      "step": 684500
+    },
+    {
+      "epoch": 1.1788373737912614,
+      "grad_norm": 2.287731170654297,
+      "learning_rate": 3.0352710436812308e-05,
+      "loss": 0.8078,
+      "step": 685000
+    },
+    {
+      "epoch": 1.1796978390276054,
+      "grad_norm": 1.9522863626480103,
+      "learning_rate": 3.033836934953991e-05,
+      "loss": 0.8051,
+      "step": 685500
+    },
+    {
+      "epoch": 1.1805583042639494,
+      "grad_norm": 1.9352096319198608,
+      "learning_rate": 3.0324028262267512e-05,
+      "loss": 0.8016,
+      "step": 686000
+    },
+    {
+      "epoch": 1.1814187695002933,
+      "grad_norm": 1.9841794967651367,
+      "learning_rate": 3.030968717499511e-05,
+      "loss": 0.8011,
+      "step": 686500
+    },
+    {
+      "epoch": 1.1822792347366373,
+      "grad_norm": 1.9001405239105225,
+      "learning_rate": 3.029534608772271e-05,
+      "loss": 0.8056,
+      "step": 687000
+    },
+    {
+      "epoch": 1.1831396999729813,
+      "grad_norm": 1.9476912021636963,
+      "learning_rate": 3.0281005000450314e-05,
+      "loss": 0.8004,
+      "step": 687500
+    },
+    {
+      "epoch": 1.1840001652093255,
+      "grad_norm": 2.282691717147827,
+      "learning_rate": 3.026666391317791e-05,
+      "loss": 0.806,
+      "step": 688000
+    },
+    {
+      "epoch": 1.1848606304456695,
+      "grad_norm": 2.1164588928222656,
+      "learning_rate": 3.025232282590551e-05,
+      "loss": 0.8096,
+      "step": 688500
+    },
+    {
+      "epoch": 1.1857210956820134,
+      "grad_norm": 2.249427318572998,
+      "learning_rate": 3.0237981738633115e-05,
+      "loss": 0.8067,
+      "step": 689000
+    },
+    {
+      "epoch": 1.1865815609183574,
+      "grad_norm": 2.1805477142333984,
+      "learning_rate": 3.0223640651360712e-05,
+      "loss": 0.8049,
+      "step": 689500
+    },
+    {
+      "epoch": 1.1874420261547014,
+      "grad_norm": 2.3295178413391113,
+      "learning_rate": 3.0209299564088313e-05,
+      "loss": 0.8091,
+      "step": 690000
+    },
+    {
+      "epoch": 1.1883024913910454,
+      "grad_norm": 2.17014741897583,
+      "learning_rate": 3.0194958476815917e-05,
+      "loss": 0.8094,
+      "step": 690500
+    },
+    {
+      "epoch": 1.1891629566273894,
+      "grad_norm": 1.8217567205429077,
+      "learning_rate": 3.0180617389543514e-05,
+      "loss": 0.8088,
+      "step": 691000
+    },
+    {
+      "epoch": 1.1900234218637333,
+      "grad_norm": 2.156625747680664,
+      "learning_rate": 3.0166276302271114e-05,
+      "loss": 0.8039,
+      "step": 691500
+    },
+    {
+      "epoch": 1.1908838871000773,
+      "grad_norm": 1.924224853515625,
+      "learning_rate": 3.015193521499871e-05,
+      "loss": 0.7991,
+      "step": 692000
+    },
+    {
+      "epoch": 1.1917443523364213,
+      "grad_norm": 2.1445975303649902,
+      "learning_rate": 3.0137594127726315e-05,
+      "loss": 0.803,
+      "step": 692500
+    },
+    {
+      "epoch": 1.1926048175727653,
+      "grad_norm": 2.235790729522705,
+      "learning_rate": 3.0123253040453912e-05,
+      "loss": 0.8096,
+      "step": 693000
+    },
+    {
+      "epoch": 1.1934652828091092,
+      "grad_norm": 1.9758915901184082,
+      "learning_rate": 3.0108911953181513e-05,
+      "loss": 0.8016,
+      "step": 693500
+    },
+    {
+      "epoch": 1.1943257480454532,
+      "grad_norm": 2.021066427230835,
+      "learning_rate": 3.0094570865909117e-05,
+      "loss": 0.8035,
+      "step": 694000
+    },
+    {
+      "epoch": 1.1951862132817972,
+      "grad_norm": 2.056852340698242,
+      "learning_rate": 3.0080229778636714e-05,
+      "loss": 0.806,
+      "step": 694500
+    },
+    {
+      "epoch": 1.1960466785181412,
+      "grad_norm": 2.177889347076416,
+      "learning_rate": 3.0065888691364314e-05,
+      "loss": 0.8104,
+      "step": 695000
+    },
+    {
+      "epoch": 1.1969071437544851,
+      "grad_norm": 2.194106340408325,
+      "learning_rate": 3.0051547604091918e-05,
+      "loss": 0.8062,
+      "step": 695500
+    },
+    {
+      "epoch": 1.1977676089908291,
+      "grad_norm": 2.0299899578094482,
+      "learning_rate": 3.0037206516819515e-05,
+      "loss": 0.801,
+      "step": 696000
+    },
+    {
+      "epoch": 1.198628074227173,
+      "grad_norm": 2.2731423377990723,
+      "learning_rate": 3.0022865429547116e-05,
+      "loss": 0.8099,
+      "step": 696500
+    },
+    {
+      "epoch": 1.199488539463517,
+      "grad_norm": 2.029911756515503,
+      "learning_rate": 3.0008524342274713e-05,
+      "loss": 0.807,
+      "step": 697000
+    },
+    {
+      "epoch": 1.200349004699861,
+      "grad_norm": 2.0720105171203613,
+      "learning_rate": 2.9994183255002316e-05,
+      "loss": 0.8114,
+      "step": 697500
+    },
+    {
+      "epoch": 1.201209469936205,
+      "grad_norm": 1.9225728511810303,
+      "learning_rate": 2.9979842167729917e-05,
+      "loss": 0.8052,
+      "step": 698000
+    },
+    {
+      "epoch": 1.202069935172549,
+      "grad_norm": 2.078575849533081,
+      "learning_rate": 2.9965501080457514e-05,
+      "loss": 0.8064,
+      "step": 698500
+    },
+    {
+      "epoch": 1.202930400408893,
+      "grad_norm": 3.18145751953125,
+      "learning_rate": 2.9951159993185118e-05,
+      "loss": 0.8038,
+      "step": 699000
+    },
+    {
+      "epoch": 1.203790865645237,
+      "grad_norm": 2.0273611545562744,
+      "learning_rate": 2.993681890591272e-05,
+      "loss": 0.807,
+      "step": 699500
+    },
+    {
+      "epoch": 1.204651330881581,
+      "grad_norm": 2.17539119720459,
+      "learning_rate": 2.9922477818640315e-05,
+      "loss": 0.8059,
+      "step": 700000
+    },
+    {
+      "epoch": 1.205511796117925,
+      "grad_norm": 2.16237211227417,
+      "learning_rate": 2.990813673136792e-05,
+      "loss": 0.8083,
+      "step": 700500
+    },
+    {
+      "epoch": 1.2063722613542691,
+      "grad_norm": 2.004716634750366,
+      "learning_rate": 2.9893795644095516e-05,
+      "loss": 0.8029,
+      "step": 701000
+    },
+    {
+      "epoch": 1.207232726590613,
+      "grad_norm": 2.149090528488159,
+      "learning_rate": 2.9879454556823117e-05,
+      "loss": 0.8038,
+      "step": 701500
+    },
+    {
+      "epoch": 1.208093191826957,
+      "grad_norm": 2.193436622619629,
+      "learning_rate": 2.986511346955072e-05,
+      "loss": 0.8105,
+      "step": 702000
+    },
+    {
+      "epoch": 1.208953657063301,
+      "grad_norm": 2.0396177768707275,
+      "learning_rate": 2.9850772382278318e-05,
+      "loss": 0.8075,
+      "step": 702500
+    },
+    {
+      "epoch": 1.209814122299645,
+      "grad_norm": 2.060792922973633,
+      "learning_rate": 2.9836431295005918e-05,
+      "loss": 0.8064,
+      "step": 703000
+    },
+    {
+      "epoch": 1.210674587535989,
+      "grad_norm": 2.02762508392334,
+      "learning_rate": 2.9822090207733515e-05,
+      "loss": 0.8028,
+      "step": 703500
+    },
+    {
+      "epoch": 1.211535052772333,
+      "grad_norm": 2.1213901042938232,
+      "learning_rate": 2.980774912046112e-05,
+      "loss": 0.8069,
+      "step": 704000
+    },
+    {
+      "epoch": 1.212395518008677,
+      "grad_norm": 2.2344391345977783,
+      "learning_rate": 2.979340803318872e-05,
+      "loss": 0.7963,
+      "step": 704500
+    },
+    {
+      "epoch": 1.213255983245021,
+      "grad_norm": 1.9241515398025513,
+      "learning_rate": 2.9779066945916317e-05,
+      "loss": 0.803,
+      "step": 705000
+    },
+    {
+      "epoch": 1.214116448481365,
+      "grad_norm": 2.2126944065093994,
+      "learning_rate": 2.976472585864392e-05,
+      "loss": 0.8061,
+      "step": 705500
+    },
+    {
+      "epoch": 1.2149769137177089,
+      "grad_norm": 2.169583320617676,
+      "learning_rate": 2.975038477137152e-05,
+      "loss": 0.8037,
+      "step": 706000
+    },
+    {
+      "epoch": 1.2158373789540529,
+      "grad_norm": 2.132267475128174,
+      "learning_rate": 2.9736043684099118e-05,
+      "loss": 0.807,
+      "step": 706500
+    },
+    {
+      "epoch": 1.2166978441903968,
+      "grad_norm": 2.0939061641693115,
+      "learning_rate": 2.9721702596826722e-05,
+      "loss": 0.8023,
+      "step": 707000
+    },
+    {
+      "epoch": 1.2175583094267408,
+      "grad_norm": 2.0683717727661133,
+      "learning_rate": 2.9707361509554322e-05,
+      "loss": 0.8027,
+      "step": 707500
+    },
+    {
+      "epoch": 1.2184187746630848,
+      "grad_norm": 2.210061550140381,
+      "learning_rate": 2.969302042228192e-05,
+      "loss": 0.8008,
+      "step": 708000
+    },
+    {
+      "epoch": 1.2192792398994288,
+      "grad_norm": 2.072613000869751,
+      "learning_rate": 2.9678679335009523e-05,
+      "loss": 0.8027,
+      "step": 708500
+    },
+    {
+      "epoch": 1.2201397051357727,
+      "grad_norm": 2.157541275024414,
+      "learning_rate": 2.9664338247737124e-05,
+      "loss": 0.8038,
+      "step": 709000
+    },
+    {
+      "epoch": 1.2210001703721167,
+      "grad_norm": 2.2376465797424316,
+      "learning_rate": 2.964999716046472e-05,
+      "loss": 0.799,
+      "step": 709500
+    },
+    {
+      "epoch": 1.2218606356084607,
+      "grad_norm": 2.337463855743408,
+      "learning_rate": 2.9635656073192318e-05,
+      "loss": 0.8016,
+      "step": 710000
+    },
+    {
+      "epoch": 1.2227211008448047,
+      "grad_norm": 2.2762887477874756,
+      "learning_rate": 2.9621314985919922e-05,
+      "loss": 0.8053,
+      "step": 710500
+    },
+    {
+      "epoch": 1.2235815660811489,
+      "grad_norm": 2.1711015701293945,
+      "learning_rate": 2.9606973898647522e-05,
+      "loss": 0.8064,
+      "step": 711000
+    },
+    {
+      "epoch": 1.2244420313174929,
+      "grad_norm": 2.0335445404052734,
+      "learning_rate": 2.959263281137512e-05,
+      "loss": 0.8029,
+      "step": 711500
+    },
+    {
+      "epoch": 1.2253024965538368,
+      "grad_norm": 2.1329965591430664,
+      "learning_rate": 2.9578291724102723e-05,
+      "loss": 0.8068,
+      "step": 712000
+    },
+    {
+      "epoch": 1.2261629617901808,
+      "grad_norm": 2.1596531867980957,
+      "learning_rate": 2.9563950636830324e-05,
+      "loss": 0.8075,
+      "step": 712500
+    },
+    {
+      "epoch": 1.2270234270265248,
+      "grad_norm": 1.909252643585205,
+      "learning_rate": 2.954960954955792e-05,
+      "loss": 0.8048,
+      "step": 713000
+    },
+    {
+      "epoch": 1.2278838922628688,
+      "grad_norm": 1.9895068407058716,
+      "learning_rate": 2.9535268462285525e-05,
+      "loss": 0.8019,
+      "step": 713500
+    },
+    {
+      "epoch": 1.2287443574992127,
+      "grad_norm": 2.0683231353759766,
+      "learning_rate": 2.9520927375013125e-05,
+      "loss": 0.8019,
+      "step": 714000
+    },
+    {
+      "epoch": 1.2296048227355567,
+      "grad_norm": 2.2134108543395996,
+      "learning_rate": 2.9506586287740722e-05,
+      "loss": 0.8018,
+      "step": 714500
+    },
+    {
+      "epoch": 1.2304652879719007,
+      "grad_norm": 1.9018226861953735,
+      "learning_rate": 2.9492245200468323e-05,
+      "loss": 0.8085,
+      "step": 715000
+    },
+    {
+      "epoch": 1.2313257532082447,
+      "grad_norm": 2.1485302448272705,
+      "learning_rate": 2.9477904113195927e-05,
+      "loss": 0.7978,
+      "step": 715500
+    },
+    {
+      "epoch": 1.2321862184445886,
+      "grad_norm": 2.0110764503479004,
+      "learning_rate": 2.9463563025923524e-05,
+      "loss": 0.7995,
+      "step": 716000
+    },
+    {
+      "epoch": 1.2330466836809326,
+      "grad_norm": 1.984613299369812,
+      "learning_rate": 2.9449221938651124e-05,
+      "loss": 0.8049,
+      "step": 716500
+    },
+    {
+      "epoch": 1.2339071489172766,
+      "grad_norm": 2.059112787246704,
+      "learning_rate": 2.9434880851378728e-05,
+      "loss": 0.807,
+      "step": 717000
+    },
+    {
+      "epoch": 1.2347676141536206,
+      "grad_norm": 2.2094051837921143,
+      "learning_rate": 2.9420539764106325e-05,
+      "loss": 0.8006,
+      "step": 717500
+    },
+    {
+      "epoch": 1.2356280793899646,
+      "grad_norm": 2.09016752243042,
+      "learning_rate": 2.9406198676833922e-05,
+      "loss": 0.8031,
+      "step": 718000
+    },
+    {
+      "epoch": 1.2364885446263085,
+      "grad_norm": 1.952208399772644,
+      "learning_rate": 2.939185758956153e-05,
+      "loss": 0.8045,
+      "step": 718500
+    },
+    {
+      "epoch": 1.2373490098626525,
+      "grad_norm": 2.049285411834717,
+      "learning_rate": 2.9377516502289127e-05,
+      "loss": 0.8032,
+      "step": 719000
+    },
+    {
+      "epoch": 1.2382094750989965,
+      "grad_norm": 2.2089593410491943,
+      "learning_rate": 2.9363175415016724e-05,
+      "loss": 0.8082,
+      "step": 719500
+    },
+    {
+      "epoch": 1.2390699403353405,
+      "grad_norm": 2.038893699645996,
+      "learning_rate": 2.9348834327744327e-05,
+      "loss": 0.8035,
+      "step": 720000
+    },
+    {
+      "epoch": 1.2399304055716844,
+      "grad_norm": 1.9960216283798218,
+      "learning_rate": 2.9334493240471928e-05,
+      "loss": 0.8006,
+      "step": 720500
+    },
+    {
+      "epoch": 1.2407908708080284,
+      "grad_norm": 1.9117454290390015,
+      "learning_rate": 2.9320152153199525e-05,
+      "loss": 0.8069,
+      "step": 721000
+    },
+    {
+      "epoch": 1.2416513360443724,
+      "grad_norm": 1.9376147985458374,
+      "learning_rate": 2.9305811065927125e-05,
+      "loss": 0.7984,
+      "step": 721500
+    },
+    {
+      "epoch": 1.2425118012807164,
+      "grad_norm": 2.009612798690796,
+      "learning_rate": 2.929146997865473e-05,
+      "loss": 0.7956,
+      "step": 722000
+    },
+    {
+      "epoch": 1.2433722665170603,
+      "grad_norm": 2.106132745742798,
+      "learning_rate": 2.9277128891382326e-05,
+      "loss": 0.8042,
+      "step": 722500
+    },
+    {
+      "epoch": 1.2442327317534043,
+      "grad_norm": 2.345006227493286,
+      "learning_rate": 2.9262787804109927e-05,
+      "loss": 0.8082,
+      "step": 723000
+    },
+    {
+      "epoch": 1.2450931969897483,
+      "grad_norm": 2.0162105560302734,
+      "learning_rate": 2.924844671683753e-05,
+      "loss": 0.7961,
+      "step": 723500
+    },
+    {
+      "epoch": 1.2459536622260925,
+      "grad_norm": 2.337620735168457,
+      "learning_rate": 2.9234105629565128e-05,
+      "loss": 0.8081,
+      "step": 724000
+    },
+    {
+      "epoch": 1.2468141274624365,
+      "grad_norm": 1.895340085029602,
+      "learning_rate": 2.921976454229273e-05,
+      "loss": 0.7943,
+      "step": 724500
+    },
+    {
+      "epoch": 1.2476745926987804,
+      "grad_norm": 2.0250372886657715,
+      "learning_rate": 2.9205423455020332e-05,
+      "loss": 0.8062,
+      "step": 725000
+    },
+    {
+      "epoch": 1.2485350579351244,
+      "grad_norm": 2.109849691390991,
+      "learning_rate": 2.919108236774793e-05,
+      "loss": 0.8067,
+      "step": 725500
+    },
+    {
+      "epoch": 1.2493955231714684,
+      "grad_norm": 1.9787213802337646,
+      "learning_rate": 2.9176741280475526e-05,
+      "loss": 0.7954,
+      "step": 726000
+    },
+    {
+      "epoch": 1.2502559884078124,
+      "grad_norm": 1.98014235496521,
+      "learning_rate": 2.9162400193203127e-05,
+      "loss": 0.8023,
+      "step": 726500
+    },
+    {
+      "epoch": 1.2511164536441564,
+      "grad_norm": 2.0103554725646973,
+      "learning_rate": 2.914805910593073e-05,
+      "loss": 0.799,
+      "step": 727000
+    },
+    {
+      "epoch": 1.2519769188805003,
+      "grad_norm": 2.017174005508423,
+      "learning_rate": 2.9133718018658328e-05,
+      "loss": 0.7974,
+      "step": 727500
+    },
+    {
+      "epoch": 1.2528373841168443,
+      "grad_norm": 2.1232900619506836,
+      "learning_rate": 2.9119376931385928e-05,
+      "loss": 0.8008,
+      "step": 728000
+    },
+    {
+      "epoch": 1.2536978493531883,
+      "grad_norm": 1.8166762590408325,
+      "learning_rate": 2.9105035844113532e-05,
+      "loss": 0.7992,
+      "step": 728500
+    },
+    {
+      "epoch": 1.2545583145895323,
+      "grad_norm": 2.131291627883911,
+      "learning_rate": 2.909069475684113e-05,
+      "loss": 0.7965,
+      "step": 729000
+    },
+    {
+      "epoch": 1.2554187798258762,
+      "grad_norm": 1.8457263708114624,
+      "learning_rate": 2.907635366956873e-05,
+      "loss": 0.7992,
+      "step": 729500
+    },
+    {
+      "epoch": 1.2562792450622202,
+      "grad_norm": 2.3724331855773926,
+      "learning_rate": 2.9062012582296333e-05,
+      "loss": 0.7944,
+      "step": 730000
+    },
+    {
+      "epoch": 1.2571397102985642,
+      "grad_norm": 1.9103881120681763,
+      "learning_rate": 2.904767149502393e-05,
+      "loss": 0.8008,
+      "step": 730500
+    },
+    {
+      "epoch": 1.2580001755349082,
+      "grad_norm": 2.1090755462646484,
+      "learning_rate": 2.903333040775153e-05,
+      "loss": 0.7935,
+      "step": 731000
+    },
+    {
+      "epoch": 1.2588606407712521,
+      "grad_norm": 2.2821147441864014,
+      "learning_rate": 2.9018989320479135e-05,
+      "loss": 0.8027,
+      "step": 731500
+    },
+    {
+      "epoch": 1.2597211060075961,
+      "grad_norm": 2.159932851791382,
+      "learning_rate": 2.9004648233206732e-05,
+      "loss": 0.8053,
+      "step": 732000
+    },
+    {
+      "epoch": 1.26058157124394,
+      "grad_norm": 2.012387752532959,
+      "learning_rate": 2.8990307145934332e-05,
+      "loss": 0.8017,
+      "step": 732500
+    },
+    {
+      "epoch": 1.261442036480284,
+      "grad_norm": 2.0438127517700195,
+      "learning_rate": 2.897596605866193e-05,
+      "loss": 0.7992,
+      "step": 733000
+    },
+    {
+      "epoch": 1.2623025017166283,
+      "grad_norm": 2.1479904651641846,
+      "learning_rate": 2.8961624971389533e-05,
+      "loss": 0.8013,
+      "step": 733500
+    },
+    {
+      "epoch": 1.2631629669529723,
+      "grad_norm": 2.248114585876465,
+      "learning_rate": 2.8947283884117134e-05,
+      "loss": 0.7976,
+      "step": 734000
+    },
+    {
+      "epoch": 1.2640234321893162,
+      "grad_norm": 1.8806530237197876,
+      "learning_rate": 2.893294279684473e-05,
+      "loss": 0.8029,
+      "step": 734500
+    },
+    {
+      "epoch": 1.2648838974256602,
+      "grad_norm": 1.8940871953964233,
+      "learning_rate": 2.8918601709572335e-05,
+      "loss": 0.797,
+      "step": 735000
+    },
+    {
+      "epoch": 1.2657443626620042,
+      "grad_norm": 2.128809928894043,
+      "learning_rate": 2.8904260622299932e-05,
+      "loss": 0.7958,
+      "step": 735500
+    },
+    {
+      "epoch": 1.2666048278983482,
+      "grad_norm": 2.1618053913116455,
+      "learning_rate": 2.8889919535027532e-05,
+      "loss": 0.7974,
+      "step": 736000
+    },
+    {
+      "epoch": 1.2674652931346921,
+      "grad_norm": 2.073007345199585,
+      "learning_rate": 2.8875578447755136e-05,
+      "loss": 0.8066,
+      "step": 736500
+    },
+    {
+      "epoch": 1.2683257583710361,
+      "grad_norm": 1.8977770805358887,
+      "learning_rate": 2.8861237360482733e-05,
+      "loss": 0.7984,
+      "step": 737000
+    },
+    {
+      "epoch": 1.26918622360738,
+      "grad_norm": 2.1680843830108643,
+      "learning_rate": 2.8846896273210334e-05,
+      "loss": 0.799,
+      "step": 737500
+    },
+    {
+      "epoch": 1.270046688843724,
+      "grad_norm": 1.998214602470398,
+      "learning_rate": 2.883255518593793e-05,
+      "loss": 0.7951,
+      "step": 738000
+    },
+    {
+      "epoch": 1.270907154080068,
+      "grad_norm": 2.0669476985931396,
+      "learning_rate": 2.8818214098665535e-05,
+      "loss": 0.7962,
+      "step": 738500
+    },
+    {
+      "epoch": 1.271767619316412,
+      "grad_norm": 2.1200294494628906,
+      "learning_rate": 2.8803873011393135e-05,
+      "loss": 0.795,
+      "step": 739000
+    },
+    {
+      "epoch": 1.272628084552756,
+      "grad_norm": 2.27681040763855,
+      "learning_rate": 2.8789531924120732e-05,
+      "loss": 0.7988,
+      "step": 739500
+    },
+    {
+      "epoch": 1.2734885497891,
+      "grad_norm": 2.213317394256592,
+      "learning_rate": 2.8775190836848336e-05,
+      "loss": 0.7966,
+      "step": 740000
+    },
+    {
+      "epoch": 1.274349015025444,
+      "grad_norm": 2.038018226623535,
+      "learning_rate": 2.8760849749575937e-05,
+      "loss": 0.7996,
+      "step": 740500
+    },
+    {
+      "epoch": 1.275209480261788,
+      "grad_norm": 2.1453094482421875,
+      "learning_rate": 2.8746508662303534e-05,
+      "loss": 0.7944,
+      "step": 741000
+    },
+    {
+      "epoch": 1.276069945498132,
+      "grad_norm": 2.005464553833008,
+      "learning_rate": 2.8732167575031138e-05,
+      "loss": 0.7977,
+      "step": 741500
+    },
+    {
+      "epoch": 1.2769304107344759,
+      "grad_norm": 2.018033027648926,
+      "learning_rate": 2.8717826487758738e-05,
+      "loss": 0.8066,
+      "step": 742000
+    },
+    {
+      "epoch": 1.2777908759708199,
+      "grad_norm": 2.1077606678009033,
+      "learning_rate": 2.8703485400486335e-05,
+      "loss": 0.8033,
+      "step": 742500
+    },
+    {
+      "epoch": 1.2786513412071638,
+      "grad_norm": 2.3417065143585205,
+      "learning_rate": 2.868914431321394e-05,
+      "loss": 0.8019,
+      "step": 743000
+    },
+    {
+      "epoch": 1.2795118064435078,
+      "grad_norm": 2.445237636566162,
+      "learning_rate": 2.867480322594154e-05,
+      "loss": 0.7984,
+      "step": 743500
+    },
+    {
+      "epoch": 1.2803722716798518,
+      "grad_norm": 2.2032246589660645,
+      "learning_rate": 2.8660462138669136e-05,
+      "loss": 0.7971,
+      "step": 744000
+    },
+    {
+      "epoch": 1.2812327369161958,
+      "grad_norm": 2.074751615524292,
+      "learning_rate": 2.8646121051396734e-05,
+      "loss": 0.799,
+      "step": 744500
+    },
+    {
+      "epoch": 1.2820932021525397,
+      "grad_norm": 2.1933205127716064,
+      "learning_rate": 2.8631779964124337e-05,
+      "loss": 0.7944,
+      "step": 745000
+    },
+    {
+      "epoch": 1.2829536673888837,
+      "grad_norm": 2.374495029449463,
+      "learning_rate": 2.8617438876851938e-05,
+      "loss": 0.7947,
+      "step": 745500
+    },
+    {
+      "epoch": 1.2838141326252277,
+      "grad_norm": 2.0120351314544678,
+      "learning_rate": 2.8603097789579535e-05,
+      "loss": 0.7951,
+      "step": 746000
+    },
+    {
+      "epoch": 1.2846745978615717,
+      "grad_norm": 2.224721908569336,
+      "learning_rate": 2.858875670230714e-05,
+      "loss": 0.7985,
+      "step": 746500
+    },
+    {
+      "epoch": 1.2855350630979157,
+      "grad_norm": 1.875423789024353,
+      "learning_rate": 2.857441561503474e-05,
+      "loss": 0.7913,
+      "step": 747000
+    },
+    {
+      "epoch": 1.2863955283342596,
+      "grad_norm": 2.116672992706299,
+      "learning_rate": 2.8560074527762336e-05,
+      "loss": 0.8006,
+      "step": 747500
+    },
+    {
+      "epoch": 1.2872559935706038,
+      "grad_norm": 2.1012139320373535,
+      "learning_rate": 2.854573344048994e-05,
+      "loss": 0.8027,
+      "step": 748000
+    },
+    {
+      "epoch": 1.2881164588069478,
+      "grad_norm": 2.454265594482422,
+      "learning_rate": 2.853139235321754e-05,
+      "loss": 0.7992,
+      "step": 748500
+    },
+    {
+      "epoch": 1.2889769240432918,
+      "grad_norm": 2.262613534927368,
+      "learning_rate": 2.8517051265945138e-05,
+      "loss": 0.7956,
+      "step": 749000
+    },
+    {
+      "epoch": 1.2898373892796358,
+      "grad_norm": 2.1560659408569336,
+      "learning_rate": 2.8502710178672738e-05,
+      "loss": 0.7982,
+      "step": 749500
+    },
+    {
+      "epoch": 1.2906978545159797,
+      "grad_norm": 2.0437734127044678,
+      "learning_rate": 2.8488369091400342e-05,
+      "loss": 0.791,
+      "step": 750000
+    },
+    {
+      "epoch": 1.2915583197523237,
+      "grad_norm": 2.1690988540649414,
+      "learning_rate": 2.847402800412794e-05,
+      "loss": 0.796,
+      "step": 750500
+    },
+    {
+      "epoch": 1.2924187849886677,
+      "grad_norm": 2.0291025638580322,
+      "learning_rate": 2.845968691685554e-05,
+      "loss": 0.7942,
+      "step": 751000
+    },
+    {
+      "epoch": 1.2932792502250117,
+      "grad_norm": 2.1441104412078857,
+      "learning_rate": 2.8445345829583144e-05,
+      "loss": 0.795,
+      "step": 751500
+    },
+    {
+      "epoch": 1.2941397154613556,
+      "grad_norm": 2.0123281478881836,
+      "learning_rate": 2.843100474231074e-05,
+      "loss": 0.8002,
+      "step": 752000
+    },
+    {
+      "epoch": 1.2950001806976996,
+      "grad_norm": 2.0292112827301025,
+      "learning_rate": 2.8416663655038338e-05,
+      "loss": 0.8004,
+      "step": 752500
+    },
+    {
+      "epoch": 1.2958606459340436,
+      "grad_norm": 2.077418565750122,
+      "learning_rate": 2.840232256776594e-05,
+      "loss": 0.79,
+      "step": 753000
+    },
+    {
+      "epoch": 1.2967211111703876,
+      "grad_norm": 1.8476754426956177,
+      "learning_rate": 2.8387981480493542e-05,
+      "loss": 0.7929,
+      "step": 753500
+    },
+    {
+      "epoch": 1.2975815764067316,
+      "grad_norm": 2.1577837467193604,
+      "learning_rate": 2.837364039322114e-05,
+      "loss": 0.7975,
+      "step": 754000
+    },
+    {
+      "epoch": 1.2984420416430755,
+      "grad_norm": 2.1313822269439697,
+      "learning_rate": 2.8359299305948743e-05,
+      "loss": 0.8,
+      "step": 754500
+    },
+    {
+      "epoch": 1.2993025068794195,
+      "grad_norm": 1.9185123443603516,
+      "learning_rate": 2.8344958218676343e-05,
+      "loss": 0.797,
+      "step": 755000
+    },
+    {
+      "epoch": 1.3001629721157635,
+      "grad_norm": 2.139822006225586,
+      "learning_rate": 2.833061713140394e-05,
+      "loss": 0.7926,
+      "step": 755500
+    },
+    {
+      "epoch": 1.3010234373521075,
+      "grad_norm": 2.4005961418151855,
+      "learning_rate": 2.831627604413154e-05,
+      "loss": 0.7925,
+      "step": 756000
+    },
+    {
+      "epoch": 1.3018839025884517,
+      "grad_norm": 2.130309820175171,
+      "learning_rate": 2.8301934956859145e-05,
+      "loss": 0.7971,
+      "step": 756500
+    },
+    {
+      "epoch": 1.3027443678247956,
+      "grad_norm": 1.9937305450439453,
+      "learning_rate": 2.8287593869586742e-05,
+      "loss": 0.7948,
+      "step": 757000
+    },
+    {
+      "epoch": 1.3036048330611396,
+      "grad_norm": 2.077946186065674,
+      "learning_rate": 2.8273252782314342e-05,
+      "loss": 0.7924,
+      "step": 757500
+    },
+    {
+      "epoch": 1.3044652982974836,
+      "grad_norm": 2.286851644515991,
+      "learning_rate": 2.8258911695041946e-05,
+      "loss": 0.7989,
+      "step": 758000
+    },
+    {
+      "epoch": 1.3053257635338276,
+      "grad_norm": 2.0541067123413086,
+      "learning_rate": 2.8244570607769543e-05,
+      "loss": 0.7943,
+      "step": 758500
+    },
+    {
+      "epoch": 1.3061862287701715,
+      "grad_norm": 2.131272315979004,
+      "learning_rate": 2.8230229520497144e-05,
+      "loss": 0.7963,
+      "step": 759000
+    },
+    {
+      "epoch": 1.3070466940065155,
+      "grad_norm": 2.097992181777954,
+      "learning_rate": 2.8215888433224748e-05,
+      "loss": 0.7926,
+      "step": 759500
+    },
+    {
+      "epoch": 1.3079071592428595,
+      "grad_norm": 2.153024196624756,
+      "learning_rate": 2.8201547345952345e-05,
+      "loss": 0.7941,
+      "step": 760000
+    },
+    {
+      "epoch": 1.3087676244792035,
+      "grad_norm": 2.235593557357788,
+      "learning_rate": 2.8187206258679942e-05,
+      "loss": 0.7963,
+      "step": 760500
+    },
+    {
+      "epoch": 1.3096280897155474,
+      "grad_norm": 2.201599359512329,
+      "learning_rate": 2.8172865171407542e-05,
+      "loss": 0.7913,
+      "step": 761000
+    },
+    {
+      "epoch": 1.3104885549518914,
+      "grad_norm": 1.854984164237976,
+      "learning_rate": 2.8158524084135146e-05,
+      "loss": 0.7949,
+      "step": 761500
+    },
+    {
+      "epoch": 1.3113490201882354,
+      "grad_norm": 2.277507781982422,
+      "learning_rate": 2.8144182996862743e-05,
+      "loss": 0.7938,
+      "step": 762000
+    },
+    {
+      "epoch": 1.3122094854245794,
+      "grad_norm": 2.180150032043457,
+      "learning_rate": 2.8129841909590344e-05,
+      "loss": 0.7941,
+      "step": 762500
+    },
+    {
+      "epoch": 1.3130699506609234,
+      "grad_norm": 2.0832903385162354,
+      "learning_rate": 2.8115500822317948e-05,
+      "loss": 0.7906,
+      "step": 763000
+    },
+    {
+      "epoch": 1.3139304158972673,
+      "grad_norm": 2.1612677574157715,
+      "learning_rate": 2.8101159735045545e-05,
+      "loss": 0.7922,
+      "step": 763500
+    },
+    {
+      "epoch": 1.3147908811336113,
+      "grad_norm": 1.9641860723495483,
+      "learning_rate": 2.8086818647773145e-05,
+      "loss": 0.7943,
+      "step": 764000
+    },
+    {
+      "epoch": 1.3156513463699553,
+      "grad_norm": 2.2745723724365234,
+      "learning_rate": 2.807247756050075e-05,
+      "loss": 0.796,
+      "step": 764500
+    },
+    {
+      "epoch": 1.3165118116062993,
+      "grad_norm": 1.9285554885864258,
+      "learning_rate": 2.8058136473228346e-05,
+      "loss": 0.7999,
+      "step": 765000
+    },
+    {
+      "epoch": 1.3173722768426432,
+      "grad_norm": 2.124898672103882,
+      "learning_rate": 2.8043795385955947e-05,
+      "loss": 0.7919,
+      "step": 765500
+    },
+    {
+      "epoch": 1.3182327420789872,
+      "grad_norm": 2.0991389751434326,
+      "learning_rate": 2.802945429868355e-05,
+      "loss": 0.7942,
+      "step": 766000
+    },
+    {
+      "epoch": 1.3190932073153312,
+      "grad_norm": 2.0024592876434326,
+      "learning_rate": 2.8015113211411148e-05,
+      "loss": 0.7937,
+      "step": 766500
+    },
+    {
+      "epoch": 1.3199536725516752,
+      "grad_norm": 2.1842150688171387,
+      "learning_rate": 2.8000772124138748e-05,
+      "loss": 0.7988,
+      "step": 767000
+    },
+    {
+      "epoch": 1.3208141377880191,
+      "grad_norm": 2.1621017456054688,
+      "learning_rate": 2.7986431036866345e-05,
+      "loss": 0.7991,
+      "step": 767500
+    },
+    {
+      "epoch": 1.3216746030243631,
+      "grad_norm": 2.237100839614868,
+      "learning_rate": 2.797208994959395e-05,
+      "loss": 0.7895,
+      "step": 768000
+    },
+    {
+      "epoch": 1.322535068260707,
+      "grad_norm": 2.050424098968506,
+      "learning_rate": 2.795774886232155e-05,
+      "loss": 0.7929,
+      "step": 768500
+    },
+    {
+      "epoch": 1.323395533497051,
+      "grad_norm": 2.0490074157714844,
+      "learning_rate": 2.7943407775049146e-05,
+      "loss": 0.7906,
+      "step": 769000
+    },
+    {
+      "epoch": 1.324255998733395,
+      "grad_norm": 2.1215157508850098,
+      "learning_rate": 2.792906668777675e-05,
+      "loss": 0.8012,
+      "step": 769500
+    },
+    {
+      "epoch": 1.325116463969739,
+      "grad_norm": 2.28594708442688,
+      "learning_rate": 2.7914725600504347e-05,
+      "loss": 0.7936,
+      "step": 770000
+    },
+    {
+      "epoch": 1.325976929206083,
+      "grad_norm": 1.9430696964263916,
+      "learning_rate": 2.7900384513231948e-05,
+      "loss": 0.7891,
+      "step": 770500
+    },
+    {
+      "epoch": 1.3268373944424272,
+      "grad_norm": 2.1309525966644287,
+      "learning_rate": 2.7886043425959552e-05,
+      "loss": 0.7881,
+      "step": 771000
+    },
+    {
+      "epoch": 1.3276978596787712,
+      "grad_norm": 1.9575910568237305,
+      "learning_rate": 2.787170233868715e-05,
+      "loss": 0.7892,
+      "step": 771500
+    },
+    {
+      "epoch": 1.3285583249151152,
+      "grad_norm": 2.077388048171997,
+      "learning_rate": 2.785736125141475e-05,
+      "loss": 0.7905,
+      "step": 772000
+    },
+    {
+      "epoch": 1.3294187901514591,
+      "grad_norm": 2.061293840408325,
+      "learning_rate": 2.7843020164142346e-05,
+      "loss": 0.796,
+      "step": 772500
+    },
+    {
+      "epoch": 1.3302792553878031,
+      "grad_norm": 2.2610721588134766,
+      "learning_rate": 2.782867907686995e-05,
+      "loss": 0.7873,
+      "step": 773000
+    },
+    {
+      "epoch": 1.331139720624147,
+      "grad_norm": 2.2062535285949707,
+      "learning_rate": 2.781433798959755e-05,
+      "loss": 0.7969,
+      "step": 773500
+    },
+    {
+      "epoch": 1.332000185860491,
+      "grad_norm": 2.090650796890259,
+      "learning_rate": 2.7799996902325148e-05,
+      "loss": 0.7914,
+      "step": 774000
+    },
+    {
+      "epoch": 1.332860651096835,
+      "grad_norm": 2.083618402481079,
+      "learning_rate": 2.778565581505275e-05,
+      "loss": 0.7907,
+      "step": 774500
+    },
+    {
+      "epoch": 1.333721116333179,
+      "grad_norm": 1.9934265613555908,
+      "learning_rate": 2.7771314727780352e-05,
+      "loss": 0.7986,
+      "step": 775000
+    },
+    {
+      "epoch": 1.334581581569523,
+      "grad_norm": 1.9388866424560547,
+      "learning_rate": 2.775697364050795e-05,
+      "loss": 0.7874,
+      "step": 775500
+    },
+    {
+      "epoch": 1.335442046805867,
+      "grad_norm": 2.0379679203033447,
+      "learning_rate": 2.7742632553235553e-05,
+      "loss": 0.7916,
+      "step": 776000
+    },
+    {
+      "epoch": 1.336302512042211,
+      "grad_norm": 2.174241542816162,
+      "learning_rate": 2.7728291465963154e-05,
+      "loss": 0.7892,
+      "step": 776500
+    },
+    {
+      "epoch": 1.337162977278555,
+      "grad_norm": 2.243471384048462,
+      "learning_rate": 2.771395037869075e-05,
+      "loss": 0.7976,
+      "step": 777000
+    },
+    {
+      "epoch": 1.338023442514899,
+      "grad_norm": 2.231506824493408,
+      "learning_rate": 2.7699609291418354e-05,
+      "loss": 0.7988,
+      "step": 777500
+    },
+    {
+      "epoch": 1.3388839077512429,
+      "grad_norm": 1.9651925563812256,
+      "learning_rate": 2.768526820414595e-05,
+      "loss": 0.7943,
+      "step": 778000
+    },
+    {
+      "epoch": 1.3397443729875869,
+      "grad_norm": 2.184065818786621,
+      "learning_rate": 2.7670927116873552e-05,
+      "loss": 0.7884,
+      "step": 778500
+    },
+    {
+      "epoch": 1.3406048382239308,
+      "grad_norm": 2.0603458881378174,
+      "learning_rate": 2.765658602960115e-05,
+      "loss": 0.7952,
+      "step": 779000
+    },
+    {
+      "epoch": 1.341465303460275,
+      "grad_norm": 2.1272175312042236,
+      "learning_rate": 2.7642244942328753e-05,
+      "loss": 0.7968,
+      "step": 779500
+    },
+    {
+      "epoch": 1.342325768696619,
+      "grad_norm": 2.0302109718322754,
+      "learning_rate": 2.7627903855056353e-05,
+      "loss": 0.7892,
+      "step": 780000
+    },
+    {
+      "epoch": 1.343186233932963,
+      "grad_norm": 2.089341163635254,
+      "learning_rate": 2.761356276778395e-05,
+      "loss": 0.7902,
+      "step": 780500
+    },
+    {
+      "epoch": 1.344046699169307,
+      "grad_norm": 2.258023500442505,
+      "learning_rate": 2.7599221680511554e-05,
+      "loss": 0.7874,
+      "step": 781000
+    },
+    {
+      "epoch": 1.344907164405651,
+      "grad_norm": 2.0876317024230957,
+      "learning_rate": 2.7584880593239155e-05,
+      "loss": 0.7919,
+      "step": 781500
+    },
+    {
+      "epoch": 1.345767629641995,
+      "grad_norm": 1.8872166872024536,
+      "learning_rate": 2.7570539505966752e-05,
+      "loss": 0.8001,
+      "step": 782000
+    },
+    {
+      "epoch": 1.346628094878339,
+      "grad_norm": 2.0026206970214844,
+      "learning_rate": 2.7556198418694356e-05,
+      "loss": 0.7963,
+      "step": 782500
+    },
+    {
+      "epoch": 1.3474885601146829,
+      "grad_norm": 1.90133535861969,
+      "learning_rate": 2.7541857331421956e-05,
+      "loss": 0.7928,
+      "step": 783000
+    },
+    {
+      "epoch": 1.3483490253510269,
+      "grad_norm": 2.0178732872009277,
+      "learning_rate": 2.7527516244149553e-05,
+      "loss": 0.7958,
+      "step": 783500
+    },
+    {
+      "epoch": 1.3492094905873708,
+      "grad_norm": 2.0105807781219482,
+      "learning_rate": 2.7513175156877154e-05,
+      "loss": 0.7899,
+      "step": 784000
+    },
+    {
+      "epoch": 1.3500699558237148,
+      "grad_norm": 2.131472587585449,
+      "learning_rate": 2.7498834069604758e-05,
+      "loss": 0.7949,
+      "step": 784500
+    },
+    {
+      "epoch": 1.3509304210600588,
+      "grad_norm": 2.2190399169921875,
+      "learning_rate": 2.7484492982332355e-05,
+      "loss": 0.8001,
+      "step": 785000
+    },
+    {
+      "epoch": 1.3517908862964028,
+      "grad_norm": 2.1613001823425293,
+      "learning_rate": 2.7470151895059952e-05,
+      "loss": 0.7914,
+      "step": 785500
+    },
+    {
+      "epoch": 1.3526513515327467,
+      "grad_norm": 2.416355609893799,
+      "learning_rate": 2.745581080778756e-05,
+      "loss": 0.7977,
+      "step": 786000
+    },
+    {
+      "epoch": 1.3535118167690907,
+      "grad_norm": 2.159780740737915,
+      "learning_rate": 2.7441469720515156e-05,
+      "loss": 0.7909,
+      "step": 786500
+    },
+    {
+      "epoch": 1.3543722820054347,
+      "grad_norm": 2.2719836235046387,
+      "learning_rate": 2.7427128633242753e-05,
+      "loss": 0.7887,
+      "step": 787000
+    },
+    {
+      "epoch": 1.3552327472417787,
+      "grad_norm": 2.027290105819702,
+      "learning_rate": 2.7412787545970357e-05,
+      "loss": 0.7936,
+      "step": 787500
+    },
+    {
+      "epoch": 1.3560932124781226,
+      "grad_norm": 2.148388624191284,
+      "learning_rate": 2.7398446458697958e-05,
+      "loss": 0.7945,
+      "step": 788000
+    },
+    {
+      "epoch": 1.3569536777144666,
+      "grad_norm": 2.142289161682129,
+      "learning_rate": 2.7384105371425555e-05,
+      "loss": 0.7937,
+      "step": 788500
+    },
+    {
+      "epoch": 1.3578141429508106,
+      "grad_norm": 2.2423949241638184,
+      "learning_rate": 2.736976428415316e-05,
+      "loss": 0.7912,
+      "step": 789000
+    },
+    {
+      "epoch": 1.3586746081871546,
+      "grad_norm": 2.101126194000244,
+      "learning_rate": 2.735542319688076e-05,
+      "loss": 0.7911,
+      "step": 789500
+    },
+    {
+      "epoch": 1.3595350734234986,
+      "grad_norm": 2.2099528312683105,
+      "learning_rate": 2.7341082109608356e-05,
+      "loss": 0.7948,
+      "step": 790000
+    },
+    {
+      "epoch": 1.3603955386598425,
+      "grad_norm": 1.9013993740081787,
+      "learning_rate": 2.7326741022335957e-05,
+      "loss": 0.7987,
+      "step": 790500
+    },
+    {
+      "epoch": 1.3612560038961865,
+      "grad_norm": 2.013784646987915,
+      "learning_rate": 2.731239993506356e-05,
+      "loss": 0.7913,
+      "step": 791000
+    },
+    {
+      "epoch": 1.3621164691325305,
+      "grad_norm": 2.1542694568634033,
+      "learning_rate": 2.7298058847791157e-05,
+      "loss": 0.7947,
+      "step": 791500
+    },
+    {
+      "epoch": 1.3629769343688745,
+      "grad_norm": 2.124849796295166,
+      "learning_rate": 2.7283717760518758e-05,
+      "loss": 0.79,
+      "step": 792000
+    },
+    {
+      "epoch": 1.3638373996052184,
+      "grad_norm": 2.034287691116333,
+      "learning_rate": 2.7269376673246362e-05,
+      "loss": 0.7918,
+      "step": 792500
+    },
+    {
+      "epoch": 1.3646978648415624,
+      "grad_norm": 2.1357600688934326,
+      "learning_rate": 2.725503558597396e-05,
+      "loss": 0.7942,
+      "step": 793000
+    },
+    {
+      "epoch": 1.3655583300779064,
+      "grad_norm": 2.0473575592041016,
+      "learning_rate": 2.724069449870156e-05,
+      "loss": 0.7944,
+      "step": 793500
+    },
+    {
+      "epoch": 1.3664187953142506,
+      "grad_norm": 2.184004545211792,
+      "learning_rate": 2.7226353411429163e-05,
+      "loss": 0.7896,
+      "step": 794000
+    },
+    {
+      "epoch": 1.3672792605505946,
+      "grad_norm": 2.0148677825927734,
+      "learning_rate": 2.721201232415676e-05,
+      "loss": 0.7834,
+      "step": 794500
+    },
+    {
+      "epoch": 1.3681397257869385,
+      "grad_norm": 2.2517378330230713,
+      "learning_rate": 2.7197671236884357e-05,
+      "loss": 0.7933,
+      "step": 795000
+    },
+    {
+      "epoch": 1.3690001910232825,
+      "grad_norm": 2.022488594055176,
+      "learning_rate": 2.7183330149611958e-05,
+      "loss": 0.7912,
+      "step": 795500
+    },
+    {
+      "epoch": 1.3698606562596265,
+      "grad_norm": 2.242121458053589,
+      "learning_rate": 2.7168989062339562e-05,
+      "loss": 0.7959,
+      "step": 796000
+    },
+    {
+      "epoch": 1.3707211214959705,
+      "grad_norm": 1.8660246133804321,
+      "learning_rate": 2.715464797506716e-05,
+      "loss": 0.7889,
+      "step": 796500
+    },
+    {
+      "epoch": 1.3715815867323145,
+      "grad_norm": 1.893210768699646,
+      "learning_rate": 2.714030688779476e-05,
+      "loss": 0.7919,
+      "step": 797000
+    },
+    {
+      "epoch": 1.3724420519686584,
+      "grad_norm": 1.7890762090682983,
+      "learning_rate": 2.7125965800522363e-05,
+      "loss": 0.7938,
+      "step": 797500
+    },
+    {
+      "epoch": 1.3733025172050024,
+      "grad_norm": 2.115593910217285,
+      "learning_rate": 2.711162471324996e-05,
+      "loss": 0.7844,
+      "step": 798000
+    },
+    {
+      "epoch": 1.3741629824413464,
+      "grad_norm": 2.0752272605895996,
+      "learning_rate": 2.709728362597756e-05,
+      "loss": 0.7849,
+      "step": 798500
+    },
+    {
+      "epoch": 1.3750234476776904,
+      "grad_norm": 2.2504701614379883,
+      "learning_rate": 2.7082942538705165e-05,
+      "loss": 0.7883,
+      "step": 799000
+    },
+    {
+      "epoch": 1.3758839129140343,
+      "grad_norm": 1.9834873676300049,
+      "learning_rate": 2.706860145143276e-05,
+      "loss": 0.7902,
+      "step": 799500
+    },
+    {
+      "epoch": 1.3767443781503783,
+      "grad_norm": 2.212036609649658,
+      "learning_rate": 2.7054260364160362e-05,
+      "loss": 0.788,
+      "step": 800000
+    },
+    {
+      "epoch": 1.3776048433867223,
+      "grad_norm": 2.212900161743164,
+      "learning_rate": 2.7039919276887966e-05,
+      "loss": 0.7913,
+      "step": 800500
+    },
+    {
+      "epoch": 1.3784653086230663,
+      "grad_norm": 2.1146013736724854,
+      "learning_rate": 2.7025578189615563e-05,
+      "loss": 0.786,
+      "step": 801000
+    },
+    {
+      "epoch": 1.3793257738594102,
+      "grad_norm": 1.9993088245391846,
+      "learning_rate": 2.7011237102343163e-05,
+      "loss": 0.7874,
+      "step": 801500
+    },
+    {
+      "epoch": 1.3801862390957542,
+      "grad_norm": 2.087582588195801,
+      "learning_rate": 2.699689601507076e-05,
+      "loss": 0.7959,
+      "step": 802000
+    },
+    {
+      "epoch": 1.3810467043320982,
+      "grad_norm": 2.0022664070129395,
+      "learning_rate": 2.6982554927798364e-05,
+      "loss": 0.788,
+      "step": 802500
+    },
+    {
+      "epoch": 1.3819071695684424,
+      "grad_norm": 2.188950777053833,
+      "learning_rate": 2.6968213840525965e-05,
+      "loss": 0.7871,
+      "step": 803000
+    },
+    {
+      "epoch": 1.3827676348047864,
+      "grad_norm": 2.002671003341675,
+      "learning_rate": 2.6953872753253562e-05,
+      "loss": 0.7899,
+      "step": 803500
+    },
+    {
+      "epoch": 1.3836281000411303,
+      "grad_norm": 2.184645175933838,
+      "learning_rate": 2.6939531665981166e-05,
+      "loss": 0.791,
+      "step": 804000
+    },
+    {
+      "epoch": 1.3844885652774743,
+      "grad_norm": 2.0297811031341553,
+      "learning_rate": 2.6925190578708763e-05,
+      "loss": 0.7846,
+      "step": 804500
+    },
+    {
+      "epoch": 1.3853490305138183,
+      "grad_norm": 2.2588882446289062,
+      "learning_rate": 2.6910849491436363e-05,
+      "loss": 0.7898,
+      "step": 805000
+    },
+    {
+      "epoch": 1.3862094957501623,
+      "grad_norm": 1.9149460792541504,
+      "learning_rate": 2.6896508404163967e-05,
+      "loss": 0.7891,
+      "step": 805500
+    },
+    {
+      "epoch": 1.3870699609865063,
+      "grad_norm": 2.145394802093506,
+      "learning_rate": 2.6882167316891564e-05,
+      "loss": 0.7866,
+      "step": 806000
+    },
+    {
+      "epoch": 1.3879304262228502,
+      "grad_norm": 2.0832836627960205,
+      "learning_rate": 2.6867826229619165e-05,
+      "loss": 0.7943,
+      "step": 806500
+    },
+    {
+      "epoch": 1.3887908914591942,
+      "grad_norm": 2.0837013721466064,
+      "learning_rate": 2.6853485142346762e-05,
+      "loss": 0.7868,
+      "step": 807000
+    },
+    {
+      "epoch": 1.3896513566955382,
+      "grad_norm": 2.2829363346099854,
+      "learning_rate": 2.6839144055074366e-05,
+      "loss": 0.7926,
+      "step": 807500
+    },
+    {
+      "epoch": 1.3905118219318822,
+      "grad_norm": 1.8782167434692383,
+      "learning_rate": 2.6824802967801966e-05,
+      "loss": 0.7989,
+      "step": 808000
+    },
+    {
+      "epoch": 1.3913722871682261,
+      "grad_norm": 2.088763952255249,
+      "learning_rate": 2.6810461880529563e-05,
+      "loss": 0.7846,
+      "step": 808500
+    },
+    {
+      "epoch": 1.3922327524045701,
+      "grad_norm": 2.1010475158691406,
+      "learning_rate": 2.6796120793257167e-05,
+      "loss": 0.7855,
+      "step": 809000
+    },
+    {
+      "epoch": 1.393093217640914,
+      "grad_norm": 2.3863370418548584,
+      "learning_rate": 2.6781779705984768e-05,
+      "loss": 0.7866,
+      "step": 809500
+    },
+    {
+      "epoch": 1.393953682877258,
+      "grad_norm": 2.236123561859131,
+      "learning_rate": 2.6767438618712365e-05,
+      "loss": 0.7878,
+      "step": 810000
+    },
+    {
+      "epoch": 1.394814148113602,
+      "grad_norm": 2.2748334407806396,
+      "learning_rate": 2.675309753143997e-05,
+      "loss": 0.7925,
+      "step": 810500
+    },
+    {
+      "epoch": 1.395674613349946,
+      "grad_norm": 2.1135990619659424,
+      "learning_rate": 2.673875644416757e-05,
+      "loss": 0.7934,
+      "step": 811000
+    },
+    {
+      "epoch": 1.39653507858629,
+      "grad_norm": 2.133629083633423,
+      "learning_rate": 2.6724415356895166e-05,
+      "loss": 0.7862,
+      "step": 811500
+    },
+    {
+      "epoch": 1.397395543822634,
+      "grad_norm": 1.9208927154541016,
+      "learning_rate": 2.671007426962277e-05,
+      "loss": 0.7879,
+      "step": 812000
+    },
+    {
+      "epoch": 1.398256009058978,
+      "grad_norm": 1.9107301235198975,
+      "learning_rate": 2.6695733182350367e-05,
+      "loss": 0.7919,
+      "step": 812500
+    },
+    {
+      "epoch": 1.399116474295322,
+      "grad_norm": 2.0901243686676025,
+      "learning_rate": 2.6681392095077968e-05,
+      "loss": 0.7935,
+      "step": 813000
+    },
+    {
+      "epoch": 1.399976939531666,
+      "grad_norm": 2.2657084465026855,
+      "learning_rate": 2.6667051007805565e-05,
+      "loss": 0.7869,
+      "step": 813500
+    },
+    {
+      "epoch": 1.4008374047680099,
+      "grad_norm": 1.8893272876739502,
+      "learning_rate": 2.665270992053317e-05,
+      "loss": 0.7882,
+      "step": 814000
+    },
+    {
+      "epoch": 1.4016978700043539,
+      "grad_norm": 2.058760404586792,
+      "learning_rate": 2.663836883326077e-05,
+      "loss": 0.7915,
+      "step": 814500
+    },
+    {
+      "epoch": 1.4025583352406978,
+      "grad_norm": 1.98762845993042,
+      "learning_rate": 2.6624027745988366e-05,
+      "loss": 0.7847,
+      "step": 815000
+    },
+    {
+      "epoch": 1.4034188004770418,
+      "grad_norm": 1.9851107597351074,
+      "learning_rate": 2.660968665871597e-05,
+      "loss": 0.7921,
+      "step": 815500
+    },
+    {
+      "epoch": 1.4042792657133858,
+      "grad_norm": 2.0246217250823975,
+      "learning_rate": 2.659534557144357e-05,
+      "loss": 0.7882,
+      "step": 816000
+    },
+    {
+      "epoch": 1.4051397309497298,
+      "grad_norm": 2.0061659812927246,
+      "learning_rate": 2.6581004484171167e-05,
+      "loss": 0.7917,
+      "step": 816500
+    },
+    {
+      "epoch": 1.406000196186074,
+      "grad_norm": 2.1723735332489014,
+      "learning_rate": 2.656666339689877e-05,
+      "loss": 0.7853,
+      "step": 817000
+    },
+    {
+      "epoch": 1.406860661422418,
+      "grad_norm": 2.245793342590332,
+      "learning_rate": 2.6552322309626372e-05,
+      "loss": 0.7894,
+      "step": 817500
+    },
+    {
+      "epoch": 1.407721126658762,
+      "grad_norm": 1.9718598127365112,
+      "learning_rate": 2.653798122235397e-05,
+      "loss": 0.7898,
+      "step": 818000
+    },
+    {
+      "epoch": 1.408581591895106,
+      "grad_norm": 2.3851137161254883,
+      "learning_rate": 2.652364013508157e-05,
+      "loss": 0.7886,
+      "step": 818500
+    },
+    {
+      "epoch": 1.4094420571314499,
+      "grad_norm": 2.107008218765259,
+      "learning_rate": 2.6509299047809173e-05,
+      "loss": 0.7865,
+      "step": 819000
+    },
+    {
+      "epoch": 1.4103025223677939,
+      "grad_norm": 2.15132999420166,
+      "learning_rate": 2.649495796053677e-05,
+      "loss": 0.7899,
+      "step": 819500
+    },
+    {
+      "epoch": 1.4111629876041378,
+      "grad_norm": 2.262600898742676,
+      "learning_rate": 2.6480616873264367e-05,
+      "loss": 0.7867,
+      "step": 820000
+    },
+    {
+      "epoch": 1.4120234528404818,
+      "grad_norm": 2.184628486633301,
+      "learning_rate": 2.6466275785991975e-05,
+      "loss": 0.7841,
+      "step": 820500
+    },
+    {
+      "epoch": 1.4128839180768258,
+      "grad_norm": 2.1769585609436035,
+      "learning_rate": 2.645193469871957e-05,
+      "loss": 0.7902,
+      "step": 821000
+    },
+    {
+      "epoch": 1.4137443833131698,
+      "grad_norm": 1.972236156463623,
+      "learning_rate": 2.643759361144717e-05,
+      "loss": 0.7807,
+      "step": 821500
+    },
+    {
+      "epoch": 1.4146048485495137,
+      "grad_norm": 2.0056240558624268,
+      "learning_rate": 2.6423252524174773e-05,
+      "loss": 0.7876,
+      "step": 822000
+    },
+    {
+      "epoch": 1.4154653137858577,
+      "grad_norm": 1.9562522172927856,
+      "learning_rate": 2.6408911436902373e-05,
+      "loss": 0.7888,
+      "step": 822500
+    },
+    {
+      "epoch": 1.4163257790222017,
+      "grad_norm": 1.9554847478866577,
+      "learning_rate": 2.639457034962997e-05,
+      "loss": 0.7817,
+      "step": 823000
+    },
+    {
+      "epoch": 1.4171862442585457,
+      "grad_norm": 2.1087417602539062,
+      "learning_rate": 2.6380229262357574e-05,
+      "loss": 0.7848,
+      "step": 823500
+    },
+    {
+      "epoch": 1.4180467094948896,
+      "grad_norm": 2.0702404975891113,
+      "learning_rate": 2.6365888175085174e-05,
+      "loss": 0.7847,
+      "step": 824000
+    },
+    {
+      "epoch": 1.4189071747312336,
+      "grad_norm": 2.1107707023620605,
+      "learning_rate": 2.635154708781277e-05,
+      "loss": 0.7875,
+      "step": 824500
+    },
+    {
+      "epoch": 1.4197676399675776,
+      "grad_norm": 2.111306667327881,
+      "learning_rate": 2.6337206000540372e-05,
+      "loss": 0.7847,
+      "step": 825000
+    },
+    {
+      "epoch": 1.4206281052039216,
+      "grad_norm": 2.0543792247772217,
+      "learning_rate": 2.6322864913267976e-05,
+      "loss": 0.7801,
+      "step": 825500
+    },
+    {
+      "epoch": 1.4214885704402658,
+      "grad_norm": 2.030547857284546,
+      "learning_rate": 2.6308523825995573e-05,
+      "loss": 0.788,
+      "step": 826000
+    },
+    {
+      "epoch": 1.4223490356766098,
+      "grad_norm": 2.2248077392578125,
+      "learning_rate": 2.6294182738723173e-05,
+      "loss": 0.7876,
+      "step": 826500
+    },
+    {
+      "epoch": 1.4232095009129537,
+      "grad_norm": 1.9383457899093628,
+      "learning_rate": 2.6279841651450777e-05,
+      "loss": 0.7853,
+      "step": 827000
+    },
+    {
+      "epoch": 1.4240699661492977,
+      "grad_norm": 2.11863112449646,
+      "learning_rate": 2.6265500564178374e-05,
+      "loss": 0.7822,
+      "step": 827500
+    },
+    {
+      "epoch": 1.4249304313856417,
+      "grad_norm": 2.0522687435150146,
+      "learning_rate": 2.6251159476905975e-05,
+      "loss": 0.7876,
+      "step": 828000
+    },
+    {
+      "epoch": 1.4257908966219857,
+      "grad_norm": 2.1168301105499268,
+      "learning_rate": 2.623681838963358e-05,
+      "loss": 0.7864,
+      "step": 828500
+    },
+    {
+      "epoch": 1.4266513618583296,
+      "grad_norm": 2.140026569366455,
+      "learning_rate": 2.6222477302361176e-05,
+      "loss": 0.7795,
+      "step": 829000
+    },
+    {
+      "epoch": 1.4275118270946736,
+      "grad_norm": 1.9109783172607422,
+      "learning_rate": 2.6208136215088773e-05,
+      "loss": 0.7868,
+      "step": 829500
+    },
+    {
+      "epoch": 1.4283722923310176,
+      "grad_norm": 2.115460157394409,
+      "learning_rate": 2.619379512781638e-05,
+      "loss": 0.7908,
+      "step": 830000
+    },
+    {
+      "epoch": 1.4292327575673616,
+      "grad_norm": 2.587132692337036,
+      "learning_rate": 2.6179454040543977e-05,
+      "loss": 0.7954,
+      "step": 830500
+    },
+    {
+      "epoch": 1.4300932228037055,
+      "grad_norm": 1.930621862411499,
+      "learning_rate": 2.6165112953271574e-05,
+      "loss": 0.7875,
+      "step": 831000
+    },
+    {
+      "epoch": 1.4309536880400495,
+      "grad_norm": 2.050018072128296,
+      "learning_rate": 2.6150771865999175e-05,
+      "loss": 0.7929,
+      "step": 831500
+    },
+    {
+      "epoch": 1.4318141532763935,
+      "grad_norm": 2.272819995880127,
+      "learning_rate": 2.613643077872678e-05,
+      "loss": 0.7858,
+      "step": 832000
+    },
+    {
+      "epoch": 1.4326746185127375,
+      "grad_norm": 1.9963629245758057,
+      "learning_rate": 2.6122089691454376e-05,
+      "loss": 0.7816,
+      "step": 832500
+    },
+    {
+      "epoch": 1.4335350837490815,
+      "grad_norm": 2.0288803577423096,
+      "learning_rate": 2.6107748604181976e-05,
+      "loss": 0.7887,
+      "step": 833000
+    },
+    {
+      "epoch": 1.4343955489854254,
+      "grad_norm": 2.02726411819458,
+      "learning_rate": 2.609340751690958e-05,
+      "loss": 0.7814,
+      "step": 833500
+    },
+    {
+      "epoch": 1.4352560142217694,
+      "grad_norm": 2.012253999710083,
+      "learning_rate": 2.6079066429637177e-05,
+      "loss": 0.7818,
+      "step": 834000
+    },
+    {
+      "epoch": 1.4361164794581134,
+      "grad_norm": 2.117372512817383,
+      "learning_rate": 2.6064725342364778e-05,
+      "loss": 0.792,
+      "step": 834500
+    },
+    {
+      "epoch": 1.4369769446944574,
+      "grad_norm": 2.0443718433380127,
+      "learning_rate": 2.605038425509238e-05,
+      "loss": 0.7854,
+      "step": 835000
+    },
+    {
+      "epoch": 1.4378374099308013,
+      "grad_norm": 2.198227882385254,
+      "learning_rate": 2.603604316781998e-05,
+      "loss": 0.7855,
+      "step": 835500
+    },
+    {
+      "epoch": 1.4386978751671453,
+      "grad_norm": 2.020047187805176,
+      "learning_rate": 2.602170208054758e-05,
+      "loss": 0.7972,
+      "step": 836000
+    },
+    {
+      "epoch": 1.4395583404034893,
+      "grad_norm": 2.201357841491699,
+      "learning_rate": 2.6007360993275176e-05,
+      "loss": 0.7895,
+      "step": 836500
+    },
+    {
+      "epoch": 1.4404188056398333,
+      "grad_norm": 1.8877038955688477,
+      "learning_rate": 2.599301990600278e-05,
+      "loss": 0.7868,
+      "step": 837000
+    },
+    {
+      "epoch": 1.4412792708761772,
+      "grad_norm": 2.0734703540802,
+      "learning_rate": 2.5978678818730377e-05,
+      "loss": 0.7822,
+      "step": 837500
+    },
+    {
+      "epoch": 1.4421397361125212,
+      "grad_norm": 2.0934813022613525,
+      "learning_rate": 2.5964337731457978e-05,
+      "loss": 0.7779,
+      "step": 838000
+    },
+    {
+      "epoch": 1.4430002013488652,
+      "grad_norm": 2.2466344833374023,
+      "learning_rate": 2.594999664418558e-05,
+      "loss": 0.7773,
+      "step": 838500
+    },
+    {
+      "epoch": 1.4438606665852092,
+      "grad_norm": 2.1962478160858154,
+      "learning_rate": 2.593565555691318e-05,
+      "loss": 0.7824,
+      "step": 839000
+    },
+    {
+      "epoch": 1.4447211318215532,
+      "grad_norm": 2.0171070098876953,
+      "learning_rate": 2.592131446964078e-05,
+      "loss": 0.7868,
+      "step": 839500
+    },
+    {
+      "epoch": 1.4455815970578971,
+      "grad_norm": 2.070829153060913,
+      "learning_rate": 2.5906973382368383e-05,
+      "loss": 0.7862,
+      "step": 840000
+    },
+    {
+      "epoch": 1.4464420622942413,
+      "grad_norm": 2.067033052444458,
+      "learning_rate": 2.589263229509598e-05,
+      "loss": 0.7832,
+      "step": 840500
+    },
+    {
+      "epoch": 1.4473025275305853,
+      "grad_norm": 2.241600751876831,
+      "learning_rate": 2.587829120782358e-05,
+      "loss": 0.7822,
+      "step": 841000
+    },
+    {
+      "epoch": 1.4481629927669293,
+      "grad_norm": 2.129298210144043,
+      "learning_rate": 2.5863950120551184e-05,
+      "loss": 0.7838,
+      "step": 841500
+    },
+    {
+      "epoch": 1.4490234580032733,
+      "grad_norm": 1.9001485109329224,
+      "learning_rate": 2.584960903327878e-05,
+      "loss": 0.7836,
+      "step": 842000
+    },
+    {
+      "epoch": 1.4498839232396172,
+      "grad_norm": 2.0611612796783447,
+      "learning_rate": 2.5835267946006382e-05,
+      "loss": 0.7885,
+      "step": 842500
+    },
+    {
+      "epoch": 1.4507443884759612,
+      "grad_norm": 2.0608203411102295,
+      "learning_rate": 2.582092685873398e-05,
+      "loss": 0.7813,
+      "step": 843000
+    },
+    {
+      "epoch": 1.4516048537123052,
+      "grad_norm": 2.246284008026123,
+      "learning_rate": 2.5806585771461583e-05,
+      "loss": 0.7902,
+      "step": 843500
+    },
+    {
+      "epoch": 1.4524653189486492,
+      "grad_norm": 2.0218515396118164,
+      "learning_rate": 2.5792244684189183e-05,
+      "loss": 0.7825,
+      "step": 844000
+    },
+    {
+      "epoch": 1.4533257841849931,
+      "grad_norm": 1.9001346826553345,
+      "learning_rate": 2.577790359691678e-05,
+      "loss": 0.7847,
+      "step": 844500
+    },
+    {
+      "epoch": 1.4541862494213371,
+      "grad_norm": 2.2586870193481445,
+      "learning_rate": 2.5763562509644384e-05,
+      "loss": 0.786,
+      "step": 845000
+    },
+    {
+      "epoch": 1.455046714657681,
+      "grad_norm": 2.003862142562866,
+      "learning_rate": 2.5749221422371985e-05,
+      "loss": 0.7761,
+      "step": 845500
+    },
+    {
+      "epoch": 1.455907179894025,
+      "grad_norm": 2.1195812225341797,
+      "learning_rate": 2.573488033509958e-05,
+      "loss": 0.7877,
+      "step": 846000
+    },
+    {
+      "epoch": 1.456767645130369,
+      "grad_norm": 2.1536705493927,
+      "learning_rate": 2.5720539247827186e-05,
+      "loss": 0.7871,
+      "step": 846500
+    },
+    {
+      "epoch": 1.457628110366713,
+      "grad_norm": 2.003711700439453,
+      "learning_rate": 2.5706198160554783e-05,
+      "loss": 0.7824,
+      "step": 847000
+    },
+    {
+      "epoch": 1.458488575603057,
+      "grad_norm": 2.0670664310455322,
+      "learning_rate": 2.5691857073282383e-05,
+      "loss": 0.7894,
+      "step": 847500
+    },
+    {
+      "epoch": 1.459349040839401,
+      "grad_norm": 2.1034762859344482,
+      "learning_rate": 2.567751598600998e-05,
+      "loss": 0.7916,
+      "step": 848000
+    },
+    {
+      "epoch": 1.460209506075745,
+      "grad_norm": 2.1943678855895996,
+      "learning_rate": 2.5663174898737584e-05,
+      "loss": 0.7867,
+      "step": 848500
+    },
+    {
+      "epoch": 1.4610699713120892,
+      "grad_norm": 2.0474393367767334,
+      "learning_rate": 2.5648833811465184e-05,
+      "loss": 0.7886,
+      "step": 849000
+    },
+    {
+      "epoch": 1.4619304365484331,
+      "grad_norm": 2.041001558303833,
+      "learning_rate": 2.563449272419278e-05,
+      "loss": 0.782,
+      "step": 849500
+    },
+    {
+      "epoch": 1.462790901784777,
+      "grad_norm": 2.227505683898926,
+      "learning_rate": 2.5620151636920385e-05,
+      "loss": 0.7847,
+      "step": 850000
+    },
+    {
+      "epoch": 1.463651367021121,
+      "grad_norm": 2.0753791332244873,
+      "learning_rate": 2.5605810549647986e-05,
+      "loss": 0.7835,
+      "step": 850500
+    },
+    {
+      "epoch": 1.464511832257465,
+      "grad_norm": 2.0337870121002197,
+      "learning_rate": 2.5591469462375583e-05,
+      "loss": 0.7849,
+      "step": 851000
+    },
+    {
+      "epoch": 1.465372297493809,
+      "grad_norm": 2.089561700820923,
+      "learning_rate": 2.5577128375103187e-05,
+      "loss": 0.7793,
+      "step": 851500
+    },
+    {
+      "epoch": 1.466232762730153,
+      "grad_norm": 2.158207654953003,
+      "learning_rate": 2.5562787287830787e-05,
+      "loss": 0.784,
+      "step": 852000
+    },
+    {
+      "epoch": 1.467093227966497,
+      "grad_norm": 2.1370127201080322,
+      "learning_rate": 2.5548446200558384e-05,
+      "loss": 0.7834,
+      "step": 852500
+    },
+    {
+      "epoch": 1.467953693202841,
+      "grad_norm": 2.061452627182007,
+      "learning_rate": 2.5534105113285988e-05,
+      "loss": 0.7817,
+      "step": 853000
+    },
+    {
+      "epoch": 1.468814158439185,
+      "grad_norm": 2.0483486652374268,
+      "learning_rate": 2.551976402601359e-05,
+      "loss": 0.7869,
+      "step": 853500
+    },
+    {
+      "epoch": 1.469674623675529,
+      "grad_norm": 2.03947377204895,
+      "learning_rate": 2.5505422938741186e-05,
+      "loss": 0.7811,
+      "step": 854000
+    },
+    {
+      "epoch": 1.470535088911873,
+      "grad_norm": 2.0983335971832275,
+      "learning_rate": 2.5491081851468783e-05,
+      "loss": 0.7821,
+      "step": 854500
+    },
+    {
+      "epoch": 1.4713955541482169,
+      "grad_norm": 2.134584903717041,
+      "learning_rate": 2.547674076419639e-05,
+      "loss": 0.7827,
+      "step": 855000
+    },
+    {
+      "epoch": 1.4722560193845609,
+      "grad_norm": 2.0175790786743164,
+      "learning_rate": 2.5462399676923987e-05,
+      "loss": 0.783,
+      "step": 855500
+    },
+    {
+      "epoch": 1.4731164846209048,
+      "grad_norm": 1.9841803312301636,
+      "learning_rate": 2.5448058589651584e-05,
+      "loss": 0.7782,
+      "step": 856000
+    },
+    {
+      "epoch": 1.4739769498572488,
+      "grad_norm": 1.9247914552688599,
+      "learning_rate": 2.5433717502379188e-05,
+      "loss": 0.792,
+      "step": 856500
+    },
+    {
+      "epoch": 1.4748374150935928,
+      "grad_norm": 1.9805024862289429,
+      "learning_rate": 2.541937641510679e-05,
+      "loss": 0.7829,
+      "step": 857000
+    },
+    {
+      "epoch": 1.4756978803299368,
+      "grad_norm": 1.9660342931747437,
+      "learning_rate": 2.5405035327834386e-05,
+      "loss": 0.7824,
+      "step": 857500
+    },
+    {
+      "epoch": 1.4765583455662807,
+      "grad_norm": 2.1732521057128906,
+      "learning_rate": 2.539069424056199e-05,
+      "loss": 0.782,
+      "step": 858000
+    },
+    {
+      "epoch": 1.4774188108026247,
+      "grad_norm": 2.0135018825531006,
+      "learning_rate": 2.537635315328959e-05,
+      "loss": 0.78,
+      "step": 858500
+    },
+    {
+      "epoch": 1.4782792760389687,
+      "grad_norm": 2.021561861038208,
+      "learning_rate": 2.5362012066017187e-05,
+      "loss": 0.7863,
+      "step": 859000
+    },
+    {
+      "epoch": 1.4791397412753127,
+      "grad_norm": 2.2785863876342773,
+      "learning_rate": 2.5347670978744788e-05,
+      "loss": 0.7862,
+      "step": 859500
+    },
+    {
+      "epoch": 1.4800002065116566,
+      "grad_norm": 1.9567214250564575,
+      "learning_rate": 2.533332989147239e-05,
+      "loss": 0.7847,
+      "step": 860000
+    },
+    {
+      "epoch": 1.4808606717480006,
+      "grad_norm": 2.1036536693573,
+      "learning_rate": 2.531898880419999e-05,
+      "loss": 0.7813,
+      "step": 860500
+    },
+    {
+      "epoch": 1.4817211369843446,
+      "grad_norm": 2.144239664077759,
+      "learning_rate": 2.530464771692759e-05,
+      "loss": 0.7854,
+      "step": 861000
+    },
+    {
+      "epoch": 1.4825816022206886,
+      "grad_norm": 2.243926525115967,
+      "learning_rate": 2.5290306629655193e-05,
+      "loss": 0.7927,
+      "step": 861500
+    },
+    {
+      "epoch": 1.4834420674570326,
+      "grad_norm": 2.141397714614868,
+      "learning_rate": 2.527596554238279e-05,
+      "loss": 0.7895,
+      "step": 862000
+    },
+    {
+      "epoch": 1.4843025326933765,
+      "grad_norm": 2.1214475631713867,
+      "learning_rate": 2.5261624455110387e-05,
+      "loss": 0.7831,
+      "step": 862500
+    },
+    {
+      "epoch": 1.4851629979297205,
+      "grad_norm": 2.0760138034820557,
+      "learning_rate": 2.5247283367837994e-05,
+      "loss": 0.7839,
+      "step": 863000
+    },
+    {
+      "epoch": 1.4860234631660647,
+      "grad_norm": 2.261528491973877,
+      "learning_rate": 2.523294228056559e-05,
+      "loss": 0.7794,
+      "step": 863500
+    },
+    {
+      "epoch": 1.4868839284024087,
+      "grad_norm": 2.129002809524536,
+      "learning_rate": 2.521860119329319e-05,
+      "loss": 0.7752,
+      "step": 864000
+    },
+    {
+      "epoch": 1.4877443936387527,
+      "grad_norm": 2.004643440246582,
+      "learning_rate": 2.5204260106020792e-05,
+      "loss": 0.7842,
+      "step": 864500
+    },
+    {
+      "epoch": 1.4886048588750966,
+      "grad_norm": 1.9456123113632202,
+      "learning_rate": 2.5189919018748393e-05,
+      "loss": 0.7798,
+      "step": 865000
+    },
+    {
+      "epoch": 1.4894653241114406,
+      "grad_norm": 2.0475871562957764,
+      "learning_rate": 2.517557793147599e-05,
+      "loss": 0.7867,
+      "step": 865500
+    },
+    {
+      "epoch": 1.4903257893477846,
+      "grad_norm": 2.360131025314331,
+      "learning_rate": 2.516123684420359e-05,
+      "loss": 0.7775,
+      "step": 866000
+    },
+    {
+      "epoch": 1.4911862545841286,
+      "grad_norm": 2.2324533462524414,
+      "learning_rate": 2.5146895756931194e-05,
+      "loss": 0.7877,
+      "step": 866500
+    },
+    {
+      "epoch": 1.4920467198204725,
+      "grad_norm": 2.1139626502990723,
+      "learning_rate": 2.513255466965879e-05,
+      "loss": 0.7892,
+      "step": 867000
+    },
+    {
+      "epoch": 1.4929071850568165,
+      "grad_norm": 2.0497446060180664,
+      "learning_rate": 2.5118213582386392e-05,
+      "loss": 0.7828,
+      "step": 867500
+    },
+    {
+      "epoch": 1.4937676502931605,
+      "grad_norm": 2.2306737899780273,
+      "learning_rate": 2.5103872495113996e-05,
+      "loss": 0.7873,
+      "step": 868000
+    },
+    {
+      "epoch": 1.4946281155295045,
+      "grad_norm": 2.103174924850464,
+      "learning_rate": 2.5089531407841593e-05,
+      "loss": 0.7827,
+      "step": 868500
+    },
+    {
+      "epoch": 1.4954885807658485,
+      "grad_norm": 2.195070743560791,
+      "learning_rate": 2.5075190320569193e-05,
+      "loss": 0.7758,
+      "step": 869000
+    },
+    {
+      "epoch": 1.4963490460021924,
+      "grad_norm": 2.206838607788086,
+      "learning_rate": 2.5060849233296797e-05,
+      "loss": 0.7786,
+      "step": 869500
+    },
+    {
+      "epoch": 1.4972095112385364,
+      "grad_norm": 2.0621566772460938,
+      "learning_rate": 2.5046508146024394e-05,
+      "loss": 0.7844,
+      "step": 870000
+    },
+    {
+      "epoch": 1.4980699764748804,
+      "grad_norm": 2.2440409660339355,
+      "learning_rate": 2.5032167058751995e-05,
+      "loss": 0.7817,
+      "step": 870500
+    },
+    {
+      "epoch": 1.4989304417112244,
+      "grad_norm": 2.126237392425537,
+      "learning_rate": 2.501782597147959e-05,
+      "loss": 0.784,
+      "step": 871000
+    },
+    {
+      "epoch": 1.4997909069475683,
+      "grad_norm": 2.475597381591797,
+      "learning_rate": 2.5003484884207195e-05,
+      "loss": 0.784,
+      "step": 871500
+    },
+    {
+      "epoch": 1.5006513721839125,
+      "grad_norm": 1.868719458580017,
+      "learning_rate": 2.4989143796934793e-05,
+      "loss": 0.7921,
+      "step": 872000
+    },
+    {
+      "epoch": 1.5015118374202565,
+      "grad_norm": 2.4086437225341797,
+      "learning_rate": 2.4974802709662396e-05,
+      "loss": 0.786,
+      "step": 872500
+    },
+    {
+      "epoch": 1.5023723026566005,
+      "grad_norm": 2.0455057621002197,
+      "learning_rate": 2.4960461622389993e-05,
+      "loss": 0.7737,
+      "step": 873000
+    },
+    {
+      "epoch": 1.5032327678929445,
+      "grad_norm": 2.060215711593628,
+      "learning_rate": 2.4946120535117594e-05,
+      "loss": 0.7831,
+      "step": 873500
+    },
+    {
+      "epoch": 1.5040932331292884,
+      "grad_norm": 2.206716537475586,
+      "learning_rate": 2.4931779447845198e-05,
+      "loss": 0.7786,
+      "step": 874000
+    },
+    {
+      "epoch": 1.5049536983656324,
+      "grad_norm": 2.23823618888855,
+      "learning_rate": 2.4917438360572795e-05,
+      "loss": 0.7794,
+      "step": 874500
+    },
+    {
+      "epoch": 1.5058141636019764,
+      "grad_norm": 1.9800615310668945,
+      "learning_rate": 2.4903097273300395e-05,
+      "loss": 0.7717,
+      "step": 875000
+    },
+    {
+      "epoch": 1.5066746288383204,
+      "grad_norm": 2.373481035232544,
+      "learning_rate": 2.4888756186027996e-05,
+      "loss": 0.782,
+      "step": 875500
+    },
+    {
+      "epoch": 1.5075350940746644,
+      "grad_norm": 2.071988344192505,
+      "learning_rate": 2.4874415098755596e-05,
+      "loss": 0.7754,
+      "step": 876000
+    },
+    {
+      "epoch": 1.5083955593110083,
+      "grad_norm": 2.165459394454956,
+      "learning_rate": 2.4860074011483197e-05,
+      "loss": 0.7822,
+      "step": 876500
+    },
+    {
+      "epoch": 1.5092560245473523,
+      "grad_norm": 2.2290797233581543,
+      "learning_rate": 2.4845732924210797e-05,
+      "loss": 0.7783,
+      "step": 877000
+    },
+    {
+      "epoch": 1.5101164897836963,
+      "grad_norm": 1.7810156345367432,
+      "learning_rate": 2.4831391836938398e-05,
+      "loss": 0.7803,
+      "step": 877500
+    },
+    {
+      "epoch": 1.5109769550200403,
+      "grad_norm": 2.023090124130249,
+      "learning_rate": 2.4817050749665995e-05,
+      "loss": 0.7701,
+      "step": 878000
+    },
+    {
+      "epoch": 1.5118374202563842,
+      "grad_norm": 2.0860238075256348,
+      "learning_rate": 2.48027096623936e-05,
+      "loss": 0.7855,
+      "step": 878500
+    },
+    {
+      "epoch": 1.5126978854927282,
+      "grad_norm": 2.0972609519958496,
+      "learning_rate": 2.47883685751212e-05,
+      "loss": 0.78,
+      "step": 879000
+    },
+    {
+      "epoch": 1.5135583507290722,
+      "grad_norm": 2.0910556316375732,
+      "learning_rate": 2.4774027487848796e-05,
+      "loss": 0.7828,
+      "step": 879500
+    },
+    {
+      "epoch": 1.5144188159654162,
+      "grad_norm": 2.109569549560547,
+      "learning_rate": 2.47596864005764e-05,
+      "loss": 0.7832,
+      "step": 880000
+    },
+    {
+      "epoch": 1.5152792812017601,
+      "grad_norm": 2.0984444618225098,
+      "learning_rate": 2.4745345313304e-05,
+      "loss": 0.7821,
+      "step": 880500
+    },
+    {
+      "epoch": 1.5161397464381041,
+      "grad_norm": 2.1185643672943115,
+      "learning_rate": 2.4731004226031598e-05,
+      "loss": 0.7831,
+      "step": 881000
+    },
+    {
+      "epoch": 1.517000211674448,
+      "grad_norm": 2.2747139930725098,
+      "learning_rate": 2.4716663138759198e-05,
+      "loss": 0.7775,
+      "step": 881500
+    },
+    {
+      "epoch": 1.517860676910792,
+      "grad_norm": 2.023024320602417,
+      "learning_rate": 2.47023220514868e-05,
+      "loss": 0.7772,
+      "step": 882000
+    },
+    {
+      "epoch": 1.518721142147136,
+      "grad_norm": 2.236903429031372,
+      "learning_rate": 2.46879809642144e-05,
+      "loss": 0.7797,
+      "step": 882500
+    },
+    {
+      "epoch": 1.51958160738348,
+      "grad_norm": 2.104433298110962,
+      "learning_rate": 2.4673639876942e-05,
+      "loss": 0.7823,
+      "step": 883000
+    },
+    {
+      "epoch": 1.520442072619824,
+      "grad_norm": 2.371293544769287,
+      "learning_rate": 2.46592987896696e-05,
+      "loss": 0.7781,
+      "step": 883500
+    },
+    {
+      "epoch": 1.521302537856168,
+      "grad_norm": 2.0767617225646973,
+      "learning_rate": 2.46449577023972e-05,
+      "loss": 0.778,
+      "step": 884000
+    },
+    {
+      "epoch": 1.522163003092512,
+      "grad_norm": 1.856438159942627,
+      "learning_rate": 2.4630616615124798e-05,
+      "loss": 0.7771,
+      "step": 884500
+    },
+    {
+      "epoch": 1.523023468328856,
+      "grad_norm": 2.0989482402801514,
+      "learning_rate": 2.46162755278524e-05,
+      "loss": 0.7774,
+      "step": 885000
+    },
+    {
+      "epoch": 1.5238839335652,
+      "grad_norm": 1.8530632257461548,
+      "learning_rate": 2.4601934440580002e-05,
+      "loss": 0.7752,
+      "step": 885500
+    },
+    {
+      "epoch": 1.5247443988015439,
+      "grad_norm": 2.0262563228607178,
+      "learning_rate": 2.45875933533076e-05,
+      "loss": 0.7796,
+      "step": 886000
+    },
+    {
+      "epoch": 1.5256048640378879,
+      "grad_norm": 2.2254626750946045,
+      "learning_rate": 2.4573252266035203e-05,
+      "loss": 0.7793,
+      "step": 886500
+    },
+    {
+      "epoch": 1.5264653292742318,
+      "grad_norm": 2.0390655994415283,
+      "learning_rate": 2.45589111787628e-05,
+      "loss": 0.7805,
+      "step": 887000
+    },
+    {
+      "epoch": 1.5273257945105758,
+      "grad_norm": 2.127013683319092,
+      "learning_rate": 2.45445700914904e-05,
+      "loss": 0.7819,
+      "step": 887500
+    },
+    {
+      "epoch": 1.52818625974692,
+      "grad_norm": 2.007023572921753,
+      "learning_rate": 2.4530229004218004e-05,
+      "loss": 0.7837,
+      "step": 888000
+    },
+    {
+      "epoch": 1.529046724983264,
+      "grad_norm": 1.9689440727233887,
+      "learning_rate": 2.45158879169456e-05,
+      "loss": 0.7796,
+      "step": 888500
+    },
+    {
+      "epoch": 1.529907190219608,
+      "grad_norm": 1.9465103149414062,
+      "learning_rate": 2.4501546829673202e-05,
+      "loss": 0.7783,
+      "step": 889000
+    },
+    {
+      "epoch": 1.530767655455952,
+      "grad_norm": 2.0596694946289062,
+      "learning_rate": 2.4487205742400802e-05,
+      "loss": 0.7874,
+      "step": 889500
+    },
+    {
+      "epoch": 1.531628120692296,
+      "grad_norm": 2.059352397918701,
+      "learning_rate": 2.4472864655128403e-05,
+      "loss": 0.7789,
+      "step": 890000
+    },
+    {
+      "epoch": 1.53248858592864,
+      "grad_norm": 2.119255542755127,
+      "learning_rate": 2.4458523567856003e-05,
+      "loss": 0.782,
+      "step": 890500
+    },
+    {
+      "epoch": 1.5333490511649839,
+      "grad_norm": 2.0945546627044678,
+      "learning_rate": 2.4444182480583604e-05,
+      "loss": 0.7888,
+      "step": 891000
+    },
+    {
+      "epoch": 1.5342095164013279,
+      "grad_norm": 2.090996265411377,
+      "learning_rate": 2.4429841393311204e-05,
+      "loss": 0.7818,
+      "step": 891500
+    },
+    {
+      "epoch": 1.5350699816376718,
+      "grad_norm": 2.1802480220794678,
+      "learning_rate": 2.4415500306038805e-05,
+      "loss": 0.7745,
+      "step": 892000
+    },
+    {
+      "epoch": 1.5359304468740158,
+      "grad_norm": 1.8560463190078735,
+      "learning_rate": 2.4401159218766405e-05,
+      "loss": 0.7792,
+      "step": 892500
+    },
+    {
+      "epoch": 1.5367909121103598,
+      "grad_norm": 2.067178249359131,
+      "learning_rate": 2.4386818131494006e-05,
+      "loss": 0.7813,
+      "step": 893000
+    },
+    {
+      "epoch": 1.5376513773467038,
+      "grad_norm": 2.077035665512085,
+      "learning_rate": 2.4372477044221603e-05,
+      "loss": 0.7756,
+      "step": 893500
+    },
+    {
+      "epoch": 1.538511842583048,
+      "grad_norm": 1.9644643068313599,
+      "learning_rate": 2.4358135956949203e-05,
+      "loss": 0.775,
+      "step": 894000
+    },
+    {
+      "epoch": 1.539372307819392,
+      "grad_norm": 2.1593332290649414,
+      "learning_rate": 2.4343794869676807e-05,
+      "loss": 0.7829,
+      "step": 894500
+    },
+    {
+      "epoch": 1.540232773055736,
+      "grad_norm": 2.035702705383301,
+      "learning_rate": 2.4329453782404404e-05,
+      "loss": 0.7789,
+      "step": 895000
+    },
+    {
+      "epoch": 1.54109323829208,
+      "grad_norm": 1.9208450317382812,
+      "learning_rate": 2.4315112695132005e-05,
+      "loss": 0.7737,
+      "step": 895500
+    },
+    {
+      "epoch": 1.5419537035284239,
+      "grad_norm": 2.035158395767212,
+      "learning_rate": 2.4300771607859605e-05,
+      "loss": 0.7754,
+      "step": 896000
+    },
+    {
+      "epoch": 1.5428141687647678,
+      "grad_norm": 2.130017042160034,
+      "learning_rate": 2.4286430520587205e-05,
+      "loss": 0.7805,
+      "step": 896500
+    },
+    {
+      "epoch": 1.5436746340011118,
+      "grad_norm": 2.0422582626342773,
+      "learning_rate": 2.4272089433314806e-05,
+      "loss": 0.7746,
+      "step": 897000
+    },
+    {
+      "epoch": 1.5445350992374558,
+      "grad_norm": 1.9432518482208252,
+      "learning_rate": 2.4257748346042406e-05,
+      "loss": 0.7772,
+      "step": 897500
+    },
+    {
+      "epoch": 1.5453955644737998,
+      "grad_norm": 2.2162744998931885,
+      "learning_rate": 2.4243407258770007e-05,
+      "loss": 0.7771,
+      "step": 898000
+    },
+    {
+      "epoch": 1.5462560297101438,
+      "grad_norm": 1.968711018562317,
+      "learning_rate": 2.4229066171497604e-05,
+      "loss": 0.7763,
+      "step": 898500
+    },
+    {
+      "epoch": 1.5471164949464877,
+      "grad_norm": 2.030256748199463,
+      "learning_rate": 2.4214725084225208e-05,
+      "loss": 0.7808,
+      "step": 899000
+    },
+    {
+      "epoch": 1.5479769601828317,
+      "grad_norm": 2.1321399211883545,
+      "learning_rate": 2.4200383996952808e-05,
+      "loss": 0.7828,
+      "step": 899500
+    },
+    {
+      "epoch": 1.5488374254191757,
+      "grad_norm": 2.0849130153656006,
+      "learning_rate": 2.4186042909680405e-05,
+      "loss": 0.7749,
+      "step": 900000
+    },
+    {
+      "epoch": 1.5496978906555197,
+      "grad_norm": 2.061917543411255,
+      "learning_rate": 2.417170182240801e-05,
+      "loss": 0.7801,
+      "step": 900500
+    },
+    {
+      "epoch": 1.5505583558918636,
+      "grad_norm": 1.9755289554595947,
+      "learning_rate": 2.4157360735135606e-05,
+      "loss": 0.7789,
+      "step": 901000
+    },
+    {
+      "epoch": 1.5514188211282076,
+      "grad_norm": 2.4730238914489746,
+      "learning_rate": 2.4143019647863207e-05,
+      "loss": 0.7739,
+      "step": 901500
+    },
+    {
+      "epoch": 1.5522792863645516,
+      "grad_norm": 2.032313108444214,
+      "learning_rate": 2.4128678560590807e-05,
+      "loss": 0.7821,
+      "step": 902000
+    },
+    {
+      "epoch": 1.5531397516008956,
+      "grad_norm": 2.0177109241485596,
+      "learning_rate": 2.4114337473318408e-05,
+      "loss": 0.7789,
+      "step": 902500
+    },
+    {
+      "epoch": 1.5540002168372395,
+      "grad_norm": 2.158867120742798,
+      "learning_rate": 2.4099996386046008e-05,
+      "loss": 0.7752,
+      "step": 903000
+    },
+    {
+      "epoch": 1.5548606820735835,
+      "grad_norm": 2.1361749172210693,
+      "learning_rate": 2.408565529877361e-05,
+      "loss": 0.7755,
+      "step": 903500
+    },
+    {
+      "epoch": 1.5557211473099275,
+      "grad_norm": 2.079887628555298,
+      "learning_rate": 2.407131421150121e-05,
+      "loss": 0.7744,
+      "step": 904000
+    },
+    {
+      "epoch": 1.5565816125462715,
+      "grad_norm": 2.108428478240967,
+      "learning_rate": 2.405697312422881e-05,
+      "loss": 0.7738,
+      "step": 904500
+    },
+    {
+      "epoch": 1.5574420777826155,
+      "grad_norm": 2.295281171798706,
+      "learning_rate": 2.404263203695641e-05,
+      "loss": 0.7747,
+      "step": 905000
+    },
+    {
+      "epoch": 1.5583025430189594,
+      "grad_norm": 2.012521743774414,
+      "learning_rate": 2.402829094968401e-05,
+      "loss": 0.7816,
+      "step": 905500
+    },
+    {
+      "epoch": 1.5591630082553034,
+      "grad_norm": 2.0468950271606445,
+      "learning_rate": 2.401394986241161e-05,
+      "loss": 0.7806,
+      "step": 906000
+    },
+    {
+      "epoch": 1.5600234734916474,
+      "grad_norm": 2.070603609085083,
+      "learning_rate": 2.3999608775139208e-05,
+      "loss": 0.7759,
+      "step": 906500
+    },
+    {
+      "epoch": 1.5608839387279914,
+      "grad_norm": 2.0377748012542725,
+      "learning_rate": 2.3985267687866812e-05,
+      "loss": 0.7781,
+      "step": 907000
+    },
+    {
+      "epoch": 1.5617444039643353,
+      "grad_norm": 2.356083631515503,
+      "learning_rate": 2.397092660059441e-05,
+      "loss": 0.7772,
+      "step": 907500
+    },
+    {
+      "epoch": 1.5626048692006793,
+      "grad_norm": 2.108891010284424,
+      "learning_rate": 2.395658551332201e-05,
+      "loss": 0.7729,
+      "step": 908000
+    },
+    {
+      "epoch": 1.5634653344370233,
+      "grad_norm": 2.133230686187744,
+      "learning_rate": 2.3942244426049613e-05,
+      "loss": 0.772,
+      "step": 908500
+    },
+    {
+      "epoch": 1.5643257996733673,
+      "grad_norm": 2.0232138633728027,
+      "learning_rate": 2.392790333877721e-05,
+      "loss": 0.7725,
+      "step": 909000
+    },
+    {
+      "epoch": 1.5651862649097112,
+      "grad_norm": 2.1651556491851807,
+      "learning_rate": 2.391356225150481e-05,
+      "loss": 0.7794,
+      "step": 909500
+    },
+    {
+      "epoch": 1.5660467301460552,
+      "grad_norm": 2.0832695960998535,
+      "learning_rate": 2.389922116423241e-05,
+      "loss": 0.7836,
+      "step": 910000
+    },
+    {
+      "epoch": 1.5669071953823992,
+      "grad_norm": 2.2311220169067383,
+      "learning_rate": 2.3884880076960012e-05,
+      "loss": 0.7729,
+      "step": 910500
+    },
+    {
+      "epoch": 1.5677676606187434,
+      "grad_norm": 1.9575518369674683,
+      "learning_rate": 2.3870538989687612e-05,
+      "loss": 0.7852,
+      "step": 911000
+    },
+    {
+      "epoch": 1.5686281258550874,
+      "grad_norm": 2.2378692626953125,
+      "learning_rate": 2.3856197902415213e-05,
+      "loss": 0.775,
+      "step": 911500
+    },
+    {
+      "epoch": 1.5694885910914314,
+      "grad_norm": 2.328058958053589,
+      "learning_rate": 2.3841856815142813e-05,
+      "loss": 0.7738,
+      "step": 912000
+    },
+    {
+      "epoch": 1.5703490563277753,
+      "grad_norm": 2.132398843765259,
+      "learning_rate": 2.382751572787041e-05,
+      "loss": 0.7826,
+      "step": 912500
+    },
+    {
+      "epoch": 1.5712095215641193,
+      "grad_norm": 2.2009437084198,
+      "learning_rate": 2.3813174640598014e-05,
+      "loss": 0.7766,
+      "step": 913000
+    },
+    {
+      "epoch": 1.5720699868004633,
+      "grad_norm": 2.1141409873962402,
+      "learning_rate": 2.3798833553325615e-05,
+      "loss": 0.7776,
+      "step": 913500
+    },
+    {
+      "epoch": 1.5729304520368073,
+      "grad_norm": 2.1866352558135986,
+      "learning_rate": 2.3784492466053212e-05,
+      "loss": 0.7801,
+      "step": 914000
+    },
+    {
+      "epoch": 1.5737909172731512,
+      "grad_norm": 2.5267555713653564,
+      "learning_rate": 2.3770151378780812e-05,
+      "loss": 0.7754,
+      "step": 914500
+    },
+    {
+      "epoch": 1.5746513825094952,
+      "grad_norm": 2.0585038661956787,
+      "learning_rate": 2.3755810291508416e-05,
+      "loss": 0.7799,
+      "step": 915000
+    },
+    {
+      "epoch": 1.5755118477458392,
+      "grad_norm": 2.211273670196533,
+      "learning_rate": 2.3741469204236013e-05,
+      "loss": 0.7758,
+      "step": 915500
+    },
+    {
+      "epoch": 1.5763723129821832,
+      "grad_norm": 2.3003201484680176,
+      "learning_rate": 2.3727128116963614e-05,
+      "loss": 0.7742,
+      "step": 916000
+    },
+    {
+      "epoch": 1.5772327782185271,
+      "grad_norm": 2.176165819168091,
+      "learning_rate": 2.3712787029691214e-05,
+      "loss": 0.7792,
+      "step": 916500
+    },
+    {
+      "epoch": 1.5780932434548713,
+      "grad_norm": 2.0095202922821045,
+      "learning_rate": 2.3698445942418815e-05,
+      "loss": 0.7768,
+      "step": 917000
+    },
+    {
+      "epoch": 1.5789537086912153,
+      "grad_norm": 1.853319764137268,
+      "learning_rate": 2.3684104855146415e-05,
+      "loss": 0.7828,
+      "step": 917500
+    },
+    {
+      "epoch": 1.5798141739275593,
+      "grad_norm": 1.9998551607131958,
+      "learning_rate": 2.3669763767874016e-05,
+      "loss": 0.775,
+      "step": 918000
+    },
+    {
+      "epoch": 1.5806746391639033,
+      "grad_norm": 2.039543867111206,
+      "learning_rate": 2.3655422680601616e-05,
+      "loss": 0.7798,
+      "step": 918500
+    },
+    {
+      "epoch": 1.5815351044002472,
+      "grad_norm": 1.8610059022903442,
+      "learning_rate": 2.3641081593329213e-05,
+      "loss": 0.7767,
+      "step": 919000
+    },
+    {
+      "epoch": 1.5823955696365912,
+      "grad_norm": 2.313720703125,
+      "learning_rate": 2.3626740506056817e-05,
+      "loss": 0.7792,
+      "step": 919500
+    },
+    {
+      "epoch": 1.5832560348729352,
+      "grad_norm": 1.9850460290908813,
+      "learning_rate": 2.3612399418784417e-05,
+      "loss": 0.7795,
+      "step": 920000
+    },
+    {
+      "epoch": 1.5841165001092792,
+      "grad_norm": 2.0019094944000244,
+      "learning_rate": 2.3598058331512014e-05,
+      "loss": 0.775,
+      "step": 920500
+    },
+    {
+      "epoch": 1.5849769653456232,
+      "grad_norm": 2.1106672286987305,
+      "learning_rate": 2.358371724423962e-05,
+      "loss": 0.7683,
+      "step": 921000
+    },
+    {
+      "epoch": 1.5858374305819671,
+      "grad_norm": 2.102208137512207,
+      "learning_rate": 2.3569376156967215e-05,
+      "loss": 0.775,
+      "step": 921500
+    },
+    {
+      "epoch": 1.586697895818311,
+      "grad_norm": 2.1192538738250732,
+      "learning_rate": 2.3555035069694816e-05,
+      "loss": 0.7751,
+      "step": 922000
+    },
+    {
+      "epoch": 1.587558361054655,
+      "grad_norm": 2.1068108081817627,
+      "learning_rate": 2.354069398242242e-05,
+      "loss": 0.775,
+      "step": 922500
+    },
+    {
+      "epoch": 1.588418826290999,
+      "grad_norm": 1.947999358177185,
+      "learning_rate": 2.3526352895150017e-05,
+      "loss": 0.7753,
+      "step": 923000
+    },
+    {
+      "epoch": 1.589279291527343,
+      "grad_norm": 2.0711350440979004,
+      "learning_rate": 2.3512011807877617e-05,
+      "loss": 0.7682,
+      "step": 923500
+    },
+    {
+      "epoch": 1.590139756763687,
+      "grad_norm": 2.0397300720214844,
+      "learning_rate": 2.3497670720605218e-05,
+      "loss": 0.774,
+      "step": 924000
+    },
+    {
+      "epoch": 1.591000222000031,
+      "grad_norm": 2.084848165512085,
+      "learning_rate": 2.3483329633332818e-05,
+      "loss": 0.7811,
+      "step": 924500
+    },
+    {
+      "epoch": 1.591860687236375,
+      "grad_norm": 2.303589105606079,
+      "learning_rate": 2.346898854606042e-05,
+      "loss": 0.7776,
+      "step": 925000
+    },
+    {
+      "epoch": 1.592721152472719,
+      "grad_norm": 2.401545524597168,
+      "learning_rate": 2.345464745878802e-05,
+      "loss": 0.7766,
+      "step": 925500
+    },
+    {
+      "epoch": 1.593581617709063,
+      "grad_norm": 1.9984424114227295,
+      "learning_rate": 2.344030637151562e-05,
+      "loss": 0.7738,
+      "step": 926000
+    },
+    {
+      "epoch": 1.594442082945407,
+      "grad_norm": 2.219634532928467,
+      "learning_rate": 2.342596528424322e-05,
+      "loss": 0.7714,
+      "step": 926500
+    },
+    {
+      "epoch": 1.5953025481817509,
+      "grad_norm": 1.869746208190918,
+      "learning_rate": 2.3411624196970817e-05,
+      "loss": 0.7761,
+      "step": 927000
+    },
+    {
+      "epoch": 1.5961630134180949,
+      "grad_norm": 2.093459129333496,
+      "learning_rate": 2.339728310969842e-05,
+      "loss": 0.7746,
+      "step": 927500
+    },
+    {
+      "epoch": 1.5970234786544388,
+      "grad_norm": 2.0869009494781494,
+      "learning_rate": 2.3382942022426018e-05,
+      "loss": 0.7769,
+      "step": 928000
+    },
+    {
+      "epoch": 1.5978839438907828,
+      "grad_norm": 1.8973535299301147,
+      "learning_rate": 2.336860093515362e-05,
+      "loss": 0.7741,
+      "step": 928500
+    },
+    {
+      "epoch": 1.5987444091271268,
+      "grad_norm": 2.18300461769104,
+      "learning_rate": 2.3354259847881222e-05,
+      "loss": 0.7759,
+      "step": 929000
+    },
+    {
+      "epoch": 1.5996048743634708,
+      "grad_norm": 2.473493814468384,
+      "learning_rate": 2.333991876060882e-05,
+      "loss": 0.7796,
+      "step": 929500
+    },
+    {
+      "epoch": 1.6004653395998147,
+      "grad_norm": 1.9093661308288574,
+      "learning_rate": 2.332557767333642e-05,
+      "loss": 0.7755,
+      "step": 930000
+    },
+    {
+      "epoch": 1.6013258048361587,
+      "grad_norm": 2.159060478210449,
+      "learning_rate": 2.331123658606402e-05,
+      "loss": 0.7775,
+      "step": 930500
+    },
+    {
+      "epoch": 1.6021862700725027,
+      "grad_norm": 2.0607120990753174,
+      "learning_rate": 2.329689549879162e-05,
+      "loss": 0.769,
+      "step": 931000
+    },
+    {
+      "epoch": 1.6030467353088467,
+      "grad_norm": 2.243135929107666,
+      "learning_rate": 2.328255441151922e-05,
+      "loss": 0.7767,
+      "step": 931500
+    },
+    {
+      "epoch": 1.6039072005451906,
+      "grad_norm": 2.1867527961730957,
+      "learning_rate": 2.3268213324246822e-05,
+      "loss": 0.7775,
+      "step": 932000
+    },
+    {
+      "epoch": 1.6047676657815346,
+      "grad_norm": 1.9539246559143066,
+      "learning_rate": 2.3253872236974422e-05,
+      "loss": 0.7721,
+      "step": 932500
+    },
+    {
+      "epoch": 1.6056281310178786,
+      "grad_norm": 2.351499557495117,
+      "learning_rate": 2.323953114970202e-05,
+      "loss": 0.7676,
+      "step": 933000
+    },
+    {
+      "epoch": 1.6064885962542226,
+      "grad_norm": 2.0803720951080322,
+      "learning_rate": 2.3225190062429623e-05,
+      "loss": 0.7755,
+      "step": 933500
+    },
+    {
+      "epoch": 1.6073490614905666,
+      "grad_norm": 2.089561700820923,
+      "learning_rate": 2.3210848975157224e-05,
+      "loss": 0.7735,
+      "step": 934000
+    },
+    {
+      "epoch": 1.6082095267269108,
+      "grad_norm": 2.2439472675323486,
+      "learning_rate": 2.319650788788482e-05,
+      "loss": 0.7818,
+      "step": 934500
+    },
+    {
+      "epoch": 1.6090699919632547,
+      "grad_norm": 2.1251161098480225,
+      "learning_rate": 2.3182166800612425e-05,
+      "loss": 0.7733,
+      "step": 935000
+    },
+    {
+      "epoch": 1.6099304571995987,
+      "grad_norm": 2.3418426513671875,
+      "learning_rate": 2.3167825713340025e-05,
+      "loss": 0.7765,
+      "step": 935500
+    },
+    {
+      "epoch": 1.6107909224359427,
+      "grad_norm": 2.1627066135406494,
+      "learning_rate": 2.3153484626067622e-05,
+      "loss": 0.7717,
+      "step": 936000
+    },
+    {
+      "epoch": 1.6116513876722867,
+      "grad_norm": 2.13486385345459,
+      "learning_rate": 2.3139143538795223e-05,
+      "loss": 0.7706,
+      "step": 936500
+    },
+    {
+      "epoch": 1.6125118529086306,
+      "grad_norm": 1.9072346687316895,
+      "learning_rate": 2.3124802451522823e-05,
+      "loss": 0.775,
+      "step": 937000
+    },
+    {
+      "epoch": 1.6133723181449746,
+      "grad_norm": 2.1810319423675537,
+      "learning_rate": 2.3110461364250424e-05,
+      "loss": 0.7739,
+      "step": 937500
+    },
+    {
+      "epoch": 1.6142327833813186,
+      "grad_norm": 2.179624080657959,
+      "learning_rate": 2.3096120276978024e-05,
+      "loss": 0.7716,
+      "step": 938000
+    },
+    {
+      "epoch": 1.6150932486176626,
+      "grad_norm": 2.184668779373169,
+      "learning_rate": 2.3081779189705625e-05,
+      "loss": 0.7727,
+      "step": 938500
+    },
+    {
+      "epoch": 1.6159537138540065,
+      "grad_norm": 1.9632959365844727,
+      "learning_rate": 2.3067438102433225e-05,
+      "loss": 0.7714,
+      "step": 939000
+    },
+    {
+      "epoch": 1.6168141790903505,
+      "grad_norm": 2.0400161743164062,
+      "learning_rate": 2.3053097015160822e-05,
+      "loss": 0.7764,
+      "step": 939500
+    },
+    {
+      "epoch": 1.6176746443266947,
+      "grad_norm": 2.226841926574707,
+      "learning_rate": 2.3038755927888426e-05,
+      "loss": 0.7689,
+      "step": 940000
+    },
+    {
+      "epoch": 1.6185351095630387,
+      "grad_norm": 2.1980738639831543,
+      "learning_rate": 2.3024414840616027e-05,
+      "loss": 0.7724,
+      "step": 940500
+    },
+    {
+      "epoch": 1.6193955747993827,
+      "grad_norm": 2.0220143795013428,
+      "learning_rate": 2.3010073753343624e-05,
+      "loss": 0.7769,
+      "step": 941000
+    },
+    {
+      "epoch": 1.6202560400357267,
+      "grad_norm": 2.0732641220092773,
+      "learning_rate": 2.2995732666071227e-05,
+      "loss": 0.7795,
+      "step": 941500
+    },
+    {
+      "epoch": 1.6211165052720706,
+      "grad_norm": 1.9810514450073242,
+      "learning_rate": 2.2981391578798825e-05,
+      "loss": 0.775,
+      "step": 942000
+    },
+    {
+      "epoch": 1.6219769705084146,
+      "grad_norm": 2.1949236392974854,
+      "learning_rate": 2.2967050491526425e-05,
+      "loss": 0.7668,
+      "step": 942500
+    },
+    {
+      "epoch": 1.6228374357447586,
+      "grad_norm": 2.2184219360351562,
+      "learning_rate": 2.295270940425403e-05,
+      "loss": 0.774,
+      "step": 943000
+    },
+    {
+      "epoch": 1.6236979009811026,
+      "grad_norm": 2.1051313877105713,
+      "learning_rate": 2.2938368316981626e-05,
+      "loss": 0.7711,
+      "step": 943500
+    },
+    {
+      "epoch": 1.6245583662174465,
+      "grad_norm": 2.0821473598480225,
+      "learning_rate": 2.2924027229709226e-05,
+      "loss": 0.7824,
+      "step": 944000
+    },
+    {
+      "epoch": 1.6254188314537905,
+      "grad_norm": 2.077939987182617,
+      "learning_rate": 2.2909686142436827e-05,
+      "loss": 0.7745,
+      "step": 944500
+    },
+    {
+      "epoch": 1.6262792966901345,
+      "grad_norm": 2.041835069656372,
+      "learning_rate": 2.2895345055164427e-05,
+      "loss": 0.777,
+      "step": 945000
+    },
+    {
+      "epoch": 1.6271397619264785,
+      "grad_norm": 2.2172017097473145,
+      "learning_rate": 2.2881003967892028e-05,
+      "loss": 0.7778,
+      "step": 945500
+    },
+    {
+      "epoch": 1.6280002271628224,
+      "grad_norm": 2.0836751461029053,
+      "learning_rate": 2.2866662880619628e-05,
+      "loss": 0.7749,
+      "step": 946000
+    },
+    {
+      "epoch": 1.6288606923991664,
+      "grad_norm": 1.9881094694137573,
+      "learning_rate": 2.285232179334723e-05,
+      "loss": 0.7735,
+      "step": 946500
+    },
+    {
+      "epoch": 1.6297211576355104,
+      "grad_norm": 2.445798873901367,
+      "learning_rate": 2.283798070607483e-05,
+      "loss": 0.7689,
+      "step": 947000
+    },
+    {
+      "epoch": 1.6305816228718544,
+      "grad_norm": 2.1970605850219727,
+      "learning_rate": 2.282363961880243e-05,
+      "loss": 0.7704,
+      "step": 947500
+    },
+    {
+      "epoch": 1.6314420881081984,
+      "grad_norm": 2.2833495140075684,
+      "learning_rate": 2.280929853153003e-05,
+      "loss": 0.7726,
+      "step": 948000
+    },
+    {
+      "epoch": 1.6323025533445423,
+      "grad_norm": 2.2963755130767822,
+      "learning_rate": 2.2794957444257627e-05,
+      "loss": 0.7721,
+      "step": 948500
+    },
+    {
+      "epoch": 1.6331630185808863,
+      "grad_norm": 2.0407752990722656,
+      "learning_rate": 2.2780616356985228e-05,
+      "loss": 0.7708,
+      "step": 949000
+    },
+    {
+      "epoch": 1.6340234838172303,
+      "grad_norm": 2.055379867553711,
+      "learning_rate": 2.276627526971283e-05,
+      "loss": 0.7735,
+      "step": 949500
+    },
+    {
+      "epoch": 1.6348839490535743,
+      "grad_norm": 1.9972501993179321,
+      "learning_rate": 2.275193418244043e-05,
+      "loss": 0.766,
+      "step": 950000
+    },
+    {
+      "epoch": 1.6357444142899182,
+      "grad_norm": 1.9890713691711426,
+      "learning_rate": 2.273759309516803e-05,
+      "loss": 0.7742,
+      "step": 950500
+    },
+    {
+      "epoch": 1.6366048795262622,
+      "grad_norm": 2.0152220726013184,
+      "learning_rate": 2.272325200789563e-05,
+      "loss": 0.7722,
+      "step": 951000
+    },
+    {
+      "epoch": 1.6374653447626062,
+      "grad_norm": 1.9014008045196533,
+      "learning_rate": 2.270891092062323e-05,
+      "loss": 0.7738,
+      "step": 951500
+    },
+    {
+      "epoch": 1.6383258099989502,
+      "grad_norm": 2.1452271938323975,
+      "learning_rate": 2.269456983335083e-05,
+      "loss": 0.7766,
+      "step": 952000
+    },
+    {
+      "epoch": 1.6391862752352941,
+      "grad_norm": 3.4031081199645996,
+      "learning_rate": 2.268022874607843e-05,
+      "loss": 0.7694,
+      "step": 952500
+    },
+    {
+      "epoch": 1.6400467404716381,
+      "grad_norm": 2.1318936347961426,
+      "learning_rate": 2.266588765880603e-05,
+      "loss": 0.7699,
+      "step": 953000
+    },
+    {
+      "epoch": 1.640907205707982,
+      "grad_norm": 1.984821081161499,
+      "learning_rate": 2.265154657153363e-05,
+      "loss": 0.7717,
+      "step": 953500
+    },
+    {
+      "epoch": 1.641767670944326,
+      "grad_norm": 2.1505789756774902,
+      "learning_rate": 2.2637205484261232e-05,
+      "loss": 0.768,
+      "step": 954000
+    },
+    {
+      "epoch": 1.64262813618067,
+      "grad_norm": 2.2152884006500244,
+      "learning_rate": 2.2622864396988833e-05,
+      "loss": 0.7697,
+      "step": 954500
+    },
+    {
+      "epoch": 1.643488601417014,
+      "grad_norm": 2.0147619247436523,
+      "learning_rate": 2.260852330971643e-05,
+      "loss": 0.7761,
+      "step": 955000
+    },
+    {
+      "epoch": 1.644349066653358,
+      "grad_norm": 1.9571027755737305,
+      "learning_rate": 2.2594182222444034e-05,
+      "loss": 0.775,
+      "step": 955500
+    },
+    {
+      "epoch": 1.645209531889702,
+      "grad_norm": 2.1475048065185547,
+      "learning_rate": 2.257984113517163e-05,
+      "loss": 0.7704,
+      "step": 956000
+    },
+    {
+      "epoch": 1.646069997126046,
+      "grad_norm": 2.149017333984375,
+      "learning_rate": 2.256550004789923e-05,
+      "loss": 0.7772,
+      "step": 956500
+    },
+    {
+      "epoch": 1.64693046236239,
+      "grad_norm": 2.056325912475586,
+      "learning_rate": 2.2551158960626835e-05,
+      "loss": 0.7673,
+      "step": 957000
+    },
+    {
+      "epoch": 1.6477909275987341,
+      "grad_norm": 2.0990519523620605,
+      "learning_rate": 2.2536817873354432e-05,
+      "loss": 0.777,
+      "step": 957500
+    },
+    {
+      "epoch": 1.6486513928350781,
+      "grad_norm": 2.128030776977539,
+      "learning_rate": 2.2522476786082033e-05,
+      "loss": 0.7731,
+      "step": 958000
+    },
+    {
+      "epoch": 1.649511858071422,
+      "grad_norm": 2.33375883102417,
+      "learning_rate": 2.2508135698809633e-05,
+      "loss": 0.767,
+      "step": 958500
+    },
+    {
+      "epoch": 1.650372323307766,
+      "grad_norm": 2.3622138500213623,
+      "learning_rate": 2.2493794611537234e-05,
+      "loss": 0.7735,
+      "step": 959000
+    },
+    {
+      "epoch": 1.65123278854411,
+      "grad_norm": 1.9758704900741577,
+      "learning_rate": 2.2479453524264834e-05,
+      "loss": 0.767,
+      "step": 959500
+    },
+    {
+      "epoch": 1.652093253780454,
+      "grad_norm": 1.9486836194992065,
+      "learning_rate": 2.2465112436992435e-05,
+      "loss": 0.7764,
+      "step": 960000
+    },
+    {
+      "epoch": 1.652953719016798,
+      "grad_norm": 2.24963641166687,
+      "learning_rate": 2.2450771349720035e-05,
+      "loss": 0.77,
+      "step": 960500
+    },
+    {
+      "epoch": 1.653814184253142,
+      "grad_norm": 2.146021842956543,
+      "learning_rate": 2.2436430262447636e-05,
+      "loss": 0.7672,
+      "step": 961000
+    },
+    {
+      "epoch": 1.654674649489486,
+      "grad_norm": 2.173103094100952,
+      "learning_rate": 2.2422089175175233e-05,
+      "loss": 0.7654,
+      "step": 961500
+    },
+    {
+      "epoch": 1.65553511472583,
+      "grad_norm": 1.993402123451233,
+      "learning_rate": 2.2407748087902837e-05,
+      "loss": 0.7746,
+      "step": 962000
+    },
+    {
+      "epoch": 1.656395579962174,
+      "grad_norm": 2.026460647583008,
+      "learning_rate": 2.2393407000630434e-05,
+      "loss": 0.7684,
+      "step": 962500
+    },
+    {
+      "epoch": 1.6572560451985179,
+      "grad_norm": 2.0561816692352295,
+      "learning_rate": 2.2379065913358034e-05,
+      "loss": 0.7761,
+      "step": 963000
+    },
+    {
+      "epoch": 1.658116510434862,
+      "grad_norm": 2.2467610836029053,
+      "learning_rate": 2.2364724826085638e-05,
+      "loss": 0.7726,
+      "step": 963500
+    },
+    {
+      "epoch": 1.658976975671206,
+      "grad_norm": 2.3235909938812256,
+      "learning_rate": 2.2350383738813235e-05,
+      "loss": 0.7747,
+      "step": 964000
+    },
+    {
+      "epoch": 1.65983744090755,
+      "grad_norm": 2.0124661922454834,
+      "learning_rate": 2.2336042651540836e-05,
+      "loss": 0.77,
+      "step": 964500
+    },
+    {
+      "epoch": 1.660697906143894,
+      "grad_norm": 2.090353012084961,
+      "learning_rate": 2.2321701564268436e-05,
+      "loss": 0.7714,
+      "step": 965000
+    },
+    {
+      "epoch": 1.661558371380238,
+      "grad_norm": 2.065913200378418,
+      "learning_rate": 2.2307360476996036e-05,
+      "loss": 0.7634,
+      "step": 965500
+    },
+    {
+      "epoch": 1.662418836616582,
+      "grad_norm": 1.965511441230774,
+      "learning_rate": 2.2293019389723637e-05,
+      "loss": 0.7725,
+      "step": 966000
+    },
+    {
+      "epoch": 1.663279301852926,
+      "grad_norm": 2.0912036895751953,
+      "learning_rate": 2.2278678302451237e-05,
+      "loss": 0.7749,
+      "step": 966500
+    },
+    {
+      "epoch": 1.66413976708927,
+      "grad_norm": 2.1004321575164795,
+      "learning_rate": 2.2264337215178838e-05,
+      "loss": 0.7674,
+      "step": 967000
+    },
+    {
+      "epoch": 1.665000232325614,
+      "grad_norm": 2.041999101638794,
+      "learning_rate": 2.2249996127906435e-05,
+      "loss": 0.7733,
+      "step": 967500
+    },
+    {
+      "epoch": 1.6658606975619579,
+      "grad_norm": 2.0694518089294434,
+      "learning_rate": 2.223565504063404e-05,
+      "loss": 0.7739,
+      "step": 968000
+    },
+    {
+      "epoch": 1.6667211627983018,
+      "grad_norm": 2.137014389038086,
+      "learning_rate": 2.222131395336164e-05,
+      "loss": 0.7684,
+      "step": 968500
+    },
+    {
+      "epoch": 1.6675816280346458,
+      "grad_norm": 2.0307774543762207,
+      "learning_rate": 2.2206972866089236e-05,
+      "loss": 0.7685,
+      "step": 969000
+    },
+    {
+      "epoch": 1.6684420932709898,
+      "grad_norm": 2.3049731254577637,
+      "learning_rate": 2.219263177881684e-05,
+      "loss": 0.771,
+      "step": 969500
+    },
+    {
+      "epoch": 1.6693025585073338,
+      "grad_norm": 2.1394450664520264,
+      "learning_rate": 2.217829069154444e-05,
+      "loss": 0.7748,
+      "step": 970000
+    },
+    {
+      "epoch": 1.6701630237436778,
+      "grad_norm": 2.100409507751465,
+      "learning_rate": 2.2163949604272038e-05,
+      "loss": 0.7703,
+      "step": 970500
+    },
+    {
+      "epoch": 1.6710234889800217,
+      "grad_norm": 2.216879367828369,
+      "learning_rate": 2.2149608516999638e-05,
+      "loss": 0.7736,
+      "step": 971000
+    },
+    {
+      "epoch": 1.6718839542163657,
+      "grad_norm": 2.220621109008789,
+      "learning_rate": 2.213526742972724e-05,
+      "loss": 0.7746,
+      "step": 971500
+    },
+    {
+      "epoch": 1.6727444194527097,
+      "grad_norm": 2.032130241394043,
+      "learning_rate": 2.212092634245484e-05,
+      "loss": 0.7708,
+      "step": 972000
+    },
+    {
+      "epoch": 1.6736048846890537,
+      "grad_norm": 2.065264940261841,
+      "learning_rate": 2.210658525518244e-05,
+      "loss": 0.772,
+      "step": 972500
+    },
+    {
+      "epoch": 1.6744653499253976,
+      "grad_norm": 2.0372793674468994,
+      "learning_rate": 2.209224416791004e-05,
+      "loss": 0.7702,
+      "step": 973000
+    },
+    {
+      "epoch": 1.6753258151617416,
+      "grad_norm": 2.292205333709717,
+      "learning_rate": 2.207790308063764e-05,
+      "loss": 0.7802,
+      "step": 973500
+    },
+    {
+      "epoch": 1.6761862803980856,
+      "grad_norm": 1.9135103225708008,
+      "learning_rate": 2.2063561993365238e-05,
+      "loss": 0.7692,
+      "step": 974000
+    },
+    {
+      "epoch": 1.6770467456344296,
+      "grad_norm": 1.9825247526168823,
+      "learning_rate": 2.204922090609284e-05,
+      "loss": 0.7747,
+      "step": 974500
+    },
+    {
+      "epoch": 1.6779072108707735,
+      "grad_norm": 1.8973615169525146,
+      "learning_rate": 2.2034879818820442e-05,
+      "loss": 0.7717,
+      "step": 975000
+    },
+    {
+      "epoch": 1.6787676761071175,
+      "grad_norm": 2.1794941425323486,
+      "learning_rate": 2.202053873154804e-05,
+      "loss": 0.7787,
+      "step": 975500
+    },
+    {
+      "epoch": 1.6796281413434615,
+      "grad_norm": 2.13614559173584,
+      "learning_rate": 2.2006197644275643e-05,
+      "loss": 0.7687,
+      "step": 976000
+    },
+    {
+      "epoch": 1.6804886065798055,
+      "grad_norm": 2.015418767929077,
+      "learning_rate": 2.199185655700324e-05,
+      "loss": 0.7716,
+      "step": 976500
+    },
+    {
+      "epoch": 1.6813490718161495,
+      "grad_norm": 2.1459038257598877,
+      "learning_rate": 2.197751546973084e-05,
+      "loss": 0.7695,
+      "step": 977000
+    },
+    {
+      "epoch": 1.6822095370524934,
+      "grad_norm": 2.061182975769043,
+      "learning_rate": 2.1963174382458444e-05,
+      "loss": 0.7659,
+      "step": 977500
+    },
+    {
+      "epoch": 1.6830700022888374,
+      "grad_norm": 2.1469383239746094,
+      "learning_rate": 2.194883329518604e-05,
+      "loss": 0.7649,
+      "step": 978000
+    },
+    {
+      "epoch": 1.6839304675251814,
+      "grad_norm": 2.144745111465454,
+      "learning_rate": 2.1934492207913642e-05,
+      "loss": 0.7725,
+      "step": 978500
+    },
+    {
+      "epoch": 1.6847909327615254,
+      "grad_norm": 2.1422934532165527,
+      "learning_rate": 2.1920151120641242e-05,
+      "loss": 0.7703,
+      "step": 979000
+    },
+    {
+      "epoch": 1.6856513979978693,
+      "grad_norm": 2.156254529953003,
+      "learning_rate": 2.1905810033368843e-05,
+      "loss": 0.7723,
+      "step": 979500
+    },
+    {
+      "epoch": 1.6865118632342133,
+      "grad_norm": 1.8768316507339478,
+      "learning_rate": 2.1891468946096443e-05,
+      "loss": 0.7693,
+      "step": 980000
+    },
+    {
+      "epoch": 1.6873723284705575,
+      "grad_norm": 2.1997392177581787,
+      "learning_rate": 2.1877127858824044e-05,
+      "loss": 0.7717,
+      "step": 980500
+    },
+    {
+      "epoch": 1.6882327937069015,
+      "grad_norm": 2.163787603378296,
+      "learning_rate": 2.1862786771551644e-05,
+      "loss": 0.7724,
+      "step": 981000
+    },
+    {
+      "epoch": 1.6890932589432455,
+      "grad_norm": 2.279360771179199,
+      "learning_rate": 2.1848445684279245e-05,
+      "loss": 0.7652,
+      "step": 981500
+    },
+    {
+      "epoch": 1.6899537241795894,
+      "grad_norm": 1.8542513847351074,
+      "learning_rate": 2.1834104597006845e-05,
+      "loss": 0.7695,
+      "step": 982000
+    },
+    {
+      "epoch": 1.6908141894159334,
+      "grad_norm": 2.083712100982666,
+      "learning_rate": 2.1819763509734446e-05,
+      "loss": 0.7726,
+      "step": 982500
+    },
+    {
+      "epoch": 1.6916746546522774,
+      "grad_norm": 2.247375965118408,
+      "learning_rate": 2.1805422422462043e-05,
+      "loss": 0.7673,
+      "step": 983000
+    },
+    {
+      "epoch": 1.6925351198886214,
+      "grad_norm": 2.1620235443115234,
+      "learning_rate": 2.1791081335189643e-05,
+      "loss": 0.7692,
+      "step": 983500
+    },
+    {
+      "epoch": 1.6933955851249654,
+      "grad_norm": 2.0795235633850098,
+      "learning_rate": 2.1776740247917247e-05,
+      "loss": 0.7718,
+      "step": 984000
+    },
+    {
+      "epoch": 1.6942560503613093,
+      "grad_norm": 2.1203017234802246,
+      "learning_rate": 2.1762399160644844e-05,
+      "loss": 0.7677,
+      "step": 984500
+    },
+    {
+      "epoch": 1.6951165155976533,
+      "grad_norm": 4.682893753051758,
+      "learning_rate": 2.1748058073372445e-05,
+      "loss": 0.766,
+      "step": 985000
+    },
+    {
+      "epoch": 1.6959769808339973,
+      "grad_norm": 2.2313523292541504,
+      "learning_rate": 2.1733716986100045e-05,
+      "loss": 0.7667,
+      "step": 985500
+    },
+    {
+      "epoch": 1.6968374460703413,
+      "grad_norm": 2.307849407196045,
+      "learning_rate": 2.1719375898827646e-05,
+      "loss": 0.7674,
+      "step": 986000
+    },
+    {
+      "epoch": 1.6976979113066855,
+      "grad_norm": 2.143303632736206,
+      "learning_rate": 2.1705034811555246e-05,
+      "loss": 0.7698,
+      "step": 986500
+    },
+    {
+      "epoch": 1.6985583765430294,
+      "grad_norm": 2.1578030586242676,
+      "learning_rate": 2.1690693724282847e-05,
+      "loss": 0.7729,
+      "step": 987000
+    },
+    {
+      "epoch": 1.6994188417793734,
+      "grad_norm": 2.201508045196533,
+      "learning_rate": 2.1676352637010447e-05,
+      "loss": 0.7677,
+      "step": 987500
+    },
+    {
+      "epoch": 1.7002793070157174,
+      "grad_norm": 2.0309183597564697,
+      "learning_rate": 2.1662011549738044e-05,
+      "loss": 0.7733,
+      "step": 988000
+    },
+    {
+      "epoch": 1.7011397722520614,
+      "grad_norm": 2.2656939029693604,
+      "learning_rate": 2.1647670462465648e-05,
+      "loss": 0.7658,
+      "step": 988500
+    },
+    {
+      "epoch": 1.7020002374884053,
+      "grad_norm": 2.0198137760162354,
+      "learning_rate": 2.163332937519325e-05,
+      "loss": 0.7708,
+      "step": 989000
+    },
+    {
+      "epoch": 1.7028607027247493,
+      "grad_norm": 2.3302905559539795,
+      "learning_rate": 2.1618988287920846e-05,
+      "loss": 0.7731,
+      "step": 989500
+    },
+    {
+      "epoch": 1.7037211679610933,
+      "grad_norm": 1.9794892072677612,
+      "learning_rate": 2.160464720064845e-05,
+      "loss": 0.7635,
+      "step": 990000
+    },
+    {
+      "epoch": 1.7045816331974373,
+      "grad_norm": 2.205071210861206,
+      "learning_rate": 2.159030611337605e-05,
+      "loss": 0.7698,
+      "step": 990500
+    },
+    {
+      "epoch": 1.7054420984337813,
+      "grad_norm": 2.046074867248535,
+      "learning_rate": 2.1575965026103647e-05,
+      "loss": 0.7729,
+      "step": 991000
+    },
+    {
+      "epoch": 1.7063025636701252,
+      "grad_norm": 2.0251247882843018,
+      "learning_rate": 2.156162393883125e-05,
+      "loss": 0.7605,
+      "step": 991500
+    },
+    {
+      "epoch": 1.7071630289064692,
+      "grad_norm": 1.9598090648651123,
+      "learning_rate": 2.1547282851558848e-05,
+      "loss": 0.7742,
+      "step": 992000
+    },
+    {
+      "epoch": 1.7080234941428132,
+      "grad_norm": 2.079505205154419,
+      "learning_rate": 2.153294176428645e-05,
+      "loss": 0.7694,
+      "step": 992500
+    },
+    {
+      "epoch": 1.7088839593791572,
+      "grad_norm": 2.1227147579193115,
+      "learning_rate": 2.151860067701405e-05,
+      "loss": 0.7683,
+      "step": 993000
+    },
+    {
+      "epoch": 1.7097444246155011,
+      "grad_norm": 1.9971901178359985,
+      "learning_rate": 2.150425958974165e-05,
+      "loss": 0.7697,
+      "step": 993500
+    },
+    {
+      "epoch": 1.7106048898518451,
+      "grad_norm": 2.0586302280426025,
+      "learning_rate": 2.148991850246925e-05,
+      "loss": 0.766,
+      "step": 994000
+    },
+    {
+      "epoch": 1.711465355088189,
+      "grad_norm": 1.9828028678894043,
+      "learning_rate": 2.147557741519685e-05,
+      "loss": 0.7666,
+      "step": 994500
+    },
+    {
+      "epoch": 1.712325820324533,
+      "grad_norm": 1.8900585174560547,
+      "learning_rate": 2.146123632792445e-05,
+      "loss": 0.7647,
+      "step": 995000
+    },
+    {
+      "epoch": 1.713186285560877,
+      "grad_norm": 2.2461373805999756,
+      "learning_rate": 2.144689524065205e-05,
+      "loss": 0.7719,
+      "step": 995500
+    },
+    {
+      "epoch": 1.714046750797221,
+      "grad_norm": 2.4255807399749756,
+      "learning_rate": 2.1432554153379648e-05,
+      "loss": 0.7637,
+      "step": 996000
+    },
+    {
+      "epoch": 1.714907216033565,
+      "grad_norm": 2.102754831314087,
+      "learning_rate": 2.1418213066107252e-05,
+      "loss": 0.7621,
+      "step": 996500
+    },
+    {
+      "epoch": 1.715767681269909,
+      "grad_norm": 1.9721736907958984,
+      "learning_rate": 2.140387197883485e-05,
+      "loss": 0.7622,
+      "step": 997000
+    },
+    {
+      "epoch": 1.716628146506253,
+      "grad_norm": 2.0540759563446045,
+      "learning_rate": 2.138953089156245e-05,
+      "loss": 0.7702,
+      "step": 997500
+    },
+    {
+      "epoch": 1.717488611742597,
+      "grad_norm": 2.1698076725006104,
+      "learning_rate": 2.1375189804290054e-05,
+      "loss": 0.769,
+      "step": 998000
+    },
+    {
+      "epoch": 1.718349076978941,
+      "grad_norm": 2.2580907344818115,
+      "learning_rate": 2.136084871701765e-05,
+      "loss": 0.7667,
+      "step": 998500
+    },
+    {
+      "epoch": 1.7192095422152849,
+      "grad_norm": 2.198620080947876,
+      "learning_rate": 2.134650762974525e-05,
+      "loss": 0.7724,
+      "step": 999000
+    },
+    {
+      "epoch": 1.7200700074516289,
+      "grad_norm": 2.2969367504119873,
+      "learning_rate": 2.133216654247285e-05,
+      "loss": 0.7674,
+      "step": 999500
+    },
+    {
+      "epoch": 1.7209304726879728,
+      "grad_norm": 2.1570844650268555,
+      "learning_rate": 2.1317825455200452e-05,
+      "loss": 0.7639,
+      "step": 1000000
+    },
+    {
+      "epoch": 1.7217909379243168,
+      "grad_norm": 1.9657459259033203,
+      "learning_rate": 2.1303484367928052e-05,
+      "loss": 0.7672,
+      "step": 1000500
+    },
+    {
+      "epoch": 1.7226514031606608,
+      "grad_norm": 1.9201055765151978,
+      "learning_rate": 2.1289143280655653e-05,
+      "loss": 0.7685,
+      "step": 1001000
+    },
+    {
+      "epoch": 1.7235118683970048,
+      "grad_norm": 2.067403793334961,
+      "learning_rate": 2.1274802193383253e-05,
+      "loss": 0.7713,
+      "step": 1001500
+    },
+    {
+      "epoch": 1.7243723336333487,
+      "grad_norm": 2.0172691345214844,
+      "learning_rate": 2.1260461106110854e-05,
+      "loss": 0.7662,
+      "step": 1002000
+    },
+    {
+      "epoch": 1.7252327988696927,
+      "grad_norm": 2.24025821685791,
+      "learning_rate": 2.1246120018838454e-05,
+      "loss": 0.7658,
+      "step": 1002500
+    },
+    {
+      "epoch": 1.7260932641060367,
+      "grad_norm": 2.196293830871582,
+      "learning_rate": 2.1231778931566055e-05,
+      "loss": 0.7639,
+      "step": 1003000
+    },
+    {
+      "epoch": 1.7269537293423807,
+      "grad_norm": 2.160950183868408,
+      "learning_rate": 2.1217437844293652e-05,
+      "loss": 0.7676,
+      "step": 1003500
+    },
+    {
+      "epoch": 1.7278141945787249,
+      "grad_norm": 2.226396083831787,
+      "learning_rate": 2.1203096757021256e-05,
+      "loss": 0.7659,
+      "step": 1004000
+    },
+    {
+      "epoch": 1.7286746598150688,
+      "grad_norm": 2.082913637161255,
+      "learning_rate": 2.1188755669748856e-05,
+      "loss": 0.7687,
+      "step": 1004500
+    },
+    {
+      "epoch": 1.7295351250514128,
+      "grad_norm": 2.2899117469787598,
+      "learning_rate": 2.1174414582476453e-05,
+      "loss": 0.7672,
+      "step": 1005000
+    },
+    {
+      "epoch": 1.7303955902877568,
+      "grad_norm": 2.0957179069519043,
+      "learning_rate": 2.1160073495204054e-05,
+      "loss": 0.7608,
+      "step": 1005500
+    },
+    {
+      "epoch": 1.7312560555241008,
+      "grad_norm": 2.2185802459716797,
+      "learning_rate": 2.1145732407931654e-05,
+      "loss": 0.7669,
+      "step": 1006000
+    },
+    {
+      "epoch": 1.7321165207604448,
+      "grad_norm": 2.14453125,
+      "learning_rate": 2.1131391320659255e-05,
+      "loss": 0.7616,
+      "step": 1006500
+    },
+    {
+      "epoch": 1.7329769859967887,
+      "grad_norm": 2.0426855087280273,
+      "learning_rate": 2.1117050233386855e-05,
+      "loss": 0.7609,
+      "step": 1007000
+    },
+    {
+      "epoch": 1.7338374512331327,
+      "grad_norm": 2.1771767139434814,
+      "learning_rate": 2.1102709146114456e-05,
+      "loss": 0.7683,
+      "step": 1007500
+    },
+    {
+      "epoch": 1.7346979164694767,
+      "grad_norm": 2.215662956237793,
+      "learning_rate": 2.1088368058842056e-05,
+      "loss": 0.7644,
+      "step": 1008000
+    },
+    {
+      "epoch": 1.7355583817058207,
+      "grad_norm": 2.010657548904419,
+      "learning_rate": 2.1074026971569653e-05,
+      "loss": 0.762,
+      "step": 1008500
+    },
+    {
+      "epoch": 1.7364188469421646,
+      "grad_norm": 2.2185544967651367,
+      "learning_rate": 2.1059685884297257e-05,
+      "loss": 0.7663,
+      "step": 1009000
+    },
+    {
+      "epoch": 1.7372793121785088,
+      "grad_norm": 1.9308372735977173,
+      "learning_rate": 2.1045344797024858e-05,
+      "loss": 0.7642,
+      "step": 1009500
+    },
+    {
+      "epoch": 1.7381397774148528,
+      "grad_norm": 2.116773843765259,
+      "learning_rate": 2.1031003709752455e-05,
+      "loss": 0.7688,
+      "step": 1010000
+    },
+    {
+      "epoch": 1.7390002426511968,
+      "grad_norm": 2.2500100135803223,
+      "learning_rate": 2.101666262248006e-05,
+      "loss": 0.7627,
+      "step": 1010500
+    },
+    {
+      "epoch": 1.7398607078875408,
+      "grad_norm": 2.222611665725708,
+      "learning_rate": 2.1002321535207656e-05,
+      "loss": 0.7697,
+      "step": 1011000
+    },
+    {
+      "epoch": 1.7407211731238847,
+      "grad_norm": 2.250838279724121,
+      "learning_rate": 2.0987980447935256e-05,
+      "loss": 0.7672,
+      "step": 1011500
+    },
+    {
+      "epoch": 1.7415816383602287,
+      "grad_norm": 2.3026421070098877,
+      "learning_rate": 2.097363936066286e-05,
+      "loss": 0.7621,
+      "step": 1012000
+    },
+    {
+      "epoch": 1.7424421035965727,
+      "grad_norm": 2.1122705936431885,
+      "learning_rate": 2.0959298273390457e-05,
+      "loss": 0.7667,
+      "step": 1012500
+    },
+    {
+      "epoch": 1.7433025688329167,
+      "grad_norm": 2.2275261878967285,
+      "learning_rate": 2.0944957186118057e-05,
+      "loss": 0.767,
+      "step": 1013000
+    },
+    {
+      "epoch": 1.7441630340692607,
+      "grad_norm": 2.136019468307495,
+      "learning_rate": 2.0930616098845658e-05,
+      "loss": 0.7684,
+      "step": 1013500
+    },
+    {
+      "epoch": 1.7450234993056046,
+      "grad_norm": 2.2779674530029297,
+      "learning_rate": 2.091627501157326e-05,
+      "loss": 0.7647,
+      "step": 1014000
+    },
+    {
+      "epoch": 1.7458839645419486,
+      "grad_norm": 2.158823013305664,
+      "learning_rate": 2.090193392430086e-05,
+      "loss": 0.7689,
+      "step": 1014500
+    },
+    {
+      "epoch": 1.7467444297782926,
+      "grad_norm": 2.0115678310394287,
+      "learning_rate": 2.088759283702846e-05,
+      "loss": 0.7649,
+      "step": 1015000
+    },
+    {
+      "epoch": 1.7476048950146366,
+      "grad_norm": 2.3010997772216797,
+      "learning_rate": 2.087325174975606e-05,
+      "loss": 0.7648,
+      "step": 1015500
+    },
+    {
+      "epoch": 1.7484653602509805,
+      "grad_norm": 2.1609082221984863,
+      "learning_rate": 2.085891066248366e-05,
+      "loss": 0.7704,
+      "step": 1016000
+    },
+    {
+      "epoch": 1.7493258254873245,
+      "grad_norm": 2.122830390930176,
+      "learning_rate": 2.084456957521126e-05,
+      "loss": 0.7663,
+      "step": 1016500
+    },
+    {
+      "epoch": 1.7501862907236685,
+      "grad_norm": 2.0455617904663086,
+      "learning_rate": 2.083022848793886e-05,
+      "loss": 0.7591,
+      "step": 1017000
+    },
+    {
+      "epoch": 1.7510467559600125,
+      "grad_norm": 2.0559451580047607,
+      "learning_rate": 2.0815887400666458e-05,
+      "loss": 0.7646,
+      "step": 1017500
+    },
+    {
+      "epoch": 1.7519072211963564,
+      "grad_norm": 2.1012825965881348,
+      "learning_rate": 2.080154631339406e-05,
+      "loss": 0.7634,
+      "step": 1018000
+    },
+    {
+      "epoch": 1.7527676864327004,
+      "grad_norm": 2.210088014602661,
+      "learning_rate": 2.0787205226121663e-05,
+      "loss": 0.7707,
+      "step": 1018500
+    },
+    {
+      "epoch": 1.7536281516690444,
+      "grad_norm": 2.0194787979125977,
+      "learning_rate": 2.077286413884926e-05,
+      "loss": 0.7653,
+      "step": 1019000
+    },
+    {
+      "epoch": 1.7544886169053884,
+      "grad_norm": 2.322936773300171,
+      "learning_rate": 2.075852305157686e-05,
+      "loss": 0.7741,
+      "step": 1019500
+    },
+    {
+      "epoch": 1.7553490821417324,
+      "grad_norm": 2.146965980529785,
+      "learning_rate": 2.074418196430446e-05,
+      "loss": 0.7677,
+      "step": 1020000
+    },
+    {
+      "epoch": 1.7562095473780763,
+      "grad_norm": 1.9335236549377441,
+      "learning_rate": 2.072984087703206e-05,
+      "loss": 0.7694,
+      "step": 1020500
+    },
+    {
+      "epoch": 1.7570700126144203,
+      "grad_norm": 1.9779301881790161,
+      "learning_rate": 2.071549978975966e-05,
+      "loss": 0.7611,
+      "step": 1021000
+    },
+    {
+      "epoch": 1.7579304778507643,
+      "grad_norm": 2.1787054538726807,
+      "learning_rate": 2.0701158702487262e-05,
+      "loss": 0.7725,
+      "step": 1021500
+    },
+    {
+      "epoch": 1.7587909430871083,
+      "grad_norm": 2.254455089569092,
+      "learning_rate": 2.0686817615214863e-05,
+      "loss": 0.7672,
+      "step": 1022000
+    },
+    {
+      "epoch": 1.7596514083234522,
+      "grad_norm": 2.1716644763946533,
+      "learning_rate": 2.067247652794246e-05,
+      "loss": 0.7624,
+      "step": 1022500
+    },
+    {
+      "epoch": 1.7605118735597962,
+      "grad_norm": 2.0068957805633545,
+      "learning_rate": 2.0658135440670063e-05,
+      "loss": 0.7635,
+      "step": 1023000
+    },
+    {
+      "epoch": 1.7613723387961402,
+      "grad_norm": 2.0849902629852295,
+      "learning_rate": 2.0643794353397664e-05,
+      "loss": 0.766,
+      "step": 1023500
+    },
+    {
+      "epoch": 1.7622328040324842,
+      "grad_norm": 2.2821147441864014,
+      "learning_rate": 2.062945326612526e-05,
+      "loss": 0.7705,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.7630932692688281,
+      "grad_norm": 2.0670814514160156,
+      "learning_rate": 2.0615112178852865e-05,
+      "loss": 0.7722,
+      "step": 1024500
+    },
+    {
+      "epoch": 1.7639537345051721,
+      "grad_norm": 2.3591463565826416,
+      "learning_rate": 2.0600771091580465e-05,
+      "loss": 0.7671,
+      "step": 1025000
+    },
+    {
+      "epoch": 1.764814199741516,
+      "grad_norm": 1.9846746921539307,
+      "learning_rate": 2.0586430004308062e-05,
+      "loss": 0.7705,
+      "step": 1025500
+    },
+    {
+      "epoch": 1.76567466497786,
+      "grad_norm": 1.9745829105377197,
+      "learning_rate": 2.0572088917035663e-05,
+      "loss": 0.7669,
+      "step": 1026000
+    },
+    {
+      "epoch": 1.766535130214204,
+      "grad_norm": 1.9543824195861816,
+      "learning_rate": 2.0557747829763263e-05,
+      "loss": 0.7687,
+      "step": 1026500
+    },
+    {
+      "epoch": 1.7673955954505483,
+      "grad_norm": 2.1610777378082275,
+      "learning_rate": 2.0543406742490864e-05,
+      "loss": 0.7659,
+      "step": 1027000
+    },
+    {
+      "epoch": 1.7682560606868922,
+      "grad_norm": 2.0030741691589355,
+      "learning_rate": 2.0529065655218464e-05,
+      "loss": 0.7676,
+      "step": 1027500
+    },
+    {
+      "epoch": 1.7691165259232362,
+      "grad_norm": 2.3656604290008545,
+      "learning_rate": 2.0514724567946065e-05,
+      "loss": 0.7666,
+      "step": 1028000
+    },
+    {
+      "epoch": 1.7699769911595802,
+      "grad_norm": 2.1445231437683105,
+      "learning_rate": 2.0500383480673665e-05,
+      "loss": 0.7697,
+      "step": 1028500
+    },
+    {
+      "epoch": 1.7708374563959242,
+      "grad_norm": 1.9734889268875122,
+      "learning_rate": 2.0486042393401266e-05,
+      "loss": 0.7618,
+      "step": 1029000
+    },
+    {
+      "epoch": 1.7716979216322681,
+      "grad_norm": 2.2043867111206055,
+      "learning_rate": 2.0471701306128866e-05,
+      "loss": 0.765,
+      "step": 1029500
+    },
+    {
+      "epoch": 1.7725583868686121,
+      "grad_norm": 2.0947585105895996,
+      "learning_rate": 2.0457360218856467e-05,
+      "loss": 0.7618,
+      "step": 1030000
+    },
+    {
+      "epoch": 1.773418852104956,
+      "grad_norm": 2.01607346534729,
+      "learning_rate": 2.0443019131584064e-05,
+      "loss": 0.7632,
+      "step": 1030500
+    },
+    {
+      "epoch": 1.7742793173413,
+      "grad_norm": 1.974541425704956,
+      "learning_rate": 2.0428678044311668e-05,
+      "loss": 0.7632,
+      "step": 1031000
+    },
+    {
+      "epoch": 1.775139782577644,
+      "grad_norm": 1.9963077306747437,
+      "learning_rate": 2.0414336957039265e-05,
+      "loss": 0.7654,
+      "step": 1031500
+    },
+    {
+      "epoch": 1.776000247813988,
+      "grad_norm": 2.0639266967773438,
+      "learning_rate": 2.0399995869766865e-05,
+      "loss": 0.7696,
+      "step": 1032000
+    },
+    {
+      "epoch": 1.7768607130503322,
+      "grad_norm": 2.184873580932617,
+      "learning_rate": 2.038565478249447e-05,
+      "loss": 0.7646,
+      "step": 1032500
+    },
+    {
+      "epoch": 1.7777211782866762,
+      "grad_norm": 2.1645946502685547,
+      "learning_rate": 2.0371313695222066e-05,
+      "loss": 0.7612,
+      "step": 1033000
+    },
+    {
+      "epoch": 1.7785816435230202,
+      "grad_norm": 2.45359206199646,
+      "learning_rate": 2.0356972607949667e-05,
+      "loss": 0.763,
+      "step": 1033500
+    },
+    {
+      "epoch": 1.7794421087593641,
+      "grad_norm": 2.0774848461151123,
+      "learning_rate": 2.0342631520677267e-05,
+      "loss": 0.7629,
+      "step": 1034000
+    },
+    {
+      "epoch": 1.7803025739957081,
+      "grad_norm": 2.261568546295166,
+      "learning_rate": 2.0328290433404868e-05,
+      "loss": 0.759,
+      "step": 1034500
+    },
+    {
+      "epoch": 1.781163039232052,
+      "grad_norm": 2.4625566005706787,
+      "learning_rate": 2.0313949346132468e-05,
+      "loss": 0.7669,
+      "step": 1035000
+    },
+    {
+      "epoch": 1.782023504468396,
+      "grad_norm": 2.1339073181152344,
+      "learning_rate": 2.029960825886007e-05,
+      "loss": 0.7582,
+      "step": 1035500
+    },
+    {
+      "epoch": 1.78288396970474,
+      "grad_norm": 2.3074862957000732,
+      "learning_rate": 2.028526717158767e-05,
+      "loss": 0.7651,
+      "step": 1036000
+    },
+    {
+      "epoch": 1.783744434941084,
+      "grad_norm": 2.2727060317993164,
+      "learning_rate": 2.027092608431527e-05,
+      "loss": 0.7684,
+      "step": 1036500
+    },
+    {
+      "epoch": 1.784604900177428,
+      "grad_norm": 2.086061716079712,
+      "learning_rate": 2.025658499704287e-05,
+      "loss": 0.7695,
+      "step": 1037000
+    },
+    {
+      "epoch": 1.785465365413772,
+      "grad_norm": 2.036452293395996,
+      "learning_rate": 2.024224390977047e-05,
+      "loss": 0.7679,
+      "step": 1037500
+    },
+    {
+      "epoch": 1.786325830650116,
+      "grad_norm": 2.2285068035125732,
+      "learning_rate": 2.0227902822498067e-05,
+      "loss": 0.7566,
+      "step": 1038000
+    },
+    {
+      "epoch": 1.78718629588646,
+      "grad_norm": 2.1766035556793213,
+      "learning_rate": 2.0213561735225668e-05,
+      "loss": 0.7636,
+      "step": 1038500
+    },
+    {
+      "epoch": 1.788046761122804,
+      "grad_norm": 2.3141143321990967,
+      "learning_rate": 2.0199220647953272e-05,
+      "loss": 0.762,
+      "step": 1039000
+    },
+    {
+      "epoch": 1.788907226359148,
+      "grad_norm": 2.2814242839813232,
+      "learning_rate": 2.018487956068087e-05,
+      "loss": 0.7609,
+      "step": 1039500
+    },
+    {
+      "epoch": 1.7897676915954919,
+      "grad_norm": 2.2914586067199707,
+      "learning_rate": 2.017053847340847e-05,
+      "loss": 0.7658,
+      "step": 1040000
+    },
+    {
+      "epoch": 1.7906281568318358,
+      "grad_norm": 2.054856300354004,
+      "learning_rate": 2.015619738613607e-05,
+      "loss": 0.765,
+      "step": 1040500
+    },
+    {
+      "epoch": 1.7914886220681798,
+      "grad_norm": 2.02543568611145,
+      "learning_rate": 2.014185629886367e-05,
+      "loss": 0.7657,
+      "step": 1041000
+    },
+    {
+      "epoch": 1.7923490873045238,
+      "grad_norm": 2.087737560272217,
+      "learning_rate": 2.012751521159127e-05,
+      "loss": 0.763,
+      "step": 1041500
+    },
+    {
+      "epoch": 1.7932095525408678,
+      "grad_norm": 2.2256555557250977,
+      "learning_rate": 2.011317412431887e-05,
+      "loss": 0.761,
+      "step": 1042000
+    },
+    {
+      "epoch": 1.7940700177772118,
+      "grad_norm": 2.0341453552246094,
+      "learning_rate": 2.009883303704647e-05,
+      "loss": 0.7623,
+      "step": 1042500
+    },
+    {
+      "epoch": 1.7949304830135557,
+      "grad_norm": 2.1734702587127686,
+      "learning_rate": 2.008449194977407e-05,
+      "loss": 0.7615,
+      "step": 1043000
+    },
+    {
+      "epoch": 1.7957909482498997,
+      "grad_norm": 1.8783676624298096,
+      "learning_rate": 2.0070150862501673e-05,
+      "loss": 0.7649,
+      "step": 1043500
+    },
+    {
+      "epoch": 1.7966514134862437,
+      "grad_norm": 2.2101192474365234,
+      "learning_rate": 2.0055809775229273e-05,
+      "loss": 0.7649,
+      "step": 1044000
+    },
+    {
+      "epoch": 1.7975118787225877,
+      "grad_norm": 2.016972780227661,
+      "learning_rate": 2.004146868795687e-05,
+      "loss": 0.7634,
+      "step": 1044500
+    },
+    {
+      "epoch": 1.7983723439589316,
+      "grad_norm": 33.173458099365234,
+      "learning_rate": 2.0027127600684474e-05,
+      "loss": 0.7701,
+      "step": 1045000
+    },
+    {
+      "epoch": 1.7992328091952756,
+      "grad_norm": 2.1872565746307373,
+      "learning_rate": 2.001278651341207e-05,
+      "loss": 0.7631,
+      "step": 1045500
+    },
+    {
+      "epoch": 1.8000932744316196,
+      "grad_norm": 2.1605005264282227,
+      "learning_rate": 1.999844542613967e-05,
+      "loss": 0.7659,
+      "step": 1046000
+    },
+    {
+      "epoch": 1.8009537396679636,
+      "grad_norm": 2.3672409057617188,
+      "learning_rate": 1.9984104338867275e-05,
+      "loss": 0.7599,
+      "step": 1046500
+    },
+    {
+      "epoch": 1.8018142049043075,
+      "grad_norm": 2.111316680908203,
+      "learning_rate": 1.9969763251594873e-05,
+      "loss": 0.7609,
+      "step": 1047000
+    },
+    {
+      "epoch": 1.8026746701406515,
+      "grad_norm": 2.227252721786499,
+      "learning_rate": 1.9955422164322473e-05,
+      "loss": 0.7632,
+      "step": 1047500
+    },
+    {
+      "epoch": 1.8035351353769955,
+      "grad_norm": 2.072519302368164,
+      "learning_rate": 1.9941081077050073e-05,
+      "loss": 0.7644,
+      "step": 1048000
+    },
+    {
+      "epoch": 1.8043956006133395,
+      "grad_norm": 2.0403823852539062,
+      "learning_rate": 1.9926739989777674e-05,
+      "loss": 0.761,
+      "step": 1048500
+    },
+    {
+      "epoch": 1.8052560658496835,
+      "grad_norm": 6.927855968475342,
+      "learning_rate": 1.9912398902505274e-05,
+      "loss": 0.7598,
+      "step": 1049000
+    },
+    {
+      "epoch": 1.8061165310860274,
+      "grad_norm": 2.0393149852752686,
+      "learning_rate": 1.9898057815232875e-05,
+      "loss": 0.7702,
+      "step": 1049500
+    },
+    {
+      "epoch": 1.8069769963223716,
+      "grad_norm": 2.066805124282837,
+      "learning_rate": 1.9883716727960475e-05,
+      "loss": 0.7655,
+      "step": 1050000
+    },
+    {
+      "epoch": 1.8078374615587156,
+      "grad_norm": 2.2270755767822266,
+      "learning_rate": 1.9869375640688076e-05,
+      "loss": 0.7641,
+      "step": 1050500
+    },
+    {
+      "epoch": 1.8086979267950596,
+      "grad_norm": 2.265721082687378,
+      "learning_rate": 1.9855034553415673e-05,
+      "loss": 0.7644,
+      "step": 1051000
+    },
+    {
+      "epoch": 1.8095583920314036,
+      "grad_norm": 2.069361448287964,
+      "learning_rate": 1.9840693466143277e-05,
+      "loss": 0.759,
+      "step": 1051500
+    },
+    {
+      "epoch": 1.8104188572677475,
+      "grad_norm": 2.284576177597046,
+      "learning_rate": 1.9826352378870874e-05,
+      "loss": 0.7663,
+      "step": 1052000
+    },
+    {
+      "epoch": 1.8112793225040915,
+      "grad_norm": 2.2018840312957764,
+      "learning_rate": 1.9812011291598474e-05,
+      "loss": 0.7661,
+      "step": 1052500
+    },
+    {
+      "epoch": 1.8121397877404355,
+      "grad_norm": 2.477423667907715,
+      "learning_rate": 1.9797670204326078e-05,
+      "loss": 0.7638,
+      "step": 1053000
+    },
+    {
+      "epoch": 1.8130002529767795,
+      "grad_norm": 7.185637474060059,
+      "learning_rate": 1.9783329117053675e-05,
+      "loss": 0.7734,
+      "step": 1053500
+    },
+    {
+      "epoch": 1.8138607182131234,
+      "grad_norm": 2.8006057739257812,
+      "learning_rate": 1.9768988029781276e-05,
+      "loss": 0.7602,
+      "step": 1054000
+    },
+    {
+      "epoch": 1.8147211834494674,
+      "grad_norm": 2.2117958068847656,
+      "learning_rate": 1.9754646942508876e-05,
+      "loss": 0.7638,
+      "step": 1054500
+    },
+    {
+      "epoch": 1.8155816486858114,
+      "grad_norm": 2.083202362060547,
+      "learning_rate": 1.9740305855236477e-05,
+      "loss": 0.7673,
+      "step": 1055000
+    },
+    {
+      "epoch": 1.8164421139221554,
+      "grad_norm": 2.132225751876831,
+      "learning_rate": 1.9725964767964077e-05,
+      "loss": 0.7708,
+      "step": 1055500
+    },
+    {
+      "epoch": 1.8173025791584996,
+      "grad_norm": 2.12669038772583,
+      "learning_rate": 1.9711623680691678e-05,
+      "loss": 0.7627,
+      "step": 1056000
+    },
+    {
+      "epoch": 1.8181630443948436,
+      "grad_norm": 1.9951871633529663,
+      "learning_rate": 1.9697282593419278e-05,
+      "loss": 0.7612,
+      "step": 1056500
+    },
+    {
+      "epoch": 1.8190235096311875,
+      "grad_norm": 2.106513738632202,
+      "learning_rate": 1.968294150614688e-05,
+      "loss": 0.7627,
+      "step": 1057000
+    },
+    {
+      "epoch": 1.8198839748675315,
+      "grad_norm": 2.035362482070923,
+      "learning_rate": 1.966860041887448e-05,
+      "loss": 0.765,
+      "step": 1057500
+    },
+    {
+      "epoch": 1.8207444401038755,
+      "grad_norm": 2.076943874359131,
+      "learning_rate": 1.965425933160208e-05,
+      "loss": 0.7668,
+      "step": 1058000
+    },
+    {
+      "epoch": 1.8216049053402195,
+      "grad_norm": 2.2097012996673584,
+      "learning_rate": 1.9639918244329677e-05,
+      "loss": 0.7589,
+      "step": 1058500
+    },
+    {
+      "epoch": 1.8224653705765634,
+      "grad_norm": 2.264671802520752,
+      "learning_rate": 1.962557715705728e-05,
+      "loss": 0.7628,
+      "step": 1059000
+    },
+    {
+      "epoch": 1.8233258358129074,
+      "grad_norm": 2.2609753608703613,
+      "learning_rate": 1.961123606978488e-05,
+      "loss": 0.7682,
+      "step": 1059500
+    },
+    {
+      "epoch": 1.8241863010492514,
+      "grad_norm": 2.274721384048462,
+      "learning_rate": 1.9596894982512478e-05,
+      "loss": 0.758,
+      "step": 1060000
+    },
+    {
+      "epoch": 1.8250467662855954,
+      "grad_norm": 2.301239252090454,
+      "learning_rate": 1.958255389524008e-05,
+      "loss": 0.7665,
+      "step": 1060500
+    },
+    {
+      "epoch": 1.8259072315219393,
+      "grad_norm": 2.0700104236602783,
+      "learning_rate": 1.956821280796768e-05,
+      "loss": 0.757,
+      "step": 1061000
+    },
+    {
+      "epoch": 1.8267676967582833,
+      "grad_norm": 2.252894639968872,
+      "learning_rate": 1.955387172069528e-05,
+      "loss": 0.7589,
+      "step": 1061500
+    },
+    {
+      "epoch": 1.8276281619946273,
+      "grad_norm": 2.1328372955322266,
+      "learning_rate": 1.953953063342288e-05,
+      "loss": 0.7634,
+      "step": 1062000
+    },
+    {
+      "epoch": 1.8284886272309713,
+      "grad_norm": 2.389291763305664,
+      "learning_rate": 1.952518954615048e-05,
+      "loss": 0.7658,
+      "step": 1062500
+    },
+    {
+      "epoch": 1.8293490924673153,
+      "grad_norm": 2.197857141494751,
+      "learning_rate": 1.951084845887808e-05,
+      "loss": 0.764,
+      "step": 1063000
+    },
+    {
+      "epoch": 1.8302095577036592,
+      "grad_norm": 1.887812614440918,
+      "learning_rate": 1.9496507371605678e-05,
+      "loss": 0.7606,
+      "step": 1063500
+    },
+    {
+      "epoch": 1.8310700229400032,
+      "grad_norm": 2.0983939170837402,
+      "learning_rate": 1.9482166284333282e-05,
+      "loss": 0.7591,
+      "step": 1064000
+    },
+    {
+      "epoch": 1.8319304881763472,
+      "grad_norm": 2.231890916824341,
+      "learning_rate": 1.9467825197060882e-05,
+      "loss": 0.7644,
+      "step": 1064500
+    },
+    {
+      "epoch": 1.8327909534126912,
+      "grad_norm": 2.0913729667663574,
+      "learning_rate": 1.945348410978848e-05,
+      "loss": 0.7675,
+      "step": 1065000
+    },
+    {
+      "epoch": 1.8336514186490351,
+      "grad_norm": 2.2117438316345215,
+      "learning_rate": 1.9439143022516083e-05,
+      "loss": 0.7549,
+      "step": 1065500
+    },
+    {
+      "epoch": 1.8345118838853791,
+      "grad_norm": 2.160916805267334,
+      "learning_rate": 1.942480193524368e-05,
+      "loss": 0.7601,
+      "step": 1066000
+    },
+    {
+      "epoch": 1.835372349121723,
+      "grad_norm": 2.1662871837615967,
+      "learning_rate": 1.941046084797128e-05,
+      "loss": 0.7615,
+      "step": 1066500
+    },
+    {
+      "epoch": 1.836232814358067,
+      "grad_norm": 2.103234052658081,
+      "learning_rate": 1.9396119760698885e-05,
+      "loss": 0.7571,
+      "step": 1067000
+    },
+    {
+      "epoch": 1.837093279594411,
+      "grad_norm": 2.0720338821411133,
+      "learning_rate": 1.938177867342648e-05,
+      "loss": 0.7576,
+      "step": 1067500
+    },
+    {
+      "epoch": 1.837953744830755,
+      "grad_norm": 2.079674005508423,
+      "learning_rate": 1.9367437586154082e-05,
+      "loss": 0.7621,
+      "step": 1068000
+    },
+    {
+      "epoch": 1.838814210067099,
+      "grad_norm": 2.1196000576019287,
+      "learning_rate": 1.9353096498881686e-05,
+      "loss": 0.7627,
+      "step": 1068500
+    },
+    {
+      "epoch": 1.839674675303443,
+      "grad_norm": 3.3423213958740234,
+      "learning_rate": 1.9338755411609283e-05,
+      "loss": 0.7632,
+      "step": 1069000
+    },
+    {
+      "epoch": 1.840535140539787,
+      "grad_norm": 2.1786093711853027,
+      "learning_rate": 1.9324414324336884e-05,
+      "loss": 0.7593,
+      "step": 1069500
+    },
+    {
+      "epoch": 1.841395605776131,
+      "grad_norm": 1.9805326461791992,
+      "learning_rate": 1.9310073237064484e-05,
+      "loss": 0.7664,
+      "step": 1070000
+    },
+    {
+      "epoch": 1.842256071012475,
+      "grad_norm": 2.258626699447632,
+      "learning_rate": 1.9295732149792084e-05,
+      "loss": 0.7596,
+      "step": 1070500
+    },
+    {
+      "epoch": 1.8431165362488189,
+      "grad_norm": 1.964232087135315,
+      "learning_rate": 1.9281391062519685e-05,
+      "loss": 0.763,
+      "step": 1071000
+    },
+    {
+      "epoch": 1.8439770014851629,
+      "grad_norm": 2.0664947032928467,
+      "learning_rate": 1.9267049975247285e-05,
+      "loss": 0.7645,
+      "step": 1071500
+    },
+    {
+      "epoch": 1.8448374667215068,
+      "grad_norm": 2.0968148708343506,
+      "learning_rate": 1.9252708887974886e-05,
+      "loss": 0.7673,
+      "step": 1072000
+    },
+    {
+      "epoch": 1.8456979319578508,
+      "grad_norm": 2.034278154373169,
+      "learning_rate": 1.9238367800702483e-05,
+      "loss": 0.7614,
+      "step": 1072500
+    },
+    {
+      "epoch": 1.846558397194195,
+      "grad_norm": 2.2998197078704834,
+      "learning_rate": 1.9224026713430083e-05,
+      "loss": 0.7549,
+      "step": 1073000
+    },
+    {
+      "epoch": 1.847418862430539,
+      "grad_norm": 2.016671657562256,
+      "learning_rate": 1.9209685626157687e-05,
+      "loss": 0.7577,
+      "step": 1073500
+    },
+    {
+      "epoch": 1.848279327666883,
+      "grad_norm": 1.9989349842071533,
+      "learning_rate": 1.9195344538885284e-05,
+      "loss": 0.7543,
+      "step": 1074000
+    },
+    {
+      "epoch": 1.849139792903227,
+      "grad_norm": 2.155221462249756,
+      "learning_rate": 1.9181003451612885e-05,
+      "loss": 0.7628,
+      "step": 1074500
+    },
+    {
+      "epoch": 1.850000258139571,
+      "grad_norm": 2.4140572547912598,
+      "learning_rate": 1.9166662364340485e-05,
+      "loss": 0.7603,
+      "step": 1075000
+    },
+    {
+      "epoch": 1.850860723375915,
+      "grad_norm": 1.9228155612945557,
+      "learning_rate": 1.9152321277068086e-05,
+      "loss": 0.7645,
+      "step": 1075500
+    },
+    {
+      "epoch": 1.8517211886122589,
+      "grad_norm": 2.0937721729278564,
+      "learning_rate": 1.9137980189795686e-05,
+      "loss": 0.7633,
+      "step": 1076000
+    },
+    {
+      "epoch": 1.8525816538486028,
+      "grad_norm": 2.3551013469696045,
+      "learning_rate": 1.9123639102523287e-05,
+      "loss": 0.7578,
+      "step": 1076500
+    },
+    {
+      "epoch": 1.8534421190849468,
+      "grad_norm": 2.13285231590271,
+      "learning_rate": 1.9109298015250887e-05,
+      "loss": 0.7586,
+      "step": 1077000
+    },
+    {
+      "epoch": 1.8543025843212908,
+      "grad_norm": 2.2685706615448,
+      "learning_rate": 1.9094956927978484e-05,
+      "loss": 0.7595,
+      "step": 1077500
+    },
+    {
+      "epoch": 1.8551630495576348,
+      "grad_norm": 2.0785560607910156,
+      "learning_rate": 1.9080615840706088e-05,
+      "loss": 0.7588,
+      "step": 1078000
+    },
+    {
+      "epoch": 1.8560235147939788,
+      "grad_norm": 2.3855340480804443,
+      "learning_rate": 1.906627475343369e-05,
+      "loss": 0.7569,
+      "step": 1078500
+    },
+    {
+      "epoch": 1.856883980030323,
+      "grad_norm": 2.313601016998291,
+      "learning_rate": 1.9051933666161286e-05,
+      "loss": 0.7609,
+      "step": 1079000
+    },
+    {
+      "epoch": 1.857744445266667,
+      "grad_norm": 2.230159044265747,
+      "learning_rate": 1.903759257888889e-05,
+      "loss": 0.7609,
+      "step": 1079500
+    },
+    {
+      "epoch": 1.858604910503011,
+      "grad_norm": 2.331984519958496,
+      "learning_rate": 1.902325149161649e-05,
+      "loss": 0.7627,
+      "step": 1080000
+    },
+    {
+      "epoch": 1.8594653757393549,
+      "grad_norm": 1.8024656772613525,
+      "learning_rate": 1.9008910404344087e-05,
+      "loss": 0.7533,
+      "step": 1080500
+    },
+    {
+      "epoch": 1.8603258409756989,
+      "grad_norm": 2.5354342460632324,
+      "learning_rate": 1.899456931707169e-05,
+      "loss": 0.7596,
+      "step": 1081000
+    },
+    {
+      "epoch": 1.8611863062120428,
+      "grad_norm": 2.176863193511963,
+      "learning_rate": 1.8980228229799288e-05,
+      "loss": 0.757,
+      "step": 1081500
+    },
+    {
+      "epoch": 1.8620467714483868,
+      "grad_norm": 2.0789923667907715,
+      "learning_rate": 1.896588714252689e-05,
+      "loss": 0.762,
+      "step": 1082000
+    },
+    {
+      "epoch": 1.8629072366847308,
+      "grad_norm": 2.1080963611602783,
+      "learning_rate": 1.895154605525449e-05,
+      "loss": 0.7577,
+      "step": 1082500
+    },
+    {
+      "epoch": 1.8637677019210748,
+      "grad_norm": 1.9690420627593994,
+      "learning_rate": 1.893720496798209e-05,
+      "loss": 0.7602,
+      "step": 1083000
+    },
+    {
+      "epoch": 1.8646281671574187,
+      "grad_norm": 2.039321184158325,
+      "learning_rate": 1.892286388070969e-05,
+      "loss": 0.7543,
+      "step": 1083500
+    },
+    {
+      "epoch": 1.8654886323937627,
+      "grad_norm": 2.1178793907165527,
+      "learning_rate": 1.890852279343729e-05,
+      "loss": 0.7565,
+      "step": 1084000
+    },
+    {
+      "epoch": 1.8663490976301067,
+      "grad_norm": 2.1791107654571533,
+      "learning_rate": 1.889418170616489e-05,
+      "loss": 0.7563,
+      "step": 1084500
+    },
+    {
+      "epoch": 1.8672095628664507,
+      "grad_norm": 2.110280990600586,
+      "learning_rate": 1.887984061889249e-05,
+      "loss": 0.7606,
+      "step": 1085000
+    },
+    {
+      "epoch": 1.8680700281027947,
+      "grad_norm": 2.0059497356414795,
+      "learning_rate": 1.886549953162009e-05,
+      "loss": 0.7556,
+      "step": 1085500
+    },
+    {
+      "epoch": 1.8689304933391386,
+      "grad_norm": 2.052835702896118,
+      "learning_rate": 1.8851158444347692e-05,
+      "loss": 0.763,
+      "step": 1086000
+    },
+    {
+      "epoch": 1.8697909585754826,
+      "grad_norm": 2.0899126529693604,
+      "learning_rate": 1.883681735707529e-05,
+      "loss": 0.7584,
+      "step": 1086500
+    },
+    {
+      "epoch": 1.8706514238118266,
+      "grad_norm": 2.019315481185913,
+      "learning_rate": 1.882247626980289e-05,
+      "loss": 0.7586,
+      "step": 1087000
+    },
+    {
+      "epoch": 1.8715118890481706,
+      "grad_norm": 2.061671733856201,
+      "learning_rate": 1.8808135182530494e-05,
+      "loss": 0.7551,
+      "step": 1087500
+    },
+    {
+      "epoch": 1.8723723542845145,
+      "grad_norm": 2.065790891647339,
+      "learning_rate": 1.879379409525809e-05,
+      "loss": 0.7631,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.8732328195208585,
+      "grad_norm": 2.2259931564331055,
+      "learning_rate": 1.877945300798569e-05,
+      "loss": 0.7591,
+      "step": 1088500
+    },
+    {
+      "epoch": 1.8740932847572025,
+      "grad_norm": 1.9983274936676025,
+      "learning_rate": 1.8765111920713292e-05,
+      "loss": 0.7585,
+      "step": 1089000
+    },
+    {
+      "epoch": 1.8749537499935465,
+      "grad_norm": 2.08388090133667,
+      "learning_rate": 1.8750770833440892e-05,
+      "loss": 0.7607,
+      "step": 1089500
+    },
+    {
+      "epoch": 1.8758142152298904,
+      "grad_norm": 2.1535439491271973,
+      "learning_rate": 1.8736429746168493e-05,
+      "loss": 0.7578,
+      "step": 1090000
+    },
+    {
+      "epoch": 1.8766746804662344,
+      "grad_norm": 2.4029664993286133,
+      "learning_rate": 1.8722088658896093e-05,
+      "loss": 0.7602,
+      "step": 1090500
+    },
+    {
+      "epoch": 1.8775351457025784,
+      "grad_norm": 2.107679605484009,
+      "learning_rate": 1.8707747571623694e-05,
+      "loss": 0.7621,
+      "step": 1091000
+    },
+    {
+      "epoch": 1.8783956109389224,
+      "grad_norm": 2.1485376358032227,
+      "learning_rate": 1.8693406484351294e-05,
+      "loss": 0.7541,
+      "step": 1091500
+    },
+    {
+      "epoch": 1.8792560761752664,
+      "grad_norm": 2.088392972946167,
+      "learning_rate": 1.8679065397078895e-05,
+      "loss": 0.756,
+      "step": 1092000
+    },
+    {
+      "epoch": 1.8801165414116103,
+      "grad_norm": 2.1533939838409424,
+      "learning_rate": 1.8664724309806495e-05,
+      "loss": 0.7581,
+      "step": 1092500
+    },
+    {
+      "epoch": 1.8809770066479543,
+      "grad_norm": 2.3350436687469482,
+      "learning_rate": 1.8650383222534092e-05,
+      "loss": 0.7497,
+      "step": 1093000
+    },
+    {
+      "epoch": 1.8818374718842983,
+      "grad_norm": 2.236022472381592,
+      "learning_rate": 1.8636042135261696e-05,
+      "loss": 0.7617,
+      "step": 1093500
+    },
+    {
+      "epoch": 1.8826979371206423,
+      "grad_norm": 2.3586575984954834,
+      "learning_rate": 1.8621701047989296e-05,
+      "loss": 0.759,
+      "step": 1094000
+    },
+    {
+      "epoch": 1.8835584023569862,
+      "grad_norm": 2.1718757152557373,
+      "learning_rate": 1.8607359960716893e-05,
+      "loss": 0.7598,
+      "step": 1094500
+    },
+    {
+      "epoch": 1.8844188675933302,
+      "grad_norm": 2.206676959991455,
+      "learning_rate": 1.8593018873444494e-05,
+      "loss": 0.7595,
+      "step": 1095000
+    },
+    {
+      "epoch": 1.8852793328296742,
+      "grad_norm": 2.1057584285736084,
+      "learning_rate": 1.8578677786172094e-05,
+      "loss": 0.7613,
+      "step": 1095500
+    },
+    {
+      "epoch": 1.8861397980660182,
+      "grad_norm": 2.199988842010498,
+      "learning_rate": 1.8564336698899695e-05,
+      "loss": 0.7597,
+      "step": 1096000
+    },
+    {
+      "epoch": 1.8870002633023624,
+      "grad_norm": 2.3150246143341064,
+      "learning_rate": 1.8549995611627295e-05,
+      "loss": 0.7588,
+      "step": 1096500
+    },
+    {
+      "epoch": 1.8878607285387063,
+      "grad_norm": 2.125781536102295,
+      "learning_rate": 1.8535654524354896e-05,
+      "loss": 0.7548,
+      "step": 1097000
+    },
+    {
+      "epoch": 1.8887211937750503,
+      "grad_norm": 2.0690038204193115,
+      "learning_rate": 1.8521313437082496e-05,
+      "loss": 0.7563,
+      "step": 1097500
+    },
+    {
+      "epoch": 1.8895816590113943,
+      "grad_norm": 2.1967461109161377,
+      "learning_rate": 1.8506972349810093e-05,
+      "loss": 0.7564,
+      "step": 1098000
+    },
+    {
+      "epoch": 1.8904421242477383,
+      "grad_norm": 2.0948901176452637,
+      "learning_rate": 1.8492631262537697e-05,
+      "loss": 0.7619,
+      "step": 1098500
+    },
+    {
+      "epoch": 1.8913025894840823,
+      "grad_norm": 2.2009682655334473,
+      "learning_rate": 1.8478290175265298e-05,
+      "loss": 0.7519,
+      "step": 1099000
+    },
+    {
+      "epoch": 1.8921630547204262,
+      "grad_norm": 2.0295839309692383,
+      "learning_rate": 1.8463949087992895e-05,
+      "loss": 0.7551,
+      "step": 1099500
+    },
+    {
+      "epoch": 1.8930235199567702,
+      "grad_norm": 2.0854525566101074,
+      "learning_rate": 1.84496080007205e-05,
+      "loss": 0.7591,
+      "step": 1100000
+    },
+    {
+      "epoch": 1.8938839851931142,
+      "grad_norm": 2.133030414581299,
+      "learning_rate": 1.8435266913448096e-05,
+      "loss": 0.7625,
+      "step": 1100500
+    },
+    {
+      "epoch": 1.8947444504294582,
+      "grad_norm": 1.9579893350601196,
+      "learning_rate": 1.8420925826175696e-05,
+      "loss": 0.7635,
+      "step": 1101000
+    },
+    {
+      "epoch": 1.8956049156658021,
+      "grad_norm": 2.1973876953125,
+      "learning_rate": 1.84065847389033e-05,
+      "loss": 0.7488,
+      "step": 1101500
+    },
+    {
+      "epoch": 1.8964653809021463,
+      "grad_norm": 2.209946393966675,
+      "learning_rate": 1.8392243651630897e-05,
+      "loss": 0.755,
+      "step": 1102000
+    },
+    {
+      "epoch": 1.8973258461384903,
+      "grad_norm": 2.0465190410614014,
+      "learning_rate": 1.8377902564358498e-05,
+      "loss": 0.7578,
+      "step": 1102500
+    },
+    {
+      "epoch": 1.8981863113748343,
+      "grad_norm": 2.160445213317871,
+      "learning_rate": 1.8363561477086098e-05,
+      "loss": 0.7566,
+      "step": 1103000
+    },
+    {
+      "epoch": 1.8990467766111783,
+      "grad_norm": 2.2007155418395996,
+      "learning_rate": 1.83492203898137e-05,
+      "loss": 0.7502,
+      "step": 1103500
+    },
+    {
+      "epoch": 1.8999072418475222,
+      "grad_norm": 2.144000768661499,
+      "learning_rate": 1.83348793025413e-05,
+      "loss": 0.7576,
+      "step": 1104000
+    },
+    {
+      "epoch": 1.9007677070838662,
+      "grad_norm": 2.284233808517456,
+      "learning_rate": 1.83205382152689e-05,
+      "loss": 0.7547,
+      "step": 1104500
+    },
+    {
+      "epoch": 1.9016281723202102,
+      "grad_norm": 2.2595906257629395,
+      "learning_rate": 1.83061971279965e-05,
+      "loss": 0.7582,
+      "step": 1105000
+    },
+    {
+      "epoch": 1.9024886375565542,
+      "grad_norm": 2.0282812118530273,
+      "learning_rate": 1.82918560407241e-05,
+      "loss": 0.7558,
+      "step": 1105500
+    },
+    {
+      "epoch": 1.9033491027928982,
+      "grad_norm": 2.142599105834961,
+      "learning_rate": 1.82775149534517e-05,
+      "loss": 0.7527,
+      "step": 1106000
+    },
+    {
+      "epoch": 1.9042095680292421,
+      "grad_norm": 2.237638235092163,
+      "learning_rate": 1.82631738661793e-05,
+      "loss": 0.749,
+      "step": 1106500
+    },
+    {
+      "epoch": 1.905070033265586,
+      "grad_norm": 2.2329697608947754,
+      "learning_rate": 1.82488327789069e-05,
+      "loss": 0.7585,
+      "step": 1107000
+    },
+    {
+      "epoch": 1.90593049850193,
+      "grad_norm": 2.2512173652648926,
+      "learning_rate": 1.82344916916345e-05,
+      "loss": 0.7553,
+      "step": 1107500
+    },
+    {
+      "epoch": 1.906790963738274,
+      "grad_norm": 2.0716335773468018,
+      "learning_rate": 1.8220150604362103e-05,
+      "loss": 0.7541,
+      "step": 1108000
+    },
+    {
+      "epoch": 1.907651428974618,
+      "grad_norm": 2.0387918949127197,
+      "learning_rate": 1.82058095170897e-05,
+      "loss": 0.7582,
+      "step": 1108500
+    },
+    {
+      "epoch": 1.908511894210962,
+      "grad_norm": 1.9657924175262451,
+      "learning_rate": 1.81914684298173e-05,
+      "loss": 0.7667,
+      "step": 1109000
+    },
+    {
+      "epoch": 1.909372359447306,
+      "grad_norm": 2.0124640464782715,
+      "learning_rate": 1.81771273425449e-05,
+      "loss": 0.7581,
+      "step": 1109500
+    },
+    {
+      "epoch": 1.91023282468365,
+      "grad_norm": 2.183788537979126,
+      "learning_rate": 1.81627862552725e-05,
+      "loss": 0.7488,
+      "step": 1110000
+    },
+    {
+      "epoch": 1.911093289919994,
+      "grad_norm": 2.0841352939605713,
+      "learning_rate": 1.8148445168000102e-05,
+      "loss": 0.7567,
+      "step": 1110500
+    },
+    {
+      "epoch": 1.911953755156338,
+      "grad_norm": 2.0061280727386475,
+      "learning_rate": 1.8134104080727702e-05,
+      "loss": 0.7633,
+      "step": 1111000
+    },
+    {
+      "epoch": 1.912814220392682,
+      "grad_norm": 1.9698486328125,
+      "learning_rate": 1.8119762993455303e-05,
+      "loss": 0.7533,
+      "step": 1111500
+    },
+    {
+      "epoch": 1.9136746856290259,
+      "grad_norm": 2.1442909240722656,
+      "learning_rate": 1.81054219061829e-05,
+      "loss": 0.7594,
+      "step": 1112000
+    },
+    {
+      "epoch": 1.9145351508653699,
+      "grad_norm": 2.279815196990967,
+      "learning_rate": 1.8091080818910504e-05,
+      "loss": 0.7584,
+      "step": 1112500
+    },
+    {
+      "epoch": 1.9153956161017138,
+      "grad_norm": 2.1331374645233154,
+      "learning_rate": 1.8076739731638104e-05,
+      "loss": 0.7553,
+      "step": 1113000
+    },
+    {
+      "epoch": 1.9162560813380578,
+      "grad_norm": 2.1729178428649902,
+      "learning_rate": 1.80623986443657e-05,
+      "loss": 0.7575,
+      "step": 1113500
+    },
+    {
+      "epoch": 1.9171165465744018,
+      "grad_norm": 2.0933024883270264,
+      "learning_rate": 1.8048057557093305e-05,
+      "loss": 0.7575,
+      "step": 1114000
+    },
+    {
+      "epoch": 1.9179770118107458,
+      "grad_norm": 2.0787551403045654,
+      "learning_rate": 1.8033716469820906e-05,
+      "loss": 0.7532,
+      "step": 1114500
+    },
+    {
+      "epoch": 1.9188374770470897,
+      "grad_norm": 2.139012336730957,
+      "learning_rate": 1.8019375382548503e-05,
+      "loss": 0.7597,
+      "step": 1115000
+    },
+    {
+      "epoch": 1.9196979422834337,
+      "grad_norm": 2.0884130001068115,
+      "learning_rate": 1.8005034295276103e-05,
+      "loss": 0.754,
+      "step": 1115500
+    },
+    {
+      "epoch": 1.9205584075197777,
+      "grad_norm": 2.2890164852142334,
+      "learning_rate": 1.7990693208003704e-05,
+      "loss": 0.754,
+      "step": 1116000
+    },
+    {
+      "epoch": 1.9214188727561217,
+      "grad_norm": 2.274405002593994,
+      "learning_rate": 1.7976352120731304e-05,
+      "loss": 0.7465,
+      "step": 1116500
+    },
+    {
+      "epoch": 1.9222793379924656,
+      "grad_norm": 2.3451716899871826,
+      "learning_rate": 1.7962011033458904e-05,
+      "loss": 0.7614,
+      "step": 1117000
+    },
+    {
+      "epoch": 1.9231398032288096,
+      "grad_norm": 2.0729122161865234,
+      "learning_rate": 1.7947669946186505e-05,
+      "loss": 0.7546,
+      "step": 1117500
+    },
+    {
+      "epoch": 1.9240002684651536,
+      "grad_norm": 2.164996862411499,
+      "learning_rate": 1.7933328858914105e-05,
+      "loss": 0.7596,
+      "step": 1118000
+    },
+    {
+      "epoch": 1.9248607337014976,
+      "grad_norm": 2.23533296585083,
+      "learning_rate": 1.7918987771641706e-05,
+      "loss": 0.7573,
+      "step": 1118500
+    },
+    {
+      "epoch": 1.9257211989378415,
+      "grad_norm": 2.1161954402923584,
+      "learning_rate": 1.7904646684369306e-05,
+      "loss": 0.7551,
+      "step": 1119000
+    },
+    {
+      "epoch": 1.9265816641741857,
+      "grad_norm": 2.200545072555542,
+      "learning_rate": 1.7890305597096907e-05,
+      "loss": 0.7533,
+      "step": 1119500
+    },
+    {
+      "epoch": 1.9274421294105297,
+      "grad_norm": 1.9847606420516968,
+      "learning_rate": 1.7875964509824504e-05,
+      "loss": 0.7537,
+      "step": 1120000
+    },
+    {
+      "epoch": 1.9283025946468737,
+      "grad_norm": 2.063674211502075,
+      "learning_rate": 1.7861623422552108e-05,
+      "loss": 0.7587,
+      "step": 1120500
+    },
+    {
+      "epoch": 1.9291630598832177,
+      "grad_norm": 2.1384332180023193,
+      "learning_rate": 1.7847282335279705e-05,
+      "loss": 0.7527,
+      "step": 1121000
+    },
+    {
+      "epoch": 1.9300235251195617,
+      "grad_norm": 2.213263750076294,
+      "learning_rate": 1.7832941248007305e-05,
+      "loss": 0.7555,
+      "step": 1121500
+    },
+    {
+      "epoch": 1.9308839903559056,
+      "grad_norm": 2.0558905601501465,
+      "learning_rate": 1.781860016073491e-05,
+      "loss": 0.7546,
+      "step": 1122000
+    },
+    {
+      "epoch": 1.9317444555922496,
+      "grad_norm": 2.223896026611328,
+      "learning_rate": 1.7804259073462506e-05,
+      "loss": 0.7577,
+      "step": 1122500
+    },
+    {
+      "epoch": 1.9326049208285936,
+      "grad_norm": 2.3315932750701904,
+      "learning_rate": 1.7789917986190107e-05,
+      "loss": 0.7549,
+      "step": 1123000
+    },
+    {
+      "epoch": 1.9334653860649376,
+      "grad_norm": 2.0839650630950928,
+      "learning_rate": 1.777557689891771e-05,
+      "loss": 0.7558,
+      "step": 1123500
+    },
+    {
+      "epoch": 1.9343258513012815,
+      "grad_norm": 2.223665237426758,
+      "learning_rate": 1.7761235811645308e-05,
+      "loss": 0.7516,
+      "step": 1124000
+    },
+    {
+      "epoch": 1.9351863165376255,
+      "grad_norm": 2.5846803188323975,
+      "learning_rate": 1.7746894724372908e-05,
+      "loss": 0.7591,
+      "step": 1124500
+    },
+    {
+      "epoch": 1.9360467817739697,
+      "grad_norm": 2.2104952335357666,
+      "learning_rate": 1.773255363710051e-05,
+      "loss": 0.754,
+      "step": 1125000
+    },
+    {
+      "epoch": 1.9369072470103137,
+      "grad_norm": 1.9384921789169312,
+      "learning_rate": 1.771821254982811e-05,
+      "loss": 0.7559,
+      "step": 1125500
+    },
+    {
+      "epoch": 1.9377677122466577,
+      "grad_norm": 1.9238662719726562,
+      "learning_rate": 1.770387146255571e-05,
+      "loss": 0.7514,
+      "step": 1126000
+    },
+    {
+      "epoch": 1.9386281774830016,
+      "grad_norm": 2.1696484088897705,
+      "learning_rate": 1.768953037528331e-05,
+      "loss": 0.7535,
+      "step": 1126500
+    },
+    {
+      "epoch": 1.9394886427193456,
+      "grad_norm": 1.9859446287155151,
+      "learning_rate": 1.767518928801091e-05,
+      "loss": 0.7547,
+      "step": 1127000
+    },
+    {
+      "epoch": 1.9403491079556896,
+      "grad_norm": 2.1499342918395996,
+      "learning_rate": 1.7660848200738508e-05,
+      "loss": 0.753,
+      "step": 1127500
+    },
+    {
+      "epoch": 1.9412095731920336,
+      "grad_norm": 2.1544241905212402,
+      "learning_rate": 1.7646507113466108e-05,
+      "loss": 0.7535,
+      "step": 1128000
+    },
+    {
+      "epoch": 1.9420700384283776,
+      "grad_norm": 2.1659960746765137,
+      "learning_rate": 1.7632166026193712e-05,
+      "loss": 0.7583,
+      "step": 1128500
+    },
+    {
+      "epoch": 1.9429305036647215,
+      "grad_norm": 2.0657598972320557,
+      "learning_rate": 1.761782493892131e-05,
+      "loss": 0.7553,
+      "step": 1129000
+    },
+    {
+      "epoch": 1.9437909689010655,
+      "grad_norm": 2.312696695327759,
+      "learning_rate": 1.760348385164891e-05,
+      "loss": 0.7538,
+      "step": 1129500
+    },
+    {
+      "epoch": 1.9446514341374095,
+      "grad_norm": 2.268152952194214,
+      "learning_rate": 1.758914276437651e-05,
+      "loss": 0.7533,
+      "step": 1130000
+    },
+    {
+      "epoch": 1.9455118993737535,
+      "grad_norm": 2.2214696407318115,
+      "learning_rate": 1.757480167710411e-05,
+      "loss": 0.7567,
+      "step": 1130500
+    },
+    {
+      "epoch": 1.9463723646100974,
+      "grad_norm": 2.3125360012054443,
+      "learning_rate": 1.756046058983171e-05,
+      "loss": 0.7504,
+      "step": 1131000
+    },
+    {
+      "epoch": 1.9472328298464414,
+      "grad_norm": 2.032647132873535,
+      "learning_rate": 1.754611950255931e-05,
+      "loss": 0.7549,
+      "step": 1131500
+    },
+    {
+      "epoch": 1.9480932950827854,
+      "grad_norm": 1.9928709268569946,
+      "learning_rate": 1.7531778415286912e-05,
+      "loss": 0.7537,
+      "step": 1132000
+    },
+    {
+      "epoch": 1.9489537603191294,
+      "grad_norm": 1.8794376850128174,
+      "learning_rate": 1.751743732801451e-05,
+      "loss": 0.7503,
+      "step": 1132500
+    },
+    {
+      "epoch": 1.9498142255554733,
+      "grad_norm": 2.239978790283203,
+      "learning_rate": 1.7503096240742113e-05,
+      "loss": 0.751,
+      "step": 1133000
+    },
+    {
+      "epoch": 1.9506746907918173,
+      "grad_norm": 2.246455430984497,
+      "learning_rate": 1.7488755153469713e-05,
+      "loss": 0.7537,
+      "step": 1133500
+    },
+    {
+      "epoch": 1.9515351560281613,
+      "grad_norm": 2.074331045150757,
+      "learning_rate": 1.747441406619731e-05,
+      "loss": 0.757,
+      "step": 1134000
+    },
+    {
+      "epoch": 1.9523956212645053,
+      "grad_norm": 2.2001051902770996,
+      "learning_rate": 1.7460072978924914e-05,
+      "loss": 0.7537,
+      "step": 1134500
+    },
+    {
+      "epoch": 1.9532560865008493,
+      "grad_norm": 2.2409298419952393,
+      "learning_rate": 1.7445731891652515e-05,
+      "loss": 0.7558,
+      "step": 1135000
+    },
+    {
+      "epoch": 1.9541165517371932,
+      "grad_norm": 1.9500313997268677,
+      "learning_rate": 1.7431390804380112e-05,
+      "loss": 0.7534,
+      "step": 1135500
+    },
+    {
+      "epoch": 1.9549770169735372,
+      "grad_norm": 2.0206470489501953,
+      "learning_rate": 1.7417049717107716e-05,
+      "loss": 0.7507,
+      "step": 1136000
+    },
+    {
+      "epoch": 1.9558374822098812,
+      "grad_norm": 1.9329551458358765,
+      "learning_rate": 1.7402708629835313e-05,
+      "loss": 0.7514,
+      "step": 1136500
+    },
+    {
+      "epoch": 1.9566979474462252,
+      "grad_norm": 2.4962384700775146,
+      "learning_rate": 1.7388367542562913e-05,
+      "loss": 0.7494,
+      "step": 1137000
+    },
+    {
+      "epoch": 1.9575584126825691,
+      "grad_norm": 2.5123205184936523,
+      "learning_rate": 1.7374026455290514e-05,
+      "loss": 0.7532,
+      "step": 1137500
+    },
+    {
+      "epoch": 1.9584188779189131,
+      "grad_norm": 1.9868464469909668,
+      "learning_rate": 1.7359685368018114e-05,
+      "loss": 0.7511,
+      "step": 1138000
+    },
+    {
+      "epoch": 1.959279343155257,
+      "grad_norm": 2.332765579223633,
+      "learning_rate": 1.7345344280745715e-05,
+      "loss": 0.7476,
+      "step": 1138500
+    },
+    {
+      "epoch": 1.960139808391601,
+      "grad_norm": 2.1686367988586426,
+      "learning_rate": 1.7331003193473315e-05,
+      "loss": 0.7435,
+      "step": 1139000
+    },
+    {
+      "epoch": 1.961000273627945,
+      "grad_norm": 2.017496347427368,
+      "learning_rate": 1.7316662106200915e-05,
+      "loss": 0.7547,
+      "step": 1139500
+    },
+    {
+      "epoch": 1.961860738864289,
+      "grad_norm": 2.2544643878936768,
+      "learning_rate": 1.7302321018928516e-05,
+      "loss": 0.7589,
+      "step": 1140000
+    },
+    {
+      "epoch": 1.962721204100633,
+      "grad_norm": 2.2415223121643066,
+      "learning_rate": 1.7287979931656113e-05,
+      "loss": 0.7549,
+      "step": 1140500
+    },
+    {
+      "epoch": 1.963581669336977,
+      "grad_norm": 2.0396015644073486,
+      "learning_rate": 1.7273638844383717e-05,
+      "loss": 0.757,
+      "step": 1141000
+    },
+    {
+      "epoch": 1.964442134573321,
+      "grad_norm": 2.1235740184783936,
+      "learning_rate": 1.7259297757111314e-05,
+      "loss": 0.7537,
+      "step": 1141500
+    },
+    {
+      "epoch": 1.965302599809665,
+      "grad_norm": 2.126119613647461,
+      "learning_rate": 1.7244956669838914e-05,
+      "loss": 0.7643,
+      "step": 1142000
+    },
+    {
+      "epoch": 1.9661630650460091,
+      "grad_norm": 2.5373361110687256,
+      "learning_rate": 1.723061558256652e-05,
+      "loss": 0.7533,
+      "step": 1142500
+    },
+    {
+      "epoch": 1.967023530282353,
+      "grad_norm": 2.3410146236419678,
+      "learning_rate": 1.7216274495294115e-05,
+      "loss": 0.761,
+      "step": 1143000
+    },
+    {
+      "epoch": 1.967883995518697,
+      "grad_norm": 2.128046989440918,
+      "learning_rate": 1.7201933408021716e-05,
+      "loss": 0.7572,
+      "step": 1143500
+    },
+    {
+      "epoch": 1.968744460755041,
+      "grad_norm": 2.1182539463043213,
+      "learning_rate": 1.7187592320749316e-05,
+      "loss": 0.7556,
+      "step": 1144000
+    },
+    {
+      "epoch": 1.969604925991385,
+      "grad_norm": 2.128629446029663,
+      "learning_rate": 1.7173251233476917e-05,
+      "loss": 0.7558,
+      "step": 1144500
+    },
+    {
+      "epoch": 1.970465391227729,
+      "grad_norm": 2.249554395675659,
+      "learning_rate": 1.7158910146204517e-05,
+      "loss": 0.7547,
+      "step": 1145000
+    },
+    {
+      "epoch": 1.971325856464073,
+      "grad_norm": 2.3424324989318848,
+      "learning_rate": 1.7144569058932118e-05,
+      "loss": 0.7472,
+      "step": 1145500
+    },
+    {
+      "epoch": 1.972186321700417,
+      "grad_norm": 2.139158010482788,
+      "learning_rate": 1.7130227971659718e-05,
+      "loss": 0.7554,
+      "step": 1146000
+    },
+    {
+      "epoch": 1.973046786936761,
+      "grad_norm": 2.144340991973877,
+      "learning_rate": 1.711588688438732e-05,
+      "loss": 0.7564,
+      "step": 1146500
+    },
+    {
+      "epoch": 1.973907252173105,
+      "grad_norm": 2.1247239112854004,
+      "learning_rate": 1.710154579711492e-05,
+      "loss": 0.7548,
+      "step": 1147000
+    },
+    {
+      "epoch": 1.974767717409449,
+      "grad_norm": 2.092890501022339,
+      "learning_rate": 1.708720470984252e-05,
+      "loss": 0.7536,
+      "step": 1147500
+    },
+    {
+      "epoch": 1.9756281826457929,
+      "grad_norm": 2.0415735244750977,
+      "learning_rate": 1.7072863622570117e-05,
+      "loss": 0.7538,
+      "step": 1148000
+    },
+    {
+      "epoch": 1.976488647882137,
+      "grad_norm": 2.2634592056274414,
+      "learning_rate": 1.705852253529772e-05,
+      "loss": 0.7566,
+      "step": 1148500
+    },
+    {
+      "epoch": 1.977349113118481,
+      "grad_norm": 2.0326716899871826,
+      "learning_rate": 1.704418144802532e-05,
+      "loss": 0.7566,
+      "step": 1149000
+    },
+    {
+      "epoch": 1.978209578354825,
+      "grad_norm": 2.2424416542053223,
+      "learning_rate": 1.7029840360752918e-05,
+      "loss": 0.7538,
+      "step": 1149500
+    },
+    {
+      "epoch": 1.979070043591169,
+      "grad_norm": 2.1675209999084473,
+      "learning_rate": 1.701549927348052e-05,
+      "loss": 0.7512,
+      "step": 1150000
+    },
+    {
+      "epoch": 1.979930508827513,
+      "grad_norm": 2.072476625442505,
+      "learning_rate": 1.700115818620812e-05,
+      "loss": 0.7528,
+      "step": 1150500
+    },
+    {
+      "epoch": 1.980790974063857,
+      "grad_norm": 2.203397274017334,
+      "learning_rate": 1.698681709893572e-05,
+      "loss": 0.7522,
+      "step": 1151000
+    },
+    {
+      "epoch": 1.981651439300201,
+      "grad_norm": 2.2937443256378174,
+      "learning_rate": 1.697247601166332e-05,
+      "loss": 0.7587,
+      "step": 1151500
+    },
+    {
+      "epoch": 1.982511904536545,
+      "grad_norm": 2.0122551918029785,
+      "learning_rate": 1.695813492439092e-05,
+      "loss": 0.749,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.9833723697728889,
+      "grad_norm": 2.243720531463623,
+      "learning_rate": 1.694379383711852e-05,
+      "loss": 0.7536,
+      "step": 1152500
+    },
+    {
+      "epoch": 1.9842328350092329,
+      "grad_norm": 2.1484217643737793,
+      "learning_rate": 1.692945274984612e-05,
+      "loss": 0.7524,
+      "step": 1153000
+    },
+    {
+      "epoch": 1.9850933002455768,
+      "grad_norm": 2.1573312282562256,
+      "learning_rate": 1.6915111662573722e-05,
+      "loss": 0.753,
+      "step": 1153500
+    },
+    {
+      "epoch": 1.9859537654819208,
+      "grad_norm": 2.0914833545684814,
+      "learning_rate": 1.6900770575301322e-05,
+      "loss": 0.7529,
+      "step": 1154000
+    },
+    {
+      "epoch": 1.9868142307182648,
+      "grad_norm": 2.172123432159424,
+      "learning_rate": 1.688642948802892e-05,
+      "loss": 0.7524,
+      "step": 1154500
+    },
+    {
+      "epoch": 1.9876746959546088,
+      "grad_norm": 2.469355344772339,
+      "learning_rate": 1.6872088400756523e-05,
+      "loss": 0.753,
+      "step": 1155000
+    },
+    {
+      "epoch": 1.9885351611909527,
+      "grad_norm": 2.174379587173462,
+      "learning_rate": 1.685774731348412e-05,
+      "loss": 0.7538,
+      "step": 1155500
+    },
+    {
+      "epoch": 1.9893956264272967,
+      "grad_norm": 2.1006627082824707,
+      "learning_rate": 1.684340622621172e-05,
+      "loss": 0.7549,
+      "step": 1156000
+    },
+    {
+      "epoch": 1.9902560916636407,
+      "grad_norm": 2.183161973953247,
+      "learning_rate": 1.6829065138939325e-05,
+      "loss": 0.7535,
+      "step": 1156500
+    },
+    {
+      "epoch": 1.9911165568999847,
+      "grad_norm": 2.2185134887695312,
+      "learning_rate": 1.6814724051666922e-05,
+      "loss": 0.7555,
+      "step": 1157000
+    },
+    {
+      "epoch": 1.9919770221363287,
+      "grad_norm": 2.2173125743865967,
+      "learning_rate": 1.6800382964394522e-05,
+      "loss": 0.7455,
+      "step": 1157500
+    },
+    {
+      "epoch": 1.9928374873726726,
+      "grad_norm": 1.9993641376495361,
+      "learning_rate": 1.6786041877122126e-05,
+      "loss": 0.7502,
+      "step": 1158000
+    },
+    {
+      "epoch": 1.9936979526090166,
+      "grad_norm": 2.248117685317993,
+      "learning_rate": 1.6771700789849723e-05,
+      "loss": 0.7535,
+      "step": 1158500
+    },
+    {
+      "epoch": 1.9945584178453606,
+      "grad_norm": 2.1918869018554688,
+      "learning_rate": 1.6757359702577324e-05,
+      "loss": 0.7534,
+      "step": 1159000
+    },
+    {
+      "epoch": 1.9954188830817046,
+      "grad_norm": 2.221872091293335,
+      "learning_rate": 1.6743018615304924e-05,
+      "loss": 0.747,
+      "step": 1159500
+    },
+    {
+      "epoch": 1.9962793483180485,
+      "grad_norm": 2.3851265907287598,
+      "learning_rate": 1.6728677528032525e-05,
+      "loss": 0.7533,
+      "step": 1160000
+    },
+    {
+      "epoch": 1.9971398135543925,
+      "grad_norm": 2.5031321048736572,
+      "learning_rate": 1.6714336440760125e-05,
+      "loss": 0.7546,
+      "step": 1160500
+    },
+    {
+      "epoch": 1.9980002787907365,
+      "grad_norm": 2.7600536346435547,
+      "learning_rate": 1.6699995353487726e-05,
+      "loss": 0.7541,
+      "step": 1161000
+    },
+    {
+      "epoch": 1.9988607440270805,
+      "grad_norm": 2.3278236389160156,
+      "learning_rate": 1.6685654266215326e-05,
+      "loss": 0.7485,
+      "step": 1161500
+    },
+    {
+      "epoch": 1.9997212092634244,
+      "grad_norm": 2.1311707496643066,
+      "learning_rate": 1.6671313178942923e-05,
+      "loss": 0.7475,
+      "step": 1162000
+    },
+    {
+      "epoch": 2.0005816744997684,
+      "grad_norm": 2.3819072246551514,
+      "learning_rate": 1.6656972091670524e-05,
+      "loss": 0.7473,
+      "step": 1162500
+    },
+    {
+      "epoch": 2.0014421397361124,
+      "grad_norm": 2.184633255004883,
+      "learning_rate": 1.6642631004398127e-05,
+      "loss": 0.7514,
+      "step": 1163000
+    },
+    {
+      "epoch": 2.0023026049724564,
+      "grad_norm": 2.1459732055664062,
+      "learning_rate": 1.6628289917125725e-05,
+      "loss": 0.7483,
+      "step": 1163500
+    },
+    {
+      "epoch": 2.0031630702088004,
+      "grad_norm": 2.32151198387146,
+      "learning_rate": 1.6613948829853325e-05,
+      "loss": 0.7509,
+      "step": 1164000
+    },
+    {
+      "epoch": 2.0040235354451443,
+      "grad_norm": 2.1330671310424805,
+      "learning_rate": 1.6599607742580925e-05,
+      "loss": 0.7514,
+      "step": 1164500
+    },
+    {
+      "epoch": 2.0048840006814883,
+      "grad_norm": 2.2304253578186035,
+      "learning_rate": 1.6585266655308526e-05,
+      "loss": 0.7582,
+      "step": 1165000
+    },
+    {
+      "epoch": 2.0057444659178323,
+      "grad_norm": 2.269551992416382,
+      "learning_rate": 1.6570925568036126e-05,
+      "loss": 0.7556,
+      "step": 1165500
+    },
+    {
+      "epoch": 2.0066049311541763,
+      "grad_norm": 2.071958065032959,
+      "learning_rate": 1.6556584480763727e-05,
+      "loss": 0.7435,
+      "step": 1166000
+    },
+    {
+      "epoch": 2.0074653963905202,
+      "grad_norm": 2.009504556655884,
+      "learning_rate": 1.6542243393491327e-05,
+      "loss": 0.7503,
+      "step": 1166500
+    },
+    {
+      "epoch": 2.008325861626864,
+      "grad_norm": 2.0557193756103516,
+      "learning_rate": 1.6527902306218924e-05,
+      "loss": 0.7491,
+      "step": 1167000
+    },
+    {
+      "epoch": 2.009186326863208,
+      "grad_norm": 2.093259572982788,
+      "learning_rate": 1.6513561218946528e-05,
+      "loss": 0.7515,
+      "step": 1167500
+    },
+    {
+      "epoch": 2.010046792099552,
+      "grad_norm": 2.12383770942688,
+      "learning_rate": 1.649922013167413e-05,
+      "loss": 0.7466,
+      "step": 1168000
+    },
+    {
+      "epoch": 2.0109072573358966,
+      "grad_norm": 2.0827810764312744,
+      "learning_rate": 1.6484879044401726e-05,
+      "loss": 0.7474,
+      "step": 1168500
+    },
+    {
+      "epoch": 2.0117677225722406,
+      "grad_norm": 2.178234338760376,
+      "learning_rate": 1.647053795712933e-05,
+      "loss": 0.7499,
+      "step": 1169000
+    },
+    {
+      "epoch": 2.0126281878085845,
+      "grad_norm": 2.224212646484375,
+      "learning_rate": 1.645619686985693e-05,
+      "loss": 0.7491,
+      "step": 1169500
+    },
+    {
+      "epoch": 2.0134886530449285,
+      "grad_norm": 2.0262248516082764,
+      "learning_rate": 1.6441855782584527e-05,
+      "loss": 0.7486,
+      "step": 1170000
+    },
+    {
+      "epoch": 2.0143491182812725,
+      "grad_norm": 2.3332037925720215,
+      "learning_rate": 1.642751469531213e-05,
+      "loss": 0.7416,
+      "step": 1170500
+    },
+    {
+      "epoch": 2.0152095835176165,
+      "grad_norm": 2.1311421394348145,
+      "learning_rate": 1.6413173608039728e-05,
+      "loss": 0.7491,
+      "step": 1171000
+    },
+    {
+      "epoch": 2.0160700487539605,
+      "grad_norm": 1.9237945079803467,
+      "learning_rate": 1.639883252076733e-05,
+      "loss": 0.7487,
+      "step": 1171500
+    },
+    {
+      "epoch": 2.0169305139903044,
+      "grad_norm": 2.152676820755005,
+      "learning_rate": 1.638449143349493e-05,
+      "loss": 0.7505,
+      "step": 1172000
+    },
+    {
+      "epoch": 2.0177909792266484,
+      "grad_norm": 2.240447759628296,
+      "learning_rate": 1.637015034622253e-05,
+      "loss": 0.7487,
+      "step": 1172500
+    },
+    {
+      "epoch": 2.0186514444629924,
+      "grad_norm": 2.2949647903442383,
+      "learning_rate": 1.635580925895013e-05,
+      "loss": 0.7508,
+      "step": 1173000
+    },
+    {
+      "epoch": 2.0195119096993364,
+      "grad_norm": 2.259401321411133,
+      "learning_rate": 1.634146817167773e-05,
+      "loss": 0.745,
+      "step": 1173500
+    },
+    {
+      "epoch": 2.0203723749356803,
+      "grad_norm": 2.1563282012939453,
+      "learning_rate": 1.632712708440533e-05,
+      "loss": 0.7499,
+      "step": 1174000
+    },
+    {
+      "epoch": 2.0212328401720243,
+      "grad_norm": 2.198007822036743,
+      "learning_rate": 1.631278599713293e-05,
+      "loss": 0.7484,
+      "step": 1174500
+    },
+    {
+      "epoch": 2.0220933054083683,
+      "grad_norm": 2.1369738578796387,
+      "learning_rate": 1.629844490986053e-05,
+      "loss": 0.7545,
+      "step": 1175000
+    },
+    {
+      "epoch": 2.0229537706447123,
+      "grad_norm": 2.226137638092041,
+      "learning_rate": 1.6284103822588132e-05,
+      "loss": 0.7532,
+      "step": 1175500
+    },
+    {
+      "epoch": 2.0238142358810562,
+      "grad_norm": 2.0248351097106934,
+      "learning_rate": 1.626976273531573e-05,
+      "loss": 0.7513,
+      "step": 1176000
+    },
+    {
+      "epoch": 2.0246747011174,
+      "grad_norm": 2.1181812286376953,
+      "learning_rate": 1.625542164804333e-05,
+      "loss": 0.7445,
+      "step": 1176500
+    },
+    {
+      "epoch": 2.025535166353744,
+      "grad_norm": 2.3558504581451416,
+      "learning_rate": 1.6241080560770934e-05,
+      "loss": 0.7465,
+      "step": 1177000
+    },
+    {
+      "epoch": 2.026395631590088,
+      "grad_norm": 2.0135488510131836,
+      "learning_rate": 1.622673947349853e-05,
+      "loss": 0.7492,
+      "step": 1177500
+    },
+    {
+      "epoch": 2.027256096826432,
+      "grad_norm": 2.0925824642181396,
+      "learning_rate": 1.621239838622613e-05,
+      "loss": 0.7468,
+      "step": 1178000
+    },
+    {
+      "epoch": 2.028116562062776,
+      "grad_norm": 2.0782012939453125,
+      "learning_rate": 1.6198057298953732e-05,
+      "loss": 0.7505,
+      "step": 1178500
+    },
+    {
+      "epoch": 2.02897702729912,
+      "grad_norm": 2.0250723361968994,
+      "learning_rate": 1.6183716211681332e-05,
+      "loss": 0.7494,
+      "step": 1179000
+    },
+    {
+      "epoch": 2.029837492535464,
+      "grad_norm": 2.0154037475585938,
+      "learning_rate": 1.6169375124408933e-05,
+      "loss": 0.7478,
+      "step": 1179500
+    },
+    {
+      "epoch": 2.030697957771808,
+      "grad_norm": 2.1285109519958496,
+      "learning_rate": 1.6155034037136533e-05,
+      "loss": 0.7462,
+      "step": 1180000
+    },
+    {
+      "epoch": 2.031558423008152,
+      "grad_norm": 2.2587168216705322,
+      "learning_rate": 1.6140692949864134e-05,
+      "loss": 0.7456,
+      "step": 1180500
+    },
+    {
+      "epoch": 2.032418888244496,
+      "grad_norm": 2.261652946472168,
+      "learning_rate": 1.6126351862591734e-05,
+      "loss": 0.7484,
+      "step": 1181000
+    },
+    {
+      "epoch": 2.03327935348084,
+      "grad_norm": 2.2460522651672363,
+      "learning_rate": 1.6112010775319335e-05,
+      "loss": 0.7486,
+      "step": 1181500
+    },
+    {
+      "epoch": 2.034139818717184,
+      "grad_norm": 2.511417865753174,
+      "learning_rate": 1.6097669688046935e-05,
+      "loss": 0.7472,
+      "step": 1182000
+    },
+    {
+      "epoch": 2.035000283953528,
+      "grad_norm": 2.14890193939209,
+      "learning_rate": 1.6083328600774532e-05,
+      "loss": 0.7476,
+      "step": 1182500
+    },
+    {
+      "epoch": 2.035860749189872,
+      "grad_norm": 2.20908260345459,
+      "learning_rate": 1.6068987513502136e-05,
+      "loss": 0.7534,
+      "step": 1183000
+    },
+    {
+      "epoch": 2.036721214426216,
+      "grad_norm": 2.3535568714141846,
+      "learning_rate": 1.6054646426229737e-05,
+      "loss": 0.7505,
+      "step": 1183500
+    },
+    {
+      "epoch": 2.03758167966256,
+      "grad_norm": 2.2410316467285156,
+      "learning_rate": 1.6040305338957334e-05,
+      "loss": 0.7467,
+      "step": 1184000
+    },
+    {
+      "epoch": 2.038442144898904,
+      "grad_norm": 2.1737914085388184,
+      "learning_rate": 1.6025964251684934e-05,
+      "loss": 0.747,
+      "step": 1184500
+    },
+    {
+      "epoch": 2.039302610135248,
+      "grad_norm": 2.2757697105407715,
+      "learning_rate": 1.6011623164412535e-05,
+      "loss": 0.7423,
+      "step": 1185000
+    },
+    {
+      "epoch": 2.040163075371592,
+      "grad_norm": 2.2338526248931885,
+      "learning_rate": 1.5997282077140135e-05,
+      "loss": 0.7474,
+      "step": 1185500
+    },
+    {
+      "epoch": 2.041023540607936,
+      "grad_norm": 2.1928253173828125,
+      "learning_rate": 1.5982940989867736e-05,
+      "loss": 0.7538,
+      "step": 1186000
+    },
+    {
+      "epoch": 2.0418840058442798,
+      "grad_norm": 2.32772159576416,
+      "learning_rate": 1.5968599902595336e-05,
+      "loss": 0.7449,
+      "step": 1186500
+    },
+    {
+      "epoch": 2.0427444710806237,
+      "grad_norm": 2.1408066749572754,
+      "learning_rate": 1.5954258815322936e-05,
+      "loss": 0.7463,
+      "step": 1187000
+    },
+    {
+      "epoch": 2.0436049363169677,
+      "grad_norm": 2.2051782608032227,
+      "learning_rate": 1.5939917728050534e-05,
+      "loss": 0.7463,
+      "step": 1187500
+    },
+    {
+      "epoch": 2.0444654015533117,
+      "grad_norm": 2.259019136428833,
+      "learning_rate": 1.5925576640778137e-05,
+      "loss": 0.7451,
+      "step": 1188000
+    },
+    {
+      "epoch": 2.0453258667896557,
+      "grad_norm": 2.147264242172241,
+      "learning_rate": 1.5911235553505738e-05,
+      "loss": 0.7554,
+      "step": 1188500
+    },
+    {
+      "epoch": 2.0461863320259996,
+      "grad_norm": 2.019881248474121,
+      "learning_rate": 1.5896894466233335e-05,
+      "loss": 0.743,
+      "step": 1189000
+    },
+    {
+      "epoch": 2.0470467972623436,
+      "grad_norm": 2.1331872940063477,
+      "learning_rate": 1.588255337896094e-05,
+      "loss": 0.7454,
+      "step": 1189500
+    },
+    {
+      "epoch": 2.0479072624986876,
+      "grad_norm": 2.1059558391571045,
+      "learning_rate": 1.586821229168854e-05,
+      "loss": 0.7452,
+      "step": 1190000
+    },
+    {
+      "epoch": 2.0487677277350316,
+      "grad_norm": 2.168329954147339,
+      "learning_rate": 1.5853871204416136e-05,
+      "loss": 0.7548,
+      "step": 1190500
+    },
+    {
+      "epoch": 2.0496281929713756,
+      "grad_norm": 2.21160888671875,
+      "learning_rate": 1.583953011714374e-05,
+      "loss": 0.7478,
+      "step": 1191000
+    },
+    {
+      "epoch": 2.05048865820772,
+      "grad_norm": 2.0780110359191895,
+      "learning_rate": 1.5825189029871337e-05,
+      "loss": 0.7444,
+      "step": 1191500
+    },
+    {
+      "epoch": 2.051349123444064,
+      "grad_norm": 2.0575709342956543,
+      "learning_rate": 1.5810847942598938e-05,
+      "loss": 0.7393,
+      "step": 1192000
+    },
+    {
+      "epoch": 2.052209588680408,
+      "grad_norm": 2.3351075649261475,
+      "learning_rate": 1.579650685532654e-05,
+      "loss": 0.7448,
+      "step": 1192500
+    },
+    {
+      "epoch": 2.053070053916752,
+      "grad_norm": 2.023301362991333,
+      "learning_rate": 1.578216576805414e-05,
+      "loss": 0.7454,
+      "step": 1193000
+    },
+    {
+      "epoch": 2.053930519153096,
+      "grad_norm": 2.1645848751068115,
+      "learning_rate": 1.576782468078174e-05,
+      "loss": 0.7443,
+      "step": 1193500
+    },
+    {
+      "epoch": 2.05479098438944,
+      "grad_norm": 2.1541600227355957,
+      "learning_rate": 1.575348359350934e-05,
+      "loss": 0.7458,
+      "step": 1194000
+    },
+    {
+      "epoch": 2.055651449625784,
+      "grad_norm": 2.260538339614868,
+      "learning_rate": 1.573914250623694e-05,
+      "loss": 0.7435,
+      "step": 1194500
+    },
+    {
+      "epoch": 2.056511914862128,
+      "grad_norm": 2.1604511737823486,
+      "learning_rate": 1.572480141896454e-05,
+      "loss": 0.7527,
+      "step": 1195000
+    },
+    {
+      "epoch": 2.057372380098472,
+      "grad_norm": 2.21130633354187,
+      "learning_rate": 1.571046033169214e-05,
+      "loss": 0.7499,
+      "step": 1195500
+    },
+    {
+      "epoch": 2.0582328453348158,
+      "grad_norm": 2.1533782482147217,
+      "learning_rate": 1.569611924441974e-05,
+      "loss": 0.7475,
+      "step": 1196000
+    },
+    {
+      "epoch": 2.0590933105711597,
+      "grad_norm": 2.1712965965270996,
+      "learning_rate": 1.568177815714734e-05,
+      "loss": 0.742,
+      "step": 1196500
+    },
+    {
+      "epoch": 2.0599537758075037,
+      "grad_norm": 2.198495388031006,
+      "learning_rate": 1.566743706987494e-05,
+      "loss": 0.7473,
+      "step": 1197000
+    },
+    {
+      "epoch": 2.0608142410438477,
+      "grad_norm": 2.3291733264923096,
+      "learning_rate": 1.5653095982602543e-05,
+      "loss": 0.7453,
+      "step": 1197500
+    },
+    {
+      "epoch": 2.0616747062801917,
+      "grad_norm": 2.1571593284606934,
+      "learning_rate": 1.563875489533014e-05,
+      "loss": 0.7467,
+      "step": 1198000
+    },
+    {
+      "epoch": 2.0625351715165356,
+      "grad_norm": 2.022718667984009,
+      "learning_rate": 1.562441380805774e-05,
+      "loss": 0.7401,
+      "step": 1198500
+    },
+    {
+      "epoch": 2.0633956367528796,
+      "grad_norm": 2.2326862812042236,
+      "learning_rate": 1.561007272078534e-05,
+      "loss": 0.7498,
+      "step": 1199000
+    },
+    {
+      "epoch": 2.0642561019892236,
+      "grad_norm": 2.4144537448883057,
+      "learning_rate": 1.559573163351294e-05,
+      "loss": 0.7467,
+      "step": 1199500
+    },
+    {
+      "epoch": 2.0651165672255676,
+      "grad_norm": 2.1190388202667236,
+      "learning_rate": 1.5581390546240542e-05,
+      "loss": 0.7462,
+      "step": 1200000
+    },
+    {
+      "epoch": 2.0659770324619116,
+      "grad_norm": 2.1224961280822754,
+      "learning_rate": 1.5567049458968142e-05,
+      "loss": 0.7461,
+      "step": 1200500
+    },
+    {
+      "epoch": 2.0668374976982555,
+      "grad_norm": 2.1967227458953857,
+      "learning_rate": 1.5552708371695743e-05,
+      "loss": 0.7479,
+      "step": 1201000
+    },
+    {
+      "epoch": 2.0676979629345995,
+      "grad_norm": 2.357408046722412,
+      "learning_rate": 1.5538367284423343e-05,
+      "loss": 0.7408,
+      "step": 1201500
+    },
+    {
+      "epoch": 2.0685584281709435,
+      "grad_norm": 2.196808099746704,
+      "learning_rate": 1.5524026197150944e-05,
+      "loss": 0.7493,
+      "step": 1202000
+    },
+    {
+      "epoch": 2.0694188934072875,
+      "grad_norm": 2.0822932720184326,
+      "learning_rate": 1.5509685109878544e-05,
+      "loss": 0.7426,
+      "step": 1202500
+    },
+    {
+      "epoch": 2.0702793586436314,
+      "grad_norm": 2.3068885803222656,
+      "learning_rate": 1.549534402260614e-05,
+      "loss": 0.747,
+      "step": 1203000
+    },
+    {
+      "epoch": 2.0711398238799754,
+      "grad_norm": 2.0930685997009277,
+      "learning_rate": 1.5481002935333745e-05,
+      "loss": 0.7487,
+      "step": 1203500
+    },
+    {
+      "epoch": 2.0720002891163194,
+      "grad_norm": 2.1573619842529297,
+      "learning_rate": 1.5466661848061346e-05,
+      "loss": 0.75,
+      "step": 1204000
+    },
+    {
+      "epoch": 2.0728607543526634,
+      "grad_norm": 2.1406896114349365,
+      "learning_rate": 1.5452320760788943e-05,
+      "loss": 0.7539,
+      "step": 1204500
+    },
+    {
+      "epoch": 2.0737212195890073,
+      "grad_norm": 2.2356812953948975,
+      "learning_rate": 1.5437979673516547e-05,
+      "loss": 0.7456,
+      "step": 1205000
+    },
+    {
+      "epoch": 2.0745816848253513,
+      "grad_norm": 2.211362361907959,
+      "learning_rate": 1.5423638586244144e-05,
+      "loss": 0.7448,
+      "step": 1205500
+    },
+    {
+      "epoch": 2.0754421500616953,
+      "grad_norm": 1.9883774518966675,
+      "learning_rate": 1.5409297498971744e-05,
+      "loss": 0.7448,
+      "step": 1206000
+    },
+    {
+      "epoch": 2.0763026152980393,
+      "grad_norm": 2.2814600467681885,
+      "learning_rate": 1.5394956411699345e-05,
+      "loss": 0.7479,
+      "step": 1206500
+    },
+    {
+      "epoch": 2.0771630805343833,
+      "grad_norm": 1.9190709590911865,
+      "learning_rate": 1.5380615324426945e-05,
+      "loss": 0.7465,
+      "step": 1207000
+    },
+    {
+      "epoch": 2.0780235457707272,
+      "grad_norm": 2.15561580657959,
+      "learning_rate": 1.5366274237154546e-05,
+      "loss": 0.7414,
+      "step": 1207500
+    },
+    {
+      "epoch": 2.078884011007071,
+      "grad_norm": 2.1482765674591064,
+      "learning_rate": 1.5351933149882146e-05,
+      "loss": 0.7477,
+      "step": 1208000
+    },
+    {
+      "epoch": 2.079744476243415,
+      "grad_norm": 2.2311723232269287,
+      "learning_rate": 1.5337592062609747e-05,
+      "loss": 0.7425,
+      "step": 1208500
+    },
+    {
+      "epoch": 2.080604941479759,
+      "grad_norm": 2.053049325942993,
+      "learning_rate": 1.5323250975337347e-05,
+      "loss": 0.7523,
+      "step": 1209000
+    },
+    {
+      "epoch": 2.081465406716103,
+      "grad_norm": 1.9471055269241333,
+      "learning_rate": 1.5308909888064944e-05,
+      "loss": 0.7486,
+      "step": 1209500
+    },
+    {
+      "epoch": 2.082325871952447,
+      "grad_norm": 2.088299512863159,
+      "learning_rate": 1.5294568800792548e-05,
+      "loss": 0.7417,
+      "step": 1210000
+    },
+    {
+      "epoch": 2.083186337188791,
+      "grad_norm": 2.083714008331299,
+      "learning_rate": 1.5280227713520145e-05,
+      "loss": 0.7433,
+      "step": 1210500
+    },
+    {
+      "epoch": 2.084046802425135,
+      "grad_norm": 2.0353481769561768,
+      "learning_rate": 1.5265886626247745e-05,
+      "loss": 0.7512,
+      "step": 1211000
+    },
+    {
+      "epoch": 2.084907267661479,
+      "grad_norm": 2.2791249752044678,
+      "learning_rate": 1.5251545538975348e-05,
+      "loss": 0.7446,
+      "step": 1211500
+    },
+    {
+      "epoch": 2.085767732897823,
+      "grad_norm": 2.114086627960205,
+      "learning_rate": 1.5237204451702946e-05,
+      "loss": 0.7454,
+      "step": 1212000
+    },
+    {
+      "epoch": 2.086628198134167,
+      "grad_norm": 2.1683554649353027,
+      "learning_rate": 1.5222863364430549e-05,
+      "loss": 0.7456,
+      "step": 1212500
+    },
+    {
+      "epoch": 2.087488663370511,
+      "grad_norm": 2.1287851333618164,
+      "learning_rate": 1.5208522277158149e-05,
+      "loss": 0.7474,
+      "step": 1213000
+    },
+    {
+      "epoch": 2.088349128606855,
+      "grad_norm": 2.2126824855804443,
+      "learning_rate": 1.5194181189885748e-05,
+      "loss": 0.7477,
+      "step": 1213500
+    },
+    {
+      "epoch": 2.089209593843199,
+      "grad_norm": 2.0747759342193604,
+      "learning_rate": 1.5179840102613348e-05,
+      "loss": 0.7482,
+      "step": 1214000
+    },
+    {
+      "epoch": 2.090070059079543,
+      "grad_norm": 2.1446080207824707,
+      "learning_rate": 1.5165499015340947e-05,
+      "loss": 0.7426,
+      "step": 1214500
+    },
+    {
+      "epoch": 2.0909305243158873,
+      "grad_norm": 2.1761319637298584,
+      "learning_rate": 1.515115792806855e-05,
+      "loss": 0.7485,
+      "step": 1215000
+    },
+    {
+      "epoch": 2.0917909895522313,
+      "grad_norm": 2.060673236846924,
+      "learning_rate": 1.513681684079615e-05,
+      "loss": 0.7415,
+      "step": 1215500
+    },
+    {
+      "epoch": 2.0926514547885753,
+      "grad_norm": 2.208153009414673,
+      "learning_rate": 1.5122475753523749e-05,
+      "loss": 0.7501,
+      "step": 1216000
+    },
+    {
+      "epoch": 2.0935119200249193,
+      "grad_norm": 2.4316937923431396,
+      "learning_rate": 1.510813466625135e-05,
+      "loss": 0.7481,
+      "step": 1216500
+    },
+    {
+      "epoch": 2.0943723852612632,
+      "grad_norm": 2.358034133911133,
+      "learning_rate": 1.509379357897895e-05,
+      "loss": 0.7501,
+      "step": 1217000
+    },
+    {
+      "epoch": 2.095232850497607,
+      "grad_norm": 2.3785462379455566,
+      "learning_rate": 1.507945249170655e-05,
+      "loss": 0.7485,
+      "step": 1217500
+    },
+    {
+      "epoch": 2.096093315733951,
+      "grad_norm": 2.0434811115264893,
+      "learning_rate": 1.5065111404434152e-05,
+      "loss": 0.7431,
+      "step": 1218000
+    },
+    {
+      "epoch": 2.096953780970295,
+      "grad_norm": 2.2284770011901855,
+      "learning_rate": 1.505077031716175e-05,
+      "loss": 0.7461,
+      "step": 1218500
+    },
+    {
+      "epoch": 2.097814246206639,
+      "grad_norm": 2.4026939868927,
+      "learning_rate": 1.5036429229889351e-05,
+      "loss": 0.7492,
+      "step": 1219000
+    },
+    {
+      "epoch": 2.098674711442983,
+      "grad_norm": 1.9877442121505737,
+      "learning_rate": 1.502208814261695e-05,
+      "loss": 0.7471,
+      "step": 1219500
+    },
+    {
+      "epoch": 2.099535176679327,
+      "grad_norm": 2.2232022285461426,
+      "learning_rate": 1.500774705534455e-05,
+      "loss": 0.7472,
+      "step": 1220000
+    },
+    {
+      "epoch": 2.100395641915671,
+      "grad_norm": 2.2752528190612793,
+      "learning_rate": 1.4993405968072153e-05,
+      "loss": 0.748,
+      "step": 1220500
+    },
+    {
+      "epoch": 2.101256107152015,
+      "grad_norm": 2.3126862049102783,
+      "learning_rate": 1.4979064880799752e-05,
+      "loss": 0.7516,
+      "step": 1221000
+    },
+    {
+      "epoch": 2.102116572388359,
+      "grad_norm": 2.171881914138794,
+      "learning_rate": 1.4964723793527352e-05,
+      "loss": 0.7539,
+      "step": 1221500
+    },
+    {
+      "epoch": 2.102977037624703,
+      "grad_norm": 2.145996570587158,
+      "learning_rate": 1.495038270625495e-05,
+      "loss": 0.7455,
+      "step": 1222000
+    },
+    {
+      "epoch": 2.103837502861047,
+      "grad_norm": 2.396362781524658,
+      "learning_rate": 1.4936041618982551e-05,
+      "loss": 0.7477,
+      "step": 1222500
+    },
+    {
+      "epoch": 2.104697968097391,
+      "grad_norm": 2.1479058265686035,
+      "learning_rate": 1.4921700531710153e-05,
+      "loss": 0.7431,
+      "step": 1223000
+    },
+    {
+      "epoch": 2.105558433333735,
+      "grad_norm": 2.0262203216552734,
+      "learning_rate": 1.4907359444437752e-05,
+      "loss": 0.7443,
+      "step": 1223500
+    },
+    {
+      "epoch": 2.106418898570079,
+      "grad_norm": 2.2727341651916504,
+      "learning_rate": 1.4893018357165353e-05,
+      "loss": 0.7525,
+      "step": 1224000
+    },
+    {
+      "epoch": 2.107279363806423,
+      "grad_norm": 2.1716670989990234,
+      "learning_rate": 1.4878677269892955e-05,
+      "loss": 0.7483,
+      "step": 1224500
+    },
+    {
+      "epoch": 2.108139829042767,
+      "grad_norm": 2.2675817012786865,
+      "learning_rate": 1.4864336182620554e-05,
+      "loss": 0.7395,
+      "step": 1225000
+    },
+    {
+      "epoch": 2.109000294279111,
+      "grad_norm": 2.3223788738250732,
+      "learning_rate": 1.4849995095348154e-05,
+      "loss": 0.7452,
+      "step": 1225500
+    },
+    {
+      "epoch": 2.109860759515455,
+      "grad_norm": 2.005021572113037,
+      "learning_rate": 1.4835654008075753e-05,
+      "loss": 0.7466,
+      "step": 1226000
+    },
+    {
+      "epoch": 2.110721224751799,
+      "grad_norm": 2.136507749557495,
+      "learning_rate": 1.4821312920803353e-05,
+      "loss": 0.7549,
+      "step": 1226500
+    },
+    {
+      "epoch": 2.1115816899881428,
+      "grad_norm": 2.095691442489624,
+      "learning_rate": 1.4806971833530955e-05,
+      "loss": 0.7451,
+      "step": 1227000
+    },
+    {
+      "epoch": 2.1124421552244868,
+      "grad_norm": 2.3151662349700928,
+      "learning_rate": 1.4792630746258554e-05,
+      "loss": 0.7487,
+      "step": 1227500
+    },
+    {
+      "epoch": 2.1133026204608307,
+      "grad_norm": 1.9382226467132568,
+      "learning_rate": 1.4778289658986155e-05,
+      "loss": 0.743,
+      "step": 1228000
+    },
+    {
+      "epoch": 2.1141630856971747,
+      "grad_norm": 2.3391473293304443,
+      "learning_rate": 1.4763948571713753e-05,
+      "loss": 0.7511,
+      "step": 1228500
+    },
+    {
+      "epoch": 2.1150235509335187,
+      "grad_norm": 2.2001867294311523,
+      "learning_rate": 1.4749607484441356e-05,
+      "loss": 0.7451,
+      "step": 1229000
+    },
+    {
+      "epoch": 2.1158840161698627,
+      "grad_norm": 2.054943084716797,
+      "learning_rate": 1.4735266397168956e-05,
+      "loss": 0.744,
+      "step": 1229500
+    },
+    {
+      "epoch": 2.1167444814062066,
+      "grad_norm": 2.11088490486145,
+      "learning_rate": 1.4720925309896555e-05,
+      "loss": 0.7486,
+      "step": 1230000
+    },
+    {
+      "epoch": 2.1176049466425506,
+      "grad_norm": 2.307340383529663,
+      "learning_rate": 1.4706584222624157e-05,
+      "loss": 0.7465,
+      "step": 1230500
+    },
+    {
+      "epoch": 2.1184654118788946,
+      "grad_norm": 2.4765310287475586,
+      "learning_rate": 1.4692243135351754e-05,
+      "loss": 0.7438,
+      "step": 1231000
+    },
+    {
+      "epoch": 2.1193258771152386,
+      "grad_norm": 2.0979506969451904,
+      "learning_rate": 1.4677902048079356e-05,
+      "loss": 0.7455,
+      "step": 1231500
+    },
+    {
+      "epoch": 2.1201863423515825,
+      "grad_norm": 2.199089765548706,
+      "learning_rate": 1.4663560960806957e-05,
+      "loss": 0.7432,
+      "step": 1232000
+    },
+    {
+      "epoch": 2.1210468075879265,
+      "grad_norm": 2.230161428451538,
+      "learning_rate": 1.4649219873534556e-05,
+      "loss": 0.7446,
+      "step": 1232500
+    },
+    {
+      "epoch": 2.1219072728242705,
+      "grad_norm": 2.351138114929199,
+      "learning_rate": 1.4634878786262158e-05,
+      "loss": 0.7419,
+      "step": 1233000
+    },
+    {
+      "epoch": 2.1227677380606145,
+      "grad_norm": 2.1402597427368164,
+      "learning_rate": 1.4620537698989757e-05,
+      "loss": 0.7454,
+      "step": 1233500
+    },
+    {
+      "epoch": 2.1236282032969584,
+      "grad_norm": 2.3817059993743896,
+      "learning_rate": 1.4606196611717357e-05,
+      "loss": 0.7449,
+      "step": 1234000
+    },
+    {
+      "epoch": 2.1244886685333024,
+      "grad_norm": 2.16794753074646,
+      "learning_rate": 1.4591855524444959e-05,
+      "loss": 0.7396,
+      "step": 1234500
+    },
+    {
+      "epoch": 2.1253491337696464,
+      "grad_norm": 2.18027925491333,
+      "learning_rate": 1.4577514437172556e-05,
+      "loss": 0.7466,
+      "step": 1235000
+    },
+    {
+      "epoch": 2.1262095990059904,
+      "grad_norm": 2.0430901050567627,
+      "learning_rate": 1.4563173349900158e-05,
+      "loss": 0.7452,
+      "step": 1235500
+    },
+    {
+      "epoch": 2.1270700642423344,
+      "grad_norm": 2.072225332260132,
+      "learning_rate": 1.4548832262627759e-05,
+      "loss": 0.7422,
+      "step": 1236000
+    },
+    {
+      "epoch": 2.1279305294786783,
+      "grad_norm": 2.125211715698242,
+      "learning_rate": 1.4534491175355358e-05,
+      "loss": 0.7458,
+      "step": 1236500
+    },
+    {
+      "epoch": 2.1287909947150223,
+      "grad_norm": 2.3344390392303467,
+      "learning_rate": 1.452015008808296e-05,
+      "loss": 0.7456,
+      "step": 1237000
+    },
+    {
+      "epoch": 2.1296514599513667,
+      "grad_norm": 2.0346381664276123,
+      "learning_rate": 1.4505809000810559e-05,
+      "loss": 0.7432,
+      "step": 1237500
+    },
+    {
+      "epoch": 2.1305119251877107,
+      "grad_norm": 2.1993465423583984,
+      "learning_rate": 1.4491467913538159e-05,
+      "loss": 0.7481,
+      "step": 1238000
+    },
+    {
+      "epoch": 2.1313723904240547,
+      "grad_norm": 2.192777633666992,
+      "learning_rate": 1.4477126826265761e-05,
+      "loss": 0.747,
+      "step": 1238500
+    },
+    {
+      "epoch": 2.1322328556603987,
+      "grad_norm": 2.124511480331421,
+      "learning_rate": 1.4462785738993358e-05,
+      "loss": 0.7459,
+      "step": 1239000
+    },
+    {
+      "epoch": 2.1330933208967426,
+      "grad_norm": 2.1302876472473145,
+      "learning_rate": 1.444844465172096e-05,
+      "loss": 0.7431,
+      "step": 1239500
+    },
+    {
+      "epoch": 2.1339537861330866,
+      "grad_norm": 2.170459747314453,
+      "learning_rate": 1.443410356444856e-05,
+      "loss": 0.739,
+      "step": 1240000
+    },
+    {
+      "epoch": 2.1348142513694306,
+      "grad_norm": 2.136836528778076,
+      "learning_rate": 1.441976247717616e-05,
+      "loss": 0.7445,
+      "step": 1240500
+    },
+    {
+      "epoch": 2.1356747166057746,
+      "grad_norm": 2.169245719909668,
+      "learning_rate": 1.4405421389903762e-05,
+      "loss": 0.7391,
+      "step": 1241000
+    },
+    {
+      "epoch": 2.1365351818421185,
+      "grad_norm": 2.053624391555786,
+      "learning_rate": 1.439108030263136e-05,
+      "loss": 0.7415,
+      "step": 1241500
+    },
+    {
+      "epoch": 2.1373956470784625,
+      "grad_norm": 2.225248336791992,
+      "learning_rate": 1.4376739215358961e-05,
+      "loss": 0.7459,
+      "step": 1242000
+    },
+    {
+      "epoch": 2.1382561123148065,
+      "grad_norm": 2.267460584640503,
+      "learning_rate": 1.436239812808656e-05,
+      "loss": 0.7454,
+      "step": 1242500
+    },
+    {
+      "epoch": 2.1391165775511505,
+      "grad_norm": 2.054811716079712,
+      "learning_rate": 1.4348057040814162e-05,
+      "loss": 0.7435,
+      "step": 1243000
+    },
+    {
+      "epoch": 2.1399770427874945,
+      "grad_norm": 2.2870633602142334,
+      "learning_rate": 1.4333715953541763e-05,
+      "loss": 0.7454,
+      "step": 1243500
+    },
+    {
+      "epoch": 2.1408375080238384,
+      "grad_norm": 2.0539770126342773,
+      "learning_rate": 1.4319374866269361e-05,
+      "loss": 0.7471,
+      "step": 1244000
+    },
+    {
+      "epoch": 2.1416979732601824,
+      "grad_norm": 2.073671817779541,
+      "learning_rate": 1.4305033778996962e-05,
+      "loss": 0.7416,
+      "step": 1244500
+    },
+    {
+      "epoch": 2.1425584384965264,
+      "grad_norm": 2.143728017807007,
+      "learning_rate": 1.4290692691724564e-05,
+      "loss": 0.7441,
+      "step": 1245000
+    },
+    {
+      "epoch": 2.1434189037328704,
+      "grad_norm": 2.2988219261169434,
+      "learning_rate": 1.4276351604452163e-05,
+      "loss": 0.7457,
+      "step": 1245500
+    },
+    {
+      "epoch": 2.1442793689692143,
+      "grad_norm": 2.397913694381714,
+      "learning_rate": 1.4262010517179763e-05,
+      "loss": 0.7459,
+      "step": 1246000
+    },
+    {
+      "epoch": 2.1451398342055583,
+      "grad_norm": 2.177351474761963,
+      "learning_rate": 1.4247669429907362e-05,
+      "loss": 0.7409,
+      "step": 1246500
+    },
+    {
+      "epoch": 2.1460002994419023,
+      "grad_norm": 2.097346305847168,
+      "learning_rate": 1.4233328342634964e-05,
+      "loss": 0.737,
+      "step": 1247000
+    },
+    {
+      "epoch": 2.1468607646782463,
+      "grad_norm": 2.0776567459106445,
+      "learning_rate": 1.4218987255362565e-05,
+      "loss": 0.738,
+      "step": 1247500
+    },
+    {
+      "epoch": 2.1477212299145902,
+      "grad_norm": 2.189054250717163,
+      "learning_rate": 1.4204646168090163e-05,
+      "loss": 0.7446,
+      "step": 1248000
+    },
+    {
+      "epoch": 2.1485816951509342,
+      "grad_norm": 2.0094523429870605,
+      "learning_rate": 1.4190305080817764e-05,
+      "loss": 0.7351,
+      "step": 1248500
+    },
+    {
+      "epoch": 2.149442160387278,
+      "grad_norm": 2.0026373863220215,
+      "learning_rate": 1.4175963993545363e-05,
+      "loss": 0.7406,
+      "step": 1249000
+    },
+    {
+      "epoch": 2.150302625623622,
+      "grad_norm": 2.2804622650146484,
+      "learning_rate": 1.4161622906272965e-05,
+      "loss": 0.7421,
+      "step": 1249500
+    },
+    {
+      "epoch": 2.151163090859966,
+      "grad_norm": 2.289987325668335,
+      "learning_rate": 1.4147281819000565e-05,
+      "loss": 0.745,
+      "step": 1250000
+    },
+    {
+      "epoch": 2.15202355609631,
+      "grad_norm": 2.009991407394409,
+      "learning_rate": 1.4132940731728164e-05,
+      "loss": 0.7475,
+      "step": 1250500
+    },
+    {
+      "epoch": 2.152884021332654,
+      "grad_norm": 2.136585235595703,
+      "learning_rate": 1.4118599644455766e-05,
+      "loss": 0.7396,
+      "step": 1251000
+    },
+    {
+      "epoch": 2.153744486568998,
+      "grad_norm": 2.2174477577209473,
+      "learning_rate": 1.4104258557183363e-05,
+      "loss": 0.7403,
+      "step": 1251500
+    },
+    {
+      "epoch": 2.154604951805342,
+      "grad_norm": 2.1195108890533447,
+      "learning_rate": 1.4089917469910965e-05,
+      "loss": 0.7475,
+      "step": 1252000
+    },
+    {
+      "epoch": 2.155465417041686,
+      "grad_norm": 2.339773178100586,
+      "learning_rate": 1.4075576382638566e-05,
+      "loss": 0.7345,
+      "step": 1252500
+    },
+    {
+      "epoch": 2.15632588227803,
+      "grad_norm": 2.186708450317383,
+      "learning_rate": 1.4061235295366165e-05,
+      "loss": 0.7432,
+      "step": 1253000
+    },
+    {
+      "epoch": 2.157186347514374,
+      "grad_norm": 2.0798115730285645,
+      "learning_rate": 1.4046894208093767e-05,
+      "loss": 0.7439,
+      "step": 1253500
+    },
+    {
+      "epoch": 2.158046812750718,
+      "grad_norm": 2.256768226623535,
+      "learning_rate": 1.4032553120821366e-05,
+      "loss": 0.7389,
+      "step": 1254000
+    },
+    {
+      "epoch": 2.158907277987062,
+      "grad_norm": 2.122495174407959,
+      "learning_rate": 1.4018212033548966e-05,
+      "loss": 0.7401,
+      "step": 1254500
+    },
+    {
+      "epoch": 2.159767743223406,
+      "grad_norm": 2.2577829360961914,
+      "learning_rate": 1.4003870946276568e-05,
+      "loss": 0.7393,
+      "step": 1255000
+    },
+    {
+      "epoch": 2.16062820845975,
+      "grad_norm": 2.040308952331543,
+      "learning_rate": 1.3989529859004167e-05,
+      "loss": 0.7419,
+      "step": 1255500
+    },
+    {
+      "epoch": 2.161488673696094,
+      "grad_norm": 2.253164291381836,
+      "learning_rate": 1.3975188771731768e-05,
+      "loss": 0.7388,
+      "step": 1256000
+    },
+    {
+      "epoch": 2.162349138932438,
+      "grad_norm": 2.166598320007324,
+      "learning_rate": 1.396084768445937e-05,
+      "loss": 0.743,
+      "step": 1256500
+    },
+    {
+      "epoch": 2.163209604168782,
+      "grad_norm": 2.025723934173584,
+      "learning_rate": 1.3946506597186967e-05,
+      "loss": 0.7392,
+      "step": 1257000
+    },
+    {
+      "epoch": 2.164070069405126,
+      "grad_norm": 2.3602654933929443,
+      "learning_rate": 1.3932165509914569e-05,
+      "loss": 0.7418,
+      "step": 1257500
+    },
+    {
+      "epoch": 2.16493053464147,
+      "grad_norm": 2.2957653999328613,
+      "learning_rate": 1.3917824422642168e-05,
+      "loss": 0.741,
+      "step": 1258000
+    },
+    {
+      "epoch": 2.1657909998778138,
+      "grad_norm": 2.3943967819213867,
+      "learning_rate": 1.3903483335369768e-05,
+      "loss": 0.7481,
+      "step": 1258500
+    },
+    {
+      "epoch": 2.1666514651141577,
+      "grad_norm": 2.1326005458831787,
+      "learning_rate": 1.388914224809737e-05,
+      "loss": 0.7392,
+      "step": 1259000
+    },
+    {
+      "epoch": 2.1675119303505017,
+      "grad_norm": 2.3093254566192627,
+      "learning_rate": 1.3874801160824969e-05,
+      "loss": 0.7385,
+      "step": 1259500
+    },
+    {
+      "epoch": 2.1683723955868457,
+      "grad_norm": 2.1691677570343018,
+      "learning_rate": 1.386046007355257e-05,
+      "loss": 0.7347,
+      "step": 1260000
+    },
+    {
+      "epoch": 2.1692328608231897,
+      "grad_norm": 2.205759048461914,
+      "learning_rate": 1.3846118986280168e-05,
+      "loss": 0.737,
+      "step": 1260500
+    },
+    {
+      "epoch": 2.1700933260595336,
+      "grad_norm": 2.23650860786438,
+      "learning_rate": 1.3831777899007769e-05,
+      "loss": 0.7378,
+      "step": 1261000
+    },
+    {
+      "epoch": 2.1709537912958776,
+      "grad_norm": 2.0677170753479004,
+      "learning_rate": 1.3817436811735371e-05,
+      "loss": 0.7417,
+      "step": 1261500
+    },
+    {
+      "epoch": 2.171814256532222,
+      "grad_norm": 1.9268617630004883,
+      "learning_rate": 1.380309572446297e-05,
+      "loss": 0.7399,
+      "step": 1262000
+    },
+    {
+      "epoch": 2.172674721768566,
+      "grad_norm": 2.105142831802368,
+      "learning_rate": 1.378875463719057e-05,
+      "loss": 0.7337,
+      "step": 1262500
+    },
+    {
+      "epoch": 2.17353518700491,
+      "grad_norm": 2.2743546962738037,
+      "learning_rate": 1.3774413549918169e-05,
+      "loss": 0.7447,
+      "step": 1263000
+    },
+    {
+      "epoch": 2.174395652241254,
+      "grad_norm": 2.34230375289917,
+      "learning_rate": 1.3760072462645771e-05,
+      "loss": 0.7413,
+      "step": 1263500
+    },
+    {
+      "epoch": 2.175256117477598,
+      "grad_norm": 2.048245668411255,
+      "learning_rate": 1.3745731375373372e-05,
+      "loss": 0.7435,
+      "step": 1264000
+    },
+    {
+      "epoch": 2.176116582713942,
+      "grad_norm": 2.1016547679901123,
+      "learning_rate": 1.373139028810097e-05,
+      "loss": 0.7411,
+      "step": 1264500
+    },
+    {
+      "epoch": 2.176977047950286,
+      "grad_norm": 1.9493845701217651,
+      "learning_rate": 1.3717049200828571e-05,
+      "loss": 0.7403,
+      "step": 1265000
+    },
+    {
+      "epoch": 2.17783751318663,
+      "grad_norm": 2.23221755027771,
+      "learning_rate": 1.370270811355617e-05,
+      "loss": 0.7385,
+      "step": 1265500
+    },
+    {
+      "epoch": 2.178697978422974,
+      "grad_norm": 2.298320770263672,
+      "learning_rate": 1.3688367026283772e-05,
+      "loss": 0.7458,
+      "step": 1266000
+    },
+    {
+      "epoch": 2.179558443659318,
+      "grad_norm": 2.093395709991455,
+      "learning_rate": 1.3674025939011372e-05,
+      "loss": 0.7366,
+      "step": 1266500
+    },
+    {
+      "epoch": 2.180418908895662,
+      "grad_norm": 2.1902594566345215,
+      "learning_rate": 1.3659684851738971e-05,
+      "loss": 0.7397,
+      "step": 1267000
+    },
+    {
+      "epoch": 2.181279374132006,
+      "grad_norm": 2.1553637981414795,
+      "learning_rate": 1.3645343764466573e-05,
+      "loss": 0.7438,
+      "step": 1267500
+    },
+    {
+      "epoch": 2.1821398393683498,
+      "grad_norm": 2.153245449066162,
+      "learning_rate": 1.3631002677194174e-05,
+      "loss": 0.7415,
+      "step": 1268000
+    },
+    {
+      "epoch": 2.1830003046046937,
+      "grad_norm": 1.9978289604187012,
+      "learning_rate": 1.3616661589921772e-05,
+      "loss": 0.7356,
+      "step": 1268500
+    },
+    {
+      "epoch": 2.1838607698410377,
+      "grad_norm": 2.105203628540039,
+      "learning_rate": 1.3602320502649375e-05,
+      "loss": 0.7415,
+      "step": 1269000
+    },
+    {
+      "epoch": 2.1847212350773817,
+      "grad_norm": 2.2339985370635986,
+      "learning_rate": 1.3587979415376972e-05,
+      "loss": 0.7338,
+      "step": 1269500
+    },
+    {
+      "epoch": 2.1855817003137257,
+      "grad_norm": 2.240745782852173,
+      "learning_rate": 1.3573638328104574e-05,
+      "loss": 0.7433,
+      "step": 1270000
+    },
+    {
+      "epoch": 2.1864421655500696,
+      "grad_norm": 2.2821898460388184,
+      "learning_rate": 1.3559297240832174e-05,
+      "loss": 0.7429,
+      "step": 1270500
+    },
+    {
+      "epoch": 2.1873026307864136,
+      "grad_norm": 2.1928908824920654,
+      "learning_rate": 1.3544956153559773e-05,
+      "loss": 0.7424,
+      "step": 1271000
+    },
+    {
+      "epoch": 2.1881630960227576,
+      "grad_norm": 2.1943445205688477,
+      "learning_rate": 1.3530615066287375e-05,
+      "loss": 0.743,
+      "step": 1271500
+    },
+    {
+      "epoch": 2.1890235612591016,
+      "grad_norm": 2.0637905597686768,
+      "learning_rate": 1.3516273979014974e-05,
+      "loss": 0.7384,
+      "step": 1272000
+    },
+    {
+      "epoch": 2.1898840264954456,
+      "grad_norm": 2.1010494232177734,
+      "learning_rate": 1.3501932891742575e-05,
+      "loss": 0.7392,
+      "step": 1272500
+    },
+    {
+      "epoch": 2.1907444917317895,
+      "grad_norm": 1.9924651384353638,
+      "learning_rate": 1.3487591804470177e-05,
+      "loss": 0.7347,
+      "step": 1273000
+    },
+    {
+      "epoch": 2.1916049569681335,
+      "grad_norm": 2.502352476119995,
+      "learning_rate": 1.3473250717197774e-05,
+      "loss": 0.7418,
+      "step": 1273500
+    },
+    {
+      "epoch": 2.1924654222044775,
+      "grad_norm": 2.1891558170318604,
+      "learning_rate": 1.3458909629925376e-05,
+      "loss": 0.738,
+      "step": 1274000
+    },
+    {
+      "epoch": 2.1933258874408215,
+      "grad_norm": 2.170668601989746,
+      "learning_rate": 1.3444568542652975e-05,
+      "loss": 0.7384,
+      "step": 1274500
+    },
+    {
+      "epoch": 2.1941863526771654,
+      "grad_norm": 2.1300723552703857,
+      "learning_rate": 1.3430227455380575e-05,
+      "loss": 0.7431,
+      "step": 1275000
+    },
+    {
+      "epoch": 2.1950468179135094,
+      "grad_norm": 2.207932472229004,
+      "learning_rate": 1.3415886368108177e-05,
+      "loss": 0.7381,
+      "step": 1275500
+    },
+    {
+      "epoch": 2.1959072831498534,
+      "grad_norm": 2.3557584285736084,
+      "learning_rate": 1.3401545280835776e-05,
+      "loss": 0.7448,
+      "step": 1276000
+    },
+    {
+      "epoch": 2.1967677483861974,
+      "grad_norm": 2.1110777854919434,
+      "learning_rate": 1.3387204193563377e-05,
+      "loss": 0.7396,
+      "step": 1276500
+    },
+    {
+      "epoch": 2.1976282136225413,
+      "grad_norm": 2.0678913593292236,
+      "learning_rate": 1.3372863106290975e-05,
+      "loss": 0.7408,
+      "step": 1277000
+    },
+    {
+      "epoch": 2.1984886788588853,
+      "grad_norm": 2.3854904174804688,
+      "learning_rate": 1.3358522019018576e-05,
+      "loss": 0.7485,
+      "step": 1277500
+    },
+    {
+      "epoch": 2.1993491440952293,
+      "grad_norm": 2.1214091777801514,
+      "learning_rate": 1.3344180931746178e-05,
+      "loss": 0.7409,
+      "step": 1278000
+    },
+    {
+      "epoch": 2.2002096093315733,
+      "grad_norm": 2.177196741104126,
+      "learning_rate": 1.3329839844473777e-05,
+      "loss": 0.748,
+      "step": 1278500
+    },
+    {
+      "epoch": 2.2010700745679173,
+      "grad_norm": 2.102724552154541,
+      "learning_rate": 1.3315498757201377e-05,
+      "loss": 0.7416,
+      "step": 1279000
+    },
+    {
+      "epoch": 2.2019305398042612,
+      "grad_norm": 2.392582893371582,
+      "learning_rate": 1.330115766992898e-05,
+      "loss": 0.7376,
+      "step": 1279500
+    },
+    {
+      "epoch": 2.202791005040605,
+      "grad_norm": 2.1870553493499756,
+      "learning_rate": 1.3286816582656578e-05,
+      "loss": 0.7355,
+      "step": 1280000
+    },
+    {
+      "epoch": 2.203651470276949,
+      "grad_norm": 2.246654987335205,
+      "learning_rate": 1.3272475495384179e-05,
+      "loss": 0.7366,
+      "step": 1280500
+    },
+    {
+      "epoch": 2.204511935513293,
+      "grad_norm": 2.104764461517334,
+      "learning_rate": 1.3258134408111777e-05,
+      "loss": 0.7377,
+      "step": 1281000
+    },
+    {
+      "epoch": 2.205372400749637,
+      "grad_norm": 2.2333292961120605,
+      "learning_rate": 1.324379332083938e-05,
+      "loss": 0.7355,
+      "step": 1281500
+    },
+    {
+      "epoch": 2.206232865985981,
+      "grad_norm": 2.2619197368621826,
+      "learning_rate": 1.322945223356698e-05,
+      "loss": 0.7392,
+      "step": 1282000
+    },
+    {
+      "epoch": 2.207093331222325,
+      "grad_norm": 2.268139600753784,
+      "learning_rate": 1.3215111146294579e-05,
+      "loss": 0.7459,
+      "step": 1282500
+    },
+    {
+      "epoch": 2.207953796458669,
+      "grad_norm": 2.0993845462799072,
+      "learning_rate": 1.320077005902218e-05,
+      "loss": 0.7407,
+      "step": 1283000
+    },
+    {
+      "epoch": 2.2088142616950135,
+      "grad_norm": 2.396580219268799,
+      "learning_rate": 1.3186428971749778e-05,
+      "loss": 0.7344,
+      "step": 1283500
+    },
+    {
+      "epoch": 2.2096747269313575,
+      "grad_norm": 2.1309633255004883,
+      "learning_rate": 1.317208788447738e-05,
+      "loss": 0.736,
+      "step": 1284000
+    },
+    {
+      "epoch": 2.2105351921677014,
+      "grad_norm": 2.0994322299957275,
+      "learning_rate": 1.315774679720498e-05,
+      "loss": 0.7407,
+      "step": 1284500
+    },
+    {
+      "epoch": 2.2113956574040454,
+      "grad_norm": 2.230180025100708,
+      "learning_rate": 1.314340570993258e-05,
+      "loss": 0.7395,
+      "step": 1285000
+    },
+    {
+      "epoch": 2.2122561226403894,
+      "grad_norm": 2.205920457839966,
+      "learning_rate": 1.3129064622660182e-05,
+      "loss": 0.7408,
+      "step": 1285500
+    },
+    {
+      "epoch": 2.2131165878767334,
+      "grad_norm": 2.1964640617370605,
+      "learning_rate": 1.3114723535387779e-05,
+      "loss": 0.7378,
+      "step": 1286000
+    },
+    {
+      "epoch": 2.2139770531130774,
+      "grad_norm": 2.0684380531311035,
+      "learning_rate": 1.3100382448115381e-05,
+      "loss": 0.7337,
+      "step": 1286500
+    },
+    {
+      "epoch": 2.2148375183494213,
+      "grad_norm": 2.1438052654266357,
+      "learning_rate": 1.3086041360842981e-05,
+      "loss": 0.735,
+      "step": 1287000
+    },
+    {
+      "epoch": 2.2156979835857653,
+      "grad_norm": 2.076854944229126,
+      "learning_rate": 1.307170027357058e-05,
+      "loss": 0.743,
+      "step": 1287500
+    },
+    {
+      "epoch": 2.2165584488221093,
+      "grad_norm": 2.2474782466888428,
+      "learning_rate": 1.3057359186298182e-05,
+      "loss": 0.7349,
+      "step": 1288000
+    },
+    {
+      "epoch": 2.2174189140584533,
+      "grad_norm": 2.067883253097534,
+      "learning_rate": 1.3043018099025781e-05,
+      "loss": 0.7308,
+      "step": 1288500
+    },
+    {
+      "epoch": 2.2182793792947972,
+      "grad_norm": 2.1735494136810303,
+      "learning_rate": 1.3028677011753382e-05,
+      "loss": 0.7308,
+      "step": 1289000
+    },
+    {
+      "epoch": 2.219139844531141,
+      "grad_norm": 1.9676673412322998,
+      "learning_rate": 1.3014335924480984e-05,
+      "loss": 0.7366,
+      "step": 1289500
+    },
+    {
+      "epoch": 2.220000309767485,
+      "grad_norm": 2.2277417182922363,
+      "learning_rate": 1.2999994837208581e-05,
+      "loss": 0.7395,
+      "step": 1290000
+    },
+    {
+      "epoch": 2.220860775003829,
+      "grad_norm": 2.261632204055786,
+      "learning_rate": 1.2985653749936183e-05,
+      "loss": 0.738,
+      "step": 1290500
+    },
+    {
+      "epoch": 2.221721240240173,
+      "grad_norm": 2.178764820098877,
+      "learning_rate": 1.2971312662663783e-05,
+      "loss": 0.7406,
+      "step": 1291000
+    },
+    {
+      "epoch": 2.222581705476517,
+      "grad_norm": 2.2707815170288086,
+      "learning_rate": 1.2956971575391382e-05,
+      "loss": 0.735,
+      "step": 1291500
+    },
+    {
+      "epoch": 2.223442170712861,
+      "grad_norm": 2.0521087646484375,
+      "learning_rate": 1.2942630488118984e-05,
+      "loss": 0.7375,
+      "step": 1292000
+    },
+    {
+      "epoch": 2.224302635949205,
+      "grad_norm": 2.2074644565582275,
+      "learning_rate": 1.2928289400846583e-05,
+      "loss": 0.7376,
+      "step": 1292500
+    },
+    {
+      "epoch": 2.225163101185549,
+      "grad_norm": 2.157623291015625,
+      "learning_rate": 1.2913948313574184e-05,
+      "loss": 0.7285,
+      "step": 1293000
+    },
+    {
+      "epoch": 2.226023566421893,
+      "grad_norm": 2.132406711578369,
+      "learning_rate": 1.2899607226301786e-05,
+      "loss": 0.7356,
+      "step": 1293500
+    },
+    {
+      "epoch": 2.226884031658237,
+      "grad_norm": 2.0070013999938965,
+      "learning_rate": 1.2885266139029385e-05,
+      "loss": 0.7315,
+      "step": 1294000
+    },
+    {
+      "epoch": 2.227744496894581,
+      "grad_norm": 2.2132139205932617,
+      "learning_rate": 1.2870925051756985e-05,
+      "loss": 0.7411,
+      "step": 1294500
+    },
+    {
+      "epoch": 2.228604962130925,
+      "grad_norm": 2.1949164867401123,
+      "learning_rate": 1.2856583964484584e-05,
+      "loss": 0.7316,
+      "step": 1295000
+    },
+    {
+      "epoch": 2.229465427367269,
+      "grad_norm": 2.2194552421569824,
+      "learning_rate": 1.2842242877212184e-05,
+      "loss": 0.7381,
+      "step": 1295500
+    },
+    {
+      "epoch": 2.230325892603613,
+      "grad_norm": 2.1448280811309814,
+      "learning_rate": 1.2827901789939787e-05,
+      "loss": 0.7388,
+      "step": 1296000
+    },
+    {
+      "epoch": 2.231186357839957,
+      "grad_norm": 2.047916889190674,
+      "learning_rate": 1.2813560702667385e-05,
+      "loss": 0.7358,
+      "step": 1296500
+    },
+    {
+      "epoch": 2.232046823076301,
+      "grad_norm": 2.1904945373535156,
+      "learning_rate": 1.2799219615394986e-05,
+      "loss": 0.7396,
+      "step": 1297000
+    },
+    {
+      "epoch": 2.232907288312645,
+      "grad_norm": 2.256561040878296,
+      "learning_rate": 1.2784878528122585e-05,
+      "loss": 0.7396,
+      "step": 1297500
+    },
+    {
+      "epoch": 2.233767753548989,
+      "grad_norm": 2.319228172302246,
+      "learning_rate": 1.2770537440850187e-05,
+      "loss": 0.7398,
+      "step": 1298000
+    },
+    {
+      "epoch": 2.234628218785333,
+      "grad_norm": 2.2488300800323486,
+      "learning_rate": 1.2756196353577787e-05,
+      "loss": 0.7365,
+      "step": 1298500
+    },
+    {
+      "epoch": 2.2354886840216768,
+      "grad_norm": 2.3413503170013428,
+      "learning_rate": 1.2741855266305386e-05,
+      "loss": 0.7369,
+      "step": 1299000
+    },
+    {
+      "epoch": 2.2363491492580208,
+      "grad_norm": 2.1444482803344727,
+      "learning_rate": 1.2727514179032986e-05,
+      "loss": 0.7332,
+      "step": 1299500
+    },
+    {
+      "epoch": 2.2372096144943647,
+      "grad_norm": 1.9140009880065918,
+      "learning_rate": 1.2713173091760585e-05,
+      "loss": 0.738,
+      "step": 1300000
+    },
+    {
+      "epoch": 2.2380700797307087,
+      "grad_norm": 1.9779835939407349,
+      "learning_rate": 1.2698832004488187e-05,
+      "loss": 0.7373,
+      "step": 1300500
+    },
+    {
+      "epoch": 2.2389305449670527,
+      "grad_norm": 2.290299415588379,
+      "learning_rate": 1.2684490917215788e-05,
+      "loss": 0.7372,
+      "step": 1301000
+    },
+    {
+      "epoch": 2.2397910102033967,
+      "grad_norm": 2.2834126949310303,
+      "learning_rate": 1.2670149829943387e-05,
+      "loss": 0.7294,
+      "step": 1301500
+    },
+    {
+      "epoch": 2.2406514754397406,
+      "grad_norm": 2.0117619037628174,
+      "learning_rate": 1.2655808742670989e-05,
+      "loss": 0.7354,
+      "step": 1302000
+    },
+    {
+      "epoch": 2.2415119406760846,
+      "grad_norm": 2.1566827297210693,
+      "learning_rate": 1.264146765539859e-05,
+      "loss": 0.734,
+      "step": 1302500
+    },
+    {
+      "epoch": 2.2423724059124286,
+      "grad_norm": 1.8935664892196655,
+      "learning_rate": 1.2627126568126188e-05,
+      "loss": 0.7403,
+      "step": 1303000
+    },
+    {
+      "epoch": 2.2432328711487726,
+      "grad_norm": 2.1429378986358643,
+      "learning_rate": 1.2612785480853788e-05,
+      "loss": 0.7323,
+      "step": 1303500
+    },
+    {
+      "epoch": 2.2440933363851165,
+      "grad_norm": 2.3545150756835938,
+      "learning_rate": 1.2598444393581387e-05,
+      "loss": 0.7439,
+      "step": 1304000
+    },
+    {
+      "epoch": 2.2449538016214605,
+      "grad_norm": 2.245781660079956,
+      "learning_rate": 1.258410330630899e-05,
+      "loss": 0.7394,
+      "step": 1304500
+    },
+    {
+      "epoch": 2.2458142668578045,
+      "grad_norm": 2.2469215393066406,
+      "learning_rate": 1.256976221903659e-05,
+      "loss": 0.7403,
+      "step": 1305000
+    },
+    {
+      "epoch": 2.2466747320941485,
+      "grad_norm": 2.2666895389556885,
+      "learning_rate": 1.2555421131764189e-05,
+      "loss": 0.7347,
+      "step": 1305500
+    },
+    {
+      "epoch": 2.2475351973304925,
+      "grad_norm": 2.2507004737854004,
+      "learning_rate": 1.254108004449179e-05,
+      "loss": 0.7326,
+      "step": 1306000
+    },
+    {
+      "epoch": 2.2483956625668364,
+      "grad_norm": 2.1795663833618164,
+      "learning_rate": 1.252673895721939e-05,
+      "loss": 0.7323,
+      "step": 1306500
+    },
+    {
+      "epoch": 2.2492561278031804,
+      "grad_norm": 2.247745990753174,
+      "learning_rate": 1.251239786994699e-05,
+      "loss": 0.7365,
+      "step": 1307000
+    },
+    {
+      "epoch": 2.2501165930395244,
+      "grad_norm": 2.1461105346679688,
+      "learning_rate": 1.249805678267459e-05,
+      "loss": 0.7374,
+      "step": 1307500
+    },
+    {
+      "epoch": 2.2509770582758684,
+      "grad_norm": 1.8984283208847046,
+      "learning_rate": 1.248371569540219e-05,
+      "loss": 0.7335,
+      "step": 1308000
+    },
+    {
+      "epoch": 2.2518375235122123,
+      "grad_norm": 2.113375186920166,
+      "learning_rate": 1.2469374608129791e-05,
+      "loss": 0.7356,
+      "step": 1308500
+    },
+    {
+      "epoch": 2.2526979887485568,
+      "grad_norm": 2.187218427658081,
+      "learning_rate": 1.2455033520857392e-05,
+      "loss": 0.7368,
+      "step": 1309000
+    },
+    {
+      "epoch": 2.2535584539849007,
+      "grad_norm": 2.347452402114868,
+      "learning_rate": 1.244069243358499e-05,
+      "loss": 0.7382,
+      "step": 1309500
+    },
+    {
+      "epoch": 2.2544189192212447,
+      "grad_norm": 2.2602875232696533,
+      "learning_rate": 1.2426351346312591e-05,
+      "loss": 0.7421,
+      "step": 1310000
+    },
+    {
+      "epoch": 2.2552793844575887,
+      "grad_norm": 2.1275899410247803,
+      "learning_rate": 1.2412010259040192e-05,
+      "loss": 0.7339,
+      "step": 1310500
+    },
+    {
+      "epoch": 2.2561398496939327,
+      "grad_norm": 2.11209774017334,
+      "learning_rate": 1.2397669171767792e-05,
+      "loss": 0.7365,
+      "step": 1311000
+    },
+    {
+      "epoch": 2.2570003149302766,
+      "grad_norm": 2.1969621181488037,
+      "learning_rate": 1.2383328084495393e-05,
+      "loss": 0.739,
+      "step": 1311500
+    },
+    {
+      "epoch": 2.2578607801666206,
+      "grad_norm": 2.293888807296753,
+      "learning_rate": 1.2368986997222991e-05,
+      "loss": 0.7302,
+      "step": 1312000
+    },
+    {
+      "epoch": 2.2587212454029646,
+      "grad_norm": 2.102910280227661,
+      "learning_rate": 1.2354645909950592e-05,
+      "loss": 0.7342,
+      "step": 1312500
+    },
+    {
+      "epoch": 2.2595817106393086,
+      "grad_norm": 2.209803342819214,
+      "learning_rate": 1.2340304822678194e-05,
+      "loss": 0.7357,
+      "step": 1313000
+    },
+    {
+      "epoch": 2.2604421758756525,
+      "grad_norm": 2.1391923427581787,
+      "learning_rate": 1.2325963735405793e-05,
+      "loss": 0.7366,
+      "step": 1313500
+    },
+    {
+      "epoch": 2.2613026411119965,
+      "grad_norm": 2.047585964202881,
+      "learning_rate": 1.2311622648133393e-05,
+      "loss": 0.7455,
+      "step": 1314000
+    },
+    {
+      "epoch": 2.2621631063483405,
+      "grad_norm": 2.154750347137451,
+      "learning_rate": 1.2297281560860994e-05,
+      "loss": 0.7318,
+      "step": 1314500
+    },
+    {
+      "epoch": 2.2630235715846845,
+      "grad_norm": 2.435215473175049,
+      "learning_rate": 1.2282940473588594e-05,
+      "loss": 0.7387,
+      "step": 1315000
+    },
+    {
+      "epoch": 2.2638840368210285,
+      "grad_norm": 2.0441153049468994,
+      "learning_rate": 1.2268599386316195e-05,
+      "loss": 0.7321,
+      "step": 1315500
+    },
+    {
+      "epoch": 2.2647445020573724,
+      "grad_norm": 2.128629446029663,
+      "learning_rate": 1.2254258299043793e-05,
+      "loss": 0.7315,
+      "step": 1316000
+    },
+    {
+      "epoch": 2.2656049672937164,
+      "grad_norm": 2.077057123184204,
+      "learning_rate": 1.2239917211771394e-05,
+      "loss": 0.7375,
+      "step": 1316500
+    },
+    {
+      "epoch": 2.2664654325300604,
+      "grad_norm": 2.0758187770843506,
+      "learning_rate": 1.2225576124498994e-05,
+      "loss": 0.735,
+      "step": 1317000
+    },
+    {
+      "epoch": 2.2673258977664044,
+      "grad_norm": 1.947325587272644,
+      "learning_rate": 1.2211235037226595e-05,
+      "loss": 0.7337,
+      "step": 1317500
+    },
+    {
+      "epoch": 2.2681863630027483,
+      "grad_norm": 2.228104591369629,
+      "learning_rate": 1.2196893949954195e-05,
+      "loss": 0.7334,
+      "step": 1318000
+    },
+    {
+      "epoch": 2.2690468282390923,
+      "grad_norm": 2.0669751167297363,
+      "learning_rate": 1.2182552862681796e-05,
+      "loss": 0.7388,
+      "step": 1318500
+    },
+    {
+      "epoch": 2.2699072934754363,
+      "grad_norm": 2.1029419898986816,
+      "learning_rate": 1.2168211775409395e-05,
+      "loss": 0.7338,
+      "step": 1319000
+    },
+    {
+      "epoch": 2.2707677587117803,
+      "grad_norm": 2.1721904277801514,
+      "learning_rate": 1.2153870688136995e-05,
+      "loss": 0.7384,
+      "step": 1319500
+    },
+    {
+      "epoch": 2.2716282239481242,
+      "grad_norm": 2.2310101985931396,
+      "learning_rate": 1.2139529600864597e-05,
+      "loss": 0.7364,
+      "step": 1320000
+    },
+    {
+      "epoch": 2.2724886891844682,
+      "grad_norm": 2.108856201171875,
+      "learning_rate": 1.2125188513592196e-05,
+      "loss": 0.7382,
+      "step": 1320500
+    },
+    {
+      "epoch": 2.273349154420812,
+      "grad_norm": 2.3644378185272217,
+      "learning_rate": 1.2110847426319796e-05,
+      "loss": 0.7382,
+      "step": 1321000
+    },
+    {
+      "epoch": 2.274209619657156,
+      "grad_norm": 2.2770323753356934,
+      "learning_rate": 1.2096506339047397e-05,
+      "loss": 0.7339,
+      "step": 1321500
+    },
+    {
+      "epoch": 2.2750700848935,
+      "grad_norm": 2.357496738433838,
+      "learning_rate": 1.2082165251774997e-05,
+      "loss": 0.7354,
+      "step": 1322000
+    },
+    {
+      "epoch": 2.275930550129844,
+      "grad_norm": 1.9793803691864014,
+      "learning_rate": 1.2067824164502598e-05,
+      "loss": 0.7329,
+      "step": 1322500
+    },
+    {
+      "epoch": 2.276791015366188,
+      "grad_norm": 2.253159284591675,
+      "learning_rate": 1.2053483077230197e-05,
+      "loss": 0.7369,
+      "step": 1323000
+    },
+    {
+      "epoch": 2.277651480602532,
+      "grad_norm": 2.310882091522217,
+      "learning_rate": 1.2039141989957797e-05,
+      "loss": 0.7387,
+      "step": 1323500
+    },
+    {
+      "epoch": 2.278511945838876,
+      "grad_norm": 2.1455655097961426,
+      "learning_rate": 1.2024800902685398e-05,
+      "loss": 0.7387,
+      "step": 1324000
+    },
+    {
+      "epoch": 2.27937241107522,
+      "grad_norm": 2.1674752235412598,
+      "learning_rate": 1.2010459815412998e-05,
+      "loss": 0.7344,
+      "step": 1324500
+    },
+    {
+      "epoch": 2.280232876311564,
+      "grad_norm": 2.193643808364868,
+      "learning_rate": 1.1996118728140599e-05,
+      "loss": 0.7328,
+      "step": 1325000
+    },
+    {
+      "epoch": 2.281093341547908,
+      "grad_norm": 2.1150364875793457,
+      "learning_rate": 1.1981777640868199e-05,
+      "loss": 0.7379,
+      "step": 1325500
+    },
+    {
+      "epoch": 2.281953806784252,
+      "grad_norm": 2.2378151416778564,
+      "learning_rate": 1.1967436553595798e-05,
+      "loss": 0.7305,
+      "step": 1326000
+    },
+    {
+      "epoch": 2.282814272020596,
+      "grad_norm": 2.3178298473358154,
+      "learning_rate": 1.19530954663234e-05,
+      "loss": 0.7361,
+      "step": 1326500
+    },
+    {
+      "epoch": 2.28367473725694,
+      "grad_norm": 2.16298770904541,
+      "learning_rate": 1.1938754379050999e-05,
+      "loss": 0.7339,
+      "step": 1327000
+    },
+    {
+      "epoch": 2.284535202493284,
+      "grad_norm": 2.0893654823303223,
+      "learning_rate": 1.19244132917786e-05,
+      "loss": 0.7347,
+      "step": 1327500
+    },
+    {
+      "epoch": 2.285395667729628,
+      "grad_norm": 2.0143134593963623,
+      "learning_rate": 1.19100722045062e-05,
+      "loss": 0.7375,
+      "step": 1328000
+    },
+    {
+      "epoch": 2.286256132965972,
+      "grad_norm": 2.1765668392181396,
+      "learning_rate": 1.1895731117233798e-05,
+      "loss": 0.7374,
+      "step": 1328500
+    },
+    {
+      "epoch": 2.2871165982023163,
+      "grad_norm": 2.2966361045837402,
+      "learning_rate": 1.18813900299614e-05,
+      "loss": 0.74,
+      "step": 1329000
+    },
+    {
+      "epoch": 2.2879770634386603,
+      "grad_norm": 2.144141674041748,
+      "learning_rate": 1.1867048942689001e-05,
+      "loss": 0.7347,
+      "step": 1329500
+    },
+    {
+      "epoch": 2.2888375286750042,
+      "grad_norm": 2.3457894325256348,
+      "learning_rate": 1.18527078554166e-05,
+      "loss": 0.7373,
+      "step": 1330000
+    },
+    {
+      "epoch": 2.289697993911348,
+      "grad_norm": 2.230165958404541,
+      "learning_rate": 1.18383667681442e-05,
+      "loss": 0.7328,
+      "step": 1330500
+    },
+    {
+      "epoch": 2.290558459147692,
+      "grad_norm": 2.159299373626709,
+      "learning_rate": 1.1824025680871802e-05,
+      "loss": 0.7313,
+      "step": 1331000
+    },
+    {
+      "epoch": 2.291418924384036,
+      "grad_norm": 2.08553409576416,
+      "learning_rate": 1.1809684593599401e-05,
+      "loss": 0.7438,
+      "step": 1331500
+    },
+    {
+      "epoch": 2.29227938962038,
+      "grad_norm": 2.318000078201294,
+      "learning_rate": 1.1795343506327002e-05,
+      "loss": 0.7357,
+      "step": 1332000
+    },
+    {
+      "epoch": 2.293139854856724,
+      "grad_norm": 2.38122296333313,
+      "learning_rate": 1.1781002419054602e-05,
+      "loss": 0.7385,
+      "step": 1332500
+    },
+    {
+      "epoch": 2.294000320093068,
+      "grad_norm": 2.157275915145874,
+      "learning_rate": 1.1766661331782201e-05,
+      "loss": 0.7319,
+      "step": 1333000
+    },
+    {
+      "epoch": 2.294860785329412,
+      "grad_norm": 2.2765510082244873,
+      "learning_rate": 1.1752320244509803e-05,
+      "loss": 0.7366,
+      "step": 1333500
+    },
+    {
+      "epoch": 2.295721250565756,
+      "grad_norm": 2.6490728855133057,
+      "learning_rate": 1.1737979157237402e-05,
+      "loss": 0.7373,
+      "step": 1334000
+    },
+    {
+      "epoch": 2.2965817158021,
+      "grad_norm": 2.020374298095703,
+      "learning_rate": 1.1723638069965002e-05,
+      "loss": 0.7359,
+      "step": 1334500
+    },
+    {
+      "epoch": 2.297442181038444,
+      "grad_norm": 1.8266252279281616,
+      "learning_rate": 1.1709296982692603e-05,
+      "loss": 0.7316,
+      "step": 1335000
+    },
+    {
+      "epoch": 2.298302646274788,
+      "grad_norm": 2.322174072265625,
+      "learning_rate": 1.1694955895420202e-05,
+      "loss": 0.7333,
+      "step": 1335500
+    },
+    {
+      "epoch": 2.299163111511132,
+      "grad_norm": 2.295354127883911,
+      "learning_rate": 1.1680614808147804e-05,
+      "loss": 0.7354,
+      "step": 1336000
+    },
+    {
+      "epoch": 2.300023576747476,
+      "grad_norm": 2.2270147800445557,
+      "learning_rate": 1.1666273720875404e-05,
+      "loss": 0.7409,
+      "step": 1336500
+    },
+    {
+      "epoch": 2.30088404198382,
+      "grad_norm": 2.2788033485412598,
+      "learning_rate": 1.1651932633603003e-05,
+      "loss": 0.7389,
+      "step": 1337000
+    },
+    {
+      "epoch": 2.301744507220164,
+      "grad_norm": 2.205354690551758,
+      "learning_rate": 1.1637591546330604e-05,
+      "loss": 0.7338,
+      "step": 1337500
+    },
+    {
+      "epoch": 2.302604972456508,
+      "grad_norm": 2.1826252937316895,
+      "learning_rate": 1.1623250459058204e-05,
+      "loss": 0.7362,
+      "step": 1338000
+    },
+    {
+      "epoch": 2.303465437692852,
+      "grad_norm": 2.2182676792144775,
+      "learning_rate": 1.1608909371785804e-05,
+      "loss": 0.7265,
+      "step": 1338500
+    },
+    {
+      "epoch": 2.304325902929196,
+      "grad_norm": 2.089738130569458,
+      "learning_rate": 1.1594568284513405e-05,
+      "loss": 0.7387,
+      "step": 1339000
+    },
+    {
+      "epoch": 2.30518636816554,
+      "grad_norm": 2.235210418701172,
+      "learning_rate": 1.1580227197241004e-05,
+      "loss": 0.7343,
+      "step": 1339500
+    },
+    {
+      "epoch": 2.3060468334018838,
+      "grad_norm": 2.5968661308288574,
+      "learning_rate": 1.1565886109968604e-05,
+      "loss": 0.7356,
+      "step": 1340000
+    },
+    {
+      "epoch": 2.3069072986382277,
+      "grad_norm": 2.1616084575653076,
+      "learning_rate": 1.1551545022696206e-05,
+      "loss": 0.7287,
+      "step": 1340500
+    },
+    {
+      "epoch": 2.3077677638745717,
+      "grad_norm": 2.295888662338257,
+      "learning_rate": 1.1537203935423805e-05,
+      "loss": 0.7356,
+      "step": 1341000
+    },
+    {
+      "epoch": 2.3086282291109157,
+      "grad_norm": 2.2833211421966553,
+      "learning_rate": 1.1522862848151406e-05,
+      "loss": 0.7337,
+      "step": 1341500
+    },
+    {
+      "epoch": 2.3094886943472597,
+      "grad_norm": 2.069758415222168,
+      "learning_rate": 1.1508521760879006e-05,
+      "loss": 0.7403,
+      "step": 1342000
+    },
+    {
+      "epoch": 2.3103491595836037,
+      "grad_norm": 2.1476922035217285,
+      "learning_rate": 1.1494180673606607e-05,
+      "loss": 0.7368,
+      "step": 1342500
+    },
+    {
+      "epoch": 2.3112096248199476,
+      "grad_norm": 2.3149638175964355,
+      "learning_rate": 1.1479839586334207e-05,
+      "loss": 0.7375,
+      "step": 1343000
+    },
+    {
+      "epoch": 2.3120700900562916,
+      "grad_norm": 2.2947609424591064,
+      "learning_rate": 1.1465498499061807e-05,
+      "loss": 0.7343,
+      "step": 1343500
+    },
+    {
+      "epoch": 2.3129305552926356,
+      "grad_norm": 2.098161458969116,
+      "learning_rate": 1.1451157411789406e-05,
+      "loss": 0.7311,
+      "step": 1344000
+    },
+    {
+      "epoch": 2.3137910205289796,
+      "grad_norm": 2.2303125858306885,
+      "learning_rate": 1.1436816324517007e-05,
+      "loss": 0.7318,
+      "step": 1344500
+    },
+    {
+      "epoch": 2.3146514857653235,
+      "grad_norm": 2.2480945587158203,
+      "learning_rate": 1.1422475237244607e-05,
+      "loss": 0.7318,
+      "step": 1345000
+    },
+    {
+      "epoch": 2.3155119510016675,
+      "grad_norm": 2.3394484519958496,
+      "learning_rate": 1.1408134149972208e-05,
+      "loss": 0.7283,
+      "step": 1345500
+    },
+    {
+      "epoch": 2.3163724162380115,
+      "grad_norm": 2.292473316192627,
+      "learning_rate": 1.1393793062699808e-05,
+      "loss": 0.735,
+      "step": 1346000
+    },
+    {
+      "epoch": 2.3172328814743555,
+      "grad_norm": 2.0187926292419434,
+      "learning_rate": 1.1379451975427407e-05,
+      "loss": 0.7329,
+      "step": 1346500
+    },
+    {
+      "epoch": 2.3180933467106994,
+      "grad_norm": 2.197573661804199,
+      "learning_rate": 1.1365110888155007e-05,
+      "loss": 0.7372,
+      "step": 1347000
+    },
+    {
+      "epoch": 2.3189538119470434,
+      "grad_norm": 2.140190601348877,
+      "learning_rate": 1.135076980088261e-05,
+      "loss": 0.7338,
+      "step": 1347500
+    },
+    {
+      "epoch": 2.3198142771833874,
+      "grad_norm": 2.119563102722168,
+      "learning_rate": 1.1336428713610208e-05,
+      "loss": 0.7285,
+      "step": 1348000
+    },
+    {
+      "epoch": 2.3206747424197314,
+      "grad_norm": 2.1876964569091797,
+      "learning_rate": 1.1322087626337809e-05,
+      "loss": 0.7345,
+      "step": 1348500
+    },
+    {
+      "epoch": 2.3215352076560754,
+      "grad_norm": 2.1912875175476074,
+      "learning_rate": 1.130774653906541e-05,
+      "loss": 0.7311,
+      "step": 1349000
+    },
+    {
+      "epoch": 2.3223956728924193,
+      "grad_norm": 2.2343170642852783,
+      "learning_rate": 1.129340545179301e-05,
+      "loss": 0.7377,
+      "step": 1349500
+    },
+    {
+      "epoch": 2.3232561381287633,
+      "grad_norm": 1.9957581758499146,
+      "learning_rate": 1.127906436452061e-05,
+      "loss": 0.7314,
+      "step": 1350000
+    },
+    {
+      "epoch": 2.3241166033651073,
+      "grad_norm": 2.154148578643799,
+      "learning_rate": 1.1264723277248209e-05,
+      "loss": 0.7268,
+      "step": 1350500
+    },
+    {
+      "epoch": 2.3249770686014513,
+      "grad_norm": 2.0333240032196045,
+      "learning_rate": 1.125038218997581e-05,
+      "loss": 0.7324,
+      "step": 1351000
+    },
+    {
+      "epoch": 2.3258375338377952,
+      "grad_norm": 1.997631311416626,
+      "learning_rate": 1.123604110270341e-05,
+      "loss": 0.7314,
+      "step": 1351500
+    },
+    {
+      "epoch": 2.326697999074139,
+      "grad_norm": 2.216383934020996,
+      "learning_rate": 1.122170001543101e-05,
+      "loss": 0.7327,
+      "step": 1352000
+    },
+    {
+      "epoch": 2.327558464310483,
+      "grad_norm": 2.4634883403778076,
+      "learning_rate": 1.1207358928158611e-05,
+      "loss": 0.7299,
+      "step": 1352500
+    },
+    {
+      "epoch": 2.328418929546827,
+      "grad_norm": 1.891774296760559,
+      "learning_rate": 1.1193017840886211e-05,
+      "loss": 0.7339,
+      "step": 1353000
+    },
+    {
+      "epoch": 2.329279394783171,
+      "grad_norm": 2.076943874359131,
+      "learning_rate": 1.117867675361381e-05,
+      "loss": 0.7351,
+      "step": 1353500
+    },
+    {
+      "epoch": 2.330139860019515,
+      "grad_norm": 2.08978533744812,
+      "learning_rate": 1.1164335666341412e-05,
+      "loss": 0.7357,
+      "step": 1354000
+    },
+    {
+      "epoch": 2.331000325255859,
+      "grad_norm": 2.38840913772583,
+      "learning_rate": 1.1149994579069011e-05,
+      "loss": 0.7343,
+      "step": 1354500
+    },
+    {
+      "epoch": 2.3318607904922035,
+      "grad_norm": 1.997729778289795,
+      "learning_rate": 1.1135653491796612e-05,
+      "loss": 0.7279,
+      "step": 1355000
+    },
+    {
+      "epoch": 2.3327212557285475,
+      "grad_norm": 2.2807769775390625,
+      "learning_rate": 1.1121312404524212e-05,
+      "loss": 0.7297,
+      "step": 1355500
+    },
+    {
+      "epoch": 2.3335817209648915,
+      "grad_norm": 2.291337251663208,
+      "learning_rate": 1.1106971317251812e-05,
+      "loss": 0.7373,
+      "step": 1356000
+    },
+    {
+      "epoch": 2.3344421862012354,
+      "grad_norm": 2.2745909690856934,
+      "learning_rate": 1.1092630229979413e-05,
+      "loss": 0.7381,
+      "step": 1356500
+    },
+    {
+      "epoch": 2.3353026514375794,
+      "grad_norm": 2.1283040046691895,
+      "learning_rate": 1.1078289142707013e-05,
+      "loss": 0.7277,
+      "step": 1357000
+    },
+    {
+      "epoch": 2.3361631166739234,
+      "grad_norm": 2.2192940711975098,
+      "learning_rate": 1.1063948055434612e-05,
+      "loss": 0.731,
+      "step": 1357500
+    },
+    {
+      "epoch": 2.3370235819102674,
+      "grad_norm": 2.1145095825195312,
+      "learning_rate": 1.1049606968162213e-05,
+      "loss": 0.7297,
+      "step": 1358000
+    },
+    {
+      "epoch": 2.3378840471466114,
+      "grad_norm": 2.089151382446289,
+      "learning_rate": 1.1035265880889815e-05,
+      "loss": 0.7357,
+      "step": 1358500
+    },
+    {
+      "epoch": 2.3387445123829553,
+      "grad_norm": 2.203192710876465,
+      "learning_rate": 1.1020924793617414e-05,
+      "loss": 0.7269,
+      "step": 1359000
+    },
+    {
+      "epoch": 2.3396049776192993,
+      "grad_norm": 2.030491352081299,
+      "learning_rate": 1.1006583706345014e-05,
+      "loss": 0.7308,
+      "step": 1359500
+    },
+    {
+      "epoch": 2.3404654428556433,
+      "grad_norm": 2.259591817855835,
+      "learning_rate": 1.0992242619072615e-05,
+      "loss": 0.7325,
+      "step": 1360000
+    },
+    {
+      "epoch": 2.3413259080919873,
+      "grad_norm": 2.142476797103882,
+      "learning_rate": 1.0977901531800213e-05,
+      "loss": 0.7313,
+      "step": 1360500
+    },
+    {
+      "epoch": 2.3421863733283312,
+      "grad_norm": 2.402512311935425,
+      "learning_rate": 1.0963560444527815e-05,
+      "loss": 0.7338,
+      "step": 1361000
+    },
+    {
+      "epoch": 2.343046838564675,
+      "grad_norm": 2.097332239151001,
+      "learning_rate": 1.0949219357255414e-05,
+      "loss": 0.7295,
+      "step": 1361500
+    },
+    {
+      "epoch": 2.343907303801019,
+      "grad_norm": 2.12955904006958,
+      "learning_rate": 1.0934878269983015e-05,
+      "loss": 0.7269,
+      "step": 1362000
+    },
+    {
+      "epoch": 2.344767769037363,
+      "grad_norm": 2.1505727767944336,
+      "learning_rate": 1.0920537182710615e-05,
+      "loss": 0.7332,
+      "step": 1362500
+    },
+    {
+      "epoch": 2.345628234273707,
+      "grad_norm": 2.206063985824585,
+      "learning_rate": 1.0906196095438214e-05,
+      "loss": 0.7299,
+      "step": 1363000
+    },
+    {
+      "epoch": 2.346488699510051,
+      "grad_norm": 2.3586273193359375,
+      "learning_rate": 1.0891855008165816e-05,
+      "loss": 0.7345,
+      "step": 1363500
+    },
+    {
+      "epoch": 2.347349164746395,
+      "grad_norm": 2.1021995544433594,
+      "learning_rate": 1.0877513920893417e-05,
+      "loss": 0.7316,
+      "step": 1364000
+    },
+    {
+      "epoch": 2.348209629982739,
+      "grad_norm": 2.3361055850982666,
+      "learning_rate": 1.0863172833621015e-05,
+      "loss": 0.7345,
+      "step": 1364500
+    },
+    {
+      "epoch": 2.349070095219083,
+      "grad_norm": 2.2210545539855957,
+      "learning_rate": 1.0848831746348616e-05,
+      "loss": 0.7385,
+      "step": 1365000
+    },
+    {
+      "epoch": 2.349930560455427,
+      "grad_norm": 2.323305368423462,
+      "learning_rate": 1.0834490659076216e-05,
+      "loss": 0.7333,
+      "step": 1365500
+    },
+    {
+      "epoch": 2.350791025691771,
+      "grad_norm": 2.095165967941284,
+      "learning_rate": 1.0820149571803817e-05,
+      "loss": 0.7328,
+      "step": 1366000
+    },
+    {
+      "epoch": 2.351651490928115,
+      "grad_norm": 2.113354206085205,
+      "learning_rate": 1.0805808484531417e-05,
+      "loss": 0.7363,
+      "step": 1366500
+    },
+    {
+      "epoch": 2.352511956164459,
+      "grad_norm": 2.207193613052368,
+      "learning_rate": 1.0791467397259018e-05,
+      "loss": 0.7302,
+      "step": 1367000
+    },
+    {
+      "epoch": 2.353372421400803,
+      "grad_norm": 2.3709208965301514,
+      "learning_rate": 1.0777126309986617e-05,
+      "loss": 0.7305,
+      "step": 1367500
+    },
+    {
+      "epoch": 2.354232886637147,
+      "grad_norm": 2.1980271339416504,
+      "learning_rate": 1.0762785222714219e-05,
+      "loss": 0.7331,
+      "step": 1368000
+    },
+    {
+      "epoch": 2.355093351873491,
+      "grad_norm": 2.143127202987671,
+      "learning_rate": 1.0748444135441817e-05,
+      "loss": 0.7289,
+      "step": 1368500
+    },
+    {
+      "epoch": 2.355953817109835,
+      "grad_norm": 2.1765642166137695,
+      "learning_rate": 1.0734103048169418e-05,
+      "loss": 0.7347,
+      "step": 1369000
+    },
+    {
+      "epoch": 2.356814282346179,
+      "grad_norm": 2.2334864139556885,
+      "learning_rate": 1.0719761960897018e-05,
+      "loss": 0.7311,
+      "step": 1369500
+    },
+    {
+      "epoch": 2.357674747582523,
+      "grad_norm": 2.210925340652466,
+      "learning_rate": 1.0705420873624619e-05,
+      "loss": 0.7312,
+      "step": 1370000
+    },
+    {
+      "epoch": 2.358535212818867,
+      "grad_norm": 2.15215802192688,
+      "learning_rate": 1.069107978635222e-05,
+      "loss": 0.73,
+      "step": 1370500
+    },
+    {
+      "epoch": 2.3593956780552108,
+      "grad_norm": 2.062427520751953,
+      "learning_rate": 1.067673869907982e-05,
+      "loss": 0.7304,
+      "step": 1371000
+    },
+    {
+      "epoch": 2.3602561432915548,
+      "grad_norm": 2.322567939758301,
+      "learning_rate": 1.0662397611807419e-05,
+      "loss": 0.7387,
+      "step": 1371500
+    },
+    {
+      "epoch": 2.3611166085278987,
+      "grad_norm": 2.153791666030884,
+      "learning_rate": 1.0648056524535019e-05,
+      "loss": 0.7327,
+      "step": 1372000
+    },
+    {
+      "epoch": 2.3619770737642427,
+      "grad_norm": 2.4358983039855957,
+      "learning_rate": 1.063371543726262e-05,
+      "loss": 0.7331,
+      "step": 1372500
+    },
+    {
+      "epoch": 2.3628375390005867,
+      "grad_norm": 2.2714719772338867,
+      "learning_rate": 1.061937434999022e-05,
+      "loss": 0.731,
+      "step": 1373000
+    },
+    {
+      "epoch": 2.3636980042369307,
+      "grad_norm": 2.271050214767456,
+      "learning_rate": 1.060503326271782e-05,
+      "loss": 0.734,
+      "step": 1373500
+    },
+    {
+      "epoch": 2.3645584694732746,
+      "grad_norm": 2.266615390777588,
+      "learning_rate": 1.059069217544542e-05,
+      "loss": 0.727,
+      "step": 1374000
+    },
+    {
+      "epoch": 2.3654189347096186,
+      "grad_norm": 2.197510004043579,
+      "learning_rate": 1.057635108817302e-05,
+      "loss": 0.7285,
+      "step": 1374500
+    },
+    {
+      "epoch": 2.3662793999459626,
+      "grad_norm": 2.2251787185668945,
+      "learning_rate": 1.0562010000900622e-05,
+      "loss": 0.7266,
+      "step": 1375000
+    },
+    {
+      "epoch": 2.367139865182307,
+      "grad_norm": 2.2310283184051514,
+      "learning_rate": 1.054766891362822e-05,
+      "loss": 0.7339,
+      "step": 1375500
+    },
+    {
+      "epoch": 2.368000330418651,
+      "grad_norm": 2.100308418273926,
+      "learning_rate": 1.0533327826355821e-05,
+      "loss": 0.7352,
+      "step": 1376000
+    },
+    {
+      "epoch": 2.368860795654995,
+      "grad_norm": 2.230976104736328,
+      "learning_rate": 1.0518986739083422e-05,
+      "loss": 0.7339,
+      "step": 1376500
+    },
+    {
+      "epoch": 2.369721260891339,
+      "grad_norm": 2.412017345428467,
+      "learning_rate": 1.0504645651811022e-05,
+      "loss": 0.7325,
+      "step": 1377000
+    },
+    {
+      "epoch": 2.370581726127683,
+      "grad_norm": 2.1025924682617188,
+      "learning_rate": 1.0490304564538623e-05,
+      "loss": 0.7349,
+      "step": 1377500
+    },
+    {
+      "epoch": 2.371442191364027,
+      "grad_norm": 2.063314914703369,
+      "learning_rate": 1.0475963477266221e-05,
+      "loss": 0.7356,
+      "step": 1378000
+    },
+    {
+      "epoch": 2.372302656600371,
+      "grad_norm": 2.093384027481079,
+      "learning_rate": 1.0461622389993822e-05,
+      "loss": 0.7277,
+      "step": 1378500
+    },
+    {
+      "epoch": 2.373163121836715,
+      "grad_norm": 2.142827033996582,
+      "learning_rate": 1.0447281302721422e-05,
+      "loss": 0.7301,
+      "step": 1379000
+    },
+    {
+      "epoch": 2.374023587073059,
+      "grad_norm": 2.0535013675689697,
+      "learning_rate": 1.0432940215449023e-05,
+      "loss": 0.7332,
+      "step": 1379500
+    },
+    {
+      "epoch": 2.374884052309403,
+      "grad_norm": 2.119863510131836,
+      "learning_rate": 1.0418599128176623e-05,
+      "loss": 0.7342,
+      "step": 1380000
+    },
+    {
+      "epoch": 2.375744517545747,
+      "grad_norm": 2.02392840385437,
+      "learning_rate": 1.0404258040904224e-05,
+      "loss": 0.7309,
+      "step": 1380500
+    },
+    {
+      "epoch": 2.3766049827820908,
+      "grad_norm": 2.1911582946777344,
+      "learning_rate": 1.0389916953631822e-05,
+      "loss": 0.7309,
+      "step": 1381000
+    },
+    {
+      "epoch": 2.3774654480184347,
+      "grad_norm": 2.183885335922241,
+      "learning_rate": 1.0375575866359425e-05,
+      "loss": 0.7327,
+      "step": 1381500
+    },
+    {
+      "epoch": 2.3783259132547787,
+      "grad_norm": 2.272047996520996,
+      "learning_rate": 1.0361234779087025e-05,
+      "loss": 0.7288,
+      "step": 1382000
+    },
+    {
+      "epoch": 2.3791863784911227,
+      "grad_norm": 2.1483206748962402,
+      "learning_rate": 1.0346893691814624e-05,
+      "loss": 0.7294,
+      "step": 1382500
+    },
+    {
+      "epoch": 2.3800468437274667,
+      "grad_norm": 2.1930770874023438,
+      "learning_rate": 1.0332552604542224e-05,
+      "loss": 0.722,
+      "step": 1383000
+    },
+    {
+      "epoch": 2.3809073089638106,
+      "grad_norm": 2.2109477519989014,
+      "learning_rate": 1.0318211517269825e-05,
+      "loss": 0.7269,
+      "step": 1383500
+    },
+    {
+      "epoch": 2.3817677742001546,
+      "grad_norm": 2.27551007270813,
+      "learning_rate": 1.0303870429997425e-05,
+      "loss": 0.7272,
+      "step": 1384000
+    },
+    {
+      "epoch": 2.3826282394364986,
+      "grad_norm": 2.266842842102051,
+      "learning_rate": 1.0289529342725026e-05,
+      "loss": 0.7267,
+      "step": 1384500
+    },
+    {
+      "epoch": 2.3834887046728426,
+      "grad_norm": 1.9027714729309082,
+      "learning_rate": 1.0275188255452625e-05,
+      "loss": 0.7346,
+      "step": 1385000
+    },
+    {
+      "epoch": 2.3843491699091865,
+      "grad_norm": 2.027723789215088,
+      "learning_rate": 1.0260847168180225e-05,
+      "loss": 0.7278,
+      "step": 1385500
+    },
+    {
+      "epoch": 2.3852096351455305,
+      "grad_norm": 2.2213375568389893,
+      "learning_rate": 1.0246506080907825e-05,
+      "loss": 0.7326,
+      "step": 1386000
+    },
+    {
+      "epoch": 2.3860701003818745,
+      "grad_norm": 2.044480085372925,
+      "learning_rate": 1.0232164993635426e-05,
+      "loss": 0.7371,
+      "step": 1386500
+    },
+    {
+      "epoch": 2.3869305656182185,
+      "grad_norm": 2.011831045150757,
+      "learning_rate": 1.0217823906363026e-05,
+      "loss": 0.7316,
+      "step": 1387000
+    },
+    {
+      "epoch": 2.3877910308545625,
+      "grad_norm": 2.323796272277832,
+      "learning_rate": 1.0203482819090627e-05,
+      "loss": 0.735,
+      "step": 1387500
+    },
+    {
+      "epoch": 2.3886514960909064,
+      "grad_norm": 2.00823974609375,
+      "learning_rate": 1.0189141731818226e-05,
+      "loss": 0.721,
+      "step": 1388000
+    },
+    {
+      "epoch": 2.3895119613272504,
+      "grad_norm": 2.2485666275024414,
+      "learning_rate": 1.0174800644545828e-05,
+      "loss": 0.7282,
+      "step": 1388500
+    },
+    {
+      "epoch": 2.3903724265635944,
+      "grad_norm": 2.062854290008545,
+      "learning_rate": 1.0160459557273427e-05,
+      "loss": 0.7298,
+      "step": 1389000
+    },
+    {
+      "epoch": 2.3912328917999384,
+      "grad_norm": 2.141507625579834,
+      "learning_rate": 1.0146118470001027e-05,
+      "loss": 0.7297,
+      "step": 1389500
+    },
+    {
+      "epoch": 2.3920933570362823,
+      "grad_norm": 2.1152637004852295,
+      "learning_rate": 1.0131777382728628e-05,
+      "loss": 0.7346,
+      "step": 1390000
+    },
+    {
+      "epoch": 2.3929538222726263,
+      "grad_norm": 2.3339223861694336,
+      "learning_rate": 1.0117436295456226e-05,
+      "loss": 0.7269,
+      "step": 1390500
+    },
+    {
+      "epoch": 2.3938142875089703,
+      "grad_norm": 2.2937827110290527,
+      "learning_rate": 1.0103095208183828e-05,
+      "loss": 0.7313,
+      "step": 1391000
+    },
+    {
+      "epoch": 2.3946747527453143,
+      "grad_norm": 2.311241626739502,
+      "learning_rate": 1.0088754120911429e-05,
+      "loss": 0.7332,
+      "step": 1391500
+    },
+    {
+      "epoch": 2.3955352179816582,
+      "grad_norm": 2.4438748359680176,
+      "learning_rate": 1.0074413033639028e-05,
+      "loss": 0.7273,
+      "step": 1392000
+    },
+    {
+      "epoch": 2.3963956832180022,
+      "grad_norm": 2.4046483039855957,
+      "learning_rate": 1.0060071946366628e-05,
+      "loss": 0.7322,
+      "step": 1392500
+    },
+    {
+      "epoch": 2.397256148454346,
+      "grad_norm": 2.233520984649658,
+      "learning_rate": 1.004573085909423e-05,
+      "loss": 0.7263,
+      "step": 1393000
+    },
+    {
+      "epoch": 2.39811661369069,
+      "grad_norm": 2.1059939861297607,
+      "learning_rate": 1.0031389771821829e-05,
+      "loss": 0.7343,
+      "step": 1393500
+    },
+    {
+      "epoch": 2.398977078927034,
+      "grad_norm": 2.5026044845581055,
+      "learning_rate": 1.001704868454943e-05,
+      "loss": 0.7294,
+      "step": 1394000
+    },
+    {
+      "epoch": 2.399837544163378,
+      "grad_norm": 2.279144287109375,
+      "learning_rate": 1.000270759727703e-05,
+      "loss": 0.7287,
+      "step": 1394500
+    },
+    {
+      "epoch": 2.400698009399722,
+      "grad_norm": 2.19842529296875,
+      "learning_rate": 9.988366510004629e-06,
+      "loss": 0.7281,
+      "step": 1395000
+    },
+    {
+      "epoch": 2.401558474636066,
+      "grad_norm": 2.2801008224487305,
+      "learning_rate": 9.974025422732231e-06,
+      "loss": 0.7302,
+      "step": 1395500
+    },
+    {
+      "epoch": 2.40241893987241,
+      "grad_norm": 1.9947917461395264,
+      "learning_rate": 9.95968433545983e-06,
+      "loss": 0.727,
+      "step": 1396000
+    },
+    {
+      "epoch": 2.403279405108754,
+      "grad_norm": 2.186469793319702,
+      "learning_rate": 9.94534324818743e-06,
+      "loss": 0.729,
+      "step": 1396500
+    },
+    {
+      "epoch": 2.404139870345098,
+      "grad_norm": 1.9610108137130737,
+      "learning_rate": 9.93100216091503e-06,
+      "loss": 0.7298,
+      "step": 1397000
+    },
+    {
+      "epoch": 2.405000335581442,
+      "grad_norm": 2.3492963314056396,
+      "learning_rate": 9.916661073642631e-06,
+      "loss": 0.732,
+      "step": 1397500
+    },
+    {
+      "epoch": 2.405860800817786,
+      "grad_norm": 2.2294130325317383,
+      "learning_rate": 9.902319986370232e-06,
+      "loss": 0.729,
+      "step": 1398000
+    },
+    {
+      "epoch": 2.40672126605413,
+      "grad_norm": 2.0163352489471436,
+      "learning_rate": 9.887978899097832e-06,
+      "loss": 0.7272,
+      "step": 1398500
+    },
+    {
+      "epoch": 2.407581731290474,
+      "grad_norm": 2.016042709350586,
+      "learning_rate": 9.873637811825431e-06,
+      "loss": 0.7286,
+      "step": 1399000
+    },
+    {
+      "epoch": 2.408442196526818,
+      "grad_norm": 2.637369155883789,
+      "learning_rate": 9.859296724553031e-06,
+      "loss": 0.7308,
+      "step": 1399500
+    },
+    {
+      "epoch": 2.409302661763162,
+      "grad_norm": 2.234714984893799,
+      "learning_rate": 9.844955637280632e-06,
+      "loss": 0.7261,
+      "step": 1400000
+    },
+    {
+      "epoch": 2.410163126999506,
+      "grad_norm": 2.139127254486084,
+      "learning_rate": 9.830614550008232e-06,
+      "loss": 0.7224,
+      "step": 1400500
+    },
+    {
+      "epoch": 2.41102359223585,
+      "grad_norm": 2.3268463611602783,
+      "learning_rate": 9.816273462735833e-06,
+      "loss": 0.7261,
+      "step": 1401000
+    },
+    {
+      "epoch": 2.4118840574721943,
+      "grad_norm": 2.230707883834839,
+      "learning_rate": 9.801932375463432e-06,
+      "loss": 0.7251,
+      "step": 1401500
+    },
+    {
+      "epoch": 2.4127445227085382,
+      "grad_norm": 2.266864061355591,
+      "learning_rate": 9.787591288191032e-06,
+      "loss": 0.7306,
+      "step": 1402000
+    },
+    {
+      "epoch": 2.413604987944882,
+      "grad_norm": 2.000060558319092,
+      "learning_rate": 9.773250200918634e-06,
+      "loss": 0.7273,
+      "step": 1402500
+    },
+    {
+      "epoch": 2.414465453181226,
+      "grad_norm": 2.3916234970092773,
+      "learning_rate": 9.758909113646233e-06,
+      "loss": 0.7296,
+      "step": 1403000
+    },
+    {
+      "epoch": 2.41532591841757,
+      "grad_norm": 2.1868340969085693,
+      "learning_rate": 9.744568026373833e-06,
+      "loss": 0.73,
+      "step": 1403500
+    },
+    {
+      "epoch": 2.416186383653914,
+      "grad_norm": 2.2738113403320312,
+      "learning_rate": 9.730226939101434e-06,
+      "loss": 0.7329,
+      "step": 1404000
+    },
+    {
+      "epoch": 2.417046848890258,
+      "grad_norm": 2.2496702671051025,
+      "learning_rate": 9.715885851829034e-06,
+      "loss": 0.7296,
+      "step": 1404500
+    },
+    {
+      "epoch": 2.417907314126602,
+      "grad_norm": 2.0994648933410645,
+      "learning_rate": 9.701544764556635e-06,
+      "loss": 0.727,
+      "step": 1405000
+    },
+    {
+      "epoch": 2.418767779362946,
+      "grad_norm": 2.1675145626068115,
+      "learning_rate": 9.687203677284235e-06,
+      "loss": 0.7325,
+      "step": 1405500
+    },
+    {
+      "epoch": 2.41962824459929,
+      "grad_norm": 2.0863234996795654,
+      "learning_rate": 9.672862590011834e-06,
+      "loss": 0.7269,
+      "step": 1406000
+    },
+    {
+      "epoch": 2.420488709835634,
+      "grad_norm": 2.3856394290924072,
+      "learning_rate": 9.658521502739435e-06,
+      "loss": 0.7242,
+      "step": 1406500
+    },
+    {
+      "epoch": 2.421349175071978,
+      "grad_norm": 2.326630115509033,
+      "learning_rate": 9.644180415467035e-06,
+      "loss": 0.7236,
+      "step": 1407000
+    },
+    {
+      "epoch": 2.422209640308322,
+      "grad_norm": 2.0575623512268066,
+      "learning_rate": 9.629839328194636e-06,
+      "loss": 0.7328,
+      "step": 1407500
+    },
+    {
+      "epoch": 2.423070105544666,
+      "grad_norm": 2.304266929626465,
+      "learning_rate": 9.615498240922236e-06,
+      "loss": 0.7255,
+      "step": 1408000
+    },
+    {
+      "epoch": 2.42393057078101,
+      "grad_norm": 2.350442409515381,
+      "learning_rate": 9.601157153649835e-06,
+      "loss": 0.7265,
+      "step": 1408500
+    },
+    {
+      "epoch": 2.424791036017354,
+      "grad_norm": 2.382349729537964,
+      "learning_rate": 9.586816066377437e-06,
+      "loss": 0.7294,
+      "step": 1409000
+    },
+    {
+      "epoch": 2.425651501253698,
+      "grad_norm": 2.092679500579834,
+      "learning_rate": 9.572474979105037e-06,
+      "loss": 0.7242,
+      "step": 1409500
+    },
+    {
+      "epoch": 2.426511966490042,
+      "grad_norm": 2.3096089363098145,
+      "learning_rate": 9.558133891832636e-06,
+      "loss": 0.7253,
+      "step": 1410000
+    },
+    {
+      "epoch": 2.427372431726386,
+      "grad_norm": 2.160024881362915,
+      "learning_rate": 9.543792804560237e-06,
+      "loss": 0.7275,
+      "step": 1410500
+    },
+    {
+      "epoch": 2.42823289696273,
+      "grad_norm": 2.258770704269409,
+      "learning_rate": 9.529451717287837e-06,
+      "loss": 0.7189,
+      "step": 1411000
+    },
+    {
+      "epoch": 2.429093362199074,
+      "grad_norm": 2.0837507247924805,
+      "learning_rate": 9.515110630015438e-06,
+      "loss": 0.7297,
+      "step": 1411500
+    },
+    {
+      "epoch": 2.4299538274354178,
+      "grad_norm": 2.198729991912842,
+      "learning_rate": 9.500769542743038e-06,
+      "loss": 0.7338,
+      "step": 1412000
+    },
+    {
+      "epoch": 2.4308142926717617,
+      "grad_norm": 2.3990306854248047,
+      "learning_rate": 9.486428455470637e-06,
+      "loss": 0.7321,
+      "step": 1412500
+    },
+    {
+      "epoch": 2.4316747579081057,
+      "grad_norm": 2.0542526245117188,
+      "learning_rate": 9.472087368198237e-06,
+      "loss": 0.7278,
+      "step": 1413000
+    },
+    {
+      "epoch": 2.4325352231444497,
+      "grad_norm": 2.1463656425476074,
+      "learning_rate": 9.457746280925838e-06,
+      "loss": 0.7346,
+      "step": 1413500
+    },
+    {
+      "epoch": 2.4333956883807937,
+      "grad_norm": 2.061422348022461,
+      "learning_rate": 9.443405193653438e-06,
+      "loss": 0.7197,
+      "step": 1414000
+    },
+    {
+      "epoch": 2.4342561536171377,
+      "grad_norm": 2.2131097316741943,
+      "learning_rate": 9.429064106381039e-06,
+      "loss": 0.7243,
+      "step": 1414500
+    },
+    {
+      "epoch": 2.4351166188534816,
+      "grad_norm": 2.2422876358032227,
+      "learning_rate": 9.41472301910864e-06,
+      "loss": 0.7338,
+      "step": 1415000
+    },
+    {
+      "epoch": 2.4359770840898256,
+      "grad_norm": 2.2505648136138916,
+      "learning_rate": 9.400381931836238e-06,
+      "loss": 0.7242,
+      "step": 1415500
+    },
+    {
+      "epoch": 2.4368375493261696,
+      "grad_norm": 2.3576719760894775,
+      "learning_rate": 9.38604084456384e-06,
+      "loss": 0.7257,
+      "step": 1416000
+    },
+    {
+      "epoch": 2.4376980145625136,
+      "grad_norm": 2.1250061988830566,
+      "learning_rate": 9.371699757291439e-06,
+      "loss": 0.7227,
+      "step": 1416500
+    },
+    {
+      "epoch": 2.4385584797988575,
+      "grad_norm": 2.289149045944214,
+      "learning_rate": 9.35735867001904e-06,
+      "loss": 0.7265,
+      "step": 1417000
+    },
+    {
+      "epoch": 2.4394189450352015,
+      "grad_norm": 2.392317533493042,
+      "learning_rate": 9.34301758274664e-06,
+      "loss": 0.7291,
+      "step": 1417500
+    },
+    {
+      "epoch": 2.4402794102715455,
+      "grad_norm": 2.255974769592285,
+      "learning_rate": 9.32867649547424e-06,
+      "loss": 0.7304,
+      "step": 1418000
+    },
+    {
+      "epoch": 2.4411398755078895,
+      "grad_norm": 2.070075035095215,
+      "learning_rate": 9.31433540820184e-06,
+      "loss": 0.7312,
+      "step": 1418500
+    },
+    {
+      "epoch": 2.4420003407442334,
+      "grad_norm": 2.2537167072296143,
+      "learning_rate": 9.299994320929441e-06,
+      "loss": 0.7321,
+      "step": 1419000
+    },
+    {
+      "epoch": 2.4428608059805774,
+      "grad_norm": 2.128941535949707,
+      "learning_rate": 9.28565323365704e-06,
+      "loss": 0.7267,
+      "step": 1419500
+    },
+    {
+      "epoch": 2.4437212712169214,
+      "grad_norm": 2.2561581134796143,
+      "learning_rate": 9.27131214638464e-06,
+      "loss": 0.7287,
+      "step": 1420000
+    },
+    {
+      "epoch": 2.4445817364532654,
+      "grad_norm": 2.095482110977173,
+      "learning_rate": 9.256971059112243e-06,
+      "loss": 0.7261,
+      "step": 1420500
+    },
+    {
+      "epoch": 2.4454422016896094,
+      "grad_norm": 2.4081618785858154,
+      "learning_rate": 9.242629971839841e-06,
+      "loss": 0.7286,
+      "step": 1421000
+    },
+    {
+      "epoch": 2.4463026669259538,
+      "grad_norm": 2.0608112812042236,
+      "learning_rate": 9.228288884567442e-06,
+      "loss": 0.7212,
+      "step": 1421500
+    },
+    {
+      "epoch": 2.4471631321622977,
+      "grad_norm": 2.0750913619995117,
+      "learning_rate": 9.213947797295042e-06,
+      "loss": 0.7248,
+      "step": 1422000
+    },
+    {
+      "epoch": 2.4480235973986417,
+      "grad_norm": 2.242100715637207,
+      "learning_rate": 9.199606710022641e-06,
+      "loss": 0.7287,
+      "step": 1422500
+    },
+    {
+      "epoch": 2.4488840626349857,
+      "grad_norm": 2.1421103477478027,
+      "learning_rate": 9.185265622750243e-06,
+      "loss": 0.7284,
+      "step": 1423000
+    },
+    {
+      "epoch": 2.4497445278713297,
+      "grad_norm": 2.3732893466949463,
+      "learning_rate": 9.170924535477842e-06,
+      "loss": 0.7303,
+      "step": 1423500
+    },
+    {
+      "epoch": 2.4506049931076737,
+      "grad_norm": 2.1787123680114746,
+      "learning_rate": 9.156583448205443e-06,
+      "loss": 0.7262,
+      "step": 1424000
+    },
+    {
+      "epoch": 2.4514654583440176,
+      "grad_norm": 2.1791365146636963,
+      "learning_rate": 9.142242360933043e-06,
+      "loss": 0.7209,
+      "step": 1424500
+    },
+    {
+      "epoch": 2.4523259235803616,
+      "grad_norm": 2.1248362064361572,
+      "learning_rate": 9.127901273660644e-06,
+      "loss": 0.7279,
+      "step": 1425000
+    },
+    {
+      "epoch": 2.4531863888167056,
+      "grad_norm": 2.3395819664001465,
+      "learning_rate": 9.113560186388244e-06,
+      "loss": 0.7291,
+      "step": 1425500
+    },
+    {
+      "epoch": 2.4540468540530496,
+      "grad_norm": 2.144453763961792,
+      "learning_rate": 9.099219099115844e-06,
+      "loss": 0.7272,
+      "step": 1426000
+    },
+    {
+      "epoch": 2.4549073192893935,
+      "grad_norm": 2.2412824630737305,
+      "learning_rate": 9.084878011843443e-06,
+      "loss": 0.7249,
+      "step": 1426500
+    },
+    {
+      "epoch": 2.4557677845257375,
+      "grad_norm": 2.4376273155212402,
+      "learning_rate": 9.070536924571044e-06,
+      "loss": 0.7204,
+      "step": 1427000
+    },
+    {
+      "epoch": 2.4566282497620815,
+      "grad_norm": 2.246067523956299,
+      "learning_rate": 9.056195837298644e-06,
+      "loss": 0.7272,
+      "step": 1427500
+    },
+    {
+      "epoch": 2.4574887149984255,
+      "grad_norm": 2.154071092605591,
+      "learning_rate": 9.041854750026245e-06,
+      "loss": 0.7192,
+      "step": 1428000
+    },
+    {
+      "epoch": 2.4583491802347694,
+      "grad_norm": 2.44376802444458,
+      "learning_rate": 9.027513662753845e-06,
+      "loss": 0.7301,
+      "step": 1428500
+    },
+    {
+      "epoch": 2.4592096454711134,
+      "grad_norm": 2.3021342754364014,
+      "learning_rate": 9.013172575481444e-06,
+      "loss": 0.7202,
+      "step": 1429000
+    },
+    {
+      "epoch": 2.4600701107074574,
+      "grad_norm": 2.472088575363159,
+      "learning_rate": 8.998831488209044e-06,
+      "loss": 0.7256,
+      "step": 1429500
+    },
+    {
+      "epoch": 2.4609305759438014,
+      "grad_norm": 2.3227767944335938,
+      "learning_rate": 8.984490400936647e-06,
+      "loss": 0.7257,
+      "step": 1430000
+    },
+    {
+      "epoch": 2.4617910411801454,
+      "grad_norm": 2.3386566638946533,
+      "learning_rate": 8.970149313664245e-06,
+      "loss": 0.727,
+      "step": 1430500
+    },
+    {
+      "epoch": 2.4626515064164893,
+      "grad_norm": 2.3084495067596436,
+      "learning_rate": 8.955808226391846e-06,
+      "loss": 0.7305,
+      "step": 1431000
+    },
+    {
+      "epoch": 2.4635119716528333,
+      "grad_norm": 2.0986111164093018,
+      "learning_rate": 8.941467139119446e-06,
+      "loss": 0.7208,
+      "step": 1431500
+    },
+    {
+      "epoch": 2.4643724368891773,
+      "grad_norm": 2.0349369049072266,
+      "learning_rate": 8.927126051847047e-06,
+      "loss": 0.7239,
+      "step": 1432000
+    },
+    {
+      "epoch": 2.4652329021255213,
+      "grad_norm": 2.507856607437134,
+      "learning_rate": 8.912784964574647e-06,
+      "loss": 0.7288,
+      "step": 1432500
+    },
+    {
+      "epoch": 2.4660933673618652,
+      "grad_norm": 2.2290287017822266,
+      "learning_rate": 8.898443877302248e-06,
+      "loss": 0.7324,
+      "step": 1433000
+    },
+    {
+      "epoch": 2.466953832598209,
+      "grad_norm": 2.147047281265259,
+      "learning_rate": 8.884102790029846e-06,
+      "loss": 0.7232,
+      "step": 1433500
+    },
+    {
+      "epoch": 2.467814297834553,
+      "grad_norm": 2.185572624206543,
+      "learning_rate": 8.869761702757447e-06,
+      "loss": 0.7264,
+      "step": 1434000
+    },
+    {
+      "epoch": 2.468674763070897,
+      "grad_norm": 2.173811435699463,
+      "learning_rate": 8.855420615485047e-06,
+      "loss": 0.7244,
+      "step": 1434500
+    },
+    {
+      "epoch": 2.469535228307241,
+      "grad_norm": 2.1698076725006104,
+      "learning_rate": 8.841079528212648e-06,
+      "loss": 0.722,
+      "step": 1435000
+    },
+    {
+      "epoch": 2.470395693543585,
+      "grad_norm": 2.0112743377685547,
+      "learning_rate": 8.826738440940248e-06,
+      "loss": 0.7231,
+      "step": 1435500
+    },
+    {
+      "epoch": 2.471256158779929,
+      "grad_norm": 2.455514669418335,
+      "learning_rate": 8.812397353667847e-06,
+      "loss": 0.7222,
+      "step": 1436000
+    },
+    {
+      "epoch": 2.472116624016273,
+      "grad_norm": 2.2176318168640137,
+      "learning_rate": 8.79805626639545e-06,
+      "loss": 0.7255,
+      "step": 1436500
+    },
+    {
+      "epoch": 2.472977089252617,
+      "grad_norm": 2.0305678844451904,
+      "learning_rate": 8.78371517912305e-06,
+      "loss": 0.7264,
+      "step": 1437000
+    },
+    {
+      "epoch": 2.473837554488961,
+      "grad_norm": 1.9838893413543701,
+      "learning_rate": 8.769374091850648e-06,
+      "loss": 0.7248,
+      "step": 1437500
+    },
+    {
+      "epoch": 2.474698019725305,
+      "grad_norm": 2.39805006980896,
+      "learning_rate": 8.755033004578249e-06,
+      "loss": 0.724,
+      "step": 1438000
+    },
+    {
+      "epoch": 2.475558484961649,
+      "grad_norm": 2.250845432281494,
+      "learning_rate": 8.74069191730585e-06,
+      "loss": 0.7266,
+      "step": 1438500
+    },
+    {
+      "epoch": 2.476418950197993,
+      "grad_norm": 2.292314291000366,
+      "learning_rate": 8.72635083003345e-06,
+      "loss": 0.719,
+      "step": 1439000
+    },
+    {
+      "epoch": 2.477279415434337,
+      "grad_norm": 2.3114101886749268,
+      "learning_rate": 8.71200974276105e-06,
+      "loss": 0.7289,
+      "step": 1439500
+    },
+    {
+      "epoch": 2.478139880670681,
+      "grad_norm": 2.1311428546905518,
+      "learning_rate": 8.697668655488649e-06,
+      "loss": 0.7294,
+      "step": 1440000
+    },
+    {
+      "epoch": 2.479000345907025,
+      "grad_norm": 2.3691234588623047,
+      "learning_rate": 8.68332756821625e-06,
+      "loss": 0.7186,
+      "step": 1440500
+    },
+    {
+      "epoch": 2.479860811143369,
+      "grad_norm": 2.3010482788085938,
+      "learning_rate": 8.66898648094385e-06,
+      "loss": 0.7329,
+      "step": 1441000
+    },
+    {
+      "epoch": 2.480721276379713,
+      "grad_norm": 2.0064890384674072,
+      "learning_rate": 8.65464539367145e-06,
+      "loss": 0.7234,
+      "step": 1441500
+    },
+    {
+      "epoch": 2.481581741616057,
+      "grad_norm": 2.1313636302948,
+      "learning_rate": 8.640304306399051e-06,
+      "loss": 0.7229,
+      "step": 1442000
+    },
+    {
+      "epoch": 2.482442206852401,
+      "grad_norm": 2.41935133934021,
+      "learning_rate": 8.625963219126652e-06,
+      "loss": 0.7206,
+      "step": 1442500
+    },
+    {
+      "epoch": 2.483302672088745,
+      "grad_norm": 2.29103684425354,
+      "learning_rate": 8.61162213185425e-06,
+      "loss": 0.7272,
+      "step": 1443000
+    },
+    {
+      "epoch": 2.4841631373250888,
+      "grad_norm": 2.1962101459503174,
+      "learning_rate": 8.597281044581852e-06,
+      "loss": 0.7324,
+      "step": 1443500
+    },
+    {
+      "epoch": 2.4850236025614327,
+      "grad_norm": 2.147658348083496,
+      "learning_rate": 8.582939957309453e-06,
+      "loss": 0.7233,
+      "step": 1444000
+    },
+    {
+      "epoch": 2.4858840677977767,
+      "grad_norm": 2.0254199504852295,
+      "learning_rate": 8.568598870037052e-06,
+      "loss": 0.7243,
+      "step": 1444500
+    },
+    {
+      "epoch": 2.4867445330341207,
+      "grad_norm": 2.3642418384552,
+      "learning_rate": 8.554257782764652e-06,
+      "loss": 0.7252,
+      "step": 1445000
+    },
+    {
+      "epoch": 2.4876049982704647,
+      "grad_norm": 2.0392184257507324,
+      "learning_rate": 8.539916695492253e-06,
+      "loss": 0.7261,
+      "step": 1445500
+    },
+    {
+      "epoch": 2.4884654635068086,
+      "grad_norm": 2.1036550998687744,
+      "learning_rate": 8.525575608219853e-06,
+      "loss": 0.7181,
+      "step": 1446000
+    },
+    {
+      "epoch": 2.4893259287431526,
+      "grad_norm": 2.117926836013794,
+      "learning_rate": 8.511234520947454e-06,
+      "loss": 0.7248,
+      "step": 1446500
+    },
+    {
+      "epoch": 2.4901863939794966,
+      "grad_norm": 2.2823784351348877,
+      "learning_rate": 8.496893433675052e-06,
+      "loss": 0.7218,
+      "step": 1447000
+    },
+    {
+      "epoch": 2.491046859215841,
+      "grad_norm": 2.1260955333709717,
+      "learning_rate": 8.482552346402653e-06,
+      "loss": 0.7291,
+      "step": 1447500
+    },
+    {
+      "epoch": 2.491907324452185,
+      "grad_norm": 2.096526861190796,
+      "learning_rate": 8.468211259130255e-06,
+      "loss": 0.7325,
+      "step": 1448000
+    },
+    {
+      "epoch": 2.492767789688529,
+      "grad_norm": 2.419377565383911,
+      "learning_rate": 8.453870171857854e-06,
+      "loss": 0.7236,
+      "step": 1448500
+    },
+    {
+      "epoch": 2.493628254924873,
+      "grad_norm": 2.5147461891174316,
+      "learning_rate": 8.439529084585454e-06,
+      "loss": 0.7247,
+      "step": 1449000
+    },
+    {
+      "epoch": 2.494488720161217,
+      "grad_norm": 2.212557792663574,
+      "learning_rate": 8.425187997313055e-06,
+      "loss": 0.7292,
+      "step": 1449500
+    },
+    {
+      "epoch": 2.495349185397561,
+      "grad_norm": 2.241730213165283,
+      "learning_rate": 8.410846910040653e-06,
+      "loss": 0.7325,
+      "step": 1450000
+    },
+    {
+      "epoch": 2.496209650633905,
+      "grad_norm": 2.205704927444458,
+      "learning_rate": 8.396505822768256e-06,
+      "loss": 0.7247,
+      "step": 1450500
+    },
+    {
+      "epoch": 2.497070115870249,
+      "grad_norm": 1.9690256118774414,
+      "learning_rate": 8.382164735495854e-06,
+      "loss": 0.7257,
+      "step": 1451000
+    },
+    {
+      "epoch": 2.497930581106593,
+      "grad_norm": 2.6139450073242188,
+      "learning_rate": 8.367823648223455e-06,
+      "loss": 0.724,
+      "step": 1451500
+    },
+    {
+      "epoch": 2.498791046342937,
+      "grad_norm": 2.0735630989074707,
+      "learning_rate": 8.353482560951055e-06,
+      "loss": 0.7255,
+      "step": 1452000
+    },
+    {
+      "epoch": 2.499651511579281,
+      "grad_norm": 2.2452523708343506,
+      "learning_rate": 8.339141473678656e-06,
+      "loss": 0.7266,
+      "step": 1452500
+    },
+    {
+      "epoch": 2.5005119768156248,
+      "grad_norm": 2.239596128463745,
+      "learning_rate": 8.324800386406256e-06,
+      "loss": 0.7227,
+      "step": 1453000
+    },
+    {
+      "epoch": 2.5013724420519687,
+      "grad_norm": 2.209899663925171,
+      "learning_rate": 8.310459299133857e-06,
+      "loss": 0.7297,
+      "step": 1453500
+    },
+    {
+      "epoch": 2.5022329072883127,
+      "grad_norm": 2.1387386322021484,
+      "learning_rate": 8.296118211861456e-06,
+      "loss": 0.7269,
+      "step": 1454000
+    },
+    {
+      "epoch": 2.5030933725246567,
+      "grad_norm": 2.203953981399536,
+      "learning_rate": 8.281777124589056e-06,
+      "loss": 0.7275,
+      "step": 1454500
+    },
+    {
+      "epoch": 2.5039538377610007,
+      "grad_norm": 2.340108871459961,
+      "learning_rate": 8.267436037316656e-06,
+      "loss": 0.7213,
+      "step": 1455000
+    },
+    {
+      "epoch": 2.5048143029973446,
+      "grad_norm": 2.1395070552825928,
+      "learning_rate": 8.253094950044257e-06,
+      "loss": 0.7235,
+      "step": 1455500
+    },
+    {
+      "epoch": 2.5056747682336886,
+      "grad_norm": 2.0611159801483154,
+      "learning_rate": 8.238753862771857e-06,
+      "loss": 0.7281,
+      "step": 1456000
+    },
+    {
+      "epoch": 2.5065352334700326,
+      "grad_norm": 2.2360119819641113,
+      "learning_rate": 8.224412775499458e-06,
+      "loss": 0.72,
+      "step": 1456500
+    },
+    {
+      "epoch": 2.5073956987063766,
+      "grad_norm": 2.186635971069336,
+      "learning_rate": 8.210071688227057e-06,
+      "loss": 0.7147,
+      "step": 1457000
+    },
+    {
+      "epoch": 2.5082561639427206,
+      "grad_norm": 2.166893482208252,
+      "learning_rate": 8.195730600954659e-06,
+      "loss": 0.7231,
+      "step": 1457500
+    },
+    {
+      "epoch": 2.5091166291790645,
+      "grad_norm": 2.1585497856140137,
+      "learning_rate": 8.181389513682258e-06,
+      "loss": 0.7254,
+      "step": 1458000
+    },
+    {
+      "epoch": 2.5099770944154085,
+      "grad_norm": 2.2699153423309326,
+      "learning_rate": 8.167048426409858e-06,
+      "loss": 0.7252,
+      "step": 1458500
+    },
+    {
+      "epoch": 2.5108375596517525,
+      "grad_norm": 2.149649143218994,
+      "learning_rate": 8.152707339137459e-06,
+      "loss": 0.7263,
+      "step": 1459000
+    },
+    {
+      "epoch": 2.5116980248880965,
+      "grad_norm": 2.4442532062530518,
+      "learning_rate": 8.138366251865059e-06,
+      "loss": 0.7213,
+      "step": 1459500
+    },
+    {
+      "epoch": 2.5125584901244404,
+      "grad_norm": 2.098191738128662,
+      "learning_rate": 8.12402516459266e-06,
+      "loss": 0.7187,
+      "step": 1460000
+    },
+    {
+      "epoch": 2.5134189553607844,
+      "grad_norm": 2.221888780593872,
+      "learning_rate": 8.10968407732026e-06,
+      "loss": 0.722,
+      "step": 1460500
+    },
+    {
+      "epoch": 2.5142794205971284,
+      "grad_norm": 2.258897542953491,
+      "learning_rate": 8.095342990047859e-06,
+      "loss": 0.7256,
+      "step": 1461000
+    },
+    {
+      "epoch": 2.5151398858334724,
+      "grad_norm": 2.0834147930145264,
+      "learning_rate": 8.08100190277546e-06,
+      "loss": 0.7232,
+      "step": 1461500
+    },
+    {
+      "epoch": 2.5160003510698163,
+      "grad_norm": 2.1934962272644043,
+      "learning_rate": 8.06666081550306e-06,
+      "loss": 0.7234,
+      "step": 1462000
+    },
+    {
+      "epoch": 2.5168608163061603,
+      "grad_norm": 2.236333131790161,
+      "learning_rate": 8.05231972823066e-06,
+      "loss": 0.7232,
+      "step": 1462500
+    },
+    {
+      "epoch": 2.5177212815425043,
+      "grad_norm": 2.2630205154418945,
+      "learning_rate": 8.03797864095826e-06,
+      "loss": 0.7188,
+      "step": 1463000
+    },
+    {
+      "epoch": 2.5185817467788483,
+      "grad_norm": 2.093230724334717,
+      "learning_rate": 8.02363755368586e-06,
+      "loss": 0.7207,
+      "step": 1463500
+    },
+    {
+      "epoch": 2.5194422120151923,
+      "grad_norm": 2.096151351928711,
+      "learning_rate": 8.009296466413462e-06,
+      "loss": 0.7179,
+      "step": 1464000
+    },
+    {
+      "epoch": 2.5203026772515362,
+      "grad_norm": 2.2687249183654785,
+      "learning_rate": 7.994955379141062e-06,
+      "loss": 0.7251,
+      "step": 1464500
+    },
+    {
+      "epoch": 2.52116314248788,
+      "grad_norm": 2.351755380630493,
+      "learning_rate": 7.98061429186866e-06,
+      "loss": 0.7242,
+      "step": 1465000
+    },
+    {
+      "epoch": 2.522023607724224,
+      "grad_norm": 2.3624746799468994,
+      "learning_rate": 7.966273204596261e-06,
+      "loss": 0.7254,
+      "step": 1465500
+    },
+    {
+      "epoch": 2.522884072960568,
+      "grad_norm": 2.23112416267395,
+      "learning_rate": 7.951932117323862e-06,
+      "loss": 0.7214,
+      "step": 1466000
+    },
+    {
+      "epoch": 2.5237445381969126,
+      "grad_norm": 2.169090747833252,
+      "learning_rate": 7.937591030051462e-06,
+      "loss": 0.7189,
+      "step": 1466500
+    },
+    {
+      "epoch": 2.5246050034332566,
+      "grad_norm": 2.4218993186950684,
+      "learning_rate": 7.923249942779063e-06,
+      "loss": 0.7271,
+      "step": 1467000
+    },
+    {
+      "epoch": 2.5254654686696005,
+      "grad_norm": 2.2031071186065674,
+      "learning_rate": 7.908908855506663e-06,
+      "loss": 0.7254,
+      "step": 1467500
+    },
+    {
+      "epoch": 2.5263259339059445,
+      "grad_norm": 2.2774407863616943,
+      "learning_rate": 7.894567768234262e-06,
+      "loss": 0.7231,
+      "step": 1468000
+    },
+    {
+      "epoch": 2.5271863991422885,
+      "grad_norm": 2.4886295795440674,
+      "learning_rate": 7.880226680961862e-06,
+      "loss": 0.7206,
+      "step": 1468500
+    },
+    {
+      "epoch": 2.5280468643786325,
+      "grad_norm": 2.159146547317505,
+      "learning_rate": 7.865885593689463e-06,
+      "loss": 0.7265,
+      "step": 1469000
+    },
+    {
+      "epoch": 2.5289073296149764,
+      "grad_norm": 2.215956926345825,
+      "learning_rate": 7.851544506417063e-06,
+      "loss": 0.7169,
+      "step": 1469500
+    },
+    {
+      "epoch": 2.5297677948513204,
+      "grad_norm": 2.1954901218414307,
+      "learning_rate": 7.837203419144664e-06,
+      "loss": 0.7195,
+      "step": 1470000
+    },
+    {
+      "epoch": 2.5306282600876644,
+      "grad_norm": 2.310394048690796,
+      "learning_rate": 7.822862331872263e-06,
+      "loss": 0.7241,
+      "step": 1470500
+    },
+    {
+      "epoch": 2.5314887253240084,
+      "grad_norm": 2.211066961288452,
+      "learning_rate": 7.808521244599865e-06,
+      "loss": 0.7218,
+      "step": 1471000
+    },
+    {
+      "epoch": 2.5323491905603523,
+      "grad_norm": 2.2090892791748047,
+      "learning_rate": 7.794180157327465e-06,
+      "loss": 0.7251,
+      "step": 1471500
+    },
+    {
+      "epoch": 2.5332096557966963,
+      "grad_norm": 2.221142053604126,
+      "learning_rate": 7.779839070055064e-06,
+      "loss": 0.7253,
+      "step": 1472000
+    },
+    {
+      "epoch": 2.5340701210330403,
+      "grad_norm": 2.2364909648895264,
+      "learning_rate": 7.765497982782664e-06,
+      "loss": 0.7286,
+      "step": 1472500
+    },
+    {
+      "epoch": 2.5349305862693843,
+      "grad_norm": 2.1668782234191895,
+      "learning_rate": 7.751156895510265e-06,
+      "loss": 0.7219,
+      "step": 1473000
+    },
+    {
+      "epoch": 2.5357910515057283,
+      "grad_norm": 2.125824451446533,
+      "learning_rate": 7.736815808237865e-06,
+      "loss": 0.7223,
+      "step": 1473500
+    },
+    {
+      "epoch": 2.5366515167420722,
+      "grad_norm": 1.9486610889434814,
+      "learning_rate": 7.722474720965466e-06,
+      "loss": 0.7259,
+      "step": 1474000
+    },
+    {
+      "epoch": 2.537511981978416,
+      "grad_norm": 2.0973780155181885,
+      "learning_rate": 7.708133633693065e-06,
+      "loss": 0.7234,
+      "step": 1474500
+    },
+    {
+      "epoch": 2.53837244721476,
+      "grad_norm": 2.1345973014831543,
+      "learning_rate": 7.693792546420665e-06,
+      "loss": 0.7221,
+      "step": 1475000
+    },
+    {
+      "epoch": 2.539232912451104,
+      "grad_norm": 2.146336078643799,
+      "learning_rate": 7.679451459148267e-06,
+      "loss": 0.7261,
+      "step": 1475500
+    },
+    {
+      "epoch": 2.540093377687448,
+      "grad_norm": 2.5509378910064697,
+      "learning_rate": 7.665110371875866e-06,
+      "loss": 0.726,
+      "step": 1476000
+    },
+    {
+      "epoch": 2.540953842923792,
+      "grad_norm": 2.422461748123169,
+      "learning_rate": 7.650769284603467e-06,
+      "loss": 0.73,
+      "step": 1476500
+    },
+    {
+      "epoch": 2.541814308160136,
+      "grad_norm": 2.1949241161346436,
+      "learning_rate": 7.636428197331067e-06,
+      "loss": 0.7229,
+      "step": 1477000
+    },
+    {
+      "epoch": 2.54267477339648,
+      "grad_norm": 2.390719413757324,
+      "learning_rate": 7.622087110058667e-06,
+      "loss": 0.7234,
+      "step": 1477500
+    },
+    {
+      "epoch": 2.543535238632824,
+      "grad_norm": 2.194368600845337,
+      "learning_rate": 7.607746022786268e-06,
+      "loss": 0.723,
+      "step": 1478000
+    },
+    {
+      "epoch": 2.544395703869168,
+      "grad_norm": 2.5566372871398926,
+      "learning_rate": 7.593404935513868e-06,
+      "loss": 0.7312,
+      "step": 1478500
+    },
+    {
+      "epoch": 2.545256169105512,
+      "grad_norm": 2.225067615509033,
+      "learning_rate": 7.579063848241467e-06,
+      "loss": 0.7157,
+      "step": 1479000
+    },
+    {
+      "epoch": 2.546116634341856,
+      "grad_norm": 2.281120777130127,
+      "learning_rate": 7.564722760969068e-06,
+      "loss": 0.7259,
+      "step": 1479500
+    },
+    {
+      "epoch": 2.5469770995782,
+      "grad_norm": 2.297755479812622,
+      "learning_rate": 7.550381673696667e-06,
+      "loss": 0.7217,
+      "step": 1480000
+    },
+    {
+      "epoch": 2.547837564814544,
+      "grad_norm": 2.464024066925049,
+      "learning_rate": 7.536040586424269e-06,
+      "loss": 0.7235,
+      "step": 1480500
+    },
+    {
+      "epoch": 2.548698030050888,
+      "grad_norm": 2.3035905361175537,
+      "learning_rate": 7.521699499151868e-06,
+      "loss": 0.722,
+      "step": 1481000
+    },
+    {
+      "epoch": 2.549558495287232,
+      "grad_norm": 2.119723320007324,
+      "learning_rate": 7.507358411879469e-06,
+      "loss": 0.7263,
+      "step": 1481500
+    },
+    {
+      "epoch": 2.550418960523576,
+      "grad_norm": 2.320289373397827,
+      "learning_rate": 7.493017324607068e-06,
+      "loss": 0.7267,
+      "step": 1482000
+    },
+    {
+      "epoch": 2.55127942575992,
+      "grad_norm": 2.233952522277832,
+      "learning_rate": 7.47867623733467e-06,
+      "loss": 0.7249,
+      "step": 1482500
+    },
+    {
+      "epoch": 2.552139890996264,
+      "grad_norm": 8.134510040283203,
+      "learning_rate": 7.464335150062269e-06,
+      "loss": 0.7205,
+      "step": 1483000
+    },
+    {
+      "epoch": 2.553000356232608,
+      "grad_norm": 2.3019843101501465,
+      "learning_rate": 7.44999406278987e-06,
+      "loss": 0.7187,
+      "step": 1483500
+    },
+    {
+      "epoch": 2.5538608214689518,
+      "grad_norm": 2.252256393432617,
+      "learning_rate": 7.435652975517469e-06,
+      "loss": 0.7246,
+      "step": 1484000
+    },
+    {
+      "epoch": 2.5547212867052957,
+      "grad_norm": 2.285295009613037,
+      "learning_rate": 7.421311888245069e-06,
+      "loss": 0.7266,
+      "step": 1484500
+    },
+    {
+      "epoch": 2.5555817519416397,
+      "grad_norm": 2.0101242065429688,
+      "learning_rate": 7.40697080097267e-06,
+      "loss": 0.7238,
+      "step": 1485000
+    },
+    {
+      "epoch": 2.5564422171779837,
+      "grad_norm": 2.1717300415039062,
+      "learning_rate": 7.392629713700271e-06,
+      "loss": 0.7274,
+      "step": 1485500
+    },
+    {
+      "epoch": 2.5573026824143277,
+      "grad_norm": 2.275235652923584,
+      "learning_rate": 7.37828862642787e-06,
+      "loss": 0.7238,
+      "step": 1486000
+    },
+    {
+      "epoch": 2.5581631476506717,
+      "grad_norm": 2.19974684715271,
+      "learning_rate": 7.363947539155471e-06,
+      "loss": 0.7286,
+      "step": 1486500
+    },
+    {
+      "epoch": 2.5590236128870156,
+      "grad_norm": 2.2823832035064697,
+      "learning_rate": 7.349606451883072e-06,
+      "loss": 0.7196,
+      "step": 1487000
+    },
+    {
+      "epoch": 2.5598840781233596,
+      "grad_norm": 2.2638237476348877,
+      "learning_rate": 7.335265364610672e-06,
+      "loss": 0.7199,
+      "step": 1487500
+    },
+    {
+      "epoch": 2.5607445433597036,
+      "grad_norm": 2.039130926132202,
+      "learning_rate": 7.3209242773382714e-06,
+      "loss": 0.7157,
+      "step": 1488000
+    },
+    {
+      "epoch": 2.5616050085960476,
+      "grad_norm": 2.1070847511291504,
+      "learning_rate": 7.306583190065872e-06,
+      "loss": 0.7229,
+      "step": 1488500
+    },
+    {
+      "epoch": 2.5624654738323915,
+      "grad_norm": 2.121931552886963,
+      "learning_rate": 7.2922421027934715e-06,
+      "loss": 0.7133,
+      "step": 1489000
+    },
+    {
+      "epoch": 2.5633259390687355,
+      "grad_norm": 2.1481940746307373,
+      "learning_rate": 7.277901015521073e-06,
+      "loss": 0.7218,
+      "step": 1489500
+    },
+    {
+      "epoch": 2.5641864043050795,
+      "grad_norm": 2.1555721759796143,
+      "learning_rate": 7.2635599282486725e-06,
+      "loss": 0.7166,
+      "step": 1490000
+    },
+    {
+      "epoch": 2.5650468695414235,
+      "grad_norm": 2.187934398651123,
+      "learning_rate": 7.249218840976273e-06,
+      "loss": 0.721,
+      "step": 1490500
+    },
+    {
+      "epoch": 2.5659073347777674,
+      "grad_norm": 2.292440176010132,
+      "learning_rate": 7.2348777537038726e-06,
+      "loss": 0.7282,
+      "step": 1491000
+    },
+    {
+      "epoch": 2.5667678000141114,
+      "grad_norm": 2.3000996112823486,
+      "learning_rate": 7.220536666431474e-06,
+      "loss": 0.7247,
+      "step": 1491500
+    },
+    {
+      "epoch": 2.5676282652504554,
+      "grad_norm": 2.1090991497039795,
+      "learning_rate": 7.2061955791590735e-06,
+      "loss": 0.7236,
+      "step": 1492000
+    },
+    {
+      "epoch": 2.5684887304867994,
+      "grad_norm": 2.2887206077575684,
+      "learning_rate": 7.191854491886674e-06,
+      "loss": 0.7212,
+      "step": 1492500
+    },
+    {
+      "epoch": 2.5693491957231434,
+      "grad_norm": 2.1364707946777344,
+      "learning_rate": 7.177513404614274e-06,
+      "loss": 0.7239,
+      "step": 1493000
+    },
+    {
+      "epoch": 2.5702096609594873,
+      "grad_norm": 2.2923684120178223,
+      "learning_rate": 7.163172317341873e-06,
+      "loss": 0.7195,
+      "step": 1493500
+    },
+    {
+      "epoch": 2.5710701261958313,
+      "grad_norm": 2.278289794921875,
+      "learning_rate": 7.1488312300694745e-06,
+      "loss": 0.7159,
+      "step": 1494000
+    },
+    {
+      "epoch": 2.5719305914321753,
+      "grad_norm": 2.3480536937713623,
+      "learning_rate": 7.134490142797075e-06,
+      "loss": 0.72,
+      "step": 1494500
+    },
+    {
+      "epoch": 2.5727910566685193,
+      "grad_norm": 2.2005200386047363,
+      "learning_rate": 7.120149055524675e-06,
+      "loss": 0.7221,
+      "step": 1495000
+    },
+    {
+      "epoch": 2.5736515219048637,
+      "grad_norm": 2.2753360271453857,
+      "learning_rate": 7.105807968252274e-06,
+      "loss": 0.7264,
+      "step": 1495500
+    },
+    {
+      "epoch": 2.5745119871412077,
+      "grad_norm": 2.1410017013549805,
+      "learning_rate": 7.091466880979875e-06,
+      "loss": 0.7216,
+      "step": 1496000
+    },
+    {
+      "epoch": 2.5753724523775516,
+      "grad_norm": 2.2372915744781494,
+      "learning_rate": 7.077125793707476e-06,
+      "loss": 0.7202,
+      "step": 1496500
+    },
+    {
+      "epoch": 2.5762329176138956,
+      "grad_norm": 2.1890013217926025,
+      "learning_rate": 7.062784706435076e-06,
+      "loss": 0.7207,
+      "step": 1497000
+    },
+    {
+      "epoch": 2.5770933828502396,
+      "grad_norm": 2.078773021697998,
+      "learning_rate": 7.048443619162675e-06,
+      "loss": 0.7203,
+      "step": 1497500
+    },
+    {
+      "epoch": 2.5779538480865836,
+      "grad_norm": 2.2205190658569336,
+      "learning_rate": 7.034102531890276e-06,
+      "loss": 0.7191,
+      "step": 1498000
+    },
+    {
+      "epoch": 2.5788143133229275,
+      "grad_norm": 2.2238242626190186,
+      "learning_rate": 7.019761444617877e-06,
+      "loss": 0.7233,
+      "step": 1498500
+    },
+    {
+      "epoch": 2.5796747785592715,
+      "grad_norm": 2.231977701187134,
+      "learning_rate": 7.005420357345477e-06,
+      "loss": 0.7199,
+      "step": 1499000
+    },
+    {
+      "epoch": 2.5805352437956155,
+      "grad_norm": 2.04296612739563,
+      "learning_rate": 6.991079270073077e-06,
+      "loss": 0.7187,
+      "step": 1499500
+    },
+    {
+      "epoch": 2.5813957090319595,
+      "grad_norm": 2.2166175842285156,
+      "learning_rate": 6.976738182800677e-06,
+      "loss": 0.7213,
+      "step": 1500000
+    },
+    {
+      "epoch": 2.5822561742683035,
+      "grad_norm": 2.2233524322509766,
+      "learning_rate": 6.9623970955282764e-06,
+      "loss": 0.7186,
+      "step": 1500500
+    },
+    {
+      "epoch": 2.5831166395046474,
+      "grad_norm": 2.268268585205078,
+      "learning_rate": 6.948056008255878e-06,
+      "loss": 0.7263,
+      "step": 1501000
+    },
+    {
+      "epoch": 2.5839771047409914,
+      "grad_norm": 2.200861930847168,
+      "learning_rate": 6.933714920983478e-06,
+      "loss": 0.7242,
+      "step": 1501500
+    },
+    {
+      "epoch": 2.5848375699773354,
+      "grad_norm": 2.327687978744507,
+      "learning_rate": 6.919373833711078e-06,
+      "loss": 0.7217,
+      "step": 1502000
+    },
+    {
+      "epoch": 2.5856980352136794,
+      "grad_norm": 2.0236458778381348,
+      "learning_rate": 6.9050327464386775e-06,
+      "loss": 0.7195,
+      "step": 1502500
+    },
+    {
+      "epoch": 2.5865585004500233,
+      "grad_norm": 2.2688241004943848,
+      "learning_rate": 6.890691659166279e-06,
+      "loss": 0.7213,
+      "step": 1503000
+    },
+    {
+      "epoch": 2.5874189656863673,
+      "grad_norm": 2.138439893722534,
+      "learning_rate": 6.876350571893879e-06,
+      "loss": 0.7229,
+      "step": 1503500
+    },
+    {
+      "epoch": 2.5882794309227113,
+      "grad_norm": 2.194458246231079,
+      "learning_rate": 6.862009484621479e-06,
+      "loss": 0.7224,
+      "step": 1504000
+    },
+    {
+      "epoch": 2.5891398961590553,
+      "grad_norm": 2.2974252700805664,
+      "learning_rate": 6.8476683973490785e-06,
+      "loss": 0.7232,
+      "step": 1504500
+    },
+    {
+      "epoch": 2.5900003613953992,
+      "grad_norm": 2.4050469398498535,
+      "learning_rate": 6.833327310076679e-06,
+      "loss": 0.7218,
+      "step": 1505000
+    },
+    {
+      "epoch": 2.590860826631743,
+      "grad_norm": 2.0813820362091064,
+      "learning_rate": 6.81898622280428e-06,
+      "loss": 0.7183,
+      "step": 1505500
+    },
+    {
+      "epoch": 2.591721291868087,
+      "grad_norm": 2.3573648929595947,
+      "learning_rate": 6.80464513553188e-06,
+      "loss": 0.7178,
+      "step": 1506000
+    },
+    {
+      "epoch": 2.592581757104431,
+      "grad_norm": 2.2958922386169434,
+      "learning_rate": 6.7903040482594795e-06,
+      "loss": 0.7174,
+      "step": 1506500
+    },
+    {
+      "epoch": 2.593442222340775,
+      "grad_norm": 2.2089388370513916,
+      "learning_rate": 6.77596296098708e-06,
+      "loss": 0.7242,
+      "step": 1507000
+    },
+    {
+      "epoch": 2.594302687577119,
+      "grad_norm": 2.1768064498901367,
+      "learning_rate": 6.76162187371468e-06,
+      "loss": 0.7232,
+      "step": 1507500
+    },
+    {
+      "epoch": 2.595163152813463,
+      "grad_norm": 2.0804152488708496,
+      "learning_rate": 6.747280786442281e-06,
+      "loss": 0.7218,
+      "step": 1508000
+    },
+    {
+      "epoch": 2.596023618049807,
+      "grad_norm": 2.265146255493164,
+      "learning_rate": 6.7329396991698806e-06,
+      "loss": 0.7233,
+      "step": 1508500
+    },
+    {
+      "epoch": 2.596884083286151,
+      "grad_norm": 2.0737407207489014,
+      "learning_rate": 6.718598611897481e-06,
+      "loss": 0.7142,
+      "step": 1509000
+    },
+    {
+      "epoch": 2.597744548522495,
+      "grad_norm": 2.1167831420898438,
+      "learning_rate": 6.704257524625081e-06,
+      "loss": 0.7221,
+      "step": 1509500
+    },
+    {
+      "epoch": 2.598605013758839,
+      "grad_norm": 2.4157352447509766,
+      "learning_rate": 6.689916437352682e-06,
+      "loss": 0.7193,
+      "step": 1510000
+    },
+    {
+      "epoch": 2.599465478995183,
+      "grad_norm": 2.2811946868896484,
+      "learning_rate": 6.675575350080282e-06,
+      "loss": 0.7244,
+      "step": 1510500
+    },
+    {
+      "epoch": 2.600325944231527,
+      "grad_norm": 2.2633402347564697,
+      "learning_rate": 6.661234262807882e-06,
+      "loss": 0.7217,
+      "step": 1511000
+    },
+    {
+      "epoch": 2.601186409467871,
+      "grad_norm": 2.3065993785858154,
+      "learning_rate": 6.646893175535482e-06,
+      "loss": 0.7236,
+      "step": 1511500
+    },
+    {
+      "epoch": 2.602046874704215,
+      "grad_norm": 2.197575569152832,
+      "learning_rate": 6.632552088263082e-06,
+      "loss": 0.7213,
+      "step": 1512000
+    },
+    {
+      "epoch": 2.602907339940559,
+      "grad_norm": 2.446925640106201,
+      "learning_rate": 6.6182110009906835e-06,
+      "loss": 0.7154,
+      "step": 1512500
+    },
+    {
+      "epoch": 2.6037678051769033,
+      "grad_norm": 2.5222103595733643,
+      "learning_rate": 6.603869913718283e-06,
+      "loss": 0.7174,
+      "step": 1513000
+    },
+    {
+      "epoch": 2.6046282704132473,
+      "grad_norm": 2.3834874629974365,
+      "learning_rate": 6.589528826445883e-06,
+      "loss": 0.7262,
+      "step": 1513500
+    },
+    {
+      "epoch": 2.6054887356495913,
+      "grad_norm": 2.1265454292297363,
+      "learning_rate": 6.575187739173483e-06,
+      "loss": 0.7213,
+      "step": 1514000
+    },
+    {
+      "epoch": 2.6063492008859352,
+      "grad_norm": 2.1744658946990967,
+      "learning_rate": 6.5608466519010845e-06,
+      "loss": 0.7266,
+      "step": 1514500
+    },
+    {
+      "epoch": 2.6072096661222792,
+      "grad_norm": 2.554331064224243,
+      "learning_rate": 6.546505564628684e-06,
+      "loss": 0.7233,
+      "step": 1515000
+    },
+    {
+      "epoch": 2.608070131358623,
+      "grad_norm": 2.343867778778076,
+      "learning_rate": 6.532164477356284e-06,
+      "loss": 0.7138,
+      "step": 1515500
+    },
+    {
+      "epoch": 2.608930596594967,
+      "grad_norm": 2.1199309825897217,
+      "learning_rate": 6.517823390083884e-06,
+      "loss": 0.7177,
+      "step": 1516000
+    },
+    {
+      "epoch": 2.609791061831311,
+      "grad_norm": 2.4904963970184326,
+      "learning_rate": 6.503482302811484e-06,
+      "loss": 0.7158,
+      "step": 1516500
+    },
+    {
+      "epoch": 2.610651527067655,
+      "grad_norm": 2.239297389984131,
+      "learning_rate": 6.489141215539085e-06,
+      "loss": 0.7218,
+      "step": 1517000
+    },
+    {
+      "epoch": 2.611511992303999,
+      "grad_norm": 2.024224042892456,
+      "learning_rate": 6.474800128266685e-06,
+      "loss": 0.7141,
+      "step": 1517500
+    },
+    {
+      "epoch": 2.612372457540343,
+      "grad_norm": 3.024536609649658,
+      "learning_rate": 6.460459040994285e-06,
+      "loss": 0.7165,
+      "step": 1518000
+    },
+    {
+      "epoch": 2.613232922776687,
+      "grad_norm": 2.288699150085449,
+      "learning_rate": 6.446117953721885e-06,
+      "loss": 0.7231,
+      "step": 1518500
+    },
+    {
+      "epoch": 2.614093388013031,
+      "grad_norm": 2.127619743347168,
+      "learning_rate": 6.431776866449486e-06,
+      "loss": 0.7144,
+      "step": 1519000
+    },
+    {
+      "epoch": 2.614953853249375,
+      "grad_norm": 2.112839460372925,
+      "learning_rate": 6.417435779177086e-06,
+      "loss": 0.7263,
+      "step": 1519500
+    },
+    {
+      "epoch": 2.615814318485719,
+      "grad_norm": 2.175316333770752,
+      "learning_rate": 6.403094691904686e-06,
+      "loss": 0.7233,
+      "step": 1520000
+    },
+    {
+      "epoch": 2.616674783722063,
+      "grad_norm": 2.203071355819702,
+      "learning_rate": 6.388753604632286e-06,
+      "loss": 0.7239,
+      "step": 1520500
+    },
+    {
+      "epoch": 2.617535248958407,
+      "grad_norm": 2.312666416168213,
+      "learning_rate": 6.3744125173598856e-06,
+      "loss": 0.7222,
+      "step": 1521000
+    },
+    {
+      "epoch": 2.618395714194751,
+      "grad_norm": 2.164508104324341,
+      "learning_rate": 6.360071430087487e-06,
+      "loss": 0.7194,
+      "step": 1521500
+    },
+    {
+      "epoch": 2.619256179431095,
+      "grad_norm": 2.3181352615356445,
+      "learning_rate": 6.345730342815087e-06,
+      "loss": 0.7183,
+      "step": 1522000
+    },
+    {
+      "epoch": 2.620116644667439,
+      "grad_norm": 2.6277377605438232,
+      "learning_rate": 6.331389255542687e-06,
+      "loss": 0.7212,
+      "step": 1522500
+    },
+    {
+      "epoch": 2.620977109903783,
+      "grad_norm": 2.370537519454956,
+      "learning_rate": 6.317048168270287e-06,
+      "loss": 0.7193,
+      "step": 1523000
+    },
+    {
+      "epoch": 2.621837575140127,
+      "grad_norm": 2.3331568241119385,
+      "learning_rate": 6.302707080997887e-06,
+      "loss": 0.7237,
+      "step": 1523500
+    },
+    {
+      "epoch": 2.622698040376471,
+      "grad_norm": 2.125333547592163,
+      "learning_rate": 6.288365993725488e-06,
+      "loss": 0.7228,
+      "step": 1524000
+    },
+    {
+      "epoch": 2.623558505612815,
+      "grad_norm": 2.0734002590179443,
+      "learning_rate": 6.274024906453088e-06,
+      "loss": 0.7272,
+      "step": 1524500
+    },
+    {
+      "epoch": 2.6244189708491588,
+      "grad_norm": 2.216912031173706,
+      "learning_rate": 6.2596838191806885e-06,
+      "loss": 0.7157,
+      "step": 1525000
+    },
+    {
+      "epoch": 2.6252794360855027,
+      "grad_norm": 2.442035436630249,
+      "learning_rate": 6.245342731908289e-06,
+      "loss": 0.7155,
+      "step": 1525500
+    },
+    {
+      "epoch": 2.6261399013218467,
+      "grad_norm": 2.1774885654449463,
+      "learning_rate": 6.2310016446358886e-06,
+      "loss": 0.7178,
+      "step": 1526000
+    },
+    {
+      "epoch": 2.6270003665581907,
+      "grad_norm": 2.2715914249420166,
+      "learning_rate": 6.216660557363489e-06,
+      "loss": 0.7141,
+      "step": 1526500
+    },
+    {
+      "epoch": 2.6278608317945347,
+      "grad_norm": 2.1633667945861816,
+      "learning_rate": 6.2023194700910895e-06,
+      "loss": 0.7173,
+      "step": 1527000
+    },
+    {
+      "epoch": 2.6287212970308786,
+      "grad_norm": 2.4417827129364014,
+      "learning_rate": 6.187978382818689e-06,
+      "loss": 0.7247,
+      "step": 1527500
+    },
+    {
+      "epoch": 2.6295817622672226,
+      "grad_norm": 2.2331953048706055,
+      "learning_rate": 6.17363729554629e-06,
+      "loss": 0.7194,
+      "step": 1528000
+    },
+    {
+      "epoch": 2.6304422275035666,
+      "grad_norm": 2.3902461528778076,
+      "learning_rate": 6.159296208273889e-06,
+      "loss": 0.7189,
+      "step": 1528500
+    },
+    {
+      "epoch": 2.6313026927399106,
+      "grad_norm": 2.0932114124298096,
+      "learning_rate": 6.1449551210014905e-06,
+      "loss": 0.719,
+      "step": 1529000
+    },
+    {
+      "epoch": 2.6321631579762546,
+      "grad_norm": 2.4338607788085938,
+      "learning_rate": 6.13061403372909e-06,
+      "loss": 0.7136,
+      "step": 1529500
+    },
+    {
+      "epoch": 2.6330236232125985,
+      "grad_norm": 2.4229514598846436,
+      "learning_rate": 6.116272946456691e-06,
+      "loss": 0.719,
+      "step": 1530000
+    },
+    {
+      "epoch": 2.6338840884489425,
+      "grad_norm": 2.219041109085083,
+      "learning_rate": 6.101931859184291e-06,
+      "loss": 0.715,
+      "step": 1530500
+    },
+    {
+      "epoch": 2.6347445536852865,
+      "grad_norm": 2.183692455291748,
+      "learning_rate": 6.087590771911891e-06,
+      "loss": 0.7195,
+      "step": 1531000
+    },
+    {
+      "epoch": 2.6356050189216305,
+      "grad_norm": 2.0069658756256104,
+      "learning_rate": 6.073249684639491e-06,
+      "loss": 0.7184,
+      "step": 1531500
+    },
+    {
+      "epoch": 2.6364654841579744,
+      "grad_norm": 2.0577995777130127,
+      "learning_rate": 6.058908597367091e-06,
+      "loss": 0.7165,
+      "step": 1532000
+    },
+    {
+      "epoch": 2.6373259493943184,
+      "grad_norm": 2.0661566257476807,
+      "learning_rate": 6.044567510094692e-06,
+      "loss": 0.7172,
+      "step": 1532500
+    },
+    {
+      "epoch": 2.6381864146306624,
+      "grad_norm": 2.362360954284668,
+      "learning_rate": 6.030226422822292e-06,
+      "loss": 0.7185,
+      "step": 1533000
+    },
+    {
+      "epoch": 2.6390468798670064,
+      "grad_norm": 2.183109998703003,
+      "learning_rate": 6.015885335549892e-06,
+      "loss": 0.7221,
+      "step": 1533500
+    },
+    {
+      "epoch": 2.6399073451033503,
+      "grad_norm": 3.1113739013671875,
+      "learning_rate": 6.001544248277492e-06,
+      "loss": 0.7196,
+      "step": 1534000
+    },
+    {
+      "epoch": 2.6407678103396943,
+      "grad_norm": 2.251563787460327,
+      "learning_rate": 5.987203161005093e-06,
+      "loss": 0.7117,
+      "step": 1534500
+    },
+    {
+      "epoch": 2.6416282755760383,
+      "grad_norm": 2.264225721359253,
+      "learning_rate": 5.972862073732693e-06,
+      "loss": 0.7211,
+      "step": 1535000
+    },
+    {
+      "epoch": 2.6424887408123823,
+      "grad_norm": 2.278864860534668,
+      "learning_rate": 5.958520986460292e-06,
+      "loss": 0.7175,
+      "step": 1535500
+    },
+    {
+      "epoch": 2.6433492060487263,
+      "grad_norm": 2.30007004737854,
+      "learning_rate": 5.944179899187893e-06,
+      "loss": 0.7235,
+      "step": 1536000
+    },
+    {
+      "epoch": 2.6442096712850702,
+      "grad_norm": 2.183664560317993,
+      "learning_rate": 5.929838811915493e-06,
+      "loss": 0.7182,
+      "step": 1536500
+    },
+    {
+      "epoch": 2.645070136521414,
+      "grad_norm": 2.0430800914764404,
+      "learning_rate": 5.915497724643094e-06,
+      "loss": 0.7184,
+      "step": 1537000
+    },
+    {
+      "epoch": 2.645930601757758,
+      "grad_norm": 2.242837429046631,
+      "learning_rate": 5.9011566373706935e-06,
+      "loss": 0.7154,
+      "step": 1537500
+    },
+    {
+      "epoch": 2.646791066994102,
+      "grad_norm": 2.0278875827789307,
+      "learning_rate": 5.886815550098295e-06,
+      "loss": 0.723,
+      "step": 1538000
+    },
+    {
+      "epoch": 2.647651532230446,
+      "grad_norm": 2.1112453937530518,
+      "learning_rate": 5.872474462825894e-06,
+      "loss": 0.7193,
+      "step": 1538500
+    },
+    {
+      "epoch": 2.64851199746679,
+      "grad_norm": 2.058110475540161,
+      "learning_rate": 5.858133375553494e-06,
+      "loss": 0.7187,
+      "step": 1539000
+    },
+    {
+      "epoch": 2.649372462703134,
+      "grad_norm": 2.3659400939941406,
+      "learning_rate": 5.8437922882810945e-06,
+      "loss": 0.714,
+      "step": 1539500
+    },
+    {
+      "epoch": 2.650232927939478,
+      "grad_norm": 1.9934048652648926,
+      "learning_rate": 5.829451201008695e-06,
+      "loss": 0.7185,
+      "step": 1540000
+    },
+    {
+      "epoch": 2.651093393175822,
+      "grad_norm": 2.47733473777771,
+      "learning_rate": 5.8151101137362954e-06,
+      "loss": 0.7158,
+      "step": 1540500
+    },
+    {
+      "epoch": 2.651953858412166,
+      "grad_norm": 2.123170852661133,
+      "learning_rate": 5.800769026463895e-06,
+      "loss": 0.7115,
+      "step": 1541000
+    },
+    {
+      "epoch": 2.6528143236485104,
+      "grad_norm": 2.4889888763427734,
+      "learning_rate": 5.7864279391914955e-06,
+      "loss": 0.7171,
+      "step": 1541500
+    },
+    {
+      "epoch": 2.6536747888848544,
+      "grad_norm": 2.187666177749634,
+      "learning_rate": 5.772086851919096e-06,
+      "loss": 0.7199,
+      "step": 1542000
+    },
+    {
+      "epoch": 2.6545352541211984,
+      "grad_norm": 2.203145742416382,
+      "learning_rate": 5.7577457646466965e-06,
+      "loss": 0.7189,
+      "step": 1542500
+    },
+    {
+      "epoch": 2.6553957193575424,
+      "grad_norm": 2.287001371383667,
+      "learning_rate": 5.743404677374296e-06,
+      "loss": 0.7182,
+      "step": 1543000
+    },
+    {
+      "epoch": 2.6562561845938863,
+      "grad_norm": 2.1928348541259766,
+      "learning_rate": 5.7290635901018966e-06,
+      "loss": 0.7168,
+      "step": 1543500
+    },
+    {
+      "epoch": 2.6571166498302303,
+      "grad_norm": 2.1668388843536377,
+      "learning_rate": 5.714722502829497e-06,
+      "loss": 0.7168,
+      "step": 1544000
+    },
+    {
+      "epoch": 2.6579771150665743,
+      "grad_norm": 2.414255380630493,
+      "learning_rate": 5.700381415557097e-06,
+      "loss": 0.7173,
+      "step": 1544500
+    },
+    {
+      "epoch": 2.6588375803029183,
+      "grad_norm": 2.3311398029327393,
+      "learning_rate": 5.686040328284697e-06,
+      "loss": 0.7201,
+      "step": 1545000
+    },
+    {
+      "epoch": 2.6596980455392623,
+      "grad_norm": 2.3980069160461426,
+      "learning_rate": 5.671699241012298e-06,
+      "loss": 0.7238,
+      "step": 1545500
+    },
+    {
+      "epoch": 2.6605585107756062,
+      "grad_norm": 2.217282772064209,
+      "learning_rate": 5.657358153739898e-06,
+      "loss": 0.7228,
+      "step": 1546000
+    },
+    {
+      "epoch": 2.66141897601195,
+      "grad_norm": 2.153686285018921,
+      "learning_rate": 5.643017066467498e-06,
+      "loss": 0.7108,
+      "step": 1546500
+    },
+    {
+      "epoch": 2.662279441248294,
+      "grad_norm": 2.188215494155884,
+      "learning_rate": 5.628675979195098e-06,
+      "loss": 0.7144,
+      "step": 1547000
+    },
+    {
+      "epoch": 2.663139906484638,
+      "grad_norm": 2.301560640335083,
+      "learning_rate": 5.614334891922699e-06,
+      "loss": 0.7198,
+      "step": 1547500
+    },
+    {
+      "epoch": 2.664000371720982,
+      "grad_norm": 2.305785894393921,
+      "learning_rate": 5.599993804650298e-06,
+      "loss": 0.7208,
+      "step": 1548000
+    },
+    {
+      "epoch": 2.664860836957326,
+      "grad_norm": 2.377692937850952,
+      "learning_rate": 5.585652717377899e-06,
+      "loss": 0.7178,
+      "step": 1548500
+    },
+    {
+      "epoch": 2.66572130219367,
+      "grad_norm": 2.112016439437866,
+      "learning_rate": 5.571311630105499e-06,
+      "loss": 0.7147,
+      "step": 1549000
+    },
+    {
+      "epoch": 2.666581767430014,
+      "grad_norm": 2.1709604263305664,
+      "learning_rate": 5.5569705428331e-06,
+      "loss": 0.7178,
+      "step": 1549500
+    },
+    {
+      "epoch": 2.667442232666358,
+      "grad_norm": 2.2401998043060303,
+      "learning_rate": 5.542629455560699e-06,
+      "loss": 0.7196,
+      "step": 1550000
+    },
+    {
+      "epoch": 2.668302697902702,
+      "grad_norm": 2.144211769104004,
+      "learning_rate": 5.5282883682883e-06,
+      "loss": 0.7187,
+      "step": 1550500
+    },
+    {
+      "epoch": 2.669163163139046,
+      "grad_norm": 2.3315768241882324,
+      "learning_rate": 5.5139472810159e-06,
+      "loss": 0.7158,
+      "step": 1551000
+    },
+    {
+      "epoch": 2.67002362837539,
+      "grad_norm": 2.303516149520874,
+      "learning_rate": 5.4996061937435e-06,
+      "loss": 0.711,
+      "step": 1551500
+    },
+    {
+      "epoch": 2.670884093611734,
+      "grad_norm": 2.337053060531616,
+      "learning_rate": 5.4852651064711e-06,
+      "loss": 0.722,
+      "step": 1552000
+    },
+    {
+      "epoch": 2.671744558848078,
+      "grad_norm": 2.1762800216674805,
+      "learning_rate": 5.470924019198701e-06,
+      "loss": 0.716,
+      "step": 1552500
+    },
+    {
+      "epoch": 2.672605024084422,
+      "grad_norm": 2.34902286529541,
+      "learning_rate": 5.456582931926301e-06,
+      "loss": 0.716,
+      "step": 1553000
+    },
+    {
+      "epoch": 2.673465489320766,
+      "grad_norm": 2.3098535537719727,
+      "learning_rate": 5.442241844653901e-06,
+      "loss": 0.7156,
+      "step": 1553500
+    },
+    {
+      "epoch": 2.67432595455711,
+      "grad_norm": 2.463733434677124,
+      "learning_rate": 5.427900757381501e-06,
+      "loss": 0.7184,
+      "step": 1554000
+    },
+    {
+      "epoch": 2.675186419793454,
+      "grad_norm": 2.3530120849609375,
+      "learning_rate": 5.413559670109102e-06,
+      "loss": 0.7143,
+      "step": 1554500
+    },
+    {
+      "epoch": 2.676046885029798,
+      "grad_norm": 2.367659330368042,
+      "learning_rate": 5.3992185828367014e-06,
+      "loss": 0.719,
+      "step": 1555000
+    },
+    {
+      "epoch": 2.676907350266142,
+      "grad_norm": 2.357699155807495,
+      "learning_rate": 5.384877495564302e-06,
+      "loss": 0.7192,
+      "step": 1555500
+    },
+    {
+      "epoch": 2.6777678155024858,
+      "grad_norm": 2.980602502822876,
+      "learning_rate": 5.370536408291902e-06,
+      "loss": 0.7164,
+      "step": 1556000
+    },
+    {
+      "epoch": 2.6786282807388297,
+      "grad_norm": 2.181764602661133,
+      "learning_rate": 5.356195321019503e-06,
+      "loss": 0.7233,
+      "step": 1556500
+    },
+    {
+      "epoch": 2.6794887459751737,
+      "grad_norm": 2.156656265258789,
+      "learning_rate": 5.3418542337471025e-06,
+      "loss": 0.7187,
+      "step": 1557000
+    },
+    {
+      "epoch": 2.6803492112115177,
+      "grad_norm": 2.2862629890441895,
+      "learning_rate": 5.327513146474703e-06,
+      "loss": 0.723,
+      "step": 1557500
+    },
+    {
+      "epoch": 2.6812096764478617,
+      "grad_norm": 2.340113639831543,
+      "learning_rate": 5.3131720592023034e-06,
+      "loss": 0.7213,
+      "step": 1558000
+    },
+    {
+      "epoch": 2.6820701416842057,
+      "grad_norm": 2.258385419845581,
+      "learning_rate": 5.298830971929903e-06,
+      "loss": 0.7208,
+      "step": 1558500
+    },
+    {
+      "epoch": 2.68293060692055,
+      "grad_norm": 2.2693705558776855,
+      "learning_rate": 5.2844898846575035e-06,
+      "loss": 0.7172,
+      "step": 1559000
+    },
+    {
+      "epoch": 2.683791072156894,
+      "grad_norm": 2.1848247051239014,
+      "learning_rate": 5.270148797385103e-06,
+      "loss": 0.7187,
+      "step": 1559500
+    },
+    {
+      "epoch": 2.684651537393238,
+      "grad_norm": 2.1628623008728027,
+      "learning_rate": 5.2558077101127045e-06,
+      "loss": 0.717,
+      "step": 1560000
+    },
+    {
+      "epoch": 2.685512002629582,
+      "grad_norm": 2.308980703353882,
+      "learning_rate": 5.241466622840304e-06,
+      "loss": 0.7178,
+      "step": 1560500
+    },
+    {
+      "epoch": 2.686372467865926,
+      "grad_norm": 2.333484649658203,
+      "learning_rate": 5.2271255355679046e-06,
+      "loss": 0.7124,
+      "step": 1561000
+    },
+    {
+      "epoch": 2.68723293310227,
+      "grad_norm": 2.2160236835479736,
+      "learning_rate": 5.212784448295505e-06,
+      "loss": 0.7181,
+      "step": 1561500
+    },
+    {
+      "epoch": 2.688093398338614,
+      "grad_norm": 2.2782230377197266,
+      "learning_rate": 5.1984433610231055e-06,
+      "loss": 0.7144,
+      "step": 1562000
+    },
+    {
+      "epoch": 2.688953863574958,
+      "grad_norm": 2.1509571075439453,
+      "learning_rate": 5.184102273750705e-06,
+      "loss": 0.7162,
+      "step": 1562500
+    },
+    {
+      "epoch": 2.689814328811302,
+      "grad_norm": 2.432591199874878,
+      "learning_rate": 5.169761186478305e-06,
+      "loss": 0.7184,
+      "step": 1563000
+    },
+    {
+      "epoch": 2.690674794047646,
+      "grad_norm": 2.318580389022827,
+      "learning_rate": 5.155420099205906e-06,
+      "loss": 0.7211,
+      "step": 1563500
+    },
+    {
+      "epoch": 2.69153525928399,
+      "grad_norm": 2.24094820022583,
+      "learning_rate": 5.141079011933506e-06,
+      "loss": 0.7178,
+      "step": 1564000
+    },
+    {
+      "epoch": 2.692395724520334,
+      "grad_norm": 2.3038113117218018,
+      "learning_rate": 5.126737924661106e-06,
+      "loss": 0.7092,
+      "step": 1564500
+    },
+    {
+      "epoch": 2.693256189756678,
+      "grad_norm": 2.427248954772949,
+      "learning_rate": 5.112396837388706e-06,
+      "loss": 0.7157,
+      "step": 1565000
+    },
+    {
+      "epoch": 2.6941166549930218,
+      "grad_norm": 2.2189440727233887,
+      "learning_rate": 5.098055750116307e-06,
+      "loss": 0.7186,
+      "step": 1565500
+    },
+    {
+      "epoch": 2.6949771202293658,
+      "grad_norm": 2.4753639698028564,
+      "learning_rate": 5.083714662843907e-06,
+      "loss": 0.7176,
+      "step": 1566000
+    },
+    {
+      "epoch": 2.6958375854657097,
+      "grad_norm": 2.243990421295166,
+      "learning_rate": 5.069373575571506e-06,
+      "loss": 0.7159,
+      "step": 1566500
+    },
+    {
+      "epoch": 2.6966980507020537,
+      "grad_norm": 2.1928725242614746,
+      "learning_rate": 5.055032488299107e-06,
+      "loss": 0.7201,
+      "step": 1567000
+    },
+    {
+      "epoch": 2.6975585159383977,
+      "grad_norm": 16.473033905029297,
+      "learning_rate": 5.040691401026707e-06,
+      "loss": 0.713,
+      "step": 1567500
+    },
+    {
+      "epoch": 2.6984189811747417,
+      "grad_norm": 2.0748605728149414,
+      "learning_rate": 5.026350313754308e-06,
+      "loss": 0.719,
+      "step": 1568000
+    },
+    {
+      "epoch": 2.6992794464110856,
+      "grad_norm": 2.2077245712280273,
+      "learning_rate": 5.012009226481907e-06,
+      "loss": 0.7133,
+      "step": 1568500
+    },
+    {
+      "epoch": 2.7001399116474296,
+      "grad_norm": 2.1680569648742676,
+      "learning_rate": 4.997668139209509e-06,
+      "loss": 0.7192,
+      "step": 1569000
+    },
+    {
+      "epoch": 2.7010003768837736,
+      "grad_norm": 2.239652156829834,
+      "learning_rate": 4.983327051937108e-06,
+      "loss": 0.7151,
+      "step": 1569500
+    },
+    {
+      "epoch": 2.7018608421201176,
+      "grad_norm": 2.357666015625,
+      "learning_rate": 4.968985964664708e-06,
+      "loss": 0.7252,
+      "step": 1570000
+    },
+    {
+      "epoch": 2.7027213073564615,
+      "grad_norm": 2.201154947280884,
+      "learning_rate": 4.954644877392308e-06,
+      "loss": 0.7147,
+      "step": 1570500
+    },
+    {
+      "epoch": 2.7035817725928055,
+      "grad_norm": 3.680764675140381,
+      "learning_rate": 4.940303790119909e-06,
+      "loss": 0.7137,
+      "step": 1571000
+    },
+    {
+      "epoch": 2.7044422378291495,
+      "grad_norm": 2.185701608657837,
+      "learning_rate": 4.925962702847509e-06,
+      "loss": 0.7142,
+      "step": 1571500
+    },
+    {
+      "epoch": 2.7053027030654935,
+      "grad_norm": 2.305809736251831,
+      "learning_rate": 4.911621615575109e-06,
+      "loss": 0.7244,
+      "step": 1572000
+    },
+    {
+      "epoch": 2.7061631683018375,
+      "grad_norm": 2.3235483169555664,
+      "learning_rate": 4.8972805283027094e-06,
+      "loss": 0.7162,
+      "step": 1572500
+    },
+    {
+      "epoch": 2.7070236335381814,
+      "grad_norm": 9.165050506591797,
+      "learning_rate": 4.88293944103031e-06,
+      "loss": 0.718,
+      "step": 1573000
+    },
+    {
+      "epoch": 2.7078840987745254,
+      "grad_norm": 2.1549155712127686,
+      "learning_rate": 4.86859835375791e-06,
+      "loss": 0.7105,
+      "step": 1573500
+    },
+    {
+      "epoch": 2.7087445640108694,
+      "grad_norm": 2.574151039123535,
+      "learning_rate": 4.85425726648551e-06,
+      "loss": 0.7169,
+      "step": 1574000
+    },
+    {
+      "epoch": 2.7096050292472134,
+      "grad_norm": 2.3129498958587646,
+      "learning_rate": 4.8399161792131105e-06,
+      "loss": 0.7116,
+      "step": 1574500
+    },
+    {
+      "epoch": 2.7104654944835573,
+      "grad_norm": 2.2015106678009033,
+      "learning_rate": 4.825575091940711e-06,
+      "loss": 0.7174,
+      "step": 1575000
+    },
+    {
+      "epoch": 2.7113259597199013,
+      "grad_norm": 2.3222100734710693,
+      "learning_rate": 4.8112340046683106e-06,
+      "loss": 0.7202,
+      "step": 1575500
+    },
+    {
+      "epoch": 2.7121864249562453,
+      "grad_norm": 2.226888418197632,
+      "learning_rate": 4.796892917395911e-06,
+      "loss": 0.718,
+      "step": 1576000
+    },
+    {
+      "epoch": 2.7130468901925893,
+      "grad_norm": 2.2705578804016113,
+      "learning_rate": 4.7825518301235115e-06,
+      "loss": 0.7136,
+      "step": 1576500
+    },
+    {
+      "epoch": 2.7139073554289332,
+      "grad_norm": 2.3102686405181885,
+      "learning_rate": 4.768210742851112e-06,
+      "loss": 0.7222,
+      "step": 1577000
+    },
+    {
+      "epoch": 2.714767820665277,
+      "grad_norm": 2.3631961345672607,
+      "learning_rate": 4.753869655578712e-06,
+      "loss": 0.7129,
+      "step": 1577500
+    },
+    {
+      "epoch": 2.715628285901621,
+      "grad_norm": 2.1715190410614014,
+      "learning_rate": 4.739528568306312e-06,
+      "loss": 0.718,
+      "step": 1578000
+    },
+    {
+      "epoch": 2.716488751137965,
+      "grad_norm": 2.1148438453674316,
+      "learning_rate": 4.7251874810339125e-06,
+      "loss": 0.7113,
+      "step": 1578500
+    },
+    {
+      "epoch": 2.717349216374309,
+      "grad_norm": 2.140204668045044,
+      "learning_rate": 4.710846393761512e-06,
+      "loss": 0.7188,
+      "step": 1579000
+    },
+    {
+      "epoch": 2.718209681610653,
+      "grad_norm": 2.269834041595459,
+      "learning_rate": 4.696505306489113e-06,
+      "loss": 0.7139,
+      "step": 1579500
+    },
+    {
+      "epoch": 2.719070146846997,
+      "grad_norm": 2.177370548248291,
+      "learning_rate": 4.682164219216713e-06,
+      "loss": 0.7139,
+      "step": 1580000
+    },
+    {
+      "epoch": 2.719930612083341,
+      "grad_norm": 2.2315120697021484,
+      "learning_rate": 4.667823131944314e-06,
+      "loss": 0.7171,
+      "step": 1580500
+    },
+    {
+      "epoch": 2.720791077319685,
+      "grad_norm": 2.628296375274658,
+      "learning_rate": 4.653482044671913e-06,
+      "loss": 0.7164,
+      "step": 1581000
+    },
+    {
+      "epoch": 2.721651542556029,
+      "grad_norm": 2.3418428897857666,
+      "learning_rate": 4.639140957399514e-06,
+      "loss": 0.7134,
+      "step": 1581500
+    },
+    {
+      "epoch": 2.722512007792373,
+      "grad_norm": 2.3924858570098877,
+      "learning_rate": 4.624799870127114e-06,
+      "loss": 0.7081,
+      "step": 1582000
+    },
+    {
+      "epoch": 2.723372473028717,
+      "grad_norm": 2.1358137130737305,
+      "learning_rate": 4.610458782854714e-06,
+      "loss": 0.7163,
+      "step": 1582500
+    },
+    {
+      "epoch": 2.724232938265061,
+      "grad_norm": 2.155128240585327,
+      "learning_rate": 4.596117695582314e-06,
+      "loss": 0.7186,
+      "step": 1583000
+    },
+    {
+      "epoch": 2.725093403501405,
+      "grad_norm": 2.1119282245635986,
+      "learning_rate": 4.581776608309915e-06,
+      "loss": 0.7141,
+      "step": 1583500
+    },
+    {
+      "epoch": 2.725953868737749,
+      "grad_norm": 2.3449249267578125,
+      "learning_rate": 4.567435521037515e-06,
+      "loss": 0.7169,
+      "step": 1584000
+    },
+    {
+      "epoch": 2.726814333974093,
+      "grad_norm": 2.5031585693359375,
+      "learning_rate": 4.553094433765115e-06,
+      "loss": 0.7203,
+      "step": 1584500
+    },
+    {
+      "epoch": 2.727674799210437,
+      "grad_norm": 2.304473400115967,
+      "learning_rate": 4.538753346492715e-06,
+      "loss": 0.7182,
+      "step": 1585000
+    },
+    {
+      "epoch": 2.728535264446781,
+      "grad_norm": 2.4063103199005127,
+      "learning_rate": 4.524412259220316e-06,
+      "loss": 0.7122,
+      "step": 1585500
+    },
+    {
+      "epoch": 2.729395729683125,
+      "grad_norm": 2.4361159801483154,
+      "learning_rate": 4.510071171947915e-06,
+      "loss": 0.7111,
+      "step": 1586000
+    },
+    {
+      "epoch": 2.730256194919469,
+      "grad_norm": 2.2181286811828613,
+      "learning_rate": 4.495730084675516e-06,
+      "loss": 0.7176,
+      "step": 1586500
+    },
+    {
+      "epoch": 2.731116660155813,
+      "grad_norm": 2.414072036743164,
+      "learning_rate": 4.481388997403116e-06,
+      "loss": 0.7226,
+      "step": 1587000
+    },
+    {
+      "epoch": 2.7319771253921568,
+      "grad_norm": 2.095151662826538,
+      "learning_rate": 4.467047910130717e-06,
+      "loss": 0.7134,
+      "step": 1587500
+    },
+    {
+      "epoch": 2.732837590628501,
+      "grad_norm": 2.0435218811035156,
+      "learning_rate": 4.452706822858316e-06,
+      "loss": 0.7117,
+      "step": 1588000
+    },
+    {
+      "epoch": 2.733698055864845,
+      "grad_norm": 2.6669647693634033,
+      "learning_rate": 4.438365735585917e-06,
+      "loss": 0.7096,
+      "step": 1588500
+    },
+    {
+      "epoch": 2.734558521101189,
+      "grad_norm": 1.955651879310608,
+      "learning_rate": 4.424024648313517e-06,
+      "loss": 0.721,
+      "step": 1589000
+    },
+    {
+      "epoch": 2.735418986337533,
+      "grad_norm": 2.1903460025787354,
+      "learning_rate": 4.409683561041118e-06,
+      "loss": 0.7173,
+      "step": 1589500
+    },
+    {
+      "epoch": 2.736279451573877,
+      "grad_norm": 2.187950372695923,
+      "learning_rate": 4.3953424737687174e-06,
+      "loss": 0.7102,
+      "step": 1590000
+    },
+    {
+      "epoch": 2.737139916810221,
+      "grad_norm": 2.1879000663757324,
+      "learning_rate": 4.381001386496317e-06,
+      "loss": 0.7166,
+      "step": 1590500
+    },
+    {
+      "epoch": 2.738000382046565,
+      "grad_norm": 2.2645974159240723,
+      "learning_rate": 4.366660299223918e-06,
+      "loss": 0.7154,
+      "step": 1591000
+    },
+    {
+      "epoch": 2.738860847282909,
+      "grad_norm": 2.1378839015960693,
+      "learning_rate": 4.352319211951518e-06,
+      "loss": 0.7153,
+      "step": 1591500
+    },
+    {
+      "epoch": 2.739721312519253,
+      "grad_norm": 2.3064563274383545,
+      "learning_rate": 4.3379781246791185e-06,
+      "loss": 0.7147,
+      "step": 1592000
+    },
+    {
+      "epoch": 2.740581777755597,
+      "grad_norm": 2.2445333003997803,
+      "learning_rate": 4.323637037406719e-06,
+      "loss": 0.707,
+      "step": 1592500
+    },
+    {
+      "epoch": 2.741442242991941,
+      "grad_norm": 2.3018760681152344,
+      "learning_rate": 4.309295950134319e-06,
+      "loss": 0.7121,
+      "step": 1593000
+    },
+    {
+      "epoch": 2.742302708228285,
+      "grad_norm": 2.269040107727051,
+      "learning_rate": 4.294954862861919e-06,
+      "loss": 0.7126,
+      "step": 1593500
+    },
+    {
+      "epoch": 2.743163173464629,
+      "grad_norm": 2.353139638900757,
+      "learning_rate": 4.280613775589519e-06,
+      "loss": 0.7233,
+      "step": 1594000
+    },
+    {
+      "epoch": 2.744023638700973,
+      "grad_norm": 2.38712215423584,
+      "learning_rate": 4.26627268831712e-06,
+      "loss": 0.7131,
+      "step": 1594500
+    },
+    {
+      "epoch": 2.744884103937317,
+      "grad_norm": 2.411133289337158,
+      "learning_rate": 4.25193160104472e-06,
+      "loss": 0.71,
+      "step": 1595000
+    },
+    {
+      "epoch": 2.745744569173661,
+      "grad_norm": 2.1307971477508545,
+      "learning_rate": 4.23759051377232e-06,
+      "loss": 0.718,
+      "step": 1595500
+    },
+    {
+      "epoch": 2.746605034410005,
+      "grad_norm": 2.3968119621276855,
+      "learning_rate": 4.22324942649992e-06,
+      "loss": 0.7162,
+      "step": 1596000
+    },
+    {
+      "epoch": 2.747465499646349,
+      "grad_norm": 2.289236068725586,
+      "learning_rate": 4.208908339227521e-06,
+      "loss": 0.7117,
+      "step": 1596500
+    },
+    {
+      "epoch": 2.7483259648826928,
+      "grad_norm": 2.1929147243499756,
+      "learning_rate": 4.194567251955121e-06,
+      "loss": 0.7153,
+      "step": 1597000
+    },
+    {
+      "epoch": 2.7491864301190367,
+      "grad_norm": 2.2217397689819336,
+      "learning_rate": 4.18022616468272e-06,
+      "loss": 0.7133,
+      "step": 1597500
+    },
+    {
+      "epoch": 2.7500468953553807,
+      "grad_norm": 2.5000901222229004,
+      "learning_rate": 4.165885077410321e-06,
+      "loss": 0.7143,
+      "step": 1598000
+    },
+    {
+      "epoch": 2.7509073605917247,
+      "grad_norm": 2.1965184211730957,
+      "learning_rate": 4.151543990137921e-06,
+      "loss": 0.711,
+      "step": 1598500
+    },
+    {
+      "epoch": 2.7517678258280687,
+      "grad_norm": 2.4667575359344482,
+      "learning_rate": 4.137202902865522e-06,
+      "loss": 0.721,
+      "step": 1599000
+    },
+    {
+      "epoch": 2.7526282910644126,
+      "grad_norm": 2.006913661956787,
+      "learning_rate": 4.122861815593121e-06,
+      "loss": 0.715,
+      "step": 1599500
+    },
+    {
+      "epoch": 2.7534887563007566,
+      "grad_norm": 2.0907866954803467,
+      "learning_rate": 4.108520728320723e-06,
+      "loss": 0.716,
+      "step": 1600000
+    },
+    {
+      "epoch": 2.7543492215371006,
+      "grad_norm": 2.113661289215088,
+      "learning_rate": 4.094179641048322e-06,
+      "loss": 0.7127,
+      "step": 1600500
+    },
+    {
+      "epoch": 2.7552096867734446,
+      "grad_norm": 2.062978744506836,
+      "learning_rate": 4.079838553775923e-06,
+      "loss": 0.7155,
+      "step": 1601000
+    },
+    {
+      "epoch": 2.7560701520097886,
+      "grad_norm": 2.0641098022460938,
+      "learning_rate": 4.065497466503522e-06,
+      "loss": 0.7156,
+      "step": 1601500
+    },
+    {
+      "epoch": 2.7569306172461325,
+      "grad_norm": 2.1433663368225098,
+      "learning_rate": 4.051156379231123e-06,
+      "loss": 0.7186,
+      "step": 1602000
+    },
+    {
+      "epoch": 2.7577910824824765,
+      "grad_norm": 2.1974167823791504,
+      "learning_rate": 4.036815291958723e-06,
+      "loss": 0.7123,
+      "step": 1602500
+    },
+    {
+      "epoch": 2.7586515477188205,
+      "grad_norm": 2.3867440223693848,
+      "learning_rate": 4.022474204686323e-06,
+      "loss": 0.7195,
+      "step": 1603000
+    },
+    {
+      "epoch": 2.7595120129551645,
+      "grad_norm": 2.26507306098938,
+      "learning_rate": 4.008133117413923e-06,
+      "loss": 0.71,
+      "step": 1603500
+    },
+    {
+      "epoch": 2.7603724781915084,
+      "grad_norm": 2.171724319458008,
+      "learning_rate": 3.993792030141524e-06,
+      "loss": 0.7135,
+      "step": 1604000
+    },
+    {
+      "epoch": 2.7612329434278524,
+      "grad_norm": 2.4163520336151123,
+      "learning_rate": 3.979450942869124e-06,
+      "loss": 0.7221,
+      "step": 1604500
+    },
+    {
+      "epoch": 2.7620934086641964,
+      "grad_norm": 2.2443559169769287,
+      "learning_rate": 3.965109855596724e-06,
+      "loss": 0.7165,
+      "step": 1605000
+    },
+    {
+      "epoch": 2.762953873900541,
+      "grad_norm": 1.999537467956543,
+      "learning_rate": 3.950768768324324e-06,
+      "loss": 0.7166,
+      "step": 1605500
+    },
+    {
+      "epoch": 2.763814339136885,
+      "grad_norm": 2.4677159786224365,
+      "learning_rate": 3.936427681051925e-06,
+      "loss": 0.7085,
+      "step": 1606000
+    },
+    {
+      "epoch": 2.7646748043732288,
+      "grad_norm": 2.335520029067993,
+      "learning_rate": 3.9220865937795245e-06,
+      "loss": 0.7078,
+      "step": 1606500
+    },
+    {
+      "epoch": 2.7655352696095727,
+      "grad_norm": 2.352250576019287,
+      "learning_rate": 3.907745506507125e-06,
+      "loss": 0.7164,
+      "step": 1607000
+    },
+    {
+      "epoch": 2.7663957348459167,
+      "grad_norm": 2.2408668994903564,
+      "learning_rate": 3.8934044192347254e-06,
+      "loss": 0.7158,
+      "step": 1607500
+    },
+    {
+      "epoch": 2.7672562000822607,
+      "grad_norm": 2.1717495918273926,
+      "learning_rate": 3.879063331962326e-06,
+      "loss": 0.7152,
+      "step": 1608000
+    },
+    {
+      "epoch": 2.7681166653186047,
+      "grad_norm": 2.263266086578369,
+      "learning_rate": 3.8647222446899255e-06,
+      "loss": 0.7139,
+      "step": 1608500
+    },
+    {
+      "epoch": 2.7689771305549487,
+      "grad_norm": 2.527707099914551,
+      "learning_rate": 3.850381157417526e-06,
+      "loss": 0.7116,
+      "step": 1609000
+    },
+    {
+      "epoch": 2.7698375957912926,
+      "grad_norm": 2.510777473449707,
+      "learning_rate": 3.8360400701451265e-06,
+      "loss": 0.7161,
+      "step": 1609500
+    },
+    {
+      "epoch": 2.7706980610276366,
+      "grad_norm": 1.8678065538406372,
+      "learning_rate": 3.821698982872726e-06,
+      "loss": 0.7149,
+      "step": 1610000
+    },
+    {
+      "epoch": 2.7715585262639806,
+      "grad_norm": 2.1824417114257812,
+      "learning_rate": 3.807357895600327e-06,
+      "loss": 0.7106,
+      "step": 1610500
+    },
+    {
+      "epoch": 2.7724189915003246,
+      "grad_norm": 2.100099802017212,
+      "learning_rate": 3.7930168083279266e-06,
+      "loss": 0.712,
+      "step": 1611000
+    },
+    {
+      "epoch": 2.7732794567366685,
+      "grad_norm": 2.1548209190368652,
+      "learning_rate": 3.7786757210555275e-06,
+      "loss": 0.7082,
+      "step": 1611500
+    },
+    {
+      "epoch": 2.7741399219730125,
+      "grad_norm": 2.0101969242095947,
+      "learning_rate": 3.764334633783127e-06,
+      "loss": 0.7105,
+      "step": 1612000
+    },
+    {
+      "epoch": 2.7750003872093565,
+      "grad_norm": 2.102904796600342,
+      "learning_rate": 3.749993546510728e-06,
+      "loss": 0.7055,
+      "step": 1612500
+    },
+    {
+      "epoch": 2.7758608524457005,
+      "grad_norm": 2.2598965167999268,
+      "learning_rate": 3.7356524592383276e-06,
+      "loss": 0.7118,
+      "step": 1613000
+    },
+    {
+      "epoch": 2.7767213176820444,
+      "grad_norm": 2.071484088897705,
+      "learning_rate": 3.7213113719659277e-06,
+      "loss": 0.709,
+      "step": 1613500
+    },
+    {
+      "epoch": 2.7775817829183884,
+      "grad_norm": 2.030569076538086,
+      "learning_rate": 3.706970284693528e-06,
+      "loss": 0.7142,
+      "step": 1614000
+    },
+    {
+      "epoch": 2.7784422481547324,
+      "grad_norm": 2.0957064628601074,
+      "learning_rate": 3.692629197421128e-06,
+      "loss": 0.7083,
+      "step": 1614500
+    },
+    {
+      "epoch": 2.7793027133910764,
+      "grad_norm": 2.3552098274230957,
+      "learning_rate": 3.678288110148729e-06,
+      "loss": 0.7138,
+      "step": 1615000
+    },
+    {
+      "epoch": 2.7801631786274204,
+      "grad_norm": 2.4001100063323975,
+      "learning_rate": 3.6639470228763287e-06,
+      "loss": 0.7178,
+      "step": 1615500
+    },
+    {
+      "epoch": 2.7810236438637643,
+      "grad_norm": 2.3316562175750732,
+      "learning_rate": 3.6496059356039296e-06,
+      "loss": 0.7125,
+      "step": 1616000
+    },
+    {
+      "epoch": 2.7818841091001083,
+      "grad_norm": 2.0596606731414795,
+      "learning_rate": 3.6352648483315292e-06,
+      "loss": 0.7096,
+      "step": 1616500
+    },
+    {
+      "epoch": 2.7827445743364523,
+      "grad_norm": 2.2260353565216064,
+      "learning_rate": 3.6209237610591293e-06,
+      "loss": 0.7193,
+      "step": 1617000
+    },
+    {
+      "epoch": 2.7836050395727963,
+      "grad_norm": 2.0922586917877197,
+      "learning_rate": 3.6065826737867298e-06,
+      "loss": 0.7118,
+      "step": 1617500
+    },
+    {
+      "epoch": 2.7844655048091402,
+      "grad_norm": 2.368934154510498,
+      "learning_rate": 3.59224158651433e-06,
+      "loss": 0.7126,
+      "step": 1618000
+    },
+    {
+      "epoch": 2.785325970045484,
+      "grad_norm": 2.3944685459136963,
+      "learning_rate": 3.5779004992419303e-06,
+      "loss": 0.7109,
+      "step": 1618500
+    },
+    {
+      "epoch": 2.786186435281828,
+      "grad_norm": 2.3955578804016113,
+      "learning_rate": 3.5635594119695303e-06,
+      "loss": 0.7081,
+      "step": 1619000
+    },
+    {
+      "epoch": 2.787046900518172,
+      "grad_norm": 2.152188301086426,
+      "learning_rate": 3.549218324697131e-06,
+      "loss": 0.7163,
+      "step": 1619500
+    },
+    {
+      "epoch": 2.787907365754516,
+      "grad_norm": 1.973374605178833,
+      "learning_rate": 3.534877237424731e-06,
+      "loss": 0.7167,
+      "step": 1620000
+    },
+    {
+      "epoch": 2.78876783099086,
+      "grad_norm": 2.142923593521118,
+      "learning_rate": 3.5205361501523313e-06,
+      "loss": 0.7101,
+      "step": 1620500
+    },
+    {
+      "epoch": 2.789628296227204,
+      "grad_norm": 2.1474125385284424,
+      "learning_rate": 3.5061950628799314e-06,
+      "loss": 0.7173,
+      "step": 1621000
+    },
+    {
+      "epoch": 2.790488761463548,
+      "grad_norm": 2.028701066970825,
+      "learning_rate": 3.4918539756075314e-06,
+      "loss": 0.7148,
+      "step": 1621500
+    },
+    {
+      "epoch": 2.791349226699892,
+      "grad_norm": 2.265554189682007,
+      "learning_rate": 3.477512888335132e-06,
+      "loss": 0.7124,
+      "step": 1622000
+    },
+    {
+      "epoch": 2.792209691936236,
+      "grad_norm": 2.309709310531616,
+      "learning_rate": 3.463171801062732e-06,
+      "loss": 0.7109,
+      "step": 1622500
+    },
+    {
+      "epoch": 2.79307015717258,
+      "grad_norm": 2.34216046333313,
+      "learning_rate": 3.4488307137903324e-06,
+      "loss": 0.7131,
+      "step": 1623000
+    },
+    {
+      "epoch": 2.793930622408924,
+      "grad_norm": 2.198592185974121,
+      "learning_rate": 3.4344896265179324e-06,
+      "loss": 0.7077,
+      "step": 1623500
+    },
+    {
+      "epoch": 2.794791087645268,
+      "grad_norm": 2.2854268550872803,
+      "learning_rate": 3.420148539245533e-06,
+      "loss": 0.7145,
+      "step": 1624000
+    },
+    {
+      "epoch": 2.795651552881612,
+      "grad_norm": 2.2535135746002197,
+      "learning_rate": 3.405807451973133e-06,
+      "loss": 0.715,
+      "step": 1624500
+    },
+    {
+      "epoch": 2.796512018117956,
+      "grad_norm": 2.2533137798309326,
+      "learning_rate": 3.391466364700733e-06,
+      "loss": 0.7105,
+      "step": 1625000
+    },
+    {
+      "epoch": 2.7973724833543,
+      "grad_norm": 2.0734574794769287,
+      "learning_rate": 3.3771252774283335e-06,
+      "loss": 0.7128,
+      "step": 1625500
+    },
+    {
+      "epoch": 2.798232948590644,
+      "grad_norm": 2.3115506172180176,
+      "learning_rate": 3.3627841901559335e-06,
+      "loss": 0.712,
+      "step": 1626000
+    },
+    {
+      "epoch": 2.799093413826988,
+      "grad_norm": 2.0518035888671875,
+      "learning_rate": 3.348443102883534e-06,
+      "loss": 0.7065,
+      "step": 1626500
+    },
+    {
+      "epoch": 2.799953879063332,
+      "grad_norm": 2.206268072128296,
+      "learning_rate": 3.334102015611134e-06,
+      "loss": 0.7109,
+      "step": 1627000
+    },
+    {
+      "epoch": 2.800814344299676,
+      "grad_norm": 2.0784969329833984,
+      "learning_rate": 3.3197609283387345e-06,
+      "loss": 0.7105,
+      "step": 1627500
+    },
+    {
+      "epoch": 2.8016748095360198,
+      "grad_norm": 2.217696189880371,
+      "learning_rate": 3.3054198410663346e-06,
+      "loss": 0.7092,
+      "step": 1628000
+    },
+    {
+      "epoch": 2.8025352747723637,
+      "grad_norm": 2.317526340484619,
+      "learning_rate": 3.291078753793935e-06,
+      "loss": 0.7107,
+      "step": 1628500
+    },
+    {
+      "epoch": 2.8033957400087077,
+      "grad_norm": 2.112042188644409,
+      "learning_rate": 3.276737666521535e-06,
+      "loss": 0.7141,
+      "step": 1629000
+    },
+    {
+      "epoch": 2.8042562052450517,
+      "grad_norm": 2.475770950317383,
+      "learning_rate": 3.262396579249135e-06,
+      "loss": 0.714,
+      "step": 1629500
+    },
+    {
+      "epoch": 2.8051166704813957,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.2480554919767356e-06,
+      "loss": 0.7104,
+      "step": 1630000
+    },
+    {
+      "epoch": 2.8059771357177397,
+      "grad_norm": 2.1881062984466553,
+      "learning_rate": 3.2337144047043356e-06,
+      "loss": 0.7104,
+      "step": 1630500
+    },
+    {
+      "epoch": 2.8068376009540836,
+      "grad_norm": 2.3878414630889893,
+      "learning_rate": 3.219373317431936e-06,
+      "loss": 0.7065,
+      "step": 1631000
+    },
+    {
+      "epoch": 2.8076980661904276,
+      "grad_norm": 2.6674928665161133,
+      "learning_rate": 3.205032230159536e-06,
+      "loss": 0.7168,
+      "step": 1631500
+    },
+    {
+      "epoch": 2.8085585314267716,
+      "grad_norm": 2.050078868865967,
+      "learning_rate": 3.1906911428871366e-06,
+      "loss": 0.7118,
+      "step": 1632000
+    },
+    {
+      "epoch": 2.8094189966631156,
+      "grad_norm": 2.17096209526062,
+      "learning_rate": 3.1763500556147367e-06,
+      "loss": 0.7136,
+      "step": 1632500
+    },
+    {
+      "epoch": 2.8102794618994595,
+      "grad_norm": 2.199841260910034,
+      "learning_rate": 3.1620089683423363e-06,
+      "loss": 0.7085,
+      "step": 1633000
+    },
+    {
+      "epoch": 2.8111399271358035,
+      "grad_norm": 2.2607522010803223,
+      "learning_rate": 3.147667881069937e-06,
+      "loss": 0.7138,
+      "step": 1633500
+    },
+    {
+      "epoch": 2.812000392372148,
+      "grad_norm": 2.1586544513702393,
+      "learning_rate": 3.1333267937975372e-06,
+      "loss": 0.7089,
+      "step": 1634000
+    },
+    {
+      "epoch": 2.812860857608492,
+      "grad_norm": 2.181818723678589,
+      "learning_rate": 3.1189857065251377e-06,
+      "loss": 0.7105,
+      "step": 1634500
+    },
+    {
+      "epoch": 2.813721322844836,
+      "grad_norm": 2.3212060928344727,
+      "learning_rate": 3.1046446192527378e-06,
+      "loss": 0.7112,
+      "step": 1635000
+    },
+    {
+      "epoch": 2.81458178808118,
+      "grad_norm": 2.3202192783355713,
+      "learning_rate": 3.090303531980338e-06,
+      "loss": 0.716,
+      "step": 1635500
+    },
+    {
+      "epoch": 2.815442253317524,
+      "grad_norm": 2.2512378692626953,
+      "learning_rate": 3.0759624447079383e-06,
+      "loss": 0.711,
+      "step": 1636000
+    },
+    {
+      "epoch": 2.816302718553868,
+      "grad_norm": 2.1948952674865723,
+      "learning_rate": 3.0616213574355383e-06,
+      "loss": 0.7098,
+      "step": 1636500
+    },
+    {
+      "epoch": 2.817163183790212,
+      "grad_norm": 2.0541694164276123,
+      "learning_rate": 3.047280270163139e-06,
+      "loss": 0.7133,
+      "step": 1637000
+    },
+    {
+      "epoch": 2.8180236490265558,
+      "grad_norm": 2.204808473587036,
+      "learning_rate": 3.032939182890739e-06,
+      "loss": 0.7149,
+      "step": 1637500
+    },
+    {
+      "epoch": 2.8188841142628998,
+      "grad_norm": 2.3189733028411865,
+      "learning_rate": 3.0185980956183393e-06,
+      "loss": 0.7067,
+      "step": 1638000
+    },
+    {
+      "epoch": 2.8197445794992437,
+      "grad_norm": 2.3678770065307617,
+      "learning_rate": 3.004257008345939e-06,
+      "loss": 0.7081,
+      "step": 1638500
+    },
+    {
+      "epoch": 2.8206050447355877,
+      "grad_norm": 2.376946210861206,
+      "learning_rate": 2.9899159210735394e-06,
+      "loss": 0.7155,
+      "step": 1639000
+    },
+    {
+      "epoch": 2.8214655099719317,
+      "grad_norm": 2.2931296825408936,
+      "learning_rate": 2.9755748338011394e-06,
+      "loss": 0.7107,
+      "step": 1639500
+    },
+    {
+      "epoch": 2.8223259752082757,
+      "grad_norm": 2.1000328063964844,
+      "learning_rate": 2.96123374652874e-06,
+      "loss": 0.7128,
+      "step": 1640000
+    },
+    {
+      "epoch": 2.8231864404446196,
+      "grad_norm": 2.423055648803711,
+      "learning_rate": 2.9468926592563404e-06,
+      "loss": 0.7117,
+      "step": 1640500
+    },
+    {
+      "epoch": 2.8240469056809636,
+      "grad_norm": 2.1777422428131104,
+      "learning_rate": 2.9325515719839404e-06,
+      "loss": 0.7121,
+      "step": 1641000
+    },
+    {
+      "epoch": 2.8249073709173076,
+      "grad_norm": 2.3314688205718994,
+      "learning_rate": 2.918210484711541e-06,
+      "loss": 0.7083,
+      "step": 1641500
+    },
+    {
+      "epoch": 2.8257678361536516,
+      "grad_norm": 2.2967236042022705,
+      "learning_rate": 2.903869397439141e-06,
+      "loss": 0.7108,
+      "step": 1642000
+    },
+    {
+      "epoch": 2.8266283013899955,
+      "grad_norm": 2.349409341812134,
+      "learning_rate": 2.889528310166741e-06,
+      "loss": 0.7097,
+      "step": 1642500
+    },
+    {
+      "epoch": 2.8274887666263395,
+      "grad_norm": 2.4915151596069336,
+      "learning_rate": 2.875187222894341e-06,
+      "loss": 0.7145,
+      "step": 1643000
+    },
+    {
+      "epoch": 2.8283492318626835,
+      "grad_norm": 2.3456923961639404,
+      "learning_rate": 2.8608461356219415e-06,
+      "loss": 0.7157,
+      "step": 1643500
+    },
+    {
+      "epoch": 2.8292096970990275,
+      "grad_norm": 2.219536781311035,
+      "learning_rate": 2.8465050483495416e-06,
+      "loss": 0.7079,
+      "step": 1644000
+    },
+    {
+      "epoch": 2.8300701623353715,
+      "grad_norm": 2.14593505859375,
+      "learning_rate": 2.832163961077142e-06,
+      "loss": 0.7116,
+      "step": 1644500
+    },
+    {
+      "epoch": 2.8309306275717154,
+      "grad_norm": 2.3239259719848633,
+      "learning_rate": 2.817822873804742e-06,
+      "loss": 0.7058,
+      "step": 1645000
+    },
+    {
+      "epoch": 2.8317910928080594,
+      "grad_norm": 2.493803024291992,
+      "learning_rate": 2.8034817865323426e-06,
+      "loss": 0.7092,
+      "step": 1645500
+    },
+    {
+      "epoch": 2.8326515580444034,
+      "grad_norm": 2.3556675910949707,
+      "learning_rate": 2.789140699259943e-06,
+      "loss": 0.713,
+      "step": 1646000
+    },
+    {
+      "epoch": 2.8335120232807474,
+      "grad_norm": 2.187157392501831,
+      "learning_rate": 2.7747996119875426e-06,
+      "loss": 0.7095,
+      "step": 1646500
+    },
+    {
+      "epoch": 2.8343724885170913,
+      "grad_norm": 2.2635531425476074,
+      "learning_rate": 2.760458524715143e-06,
+      "loss": 0.7093,
+      "step": 1647000
+    },
+    {
+      "epoch": 2.8352329537534353,
+      "grad_norm": 2.098905086517334,
+      "learning_rate": 2.746117437442743e-06,
+      "loss": 0.7116,
+      "step": 1647500
+    },
+    {
+      "epoch": 2.8360934189897793,
+      "grad_norm": 2.43005633354187,
+      "learning_rate": 2.7317763501703436e-06,
+      "loss": 0.7103,
+      "step": 1648000
+    },
+    {
+      "epoch": 2.8369538842261233,
+      "grad_norm": 2.3651890754699707,
+      "learning_rate": 2.7174352628979437e-06,
+      "loss": 0.714,
+      "step": 1648500
+    },
+    {
+      "epoch": 2.8378143494624672,
+      "grad_norm": 2.2694835662841797,
+      "learning_rate": 2.703094175625544e-06,
+      "loss": 0.7139,
+      "step": 1649000
+    },
+    {
+      "epoch": 2.838674814698811,
+      "grad_norm": 2.0746357440948486,
+      "learning_rate": 2.688753088353144e-06,
+      "loss": 0.7135,
+      "step": 1649500
+    },
+    {
+      "epoch": 2.839535279935155,
+      "grad_norm": 2.3595311641693115,
+      "learning_rate": 2.6744120010807447e-06,
+      "loss": 0.7105,
+      "step": 1650000
+    },
+    {
+      "epoch": 2.840395745171499,
+      "grad_norm": 2.122243642807007,
+      "learning_rate": 2.6600709138083447e-06,
+      "loss": 0.7106,
+      "step": 1650500
+    },
+    {
+      "epoch": 2.841256210407843,
+      "grad_norm": 2.4125218391418457,
+      "learning_rate": 2.6457298265359448e-06,
+      "loss": 0.7063,
+      "step": 1651000
+    },
+    {
+      "epoch": 2.8421166756441876,
+      "grad_norm": 2.1687426567077637,
+      "learning_rate": 2.6313887392635452e-06,
+      "loss": 0.7134,
+      "step": 1651500
+    },
+    {
+      "epoch": 2.8429771408805316,
+      "grad_norm": 2.412323474884033,
+      "learning_rate": 2.6170476519911453e-06,
+      "loss": 0.7061,
+      "step": 1652000
+    },
+    {
+      "epoch": 2.8438376061168755,
+      "grad_norm": 2.353044033050537,
+      "learning_rate": 2.6027065647187458e-06,
+      "loss": 0.7095,
+      "step": 1652500
+    },
+    {
+      "epoch": 2.8446980713532195,
+      "grad_norm": 2.2941107749938965,
+      "learning_rate": 2.588365477446346e-06,
+      "loss": 0.7082,
+      "step": 1653000
+    },
+    {
+      "epoch": 2.8455585365895635,
+      "grad_norm": 2.473214626312256,
+      "learning_rate": 2.5740243901739463e-06,
+      "loss": 0.7069,
+      "step": 1653500
+    },
+    {
+      "epoch": 2.8464190018259075,
+      "grad_norm": 2.324897527694702,
+      "learning_rate": 2.5596833029015463e-06,
+      "loss": 0.7156,
+      "step": 1654000
+    },
+    {
+      "epoch": 2.8472794670622514,
+      "grad_norm": 2.1640710830688477,
+      "learning_rate": 2.5453422156291464e-06,
+      "loss": 0.7128,
+      "step": 1654500
+    },
+    {
+      "epoch": 2.8481399322985954,
+      "grad_norm": 2.296011209487915,
+      "learning_rate": 2.5310011283567464e-06,
+      "loss": 0.7117,
+      "step": 1655000
+    },
+    {
+      "epoch": 2.8490003975349394,
+      "grad_norm": 2.151864767074585,
+      "learning_rate": 2.516660041084347e-06,
+      "loss": 0.7072,
+      "step": 1655500
+    },
+    {
+      "epoch": 2.8498608627712834,
+      "grad_norm": 2.3428564071655273,
+      "learning_rate": 2.5023189538119474e-06,
+      "loss": 0.7157,
+      "step": 1656000
+    },
+    {
+      "epoch": 2.8507213280076273,
+      "grad_norm": 2.3723461627960205,
+      "learning_rate": 2.4879778665395474e-06,
+      "loss": 0.708,
+      "step": 1656500
+    },
+    {
+      "epoch": 2.8515817932439713,
+      "grad_norm": 2.235034227371216,
+      "learning_rate": 2.473636779267148e-06,
+      "loss": 0.7027,
+      "step": 1657000
+    },
+    {
+      "epoch": 2.8524422584803153,
+      "grad_norm": 2.4615862369537354,
+      "learning_rate": 2.459295691994748e-06,
+      "loss": 0.7095,
+      "step": 1657500
+    },
+    {
+      "epoch": 2.8533027237166593,
+      "grad_norm": 2.3678109645843506,
+      "learning_rate": 2.4449546047223484e-06,
+      "loss": 0.7058,
+      "step": 1658000
+    },
+    {
+      "epoch": 2.8541631889530032,
+      "grad_norm": 2.1077773571014404,
+      "learning_rate": 2.430613517449948e-06,
+      "loss": 0.7132,
+      "step": 1658500
+    },
+    {
+      "epoch": 2.8550236541893472,
+      "grad_norm": 2.354522228240967,
+      "learning_rate": 2.4162724301775485e-06,
+      "loss": 0.7129,
+      "step": 1659000
+    },
+    {
+      "epoch": 2.855884119425691,
+      "grad_norm": 2.43855357170105,
+      "learning_rate": 2.4019313429051485e-06,
+      "loss": 0.7082,
+      "step": 1659500
+    },
+    {
+      "epoch": 2.856744584662035,
+      "grad_norm": 2.218585252761841,
+      "learning_rate": 2.387590255632749e-06,
+      "loss": 0.7151,
+      "step": 1660000
+    },
+    {
+      "epoch": 2.857605049898379,
+      "grad_norm": 2.1669533252716064,
+      "learning_rate": 2.373249168360349e-06,
+      "loss": 0.7123,
+      "step": 1660500
+    },
+    {
+      "epoch": 2.858465515134723,
+      "grad_norm": 2.2842636108398438,
+      "learning_rate": 2.3589080810879495e-06,
+      "loss": 0.7156,
+      "step": 1661000
+    },
+    {
+      "epoch": 2.859325980371067,
+      "grad_norm": 2.3778250217437744,
+      "learning_rate": 2.3445669938155496e-06,
+      "loss": 0.708,
+      "step": 1661500
+    },
+    {
+      "epoch": 2.860186445607411,
+      "grad_norm": 2.1641619205474854,
+      "learning_rate": 2.3302259065431496e-06,
+      "loss": 0.7066,
+      "step": 1662000
+    },
+    {
+      "epoch": 2.861046910843755,
+      "grad_norm": 2.343029737472534,
+      "learning_rate": 2.31588481927075e-06,
+      "loss": 0.7049,
+      "step": 1662500
+    },
+    {
+      "epoch": 2.861907376080099,
+      "grad_norm": 2.098466157913208,
+      "learning_rate": 2.30154373199835e-06,
+      "loss": 0.7094,
+      "step": 1663000
+    },
+    {
+      "epoch": 2.862767841316443,
+      "grad_norm": 2.4116451740264893,
+      "learning_rate": 2.2872026447259506e-06,
+      "loss": 0.7092,
+      "step": 1663500
+    },
+    {
+      "epoch": 2.863628306552787,
+      "grad_norm": 2.115022897720337,
+      "learning_rate": 2.2728615574535506e-06,
+      "loss": 0.712,
+      "step": 1664000
+    },
+    {
+      "epoch": 2.864488771789131,
+      "grad_norm": 2.248368978500366,
+      "learning_rate": 2.258520470181151e-06,
+      "loss": 0.7119,
+      "step": 1664500
+    },
+    {
+      "epoch": 2.865349237025475,
+      "grad_norm": 2.036473274230957,
+      "learning_rate": 2.244179382908751e-06,
+      "loss": 0.7103,
+      "step": 1665000
+    },
+    {
+      "epoch": 2.866209702261819,
+      "grad_norm": 2.222977638244629,
+      "learning_rate": 2.2298382956363516e-06,
+      "loss": 0.7106,
+      "step": 1665500
+    },
+    {
+      "epoch": 2.867070167498163,
+      "grad_norm": 2.3005199432373047,
+      "learning_rate": 2.2154972083639517e-06,
+      "loss": 0.712,
+      "step": 1666000
+    },
+    {
+      "epoch": 2.867930632734507,
+      "grad_norm": 2.1439499855041504,
+      "learning_rate": 2.2011561210915517e-06,
+      "loss": 0.7085,
+      "step": 1666500
+    },
+    {
+      "epoch": 2.868791097970851,
+      "grad_norm": 2.1440727710723877,
+      "learning_rate": 2.186815033819152e-06,
+      "loss": 0.7091,
+      "step": 1667000
+    },
+    {
+      "epoch": 2.869651563207195,
+      "grad_norm": 2.1596696376800537,
+      "learning_rate": 2.1724739465467522e-06,
+      "loss": 0.7039,
+      "step": 1667500
+    },
+    {
+      "epoch": 2.870512028443539,
+      "grad_norm": 2.039153575897217,
+      "learning_rate": 2.1581328592743527e-06,
+      "loss": 0.7029,
+      "step": 1668000
+    },
+    {
+      "epoch": 2.871372493679883,
+      "grad_norm": 2.2410709857940674,
+      "learning_rate": 2.1437917720019528e-06,
+      "loss": 0.7078,
+      "step": 1668500
+    },
+    {
+      "epoch": 2.8722329589162268,
+      "grad_norm": 2.2461483478546143,
+      "learning_rate": 2.1294506847295532e-06,
+      "loss": 0.7161,
+      "step": 1669000
+    },
+    {
+      "epoch": 2.8730934241525707,
+      "grad_norm": 2.4385788440704346,
+      "learning_rate": 2.1151095974571533e-06,
+      "loss": 0.7163,
+      "step": 1669500
+    },
+    {
+      "epoch": 2.8739538893889147,
+      "grad_norm": 2.215613842010498,
+      "learning_rate": 2.1007685101847533e-06,
+      "loss": 0.7109,
+      "step": 1670000
+    },
+    {
+      "epoch": 2.8748143546252587,
+      "grad_norm": 2.438634157180786,
+      "learning_rate": 2.0864274229123534e-06,
+      "loss": 0.7137,
+      "step": 1670500
+    },
+    {
+      "epoch": 2.8756748198616027,
+      "grad_norm": 2.2035157680511475,
+      "learning_rate": 2.072086335639954e-06,
+      "loss": 0.712,
+      "step": 1671000
+    },
+    {
+      "epoch": 2.8765352850979466,
+      "grad_norm": 2.183810234069824,
+      "learning_rate": 2.0577452483675543e-06,
+      "loss": 0.7111,
+      "step": 1671500
+    },
+    {
+      "epoch": 2.8773957503342906,
+      "grad_norm": 2.2220458984375,
+      "learning_rate": 2.0434041610951544e-06,
+      "loss": 0.7046,
+      "step": 1672000
+    },
+    {
+      "epoch": 2.8782562155706346,
+      "grad_norm": 2.3107755184173584,
+      "learning_rate": 2.029063073822755e-06,
+      "loss": 0.7065,
+      "step": 1672500
+    },
+    {
+      "epoch": 2.8791166808069786,
+      "grad_norm": 2.3349597454071045,
+      "learning_rate": 2.014721986550355e-06,
+      "loss": 0.7124,
+      "step": 1673000
+    },
+    {
+      "epoch": 2.8799771460433226,
+      "grad_norm": 2.2262802124023438,
+      "learning_rate": 2.0003808992779553e-06,
+      "loss": 0.7125,
+      "step": 1673500
+    },
+    {
+      "epoch": 2.8808376112796665,
+      "grad_norm": 2.3967578411102295,
+      "learning_rate": 1.986039812005555e-06,
+      "loss": 0.7069,
+      "step": 1674000
+    },
+    {
+      "epoch": 2.8816980765160105,
+      "grad_norm": 2.216095209121704,
+      "learning_rate": 1.9716987247331554e-06,
+      "loss": 0.7128,
+      "step": 1674500
+    },
+    {
+      "epoch": 2.8825585417523545,
+      "grad_norm": 2.380256175994873,
+      "learning_rate": 1.9573576374607555e-06,
+      "loss": 0.7121,
+      "step": 1675000
+    },
+    {
+      "epoch": 2.8834190069886985,
+      "grad_norm": 2.2062182426452637,
+      "learning_rate": 1.943016550188356e-06,
+      "loss": 0.7066,
+      "step": 1675500
+    },
+    {
+      "epoch": 2.8842794722250424,
+      "grad_norm": 2.055337905883789,
+      "learning_rate": 1.928675462915956e-06,
+      "loss": 0.7104,
+      "step": 1676000
+    },
+    {
+      "epoch": 2.8851399374613864,
+      "grad_norm": 2.2458910942077637,
+      "learning_rate": 1.9143343756435565e-06,
+      "loss": 0.7049,
+      "step": 1676500
+    },
+    {
+      "epoch": 2.8860004026977304,
+      "grad_norm": 2.4219136238098145,
+      "learning_rate": 1.8999932883711567e-06,
+      "loss": 0.7036,
+      "step": 1677000
+    },
+    {
+      "epoch": 2.8868608679340744,
+      "grad_norm": 2.331808567047119,
+      "learning_rate": 1.885652201098757e-06,
+      "loss": 0.7065,
+      "step": 1677500
+    },
+    {
+      "epoch": 2.8877213331704183,
+      "grad_norm": 2.258765459060669,
+      "learning_rate": 1.8713111138263568e-06,
+      "loss": 0.7025,
+      "step": 1678000
+    },
+    {
+      "epoch": 2.8885817984067623,
+      "grad_norm": 2.276099681854248,
+      "learning_rate": 1.856970026553957e-06,
+      "loss": 0.7109,
+      "step": 1678500
+    },
+    {
+      "epoch": 2.8894422636431063,
+      "grad_norm": 2.2920784950256348,
+      "learning_rate": 1.8426289392815576e-06,
+      "loss": 0.7092,
+      "step": 1679000
+    },
+    {
+      "epoch": 2.8903027288794503,
+      "grad_norm": 2.336284637451172,
+      "learning_rate": 1.8282878520091578e-06,
+      "loss": 0.7134,
+      "step": 1679500
+    },
+    {
+      "epoch": 2.8911631941157943,
+      "grad_norm": 2.173426389694214,
+      "learning_rate": 1.813946764736758e-06,
+      "loss": 0.6993,
+      "step": 1680000
+    },
+    {
+      "epoch": 2.8920236593521387,
+      "grad_norm": 2.4683492183685303,
+      "learning_rate": 1.7996056774643583e-06,
+      "loss": 0.7116,
+      "step": 1680500
+    },
+    {
+      "epoch": 2.8928841245884827,
+      "grad_norm": 2.2608296871185303,
+      "learning_rate": 1.7852645901919586e-06,
+      "loss": 0.7081,
+      "step": 1681000
+    },
+    {
+      "epoch": 2.8937445898248266,
+      "grad_norm": 2.368868589401245,
+      "learning_rate": 1.7709235029195588e-06,
+      "loss": 0.7101,
+      "step": 1681500
+    },
+    {
+      "epoch": 2.8946050550611706,
+      "grad_norm": 2.291165590286255,
+      "learning_rate": 1.7565824156471587e-06,
+      "loss": 0.705,
+      "step": 1682000
+    },
+    {
+      "epoch": 2.8954655202975146,
+      "grad_norm": 2.262145519256592,
+      "learning_rate": 1.742241328374759e-06,
+      "loss": 0.7079,
+      "step": 1682500
+    },
+    {
+      "epoch": 2.8963259855338586,
+      "grad_norm": 2.2170474529266357,
+      "learning_rate": 1.7279002411023592e-06,
+      "loss": 0.7126,
+      "step": 1683000
+    },
+    {
+      "epoch": 2.8971864507702025,
+      "grad_norm": 2.0943198204040527,
+      "learning_rate": 1.7135591538299595e-06,
+      "loss": 0.7084,
+      "step": 1683500
+    },
+    {
+      "epoch": 2.8980469160065465,
+      "grad_norm": 2.274672746658325,
+      "learning_rate": 1.6992180665575597e-06,
+      "loss": 0.7046,
+      "step": 1684000
+    },
+    {
+      "epoch": 2.8989073812428905,
+      "grad_norm": 2.342705011367798,
+      "learning_rate": 1.68487697928516e-06,
+      "loss": 0.7074,
+      "step": 1684500
+    },
+    {
+      "epoch": 2.8997678464792345,
+      "grad_norm": 2.3994994163513184,
+      "learning_rate": 1.6705358920127602e-06,
+      "loss": 0.7096,
+      "step": 1685000
+    },
+    {
+      "epoch": 2.9006283117155784,
+      "grad_norm": 2.6676313877105713,
+      "learning_rate": 1.6561948047403603e-06,
+      "loss": 0.7096,
+      "step": 1685500
+    },
+    {
+      "epoch": 2.9014887769519224,
+      "grad_norm": 2.506844997406006,
+      "learning_rate": 1.6418537174679605e-06,
+      "loss": 0.709,
+      "step": 1686000
+    },
+    {
+      "epoch": 2.9023492421882664,
+      "grad_norm": 2.238424301147461,
+      "learning_rate": 1.6275126301955608e-06,
+      "loss": 0.7046,
+      "step": 1686500
+    },
+    {
+      "epoch": 2.9032097074246104,
+      "grad_norm": 2.2722549438476562,
+      "learning_rate": 1.613171542923161e-06,
+      "loss": 0.7024,
+      "step": 1687000
+    },
+    {
+      "epoch": 2.9040701726609544,
+      "grad_norm": 2.783315420150757,
+      "learning_rate": 1.5988304556507613e-06,
+      "loss": 0.7166,
+      "step": 1687500
+    },
+    {
+      "epoch": 2.9049306378972983,
+      "grad_norm": 2.3448448181152344,
+      "learning_rate": 1.5844893683783616e-06,
+      "loss": 0.7065,
+      "step": 1688000
+    },
+    {
+      "epoch": 2.9057911031336423,
+      "grad_norm": 2.162137031555176,
+      "learning_rate": 1.5701482811059618e-06,
+      "loss": 0.7146,
+      "step": 1688500
+    },
+    {
+      "epoch": 2.9066515683699863,
+      "grad_norm": 2.2553751468658447,
+      "learning_rate": 1.5558071938335619e-06,
+      "loss": 0.7053,
+      "step": 1689000
+    },
+    {
+      "epoch": 2.9075120336063303,
+      "grad_norm": 2.0961170196533203,
+      "learning_rate": 1.5414661065611621e-06,
+      "loss": 0.7124,
+      "step": 1689500
+    },
+    {
+      "epoch": 2.9083724988426742,
+      "grad_norm": 2.2436742782592773,
+      "learning_rate": 1.5271250192887626e-06,
+      "loss": 0.7053,
+      "step": 1690000
+    },
+    {
+      "epoch": 2.909232964079018,
+      "grad_norm": 2.397540330886841,
+      "learning_rate": 1.5127839320163627e-06,
+      "loss": 0.7113,
+      "step": 1690500
+    },
+    {
+      "epoch": 2.910093429315362,
+      "grad_norm": 2.4580838680267334,
+      "learning_rate": 1.498442844743963e-06,
+      "loss": 0.7089,
+      "step": 1691000
+    },
+    {
+      "epoch": 2.910953894551706,
+      "grad_norm": 2.1462783813476562,
+      "learning_rate": 1.4841017574715632e-06,
+      "loss": 0.7102,
+      "step": 1691500
+    },
+    {
+      "epoch": 2.91181435978805,
+      "grad_norm": 2.124737024307251,
+      "learning_rate": 1.4697606701991634e-06,
+      "loss": 0.7063,
+      "step": 1692000
+    },
+    {
+      "epoch": 2.912674825024394,
+      "grad_norm": 2.294898271560669,
+      "learning_rate": 1.4554195829267635e-06,
+      "loss": 0.7074,
+      "step": 1692500
+    },
+    {
+      "epoch": 2.913535290260738,
+      "grad_norm": 2.4345502853393555,
+      "learning_rate": 1.4410784956543637e-06,
+      "loss": 0.7116,
+      "step": 1693000
+    },
+    {
+      "epoch": 2.914395755497082,
+      "grad_norm": 2.2579216957092285,
+      "learning_rate": 1.426737408381964e-06,
+      "loss": 0.7096,
+      "step": 1693500
+    },
+    {
+      "epoch": 2.915256220733426,
+      "grad_norm": 2.086071252822876,
+      "learning_rate": 1.4123963211095643e-06,
+      "loss": 0.7071,
+      "step": 1694000
+    },
+    {
+      "epoch": 2.91611668596977,
+      "grad_norm": 2.2941625118255615,
+      "learning_rate": 1.3980552338371645e-06,
+      "loss": 0.7117,
+      "step": 1694500
+    },
+    {
+      "epoch": 2.916977151206114,
+      "grad_norm": 2.3904547691345215,
+      "learning_rate": 1.3837141465647648e-06,
+      "loss": 0.7125,
+      "step": 1695000
+    },
+    {
+      "epoch": 2.917837616442458,
+      "grad_norm": 2.387474775314331,
+      "learning_rate": 1.369373059292365e-06,
+      "loss": 0.7047,
+      "step": 1695500
+    },
+    {
+      "epoch": 2.918698081678802,
+      "grad_norm": 2.228543281555176,
+      "learning_rate": 1.3550319720199653e-06,
+      "loss": 0.7043,
+      "step": 1696000
+    },
+    {
+      "epoch": 2.919558546915146,
+      "grad_norm": 2.063142776489258,
+      "learning_rate": 1.3406908847475653e-06,
+      "loss": 0.7069,
+      "step": 1696500
+    },
+    {
+      "epoch": 2.92041901215149,
+      "grad_norm": 2.1467058658599854,
+      "learning_rate": 1.3263497974751656e-06,
+      "loss": 0.7109,
+      "step": 1697000
+    },
+    {
+      "epoch": 2.921279477387834,
+      "grad_norm": 2.3527698516845703,
+      "learning_rate": 1.3120087102027659e-06,
+      "loss": 0.7098,
+      "step": 1697500
+    },
+    {
+      "epoch": 2.9221399426241783,
+      "grad_norm": 2.164032459259033,
+      "learning_rate": 1.2976676229303661e-06,
+      "loss": 0.7096,
+      "step": 1698000
+    },
+    {
+      "epoch": 2.9230004078605223,
+      "grad_norm": 2.1595492362976074,
+      "learning_rate": 1.2833265356579662e-06,
+      "loss": 0.7139,
+      "step": 1698500
+    },
+    {
+      "epoch": 2.9238608730968663,
+      "grad_norm": 2.340627670288086,
+      "learning_rate": 1.2689854483855664e-06,
+      "loss": 0.7066,
+      "step": 1699000
+    },
+    {
+      "epoch": 2.9247213383332102,
+      "grad_norm": 2.104126214981079,
+      "learning_rate": 1.2546443611131667e-06,
+      "loss": 0.7053,
+      "step": 1699500
+    },
+    {
+      "epoch": 2.925581803569554,
+      "grad_norm": 2.251032590866089,
+      "learning_rate": 1.240303273840767e-06,
+      "loss": 0.7069,
+      "step": 1700000
+    },
+    {
+      "epoch": 2.926442268805898,
+      "grad_norm": 2.411782741546631,
+      "learning_rate": 1.2259621865683672e-06,
+      "loss": 0.7109,
+      "step": 1700500
+    },
+    {
+      "epoch": 2.927302734042242,
+      "grad_norm": 2.136021137237549,
+      "learning_rate": 1.2116210992959675e-06,
+      "loss": 0.7092,
+      "step": 1701000
+    },
+    {
+      "epoch": 2.928163199278586,
+      "grad_norm": 2.3158106803894043,
+      "learning_rate": 1.1972800120235677e-06,
+      "loss": 0.7044,
+      "step": 1701500
+    },
+    {
+      "epoch": 2.92902366451493,
+      "grad_norm": 2.3899245262145996,
+      "learning_rate": 1.182938924751168e-06,
+      "loss": 0.7096,
+      "step": 1702000
+    },
+    {
+      "epoch": 2.929884129751274,
+      "grad_norm": 2.265352249145508,
+      "learning_rate": 1.168597837478768e-06,
+      "loss": 0.7057,
+      "step": 1702500
+    },
+    {
+      "epoch": 2.930744594987618,
+      "grad_norm": 2.140774726867676,
+      "learning_rate": 1.1542567502063683e-06,
+      "loss": 0.7097,
+      "step": 1703000
+    },
+    {
+      "epoch": 2.931605060223962,
+      "grad_norm": 2.231912136077881,
+      "learning_rate": 1.1399156629339685e-06,
+      "loss": 0.7009,
+      "step": 1703500
+    },
+    {
+      "epoch": 2.932465525460306,
+      "grad_norm": 2.278916835784912,
+      "learning_rate": 1.1255745756615688e-06,
+      "loss": 0.7149,
+      "step": 1704000
+    },
+    {
+      "epoch": 2.93332599069665,
+      "grad_norm": 2.3676211833953857,
+      "learning_rate": 1.1112334883891688e-06,
+      "loss": 0.7097,
+      "step": 1704500
+    },
+    {
+      "epoch": 2.934186455932994,
+      "grad_norm": 2.293724536895752,
+      "learning_rate": 1.096892401116769e-06,
+      "loss": 0.7083,
+      "step": 1705000
+    },
+    {
+      "epoch": 2.935046921169338,
+      "grad_norm": 2.1331515312194824,
+      "learning_rate": 1.0825513138443696e-06,
+      "loss": 0.6971,
+      "step": 1705500
+    },
+    {
+      "epoch": 2.935907386405682,
+      "grad_norm": 2.2781825065612793,
+      "learning_rate": 1.0682102265719698e-06,
+      "loss": 0.7116,
+      "step": 1706000
+    },
+    {
+      "epoch": 2.936767851642026,
+      "grad_norm": 2.1554043292999268,
+      "learning_rate": 1.0538691392995699e-06,
+      "loss": 0.7103,
+      "step": 1706500
+    },
+    {
+      "epoch": 2.93762831687837,
+      "grad_norm": 2.3552112579345703,
+      "learning_rate": 1.0395280520271701e-06,
+      "loss": 0.7088,
+      "step": 1707000
+    },
+    {
+      "epoch": 2.938488782114714,
+      "grad_norm": 2.341012954711914,
+      "learning_rate": 1.0251869647547704e-06,
+      "loss": 0.7169,
+      "step": 1707500
+    },
+    {
+      "epoch": 2.939349247351058,
+      "grad_norm": 2.056424856185913,
+      "learning_rate": 1.0108458774823704e-06,
+      "loss": 0.7039,
+      "step": 1708000
+    },
+    {
+      "epoch": 2.940209712587402,
+      "grad_norm": 2.2107203006744385,
+      "learning_rate": 9.965047902099707e-07,
+      "loss": 0.7044,
+      "step": 1708500
+    },
+    {
+      "epoch": 2.941070177823746,
+      "grad_norm": 2.317542791366577,
+      "learning_rate": 9.82163702937571e-07,
+      "loss": 0.7053,
+      "step": 1709000
+    },
+    {
+      "epoch": 2.94193064306009,
+      "grad_norm": 1.9674264192581177,
+      "learning_rate": 9.678226156651712e-07,
+      "loss": 0.7058,
+      "step": 1709500
+    },
+    {
+      "epoch": 2.9427911082964338,
+      "grad_norm": 2.1635196208953857,
+      "learning_rate": 9.534815283927714e-07,
+      "loss": 0.7025,
+      "step": 1710000
+    },
+    {
+      "epoch": 2.9436515735327777,
+      "grad_norm": 2.205508232116699,
+      "learning_rate": 9.391404411203716e-07,
+      "loss": 0.7068,
+      "step": 1710500
+    },
+    {
+      "epoch": 2.9445120387691217,
+      "grad_norm": 2.3096201419830322,
+      "learning_rate": 9.247993538479719e-07,
+      "loss": 0.7066,
+      "step": 1711000
+    },
+    {
+      "epoch": 2.9453725040054657,
+      "grad_norm": 2.1905324459075928,
+      "learning_rate": 9.104582665755721e-07,
+      "loss": 0.7035,
+      "step": 1711500
+    },
+    {
+      "epoch": 2.9462329692418097,
+      "grad_norm": 2.4388911724090576,
+      "learning_rate": 8.961171793031723e-07,
+      "loss": 0.71,
+      "step": 1712000
+    },
+    {
+      "epoch": 2.9470934344781536,
+      "grad_norm": 2.2021710872650146,
+      "learning_rate": 8.817760920307726e-07,
+      "loss": 0.7055,
+      "step": 1712500
+    },
+    {
+      "epoch": 2.9479538997144976,
+      "grad_norm": 2.2784557342529297,
+      "learning_rate": 8.674350047583728e-07,
+      "loss": 0.7062,
+      "step": 1713000
+    },
+    {
+      "epoch": 2.9488143649508416,
+      "grad_norm": 2.1617753505706787,
+      "learning_rate": 8.530939174859731e-07,
+      "loss": 0.7057,
+      "step": 1713500
+    },
+    {
+      "epoch": 2.9496748301871856,
+      "grad_norm": 2.285795211791992,
+      "learning_rate": 8.387528302135732e-07,
+      "loss": 0.7106,
+      "step": 1714000
+    },
+    {
+      "epoch": 2.9505352954235295,
+      "grad_norm": 2.445192337036133,
+      "learning_rate": 8.244117429411735e-07,
+      "loss": 0.7048,
+      "step": 1714500
+    },
+    {
+      "epoch": 2.9513957606598735,
+      "grad_norm": 2.245013475418091,
+      "learning_rate": 8.100706556687737e-07,
+      "loss": 0.7064,
+      "step": 1715000
+    },
+    {
+      "epoch": 2.9522562258962175,
+      "grad_norm": 2.259819269180298,
+      "learning_rate": 7.95729568396374e-07,
+      "loss": 0.7101,
+      "step": 1715500
+    },
+    {
+      "epoch": 2.9531166911325615,
+      "grad_norm": 2.1456029415130615,
+      "learning_rate": 7.81388481123974e-07,
+      "loss": 0.7025,
+      "step": 1716000
+    },
+    {
+      "epoch": 2.9539771563689055,
+      "grad_norm": 2.2223355770111084,
+      "learning_rate": 7.670473938515743e-07,
+      "loss": 0.7048,
+      "step": 1716500
+    },
+    {
+      "epoch": 2.9548376216052494,
+      "grad_norm": 2.179152011871338,
+      "learning_rate": 7.527063065791746e-07,
+      "loss": 0.7075,
+      "step": 1717000
+    },
+    {
+      "epoch": 2.9556980868415934,
+      "grad_norm": 2.1588294506073,
+      "learning_rate": 7.383652193067748e-07,
+      "loss": 0.7109,
+      "step": 1717500
+    },
+    {
+      "epoch": 2.9565585520779374,
+      "grad_norm": 2.1333980560302734,
+      "learning_rate": 7.240241320343751e-07,
+      "loss": 0.7111,
+      "step": 1718000
+    },
+    {
+      "epoch": 2.9574190173142814,
+      "grad_norm": 2.512148380279541,
+      "learning_rate": 7.096830447619752e-07,
+      "loss": 0.7039,
+      "step": 1718500
+    },
+    {
+      "epoch": 2.9582794825506253,
+      "grad_norm": 2.4139938354492188,
+      "learning_rate": 6.953419574895755e-07,
+      "loss": 0.7115,
+      "step": 1719000
+    },
+    {
+      "epoch": 2.9591399477869693,
+      "grad_norm": 2.1482958793640137,
+      "learning_rate": 6.810008702171758e-07,
+      "loss": 0.7001,
+      "step": 1719500
+    },
+    {
+      "epoch": 2.9600004130233133,
+      "grad_norm": 2.4225013256073,
+      "learning_rate": 6.66659782944776e-07,
+      "loss": 0.7079,
+      "step": 1720000
+    },
+    {
+      "epoch": 2.9608608782596573,
+      "grad_norm": 2.3143699169158936,
+      "learning_rate": 6.523186956723762e-07,
+      "loss": 0.7059,
+      "step": 1720500
+    },
+    {
+      "epoch": 2.9617213434960012,
+      "grad_norm": 2.1610562801361084,
+      "learning_rate": 6.379776083999764e-07,
+      "loss": 0.7112,
+      "step": 1721000
+    },
+    {
+      "epoch": 2.9625818087323452,
+      "grad_norm": 2.4261155128479004,
+      "learning_rate": 6.236365211275766e-07,
+      "loss": 0.7089,
+      "step": 1721500
+    },
+    {
+      "epoch": 2.963442273968689,
+      "grad_norm": 2.2455458641052246,
+      "learning_rate": 6.092954338551768e-07,
+      "loss": 0.7095,
+      "step": 1722000
+    },
+    {
+      "epoch": 2.964302739205033,
+      "grad_norm": 2.034914493560791,
+      "learning_rate": 5.949543465827771e-07,
+      "loss": 0.7051,
+      "step": 1722500
+    },
+    {
+      "epoch": 2.965163204441377,
+      "grad_norm": 2.3136467933654785,
+      "learning_rate": 5.806132593103774e-07,
+      "loss": 0.7055,
+      "step": 1723000
+    },
+    {
+      "epoch": 2.966023669677721,
+      "grad_norm": 2.1953635215759277,
+      "learning_rate": 5.662721720379775e-07,
+      "loss": 0.7039,
+      "step": 1723500
+    },
+    {
+      "epoch": 2.966884134914065,
+      "grad_norm": 2.405912160873413,
+      "learning_rate": 5.519310847655778e-07,
+      "loss": 0.7106,
+      "step": 1724000
+    },
+    {
+      "epoch": 2.967744600150409,
+      "grad_norm": 2.216336250305176,
+      "learning_rate": 5.375899974931779e-07,
+      "loss": 0.7082,
+      "step": 1724500
+    },
+    {
+      "epoch": 2.968605065386753,
+      "grad_norm": 2.5332186222076416,
+      "learning_rate": 5.232489102207782e-07,
+      "loss": 0.7062,
+      "step": 1725000
+    },
+    {
+      "epoch": 2.969465530623097,
+      "grad_norm": 2.1927483081817627,
+      "learning_rate": 5.089078229483784e-07,
+      "loss": 0.7146,
+      "step": 1725500
+    },
+    {
+      "epoch": 2.970325995859441,
+      "grad_norm": 2.2343924045562744,
+      "learning_rate": 4.945667356759787e-07,
+      "loss": 0.7102,
+      "step": 1726000
+    },
+    {
+      "epoch": 2.971186461095785,
+      "grad_norm": 2.3195652961730957,
+      "learning_rate": 4.802256484035788e-07,
+      "loss": 0.7045,
+      "step": 1726500
+    },
+    {
+      "epoch": 2.9720469263321294,
+      "grad_norm": 2.3527255058288574,
+      "learning_rate": 4.6588456113117916e-07,
+      "loss": 0.7085,
+      "step": 1727000
+    },
+    {
+      "epoch": 2.9729073915684734,
+      "grad_norm": 2.1666224002838135,
+      "learning_rate": 4.515434738587793e-07,
+      "loss": 0.7037,
+      "step": 1727500
+    },
+    {
+      "epoch": 2.9737678568048174,
+      "grad_norm": 2.1871511936187744,
+      "learning_rate": 4.3720238658637957e-07,
+      "loss": 0.7054,
+      "step": 1728000
+    },
+    {
+      "epoch": 2.9746283220411613,
+      "grad_norm": 2.1100893020629883,
+      "learning_rate": 4.228612993139798e-07,
+      "loss": 0.7099,
+      "step": 1728500
+    },
+    {
+      "epoch": 2.9754887872775053,
+      "grad_norm": 2.129465341567993,
+      "learning_rate": 4.0852021204158004e-07,
+      "loss": 0.7036,
+      "step": 1729000
+    },
+    {
+      "epoch": 2.9763492525138493,
+      "grad_norm": 2.361323356628418,
+      "learning_rate": 3.9417912476918024e-07,
+      "loss": 0.7105,
+      "step": 1729500
+    },
+    {
+      "epoch": 2.9772097177501933,
+      "grad_norm": 2.1633265018463135,
+      "learning_rate": 3.7983803749678045e-07,
+      "loss": 0.7035,
+      "step": 1730000
+    },
+    {
+      "epoch": 2.9780701829865373,
+      "grad_norm": 2.143930673599243,
+      "learning_rate": 3.6549695022438065e-07,
+      "loss": 0.7043,
+      "step": 1730500
+    },
+    {
+      "epoch": 2.9789306482228812,
+      "grad_norm": 2.5136213302612305,
+      "learning_rate": 3.511558629519809e-07,
+      "loss": 0.7052,
+      "step": 1731000
+    },
+    {
+      "epoch": 2.979791113459225,
+      "grad_norm": 2.3515727519989014,
+      "learning_rate": 3.368147756795811e-07,
+      "loss": 0.6972,
+      "step": 1731500
+    },
+    {
+      "epoch": 2.980651578695569,
+      "grad_norm": 2.180917501449585,
+      "learning_rate": 3.224736884071813e-07,
+      "loss": 0.7069,
+      "step": 1732000
+    },
+    {
+      "epoch": 2.981512043931913,
+      "grad_norm": 2.28267240524292,
+      "learning_rate": 3.081326011347816e-07,
+      "loss": 0.7077,
+      "step": 1732500
+    },
+    {
+      "epoch": 2.982372509168257,
+      "grad_norm": 2.4988763332366943,
+      "learning_rate": 2.937915138623818e-07,
+      "loss": 0.7109,
+      "step": 1733000
+    },
+    {
+      "epoch": 2.983232974404601,
+      "grad_norm": 1.968199610710144,
+      "learning_rate": 2.7945042658998205e-07,
+      "loss": 0.7079,
+      "step": 1733500
+    },
+    {
+      "epoch": 2.984093439640945,
+      "grad_norm": 2.156123161315918,
+      "learning_rate": 2.6510933931758225e-07,
+      "loss": 0.7057,
+      "step": 1734000
+    },
+    {
+      "epoch": 2.984953904877289,
+      "grad_norm": 2.4039177894592285,
+      "learning_rate": 2.5076825204518246e-07,
+      "loss": 0.7094,
+      "step": 1734500
+    },
+    {
+      "epoch": 2.985814370113633,
+      "grad_norm": 2.0483903884887695,
+      "learning_rate": 2.3642716477278272e-07,
+      "loss": 0.7071,
+      "step": 1735000
+    },
+    {
+      "epoch": 2.986674835349977,
+      "grad_norm": 2.2727344036102295,
+      "learning_rate": 2.2208607750038292e-07,
+      "loss": 0.706,
+      "step": 1735500
+    },
+    {
+      "epoch": 2.987535300586321,
+      "grad_norm": 2.286410093307495,
+      "learning_rate": 2.0774499022798315e-07,
+      "loss": 0.7068,
+      "step": 1736000
+    },
+    {
+      "epoch": 2.988395765822665,
+      "grad_norm": 2.280778169631958,
+      "learning_rate": 1.9340390295558336e-07,
+      "loss": 0.7066,
+      "step": 1736500
+    },
+    {
+      "epoch": 2.989256231059009,
+      "grad_norm": 2.085190534591675,
+      "learning_rate": 1.790628156831836e-07,
+      "loss": 0.7085,
+      "step": 1737000
+    },
+    {
+      "epoch": 2.990116696295353,
+      "grad_norm": 2.1531052589416504,
+      "learning_rate": 1.647217284107838e-07,
+      "loss": 0.7059,
+      "step": 1737500
+    },
+    {
+      "epoch": 2.990977161531697,
+      "grad_norm": 2.329993486404419,
+      "learning_rate": 1.5038064113838403e-07,
+      "loss": 0.703,
+      "step": 1738000
+    },
+    {
+      "epoch": 2.991837626768041,
+      "grad_norm": 2.5018150806427,
+      "learning_rate": 1.3603955386598426e-07,
+      "loss": 0.7058,
+      "step": 1738500
+    },
+    {
+      "epoch": 2.992698092004385,
+      "grad_norm": 2.285696029663086,
+      "learning_rate": 1.216984665935845e-07,
+      "loss": 0.7069,
+      "step": 1739000
+    },
+    {
+      "epoch": 2.993558557240729,
+      "grad_norm": 2.3342292308807373,
+      "learning_rate": 1.0735737932118471e-07,
+      "loss": 0.7029,
+      "step": 1739500
+    },
+    {
+      "epoch": 2.994419022477073,
+      "grad_norm": 2.3220858573913574,
+      "learning_rate": 9.301629204878495e-08,
+      "loss": 0.702,
+      "step": 1740000
+    },
+    {
+      "epoch": 2.995279487713417,
+      "grad_norm": 2.693772792816162,
+      "learning_rate": 7.867520477638517e-08,
+      "loss": 0.7053,
+      "step": 1740500
+    },
+    {
+      "epoch": 2.9961399529497608,
+      "grad_norm": 2.2459399700164795,
+      "learning_rate": 6.43341175039854e-08,
+      "loss": 0.7045,
+      "step": 1741000
+    },
+    {
+      "epoch": 2.9970004181861047,
+      "grad_norm": 2.4572110176086426,
+      "learning_rate": 4.9993030231585616e-08,
+      "loss": 0.707,
+      "step": 1741500
+    },
+    {
+      "epoch": 2.9978608834224487,
+      "grad_norm": 2.173382520675659,
+      "learning_rate": 3.565194295918584e-08,
+      "loss": 0.7097,
+      "step": 1742000
+    },
+    {
+      "epoch": 2.9987213486587927,
+      "grad_norm": 2.199479579925537,
+      "learning_rate": 2.1310855686786068e-08,
+      "loss": 0.7047,
+      "step": 1742500
+    },
+    {
+      "epoch": 2.9995818138951367,
+      "grad_norm": 2.2377700805664062,
+      "learning_rate": 6.969768414386291e-09,
+      "loss": 0.7061,
+      "step": 1743000
+    },
+    {
+      "epoch": 3.0,
+      "step": 1743243,
+      "total_flos": 1.4697804058526134e+19,
+      "train_loss": 0.807495192525475,
+      "train_runtime": 715284.2162,
+      "train_samples_per_second": 77.988,
+      "train_steps_per_second": 2.437
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 1743243,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4697804058526134e+19,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}