diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,8970 +1,2948 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9992804029743345,
-  "eval_steps": 1000,
-  "global_step": 6252,
+  "epoch": 2.9964020148716717,
+  "eval_steps": 1400,
+  "global_step": 2082,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0023986567522187576,
-      "grad_norm": 4.928963404047604,
-      "learning_rate": 7.987220447284345e-08,
-      "loss": 1.1911,
+      "epoch": 0.007195970256656272,
+      "grad_norm": 3.382328658618869,
+      "learning_rate": 1.9138755980861244e-07,
+      "loss": 1.1034,
       "step": 5
     },
     {
-      "epoch": 0.004797313504437515,
-      "grad_norm": 6.084456595330925,
-      "learning_rate": 1.597444089456869e-07,
-      "loss": 1.1865,
+      "epoch": 0.014391940513312544,
+      "grad_norm": 2.9332510464158155,
+      "learning_rate": 3.827751196172249e-07,
+      "loss": 1.1248,
       "step": 10
     },
     {
-      "epoch": 0.007195970256656272,
-      "grad_norm": 4.612988319344874,
-      "learning_rate": 2.396166134185304e-07,
-      "loss": 1.1439,
+      "epoch": 0.021587910769968816,
+      "grad_norm": 2.8021125575132175,
+      "learning_rate": 5.741626794258373e-07,
+      "loss": 1.0979,
       "step": 15
     },
     {
-      "epoch": 0.00959462700887503,
-      "grad_norm": 4.7903715576713894,
-      "learning_rate": 3.194888178913738e-07,
-      "loss": 1.1287,
+      "epoch": 0.02878388102662509,
+      "grad_norm": 2.5443691578126004,
+      "learning_rate": 7.655502392344498e-07,
+      "loss": 1.1128,
       "step": 20
     },
     {
-      "epoch": 0.011993283761093787,
-      "grad_norm": 5.511263638761197,
-      "learning_rate": 3.993610223642173e-07,
-      "loss": 1.212,
+      "epoch": 0.035979851283281364,
+      "grad_norm": 2.823918046213155,
+      "learning_rate": 9.569377990430622e-07,
+      "loss": 1.0726,
       "step": 25
     },
     {
-      "epoch": 0.014391940513312544,
-      "grad_norm": 4.364651246737439,
-      "learning_rate": 4.792332268370608e-07,
-      "loss": 1.2221,
+      "epoch": 0.04317582153993763,
+      "grad_norm": 2.0326907547127124,
+      "learning_rate": 1.1483253588516746e-06,
+      "loss": 1.0043,
       "step": 30
     },
     {
-      "epoch": 0.016790597265531303,
-      "grad_norm": 14.508142045947785,
-      "learning_rate": 5.591054313099041e-07,
-      "loss": 1.1268,
+      "epoch": 0.05037179179659391,
+      "grad_norm": 2.168998278242428,
+      "learning_rate": 1.339712918660287e-06,
+      "loss": 1.0079,
       "step": 35
     },
     {
-      "epoch": 0.01918925401775006,
-      "grad_norm": 3.618955442205435,
-      "learning_rate": 6.389776357827476e-07,
-      "loss": 1.0991,
+      "epoch": 0.05756776205325018,
+      "grad_norm": 2.152879633190319,
+      "learning_rate": 1.5311004784688995e-06,
+      "loss": 1.0086,
       "step": 40
     },
     {
-      "epoch": 0.021587910769968816,
-      "grad_norm": 3.3756865136406042,
-      "learning_rate": 7.188498402555912e-07,
-      "loss": 1.2278,
+      "epoch": 0.06476373230990645,
+      "grad_norm": 1.7771809680542394,
+      "learning_rate": 1.722488038277512e-06,
+      "loss": 0.9862,
       "step": 45
     },
     {
-      "epoch": 0.023986567522187575,
-      "grad_norm": 9.702984255699223,
-      "learning_rate": 7.987220447284346e-07,
-      "loss": 1.201,
+      "epoch": 0.07195970256656273,
+      "grad_norm": 1.6796335234145752,
+      "learning_rate": 1.9138755980861244e-06,
+      "loss": 0.9757,
       "step": 50
     },
     {
-      "epoch": 0.026385224274406333,
-      "grad_norm": 5.04285882630285,
-      "learning_rate": 8.78594249201278e-07,
-      "loss": 1.0465,
+      "epoch": 0.079155672823219,
+      "grad_norm": 1.7111004529632545,
+      "learning_rate": 2.1052631578947366e-06,
+      "loss": 0.9067,
       "step": 55
     },
     {
-      "epoch": 0.02878388102662509,
-      "grad_norm": 4.782314822096669,
-      "learning_rate": 9.584664536741215e-07,
-      "loss": 1.1928,
+      "epoch": 0.08635164307987526,
+      "grad_norm": 2.1161929062800184,
+      "learning_rate": 2.2966507177033493e-06,
+      "loss": 0.9534,
       "step": 60
     },
     {
-      "epoch": 0.031182537778843847,
-      "grad_norm": 4.72671053943802,
-      "learning_rate": 1.038338658146965e-06,
-      "loss": 1.2059,
+      "epoch": 0.09354761333653154,
+      "grad_norm": 1.739361359363285,
+      "learning_rate": 2.4880382775119615e-06,
+      "loss": 0.9368,
       "step": 65
     },
     {
-      "epoch": 0.033581194531062605,
-      "grad_norm": 6.921221031777395,
-      "learning_rate": 1.1182108626198083e-06,
-      "loss": 0.9764,
+      "epoch": 0.10074358359318782,
+      "grad_norm": 1.9645645051440501,
+      "learning_rate": 2.679425837320574e-06,
+      "loss": 0.8762,
       "step": 70
     },
     {
-      "epoch": 0.035979851283281364,
-      "grad_norm": 4.2701184749105785,
-      "learning_rate": 1.198083067092652e-06,
-      "loss": 1.091,
+      "epoch": 0.10793955384984409,
+      "grad_norm": 2.291876148211727,
+      "learning_rate": 2.8708133971291864e-06,
+      "loss": 0.9079,
       "step": 75
     },
     {
-      "epoch": 0.03837850803550012,
-      "grad_norm": 4.417378125112381,
-      "learning_rate": 1.2779552715654952e-06,
-      "loss": 0.9595,
+      "epoch": 0.11513552410650035,
+      "grad_norm": 1.9350312982517304,
+      "learning_rate": 3.062200956937799e-06,
+      "loss": 0.9063,
       "step": 80
     },
     {
-      "epoch": 0.04077716478771888,
-      "grad_norm": 6.151705197956376,
-      "learning_rate": 1.3578274760383387e-06,
-      "loss": 1.031,
+      "epoch": 0.12233149436315663,
+      "grad_norm": 1.723686001407234,
+      "learning_rate": 3.2535885167464113e-06,
+      "loss": 0.9141,
       "step": 85
     },
     {
-      "epoch": 0.04317582153993763,
-      "grad_norm": 4.20485503693391,
-      "learning_rate": 1.4376996805111824e-06,
-      "loss": 0.9779,
+      "epoch": 0.1295274646198129,
+      "grad_norm": 1.7291796951049525,
+      "learning_rate": 3.444976076555024e-06,
+      "loss": 0.8852,
       "step": 90
     },
     {
-      "epoch": 0.04557447829215639,
-      "grad_norm": 7.172834781189588,
-      "learning_rate": 1.5175718849840257e-06,
-      "loss": 1.0034,
+      "epoch": 0.13672343487646918,
+      "grad_norm": 1.5591618462436518,
+      "learning_rate": 3.636363636363636e-06,
+      "loss": 0.8812,
       "step": 95
     },
     {
-      "epoch": 0.04797313504437515,
-      "grad_norm": 3.5852162346181022,
-      "learning_rate": 1.5974440894568691e-06,
-      "loss": 1.0194,
-      "step": 100
-    },
-    {
-      "epoch": 0.04797313504437515,
-      "eval_loss": 1.0799548625946045,
-      "eval_runtime": 759.5931,
-      "eval_samples_per_second": 9.755,
-      "eval_steps_per_second": 0.611,
+      "epoch": 0.14391940513312546,
+      "grad_norm": 2.5857950863717596,
+      "learning_rate": 3.827751196172249e-06,
+      "loss": 0.8719,
       "step": 100
     },
     {
-      "epoch": 0.05037179179659391,
-      "grad_norm": 3.2060401406256314,
-      "learning_rate": 1.6773162939297128e-06,
-      "loss": 0.994,
+      "epoch": 0.15111537538978173,
+      "grad_norm": 1.882950492021297,
+      "learning_rate": 4.019138755980861e-06,
+      "loss": 0.923,
       "step": 105
     },
     {
-      "epoch": 0.052770448548812667,
-      "grad_norm": 3.8285067274755638,
-      "learning_rate": 1.757188498402556e-06,
-      "loss": 0.9704,
+      "epoch": 0.158311345646438,
+      "grad_norm": 1.9095611162862844,
+      "learning_rate": 4.210526315789473e-06,
+      "loss": 0.8466,
       "step": 110
     },
     {
-      "epoch": 0.055169105301031425,
-      "grad_norm": 4.008540740919368,
-      "learning_rate": 1.8370607028753996e-06,
-      "loss": 1.0667,
+      "epoch": 0.16550731590309425,
+      "grad_norm": 1.7349719997638062,
+      "learning_rate": 4.4019138755980855e-06,
+      "loss": 0.8433,
       "step": 115
     },
     {
-      "epoch": 0.05756776205325018,
-      "grad_norm": 4.778094922475321,
-      "learning_rate": 1.916932907348243e-06,
-      "loss": 0.9591,
+      "epoch": 0.17270328615975053,
+      "grad_norm": 1.6057643525733838,
+      "learning_rate": 4.5933014354066986e-06,
+      "loss": 0.8901,
       "step": 120
     },
     {
-      "epoch": 0.059966418805468935,
-      "grad_norm": 5.3861654284616955,
-      "learning_rate": 1.9968051118210867e-06,
-      "loss": 0.9443,
+      "epoch": 0.1798992564164068,
+      "grad_norm": 1.636440080236875,
+      "learning_rate": 4.784688995215311e-06,
+      "loss": 0.9158,
       "step": 125
     },
     {
-      "epoch": 0.062365075557687694,
-      "grad_norm": 4.224941464907675,
-      "learning_rate": 2.07667731629393e-06,
-      "loss": 0.932,
+      "epoch": 0.18709522667306308,
+      "grad_norm": 1.7680919340453238,
+      "learning_rate": 4.976076555023923e-06,
+      "loss": 0.9012,
       "step": 130
     },
     {
-      "epoch": 0.06476373230990645,
-      "grad_norm": 4.606013084447571,
-      "learning_rate": 2.1565495207667733e-06,
-      "loss": 0.9077,
+      "epoch": 0.19429119692971936,
+      "grad_norm": 1.73915276490131,
+      "learning_rate": 5.167464114832536e-06,
+      "loss": 0.8372,
       "step": 135
     },
     {
-      "epoch": 0.06716238906212521,
-      "grad_norm": 5.077386951712394,
-      "learning_rate": 2.2364217252396165e-06,
-      "loss": 0.8907,
+      "epoch": 0.20148716718637563,
+      "grad_norm": 1.4761634523053184,
+      "learning_rate": 5.358851674641148e-06,
+      "loss": 0.8631,
       "step": 140
     },
     {
-      "epoch": 0.06956104581434397,
-      "grad_norm": 3.8083760700058287,
-      "learning_rate": 2.3162939297124602e-06,
-      "loss": 0.9614,
+      "epoch": 0.2086831374430319,
+      "grad_norm": 1.7615083334015453,
+      "learning_rate": 5.5502392344497606e-06,
+      "loss": 0.8275,
       "step": 145
     },
     {
-      "epoch": 0.07195970256656273,
-      "grad_norm": 3.3687263984866314,
-      "learning_rate": 2.396166134185304e-06,
-      "loss": 0.9853,
+      "epoch": 0.21587910769968818,
+      "grad_norm": 1.5283874292360122,
+      "learning_rate": 5.741626794258373e-06,
+      "loss": 0.8625,
       "step": 150
     },
     {
-      "epoch": 0.07435835931878149,
-      "grad_norm": 3.9276886090521725,
-      "learning_rate": 2.476038338658147e-06,
-      "loss": 0.8836,
+      "epoch": 0.22307507795634446,
+      "grad_norm": 1.6114784875621726,
+      "learning_rate": 5.933014354066985e-06,
+      "loss": 0.891,
       "step": 155
     },
     {
-      "epoch": 0.07675701607100024,
-      "grad_norm": 4.504046079167171,
-      "learning_rate": 2.5559105431309904e-06,
-      "loss": 0.8545,
+      "epoch": 0.2302710482130007,
+      "grad_norm": 1.6981117694270553,
+      "learning_rate": 6.124401913875598e-06,
+      "loss": 0.8548,
       "step": 160
     },
     {
-      "epoch": 0.079155672823219,
-      "grad_norm": 3.33674905835483,
-      "learning_rate": 2.6357827476038337e-06,
-      "loss": 0.8817,
+      "epoch": 0.23746701846965698,
+      "grad_norm": 1.7093051041039082,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.8953,
       "step": 165
     },
     {
-      "epoch": 0.08155432957543776,
-      "grad_norm": 4.009856578725833,
-      "learning_rate": 2.7156549520766774e-06,
-      "loss": 0.9024,
+      "epoch": 0.24466298872631326,
+      "grad_norm": 1.6829774560988056,
+      "learning_rate": 6.5071770334928226e-06,
+      "loss": 0.7894,
       "step": 170
     },
     {
-      "epoch": 0.0839529863276565,
-      "grad_norm": 2.9677427686921694,
-      "learning_rate": 2.795527156549521e-06,
-      "loss": 0.9347,
+      "epoch": 0.25185895898296956,
+      "grad_norm": 1.7224253982363609,
+      "learning_rate": 6.698564593301436e-06,
+      "loss": 0.9342,
       "step": 175
     },
     {
-      "epoch": 0.08635164307987526,
-      "grad_norm": 6.375386023546387,
-      "learning_rate": 2.8753993610223648e-06,
-      "loss": 0.9421,
+      "epoch": 0.2590549292396258,
+      "grad_norm": 1.718825453842612,
+      "learning_rate": 6.889952153110048e-06,
+      "loss": 0.8765,
       "step": 180
     },
     {
-      "epoch": 0.08875029983209402,
-      "grad_norm": 3.33944743041201,
-      "learning_rate": 2.955271565495208e-06,
-      "loss": 0.9086,
+      "epoch": 0.26625089949628206,
+      "grad_norm": 1.8299643097592468,
+      "learning_rate": 7.081339712918659e-06,
+      "loss": 0.8562,
       "step": 185
     },
     {
-      "epoch": 0.09114895658431278,
-      "grad_norm": 4.394097556098567,
-      "learning_rate": 3.0351437699680513e-06,
-      "loss": 0.7906,
+      "epoch": 0.27344686975293836,
+      "grad_norm": 1.9650794343564884,
+      "learning_rate": 7.272727272727272e-06,
+      "loss": 0.8468,
       "step": 190
     },
     {
-      "epoch": 0.09354761333653154,
-      "grad_norm": 9.160425752039988,
-      "learning_rate": 3.1150159744408946e-06,
-      "loss": 0.9486,
+      "epoch": 0.2806428400095946,
+      "grad_norm": 1.6665409567649216,
+      "learning_rate": 7.4641148325358846e-06,
+      "loss": 0.8394,
       "step": 195
     },
     {
-      "epoch": 0.0959462700887503,
-      "grad_norm": 3.571966750853734,
-      "learning_rate": 3.1948881789137383e-06,
-      "loss": 0.8006,
-      "step": 200
-    },
-    {
-      "epoch": 0.0959462700887503,
-      "eval_loss": 0.8958361148834229,
-      "eval_runtime": 739.3588,
-      "eval_samples_per_second": 10.022,
-      "eval_steps_per_second": 0.628,
+      "epoch": 0.2878388102662509,
+      "grad_norm": 1.6178516494615538,
+      "learning_rate": 7.655502392344498e-06,
+      "loss": 0.9063,
       "step": 200
     },
     {
-      "epoch": 0.09834492684096906,
-      "grad_norm": 4.8180626358321295,
-      "learning_rate": 3.2747603833865815e-06,
-      "loss": 0.8072,
+      "epoch": 0.29503478052290716,
+      "grad_norm": 1.737750416660729,
+      "learning_rate": 7.84688995215311e-06,
+      "loss": 0.8376,
       "step": 205
     },
     {
-      "epoch": 0.10074358359318782,
-      "grad_norm": 3.7249040169973715,
-      "learning_rate": 3.3546325878594257e-06,
-      "loss": 0.9298,
+      "epoch": 0.30223075077956346,
+      "grad_norm": 1.6775958248025469,
+      "learning_rate": 7.9999943732958e-06,
+      "loss": 0.8839,
       "step": 210
     },
     {
-      "epoch": 0.10314224034540657,
-      "grad_norm": 4.440940528532116,
-      "learning_rate": 3.434504792332269e-06,
-      "loss": 0.9033,
+      "epoch": 0.3094267210362197,
+      "grad_norm": 1.6351855660641077,
+      "learning_rate": 7.999797440310976e-06,
+      "loss": 0.8138,
       "step": 215
     },
     {
-      "epoch": 0.10554089709762533,
-      "grad_norm": 2.7840742197428217,
-      "learning_rate": 3.514376996805112e-06,
-      "loss": 0.9216,
+      "epoch": 0.316622691292876,
+      "grad_norm": 1.5160913628770296,
+      "learning_rate": 7.999319187945908e-06,
+      "loss": 0.8634,
       "step": 220
     },
     {
-      "epoch": 0.10793955384984409,
-      "grad_norm": 4.303196845314569,
-      "learning_rate": 3.5942492012779555e-06,
-      "loss": 0.8446,
+      "epoch": 0.32381866154953226,
+      "grad_norm": 1.5346049341737504,
+      "learning_rate": 7.998559649837715e-06,
+      "loss": 0.8777,
       "step": 225
     },
     {
-      "epoch": 0.11033821060206285,
-      "grad_norm": 3.21207580531984,
-      "learning_rate": 3.674121405750799e-06,
-      "loss": 0.8883,
+      "epoch": 0.3310146318061885,
+      "grad_norm": 1.7202200592837338,
+      "learning_rate": 7.997518879407302e-06,
+      "loss": 0.9041,
       "step": 230
     },
     {
-      "epoch": 0.11273686735428161,
-      "grad_norm": 5.207754700223726,
-      "learning_rate": 3.7539936102236424e-06,
-      "loss": 0.8105,
+      "epoch": 0.3382106020628448,
+      "grad_norm": 1.6194787415884186,
+      "learning_rate": 7.996196949855597e-06,
+      "loss": 0.8567,
       "step": 235
     },
     {
-      "epoch": 0.11513552410650035,
-      "grad_norm": 4.057648237886164,
-      "learning_rate": 3.833865814696486e-06,
-      "loss": 0.9013,
+      "epoch": 0.34540657231950106,
+      "grad_norm": 1.5761200757620502,
+      "learning_rate": 7.994593954158409e-06,
+      "loss": 0.8683,
       "step": 240
     },
     {
-      "epoch": 0.11753418085871911,
-      "grad_norm": 3.1683543571860713,
-      "learning_rate": 3.913738019169329e-06,
-      "loss": 0.8875,
+      "epoch": 0.35260254257615736,
+      "grad_norm": 1.6216799441610352,
+      "learning_rate": 7.992710005059886e-06,
+      "loss": 0.8718,
       "step": 245
     },
     {
-      "epoch": 0.11993283761093787,
-      "grad_norm": 4.234833266114255,
-      "learning_rate": 3.9936102236421735e-06,
-      "loss": 0.8656,
+      "epoch": 0.3597985128328136,
+      "grad_norm": 1.4982517397299326,
+      "learning_rate": 7.990545235064588e-06,
+      "loss": 0.8491,
       "step": 250
     },
     {
-      "epoch": 0.12233149436315663,
-      "grad_norm": 2.909152649590802,
-      "learning_rate": 4.073482428115016e-06,
-      "loss": 0.8178,
+      "epoch": 0.3669944830894699,
+      "grad_norm": 1.7915522189289803,
+      "learning_rate": 7.988099796428161e-06,
+      "loss": 0.8546,
       "step": 255
     },
     {
-      "epoch": 0.12473015111537539,
-      "grad_norm": 2.751539094221403,
-      "learning_rate": 4.15335463258786e-06,
-      "loss": 0.9129,
+      "epoch": 0.37419045334612616,
+      "grad_norm": 1.6209807612052567,
+      "learning_rate": 7.985373861146636e-06,
+      "loss": 0.8112,
       "step": 260
     },
     {
-      "epoch": 0.12712880786759415,
-      "grad_norm": 3.0547647847167974,
-      "learning_rate": 4.233226837060703e-06,
-      "loss": 0.8499,
+      "epoch": 0.38138642360278247,
+      "grad_norm": 1.5354364067126935,
+      "learning_rate": 7.98236762094433e-06,
+      "loss": 0.8484,
       "step": 265
     },
     {
-      "epoch": 0.1295274646198129,
-      "grad_norm": 2.3819103654972023,
-      "learning_rate": 4.3130990415335465e-06,
-      "loss": 0.8459,
+      "epoch": 0.3885823938594387,
+      "grad_norm": 1.7666681382505078,
+      "learning_rate": 7.979081287260356e-06,
+      "loss": 0.8752,
       "step": 270
     },
     {
-      "epoch": 0.13192612137203166,
-      "grad_norm": 3.326228240795246,
-      "learning_rate": 4.39297124600639e-06,
-      "loss": 0.9211,
+      "epoch": 0.39577836411609496,
+      "grad_norm": 1.6458650430939799,
+      "learning_rate": 7.975515091233757e-06,
+      "loss": 0.8294,
       "step": 275
     },
     {
-      "epoch": 0.13432477812425042,
-      "grad_norm": 3.846786293390378,
-      "learning_rate": 4.472843450479233e-06,
-      "loss": 0.7904,
+      "epoch": 0.40297433437275126,
+      "grad_norm": 1.7589737639189482,
+      "learning_rate": 7.971669283687252e-06,
+      "loss": 0.8269,
       "step": 280
     },
     {
-      "epoch": 0.13672343487646918,
-      "grad_norm": 2.856338249786616,
-      "learning_rate": 4.552715654952077e-06,
-      "loss": 0.7958,
+      "epoch": 0.4101703046294075,
+      "grad_norm": 1.597742594232911,
+      "learning_rate": 7.967544135109583e-06,
+      "loss": 0.8873,
       "step": 285
     },
     {
-      "epoch": 0.13912209162868794,
-      "grad_norm": 3.452725703157698,
-      "learning_rate": 4.6325878594249205e-06,
-      "loss": 0.8353,
+      "epoch": 0.4173662748860638,
+      "grad_norm": 1.516759424294559,
+      "learning_rate": 7.963139935636505e-06,
+      "loss": 0.8162,
       "step": 290
     },
     {
-      "epoch": 0.1415207483809067,
-      "grad_norm": 2.8816531878068066,
-      "learning_rate": 4.712460063897764e-06,
-      "loss": 0.891,
+      "epoch": 0.42456224514272006,
+      "grad_norm": 1.6997534481298684,
+      "learning_rate": 7.958456995030372e-06,
+      "loss": 0.8202,
       "step": 295
     },
     {
-      "epoch": 0.14391940513312546,
-      "grad_norm": 4.1723661091130415,
-      "learning_rate": 4.792332268370608e-06,
-      "loss": 0.8183,
-      "step": 300
-    },
-    {
-      "epoch": 0.14391940513312546,
-      "eval_loss": 0.8849568963050842,
-      "eval_runtime": 741.1905,
-      "eval_samples_per_second": 9.997,
-      "eval_steps_per_second": 0.626,
+      "epoch": 0.43175821539937637,
+      "grad_norm": 1.726622337993047,
+      "learning_rate": 7.95349564265835e-06,
+      "loss": 0.8456,
       "step": 300
     },
     {
-      "epoch": 0.1463180618853442,
-      "grad_norm": 2.5376644618127506,
-      "learning_rate": 4.872204472843451e-06,
-      "loss": 0.8931,
+      "epoch": 0.4389541856560326,
+      "grad_norm": 1.647760739349251,
+      "learning_rate": 7.94825622746925e-06,
+      "loss": 0.8648,
       "step": 305
     },
     {
-      "epoch": 0.14871671863756297,
-      "grad_norm": 16.404073305191517,
-      "learning_rate": 4.952076677316294e-06,
-      "loss": 0.8317,
+      "epoch": 0.4461501559126889,
+      "grad_norm": 1.7211951131144776,
+      "learning_rate": 7.942739117968995e-06,
+      "loss": 0.8272,
       "step": 310
     },
     {
-      "epoch": 0.15111537538978173,
-      "grad_norm": 5.422899883781751,
-      "learning_rate": 5.031948881789138e-06,
-      "loss": 1.0199,
+      "epoch": 0.45334612616934516,
+      "grad_norm": 1.5335036085040443,
+      "learning_rate": 7.936944702194691e-06,
+      "loss": 0.878,
       "step": 315
     },
     {
-      "epoch": 0.1535140321420005,
-      "grad_norm": 3.618606897676806,
-      "learning_rate": 5.111821086261981e-06,
-      "loss": 0.8023,
+      "epoch": 0.4605420964260014,
+      "grad_norm": 1.4848768316508538,
+      "learning_rate": 7.93087338768734e-06,
+      "loss": 0.8456,
       "step": 320
     },
     {
-      "epoch": 0.15591268889421925,
-      "grad_norm": 3.2505122550483847,
-      "learning_rate": 5.191693290734825e-06,
-      "loss": 0.8634,
+      "epoch": 0.4677380666826577,
+      "grad_norm": 1.5394168018437981,
+      "learning_rate": 7.924525601463173e-06,
+      "loss": 0.8427,
       "step": 325
     },
     {
-      "epoch": 0.158311345646438,
-      "grad_norm": 4.168273711265947,
-      "learning_rate": 5.2715654952076674e-06,
-      "loss": 0.8512,
+      "epoch": 0.47493403693931396,
+      "grad_norm": 1.559312757501083,
+      "learning_rate": 7.91790178998362e-06,
+      "loss": 0.8202,
       "step": 330
     },
     {
-      "epoch": 0.16071000239865676,
-      "grad_norm": 3.8748834001471324,
-      "learning_rate": 5.351437699680512e-06,
-      "loss": 0.8893,
+      "epoch": 0.48213000719597027,
+      "grad_norm": 1.5792217882137543,
+      "learning_rate": 7.91100241912391e-06,
+      "loss": 0.8419,
       "step": 335
     },
     {
-      "epoch": 0.16310865915087552,
-      "grad_norm": 3.0369612897329143,
-      "learning_rate": 5.431309904153355e-06,
-      "loss": 0.8146,
+      "epoch": 0.4893259774526265,
+      "grad_norm": 1.738156211889958,
+      "learning_rate": 7.9038279741403e-06,
+      "loss": 0.8659,
       "step": 340
     },
     {
-      "epoch": 0.16550731590309425,
-      "grad_norm": 3.0242707239755164,
-      "learning_rate": 5.5111821086261985e-06,
-      "loss": 0.8655,
+      "epoch": 0.4965219477092828,
+      "grad_norm": 1.7060742398951603,
+      "learning_rate": 7.896378959635946e-06,
+      "loss": 0.8564,
       "step": 345
     },
     {
-      "epoch": 0.167905972655313,
-      "grad_norm": 2.807112546364469,
-      "learning_rate": 5.591054313099042e-06,
-      "loss": 0.8343,
+      "epoch": 0.5037179179659391,
+      "grad_norm": 1.483857852582522,
+      "learning_rate": 7.888655899525413e-06,
+      "loss": 0.8122,
       "step": 350
     },
     {
-      "epoch": 0.17030462940753177,
-      "grad_norm": 3.0445759143839974,
-      "learning_rate": 5.670926517571885e-06,
-      "loss": 0.8672,
+      "epoch": 0.5109138882225953,
+      "grad_norm": 1.468389601692516,
+      "learning_rate": 7.880659336997833e-06,
+      "loss": 0.887,
       "step": 355
     },
     {
-      "epoch": 0.17270328615975053,
-      "grad_norm": 2.533966215351572,
-      "learning_rate": 5.7507987220447296e-06,
-      "loss": 0.9204,
+      "epoch": 0.5181098584792516,
+      "grad_norm": 1.5870295716868312,
+      "learning_rate": 7.872389834478688e-06,
+      "loss": 0.8813,
       "step": 360
     },
     {
-      "epoch": 0.1751019429119693,
-      "grad_norm": 2.637089076537768,
-      "learning_rate": 5.830670926517572e-06,
-      "loss": 0.9182,
+      "epoch": 0.5253058287359079,
+      "grad_norm": 1.8895020621596357,
+      "learning_rate": 7.863847973590265e-06,
+      "loss": 0.8626,
       "step": 365
     },
     {
-      "epoch": 0.17750059966418805,
-      "grad_norm": 2.9535137171261545,
-      "learning_rate": 5.910543130990416e-06,
-      "loss": 0.9401,
+      "epoch": 0.5325017989925641,
+      "grad_norm": 1.550855518803781,
+      "learning_rate": 7.855034355110736e-06,
+      "loss": 0.8546,
       "step": 370
     },
     {
-      "epoch": 0.1798992564164068,
-      "grad_norm": 3.215701545330308,
-      "learning_rate": 5.990415335463259e-06,
-      "loss": 0.8791,
+      "epoch": 0.5396977692492204,
+      "grad_norm": 1.5611172491745864,
+      "learning_rate": 7.845949598931918e-06,
+      "loss": 0.848,
       "step": 375
     },
     {
-      "epoch": 0.18229791316862556,
-      "grad_norm": 2.7011931711589443,
-      "learning_rate": 6.070287539936103e-06,
-      "loss": 0.9199,
+      "epoch": 0.5468937395058767,
+      "grad_norm": 1.5611908671622932,
+      "learning_rate": 7.836594344015661e-06,
+      "loss": 0.8738,
       "step": 380
     },
     {
-      "epoch": 0.18469656992084432,
-      "grad_norm": 2.707790719957795,
-      "learning_rate": 6.150159744408946e-06,
-      "loss": 0.8737,
+      "epoch": 0.554089709762533,
+      "grad_norm": 1.5092812979555394,
+      "learning_rate": 7.826969248348915e-06,
+      "loss": 0.8693,
       "step": 385
     },
     {
-      "epoch": 0.18709522667306308,
-      "grad_norm": 3.597189776277077,
-      "learning_rate": 6.230031948881789e-06,
-      "loss": 0.9733,
+      "epoch": 0.5612856800191892,
+      "grad_norm": 2.804211699082011,
+      "learning_rate": 7.817074988897446e-06,
+      "loss": 0.8373,
       "step": 390
     },
     {
-      "epoch": 0.18949388342528184,
-      "grad_norm": 2.9193000070767354,
-      "learning_rate": 6.309904153354634e-06,
-      "loss": 0.8133,
+      "epoch": 0.5684816502758455,
+      "grad_norm": 1.6322846593476272,
+      "learning_rate": 7.806912261558232e-06,
+      "loss": 0.8179,
       "step": 395
     },
     {
-      "epoch": 0.1918925401775006,
-      "grad_norm": 2.7897280116321106,
-      "learning_rate": 6.3897763578274765e-06,
-      "loss": 0.7918,
-      "step": 400
-    },
-    {
-      "epoch": 0.1918925401775006,
-      "eval_loss": 0.8788505792617798,
-      "eval_runtime": 738.99,
-      "eval_samples_per_second": 10.027,
-      "eval_steps_per_second": 0.628,
+      "epoch": 0.5756776205325018,
+      "grad_norm": 1.537211750195004,
+      "learning_rate": 7.796481781110504e-06,
+      "loss": 0.8881,
       "step": 400
     },
     {
-      "epoch": 0.19429119692971936,
-      "grad_norm": 2.904956056545914,
-      "learning_rate": 6.46964856230032e-06,
-      "loss": 0.8769,
+      "epoch": 0.5828735907891581,
+      "grad_norm": 1.6424496003416038,
+      "learning_rate": 7.785784281165491e-06,
+      "loss": 0.8285,
       "step": 405
     },
     {
-      "epoch": 0.19668985368193811,
-      "grad_norm": 2.694702495778036,
-      "learning_rate": 6.549520766773163e-06,
-      "loss": 0.8796,
+      "epoch": 0.5900695610458143,
+      "grad_norm": 1.712983639226015,
+      "learning_rate": 7.774820514114804e-06,
+      "loss": 0.8471,
       "step": 410
     },
     {
-      "epoch": 0.19908851043415687,
-      "grad_norm": 2.958505594137407,
-      "learning_rate": 6.629392971246007e-06,
-      "loss": 0.7932,
+      "epoch": 0.5972655313024706,
+      "grad_norm": 1.6543355090148368,
+      "learning_rate": 7.763591251077532e-06,
+      "loss": 0.8181,
       "step": 415
     },
     {
-      "epoch": 0.20148716718637563,
-      "grad_norm": 2.954096827602939,
-      "learning_rate": 6.709265175718851e-06,
-      "loss": 0.9052,
+      "epoch": 0.6044615015591269,
+      "grad_norm": 1.7695569793499315,
+      "learning_rate": 7.752097281845998e-06,
+      "loss": 0.8317,
       "step": 420
     },
     {
-      "epoch": 0.2038858239385944,
-      "grad_norm": 4.110594691742628,
-      "learning_rate": 6.789137380191694e-06,
-      "loss": 0.7819,
+      "epoch": 0.6116574718157831,
+      "grad_norm": 1.6408594582199851,
+      "learning_rate": 7.740339414830216e-06,
+      "loss": 0.8822,
       "step": 425
     },
     {
-      "epoch": 0.20628448069081315,
-      "grad_norm": 2.8603447719001673,
-      "learning_rate": 6.869009584664538e-06,
-      "loss": 0.8848,
+      "epoch": 0.6188534420724394,
+      "grad_norm": 1.532763883790403,
+      "learning_rate": 7.72831847700103e-06,
+      "loss": 0.8858,
       "step": 430
     },
     {
-      "epoch": 0.2086831374430319,
-      "grad_norm": 3.677542996907206,
-      "learning_rate": 6.948881789137381e-06,
-      "loss": 0.8922,
+      "epoch": 0.6260494123290957,
+      "grad_norm": 1.5699531832045912,
+      "learning_rate": 7.71603531383195e-06,
+      "loss": 0.803,
       "step": 435
     },
     {
-      "epoch": 0.21108179419525067,
-      "grad_norm": 2.3643998409365006,
-      "learning_rate": 7.028753993610224e-06,
-      "loss": 0.8853,
+      "epoch": 0.633245382585752,
+      "grad_norm": 1.4023688398705096,
+      "learning_rate": 7.703490789239685e-06,
+      "loss": 0.8015,
       "step": 440
     },
     {
-      "epoch": 0.21348045094746942,
-      "grad_norm": 2.811619082486357,
-      "learning_rate": 7.108626198083067e-06,
-      "loss": 0.8435,
+      "epoch": 0.6404413528424082,
+      "grad_norm": 1.743812894669404,
+      "learning_rate": 7.690685785523388e-06,
+      "loss": 0.8398,
       "step": 445
     },
     {
-      "epoch": 0.21587910769968818,
-      "grad_norm": 2.7629535174547497,
-      "learning_rate": 7.188498402555911e-06,
-      "loss": 0.9014,
+      "epoch": 0.6476373230990645,
+      "grad_norm": 1.5621879493455977,
+      "learning_rate": 7.677621203302591e-06,
+      "loss": 0.7979,
       "step": 450
     },
     {
-      "epoch": 0.21827776445190694,
-      "grad_norm": 2.779613231443427,
-      "learning_rate": 7.268370607028754e-06,
-      "loss": 0.8528,
+      "epoch": 0.6548332933557208,
+      "grad_norm": 1.5193447362876678,
+      "learning_rate": 7.66429796145387e-06,
+      "loss": 0.8125,
       "step": 455
     },
     {
-      "epoch": 0.2206764212041257,
-      "grad_norm": 2.4729051800667845,
-      "learning_rate": 7.348242811501598e-06,
-      "loss": 0.9453,
+      "epoch": 0.662029263612377,
+      "grad_norm": 1.6795960658239877,
+      "learning_rate": 7.650716997046216e-06,
+      "loss": 0.8477,
       "step": 460
     },
     {
-      "epoch": 0.22307507795634446,
-      "grad_norm": 3.4684655010762393,
-      "learning_rate": 7.428115015974442e-06,
-      "loss": 0.9096,
+      "epoch": 0.6692252338690333,
+      "grad_norm": 1.585934440158675,
+      "learning_rate": 7.636879265275119e-06,
+      "loss": 0.845,
       "step": 465
     },
     {
-      "epoch": 0.22547373470856322,
-      "grad_norm": 3.9774609808031403,
-      "learning_rate": 7.507987220447285e-06,
-      "loss": 0.866,
+      "epoch": 0.6764212041256896,
+      "grad_norm": 1.550239333699456,
+      "learning_rate": 7.622785739395397e-06,
+      "loss": 0.8723,
       "step": 470
     },
     {
-      "epoch": 0.22787239146078198,
-      "grad_norm": 4.462097218795264,
-      "learning_rate": 7.5878594249201285e-06,
-      "loss": 0.9069,
+      "epoch": 0.6836171743823459,
+      "grad_norm": 1.6342679772831266,
+      "learning_rate": 7.608437410652739e-06,
+      "loss": 0.8237,
       "step": 475
     },
     {
-      "epoch": 0.2302710482130007,
-      "grad_norm": 2.687922202397471,
-      "learning_rate": 7.667731629392972e-06,
-      "loss": 0.8583,
+      "epoch": 0.6908131446390021,
+      "grad_norm": 1.9009580720490618,
+      "learning_rate": 7.593835288213984e-06,
+      "loss": 0.8525,
       "step": 480
     },
     {
-      "epoch": 0.23266970496521946,
-      "grad_norm": 3.03353543529184,
-      "learning_rate": 7.747603833865815e-06,
-      "loss": 0.8999,
+      "epoch": 0.6980091148956584,
+      "grad_norm": 1.4797149380660173,
+      "learning_rate": 7.578980399096153e-06,
+      "loss": 0.8343,
       "step": 485
     },
     {
-      "epoch": 0.23506836171743822,
-      "grad_norm": 3.645290023306533,
-      "learning_rate": 7.827476038338658e-06,
-      "loss": 0.9374,
+      "epoch": 0.7052050851523147,
+      "grad_norm": 1.6927806221547512,
+      "learning_rate": 7.5638737880942e-06,
+      "loss": 0.819,
       "step": 490
     },
     {
-      "epoch": 0.23746701846965698,
-      "grad_norm": 3.5934722365995095,
-      "learning_rate": 7.907348242811502e-06,
-      "loss": 0.8538,
+      "epoch": 0.712401055408971,
+      "grad_norm": 1.6132959543020267,
+      "learning_rate": 7.548516517707544e-06,
+      "loss": 0.8177,
       "step": 495
     },
     {
-      "epoch": 0.23986567522187574,
-      "grad_norm": 4.244150178459668,
-      "learning_rate": 7.987220447284347e-06,
-      "loss": 0.8661,
-      "step": 500
-    },
-    {
-      "epoch": 0.23986567522187574,
-      "eval_loss": 0.8783833980560303,
-      "eval_runtime": 739.2193,
-      "eval_samples_per_second": 10.024,
-      "eval_steps_per_second": 0.628,
+      "epoch": 0.7195970256656272,
+      "grad_norm": 1.5783090956116979,
+      "learning_rate": 7.532909668065329e-06,
+      "loss": 0.8217,
       "step": 500
     },
     {
-      "epoch": 0.2422643319740945,
-      "grad_norm": 3.363197601795703,
-      "learning_rate": 8.06709265175719e-06,
-      "loss": 0.7723,
+      "epoch": 0.7267929959222835,
+      "grad_norm": 1.439090732947757,
+      "learning_rate": 7.517054336850457e-06,
+      "loss": 0.8617,
       "step": 505
     },
     {
-      "epoch": 0.24466298872631326,
-      "grad_norm": 3.687217931851492,
-      "learning_rate": 8.146964856230033e-06,
-      "loss": 0.7713,
+      "epoch": 0.7339889661789398,
+      "grad_norm": 1.906650403594057,
+      "learning_rate": 7.500951639222389e-06,
+      "loss": 0.8427,
       "step": 510
     },
     {
-      "epoch": 0.24706164547853202,
-      "grad_norm": 3.91943742276457,
-      "learning_rate": 8.226837060702875e-06,
-      "loss": 0.9931,
+      "epoch": 0.741184936435596,
+      "grad_norm": 1.5718009905007941,
+      "learning_rate": 7.484602707738707e-06,
+      "loss": 0.9079,
       "step": 515
     },
     {
-      "epoch": 0.24946030223075077,
-      "grad_norm": 3.242784908627176,
-      "learning_rate": 8.30670926517572e-06,
-      "loss": 0.9052,
+      "epoch": 0.7483809066922523,
+      "grad_norm": 1.716197310859666,
+      "learning_rate": 7.468008692275457e-06,
+      "loss": 0.8278,
       "step": 520
     },
     {
-      "epoch": 0.25185895898296956,
-      "grad_norm": 5.452872481341522,
-      "learning_rate": 8.386581469648563e-06,
-      "loss": 0.9227,
+      "epoch": 0.7555768769489086,
+      "grad_norm": 1.5485542395076963,
+      "learning_rate": 7.45117075994628e-06,
+      "loss": 0.8326,
       "step": 525
     },
     {
-      "epoch": 0.2542576157351883,
-      "grad_norm": 2.801600778934709,
-      "learning_rate": 8.466453674121406e-06,
-      "loss": 0.9145,
+      "epoch": 0.7627728472055649,
+      "grad_norm": 1.4519889463898326,
+      "learning_rate": 7.434090095020318e-06,
+      "loss": 0.7923,
       "step": 530
     },
     {
-      "epoch": 0.2566562724874071,
-      "grad_norm": 2.80399708145586,
-      "learning_rate": 8.54632587859425e-06,
-      "loss": 0.8459,
+      "epoch": 0.7699688174622211,
+      "grad_norm": 1.4499255632268135,
+      "learning_rate": 7.416767898838926e-06,
+      "loss": 0.8449,
       "step": 535
     },
     {
-      "epoch": 0.2590549292396258,
-      "grad_norm": 3.55867842461135,
-      "learning_rate": 8.626198083067093e-06,
-      "loss": 0.9457,
+      "epoch": 0.7771647877188774,
+      "grad_norm": 1.498217506498266,
+      "learning_rate": 7.399205389731172e-06,
+      "loss": 0.8462,
       "step": 540
     },
     {
-      "epoch": 0.26145358599184454,
-      "grad_norm": 3.150113704402208,
-      "learning_rate": 8.706070287539938e-06,
-      "loss": 0.835,
+      "epoch": 0.7843607579755337,
+      "grad_norm": 1.441416826072204,
+      "learning_rate": 7.381403802928153e-06,
+      "loss": 0.7864,
       "step": 545
     },
     {
-      "epoch": 0.2638522427440633,
-      "grad_norm": 2.91320676896268,
-      "learning_rate": 8.78594249201278e-06,
-      "loss": 0.8667,
+      "epoch": 0.7915567282321899,
+      "grad_norm": 1.5728216952970062,
+      "learning_rate": 7.363364390476114e-06,
+      "loss": 0.779,
       "step": 550
     },
     {
-      "epoch": 0.26625089949628206,
-      "grad_norm": 3.2187525835358626,
-      "learning_rate": 8.865814696485623e-06,
-      "loss": 0.8478,
+      "epoch": 0.7987526984888462,
+      "grad_norm": 1.579757123001059,
+      "learning_rate": 7.34508842114839e-06,
+      "loss": 0.8342,
       "step": 555
     },
     {
-      "epoch": 0.26864955624850084,
-      "grad_norm": 3.3992782102360173,
-      "learning_rate": 8.945686900958466e-06,
-      "loss": 0.8523,
+      "epoch": 0.8059486687455025,
+      "grad_norm": 1.5204038228338386,
+      "learning_rate": 7.326577180356162e-06,
+      "loss": 0.8202,
       "step": 560
     },
     {
-      "epoch": 0.2710482130007196,
-      "grad_norm": 3.107174104616647,
-      "learning_rate": 9.02555910543131e-06,
-      "loss": 0.9199,
+      "epoch": 0.8131446390021588,
+      "grad_norm": 1.6070382704422732,
+      "learning_rate": 7.30783197005806e-06,
+      "loss": 0.7948,
       "step": 565
     },
     {
-      "epoch": 0.27344686975293836,
-      "grad_norm": 3.600873597055109,
-      "learning_rate": 9.105431309904154e-06,
-      "loss": 0.7872,
+      "epoch": 0.820340609258815,
+      "grad_norm": 1.4952361279508235,
+      "learning_rate": 7.288854108668586e-06,
+      "loss": 0.8451,
       "step": 570
     },
     {
-      "epoch": 0.2758455265051571,
-      "grad_norm": 2.9647080063438915,
-      "learning_rate": 9.185303514376998e-06,
-      "loss": 0.8401,
+      "epoch": 0.8275365795154713,
+      "grad_norm": 1.4373323975649135,
+      "learning_rate": 7.2696449309653795e-06,
+      "loss": 0.8381,
       "step": 575
     },
     {
-      "epoch": 0.2782441832573759,
-      "grad_norm": 2.7990639275693945,
-      "learning_rate": 9.265175718849841e-06,
-      "loss": 0.8267,
+      "epoch": 0.8347325497721276,
+      "grad_norm": 1.292833703910724,
+      "learning_rate": 7.250205787995353e-06,
+      "loss": 0.8286,
       "step": 580
     },
     {
-      "epoch": 0.2806428400095946,
-      "grad_norm": 3.3315727902538157,
-      "learning_rate": 9.345047923322684e-06,
-      "loss": 0.8867,
+      "epoch": 0.8419285200287839,
+      "grad_norm": 1.2885838799591653,
+      "learning_rate": 7.230538046979654e-06,
+      "loss": 0.8506,
       "step": 585
     },
     {
-      "epoch": 0.2830414967618134,
-      "grad_norm": 6.339874565170752,
-      "learning_rate": 9.424920127795528e-06,
-      "loss": 0.9029,
+      "epoch": 0.8491244902854401,
+      "grad_norm": 1.4134558176619048,
+      "learning_rate": 7.210643091217513e-06,
+      "loss": 0.8411,
       "step": 590
     },
     {
-      "epoch": 0.2854401535140321,
-      "grad_norm": 4.308904102490444,
-      "learning_rate": 9.504792332268371e-06,
-      "loss": 0.9859,
+      "epoch": 0.8563204605420964,
+      "grad_norm": 1.6657530487665673,
+      "learning_rate": 7.1905223199889425e-06,
+      "loss": 0.834,
       "step": 595
     },
     {
-      "epoch": 0.2878388102662509,
-      "grad_norm": 3.452224572712429,
-      "learning_rate": 9.584664536741216e-06,
-      "loss": 0.9071,
-      "step": 600
-    },
-    {
-      "epoch": 0.2878388102662509,
-      "eval_loss": 0.899307906627655,
-      "eval_runtime": 739.1395,
-      "eval_samples_per_second": 10.025,
-      "eval_steps_per_second": 0.628,
+      "epoch": 0.8635164307987527,
+      "grad_norm": 1.7801521918282026,
+      "learning_rate": 7.170177148456331e-06,
+      "loss": 0.8461,
       "step": 600
     },
     {
-      "epoch": 0.29023746701846964,
-      "grad_norm": 3.178382258183808,
-      "learning_rate": 9.664536741214059e-06,
-      "loss": 0.8549,
+      "epoch": 0.8707124010554089,
+      "grad_norm": 1.5712149389587902,
+      "learning_rate": 7.149609007564903e-06,
+      "loss": 0.8683,
       "step": 605
     },
     {
-      "epoch": 0.2926361237706884,
-      "grad_norm": 3.8066863018491994,
-      "learning_rate": 9.744408945686901e-06,
-      "loss": 0.8329,
+      "epoch": 0.8779083713120652,
+      "grad_norm": 1.5017107860836532,
+      "learning_rate": 7.128819343942077e-06,
+      "loss": 0.8442,
       "step": 610
     },
     {
-      "epoch": 0.29503478052290716,
-      "grad_norm": 4.217226984051162,
-      "learning_rate": 9.824281150159746e-06,
-      "loss": 0.884,
+      "epoch": 0.8851043415687215,
+      "grad_norm": 1.4727249376876361,
+      "learning_rate": 7.107809619795722e-06,
+      "loss": 0.8668,
       "step": 615
     },
     {
-      "epoch": 0.29743343727512594,
-      "grad_norm": 4.082578713375918,
-      "learning_rate": 9.904153354632589e-06,
-      "loss": 0.8972,
+      "epoch": 0.8923003118253778,
+      "grad_norm": 1.428036074094576,
+      "learning_rate": 7.086581312811309e-06,
+      "loss": 0.773,
       "step": 620
     },
     {
-      "epoch": 0.2998320940273447,
-      "grad_norm": 3.601173386435701,
-      "learning_rate": 9.984025559105433e-06,
-      "loss": 0.9211,
+      "epoch": 0.899496282082034,
+      "grad_norm": 1.459125555270072,
+      "learning_rate": 7.065135916047992e-06,
+      "loss": 0.8551,
       "step": 625
     },
     {
-      "epoch": 0.30223075077956346,
-      "grad_norm": 4.327224463154232,
-      "learning_rate": 9.999987527310802e-06,
-      "loss": 0.9326,
+      "epoch": 0.9066922523386903,
+      "grad_norm": 1.4437358399730527,
+      "learning_rate": 7.043474937833581e-06,
+      "loss": 0.8055,
       "step": 630
     },
     {
-      "epoch": 0.3046294075317822,
-      "grad_norm": 3.192949701933106,
-      "learning_rate": 9.99993685711758e-06,
-      "loss": 0.8407,
+      "epoch": 0.9138882225953466,
+      "grad_norm": 1.6487138533805716,
+      "learning_rate": 7.021599901658467e-06,
+      "loss": 0.8162,
       "step": 635
     },
     {
-      "epoch": 0.307028064284001,
-      "grad_norm": 3.7659658769393665,
-      "learning_rate": 9.999847210271954e-06,
-      "loss": 0.834,
+      "epoch": 0.9210841928520028,
+      "grad_norm": 1.5886402538030315,
+      "learning_rate": 6.999512346068467e-06,
+      "loss": 0.8472,
       "step": 640
     },
     {
-      "epoch": 0.3094267210362197,
-      "grad_norm": 2.4979512705164657,
-      "learning_rate": 9.999718587472758e-06,
-      "loss": 0.848,
+      "epoch": 0.9282801631086591,
+      "grad_norm": 1.580701659838036,
+      "learning_rate": 6.977213824556613e-06,
+      "loss": 0.8185,
       "step": 645
     },
     {
-      "epoch": 0.3118253777884385,
-      "grad_norm": 5.339110116490045,
-      "learning_rate": 9.999550989722662e-06,
-      "loss": 0.9671,
+      "epoch": 0.9354761333653154,
+      "grad_norm": 1.3791827207697653,
+      "learning_rate": 6.95470590545389e-06,
+      "loss": 0.8424,
       "step": 650
     },
     {
-      "epoch": 0.3142240345406572,
-      "grad_norm": 2.866052234022034,
-      "learning_rate": 9.999344418328161e-06,
-      "loss": 0.8325,
+      "epoch": 0.9426721036219717,
+      "grad_norm": 1.5248977943916064,
+      "learning_rate": 6.931990171818923e-06,
+      "loss": 0.8829,
       "step": 655
     },
     {
-      "epoch": 0.316622691292876,
-      "grad_norm": 3.050315691952451,
-      "learning_rate": 9.99909887489957e-06,
-      "loss": 0.8506,
+      "epoch": 0.9498680738786279,
+      "grad_norm": 1.4962085290443874,
+      "learning_rate": 6.909068221326647e-06,
+      "loss": 0.8236,
       "step": 660
     },
     {
-      "epoch": 0.31902134804509474,
-      "grad_norm": 2.906833616688605,
-      "learning_rate": 9.998814361351006e-06,
-      "loss": 0.8681,
+      "epoch": 0.9570640441352842,
+      "grad_norm": 1.6074178928012908,
+      "learning_rate": 6.88594166615593e-06,
+      "loss": 0.8165,
       "step": 665
     },
     {
-      "epoch": 0.32142000479731353,
-      "grad_norm": 3.108240781409998,
-      "learning_rate": 9.99849087990037e-06,
-      "loss": 0.9081,
+      "epoch": 0.9642600143919405,
+      "grad_norm": 1.538150422370725,
+      "learning_rate": 6.8626121328761824e-06,
+      "loss": 0.8155,
       "step": 670
     },
     {
-      "epoch": 0.32381866154953226,
-      "grad_norm": 2.772972214094703,
-      "learning_rate": 9.998128433069345e-06,
-      "loss": 0.9144,
+      "epoch": 0.9714559846485968,
+      "grad_norm": 1.4261928133080215,
+      "learning_rate": 6.839081262332957e-06,
+      "loss": 0.8271,
       "step": 675
     },
     {
-      "epoch": 0.32621731830175105,
-      "grad_norm": 4.96500038402326,
-      "learning_rate": 9.997727023683353e-06,
-      "loss": 0.987,
+      "epoch": 0.978651954905253,
+      "grad_norm": 1.4811633224353407,
+      "learning_rate": 6.815350709532544e-06,
+      "loss": 0.8417,
       "step": 680
     },
     {
-      "epoch": 0.3286159750539698,
-      "grad_norm": 2.651958496925647,
-      "learning_rate": 9.997286654871556e-06,
-      "loss": 0.8424,
+      "epoch": 0.9858479251619093,
+      "grad_norm": 1.5008044573312285,
+      "learning_rate": 6.791422143525564e-06,
+      "loss": 0.859,
       "step": 685
     },
     {
-      "epoch": 0.3310146318061885,
-      "grad_norm": 3.4588489379856484,
-      "learning_rate": 9.996807330066816e-06,
-      "loss": 0.9033,
+      "epoch": 0.9930438954185656,
+      "grad_norm": 1.5977699325206687,
+      "learning_rate": 6.767297247289585e-06,
+      "loss": 0.8663,
       "step": 690
     },
     {
-      "epoch": 0.3334132885584073,
-      "grad_norm": 2.79385907219515,
-      "learning_rate": 9.996289053005676e-06,
-      "loss": 0.8301,
+      "epoch": 1.0002398656752218,
+      "grad_norm": 2.198059715592546,
+      "learning_rate": 6.742977717610744e-06,
+      "loss": 0.8427,
       "step": 695
     },
     {
-      "epoch": 0.335811945310626,
-      "grad_norm": 3.7163749268253357,
-      "learning_rate": 9.99573182772833e-06,
-      "loss": 0.8955,
-      "step": 700
-    },
-    {
-      "epoch": 0.335811945310626,
-      "eval_loss": 0.8928011655807495,
-      "eval_runtime": 741.2881,
-      "eval_samples_per_second": 9.996,
-      "eval_steps_per_second": 0.626,
+      "epoch": 1.0074358359318782,
+      "grad_norm": 1.447486858474852,
+      "learning_rate": 6.718465264964414e-06,
+      "loss": 0.5445,
       "step": 700
     },
     {
-      "epoch": 0.3382106020628448,
-      "grad_norm": 2.5849874442305714,
-      "learning_rate": 9.995135658578587e-06,
-      "loss": 0.9328,
+      "epoch": 1.0146318061885344,
+      "grad_norm": 1.5387360085110118,
+      "learning_rate": 6.693761613394899e-06,
+      "loss": 0.5585,
       "step": 705
     },
     {
-      "epoch": 0.34060925881506354,
-      "grad_norm": 3.0610660646289705,
-      "learning_rate": 9.994500550203838e-06,
-      "loss": 0.9111,
+      "epoch": 1.0218277764451906,
+      "grad_norm": 1.3617154485025116,
+      "learning_rate": 6.668868500394172e-06,
+      "loss": 0.4605,
       "step": 710
     },
     {
-      "epoch": 0.34300791556728233,
-      "grad_norm": 3.310088400644419,
-      "learning_rate": 9.99382650755503e-06,
-      "loss": 0.8882,
+      "epoch": 1.029023746701847,
+      "grad_norm": 1.433780770938119,
+      "learning_rate": 6.643787676779671e-06,
+      "loss": 0.5254,
       "step": 715
     },
     {
-      "epoch": 0.34540657231950106,
-      "grad_norm": 2.989166571674383,
-      "learning_rate": 9.993113535886612e-06,
-      "loss": 0.9057,
+      "epoch": 1.0362197169585032,
+      "grad_norm": 1.4766138665890287,
+      "learning_rate": 6.618520906571171e-06,
+      "loss": 0.476,
       "step": 720
     },
     {
-      "epoch": 0.34780522907171985,
-      "grad_norm": 3.2839488015561638,
-      "learning_rate": 9.992361640756504e-06,
-      "loss": 0.9816,
+      "epoch": 1.0434156872151594,
+      "grad_norm": 1.4473624764279305,
+      "learning_rate": 6.593069966866694e-06,
+      "loss": 0.5404,
       "step": 725
     },
     {
-      "epoch": 0.3502038858239386,
-      "grad_norm": 2.952806180266413,
-      "learning_rate": 9.991570828026052e-06,
-      "loss": 0.9105,
+      "epoch": 1.0506116574718158,
+      "grad_norm": 1.4713255574009756,
+      "learning_rate": 6.567436647717535e-06,
+      "loss": 0.5293,
       "step": 730
     },
     {
-      "epoch": 0.35260254257615736,
-      "grad_norm": 3.0776125312419644,
-      "learning_rate": 9.99074110385998e-06,
-      "loss": 0.8911,
+      "epoch": 1.057807627728472,
+      "grad_norm": 1.3099551452546936,
+      "learning_rate": 6.541622752002355e-06,
+      "loss": 0.5168,
       "step": 735
     },
     {
-      "epoch": 0.3550011993283761,
-      "grad_norm": 2.8653507727619356,
-      "learning_rate": 9.989872474726341e-06,
-      "loss": 0.9403,
+      "epoch": 1.0650035979851282,
+      "grad_norm": 1.4756303619125208,
+      "learning_rate": 6.515630095300383e-06,
+      "loss": 0.5253,
       "step": 740
     },
     {
-      "epoch": 0.3573998560805949,
-      "grad_norm": 3.1505016143613047,
-      "learning_rate": 9.988964947396477e-06,
-      "loss": 0.8747,
+      "epoch": 1.0721995682417846,
+      "grad_norm": 1.2607903619444512,
+      "learning_rate": 6.489460505763713e-06,
+      "loss": 0.5203,
       "step": 745
     },
     {
-      "epoch": 0.3597985128328136,
-      "grad_norm": 2.6125502058810883,
-      "learning_rate": 9.98801852894495e-06,
-      "loss": 0.8193,
+      "epoch": 1.0793955384984408,
+      "grad_norm": 1.2668860008648557,
+      "learning_rate": 6.463115823988732e-06,
+      "loss": 0.5133,
       "step": 750
     },
     {
-      "epoch": 0.3621971695850324,
-      "grad_norm": 2.7512586189265003,
-      "learning_rate": 9.987033226749497e-06,
-      "loss": 0.851,
+      "epoch": 1.0865915087550972,
+      "grad_norm": 1.5445317289615172,
+      "learning_rate": 6.436597902886655e-06,
+      "loss": 0.5399,
       "step": 755
     },
     {
-      "epoch": 0.3645958263372511,
-      "grad_norm": 2.56627099994587,
-      "learning_rate": 9.986009048490971e-06,
-      "loss": 0.8504,
+      "epoch": 1.0937874790117534,
+      "grad_norm": 1.373346882134398,
+      "learning_rate": 6.409908607553217e-06,
+      "loss": 0.4742,
       "step": 760
     },
     {
-      "epoch": 0.3669944830894699,
-      "grad_norm": 3.2634625404493387,
-      "learning_rate": 9.984946002153284e-06,
-      "loss": 0.9531,
+      "epoch": 1.1009834492684096,
+      "grad_norm": 1.2423912396823111,
+      "learning_rate": 6.38304981513748e-06,
+      "loss": 0.4928,
       "step": 765
     },
     {
-      "epoch": 0.36939313984168864,
-      "grad_norm": 2.897125687255057,
-      "learning_rate": 9.983844096023334e-06,
-      "loss": 0.8097,
+      "epoch": 1.108179419525066,
+      "grad_norm": 1.564288911463839,
+      "learning_rate": 6.3560234147098155e-06,
+      "loss": 0.509,
       "step": 770
     },
     {
-      "epoch": 0.37179179659390743,
-      "grad_norm": 2.7944503893231474,
-      "learning_rate": 9.982703338690955e-06,
-      "loss": 0.9376,
+      "epoch": 1.1153753897817222,
+      "grad_norm": 1.592002787628597,
+      "learning_rate": 6.328831307129039e-06,
+      "loss": 0.5373,
       "step": 775
     },
     {
-      "epoch": 0.37419045334612616,
-      "grad_norm": 3.0897623363555455,
-      "learning_rate": 9.981523739048839e-06,
-      "loss": 0.8607,
+      "epoch": 1.1225713600383784,
+      "grad_norm": 1.2953958374191832,
+      "learning_rate": 6.30147540490871e-06,
+      "loss": 0.5053,
       "step": 780
     },
     {
-      "epoch": 0.37658911009834495,
-      "grad_norm": 2.825952708220128,
-      "learning_rate": 9.98030530629247e-06,
-      "loss": 0.9208,
+      "epoch": 1.1297673302950348,
+      "grad_norm": 1.3851707132475586,
+      "learning_rate": 6.27395763208263e-06,
+      "loss": 0.5138,
       "step": 785
     },
     {
-      "epoch": 0.3789877668505637,
-      "grad_norm": 3.21377979039193,
-      "learning_rate": 9.979048049920056e-06,
-      "loss": 0.9482,
+      "epoch": 1.136963300551691,
+      "grad_norm": 1.5083938363704992,
+      "learning_rate": 6.246279924069504e-06,
+      "loss": 0.4639,
       "step": 790
     },
     {
-      "epoch": 0.38138642360278247,
-      "grad_norm": 3.4941808752016015,
-      "learning_rate": 9.977751979732448e-06,
-      "loss": 0.8321,
+      "epoch": 1.1441592708083472,
+      "grad_norm": 1.4583984595795128,
+      "learning_rate": 6.218444227536832e-06,
+      "loss": 0.509,
       "step": 795
     },
     {
-      "epoch": 0.3837850803550012,
-      "grad_norm": 2.819630176729509,
-      "learning_rate": 9.97641710583307e-06,
-      "loss": 0.89,
-      "step": 800
-    },
-    {
-      "epoch": 0.3837850803550012,
-      "eval_loss": 0.8970605731010437,
-      "eval_runtime": 739.8413,
-      "eval_samples_per_second": 10.016,
-      "eval_steps_per_second": 0.627,
+      "epoch": 1.1513552410650036,
+      "grad_norm": 1.1915713777559744,
+      "learning_rate": 6.190452500263975e-06,
+      "loss": 0.4771,
       "step": 800
     },
     {
-      "epoch": 0.38618373710722,
-      "grad_norm": 2.6833336278666264,
-      "learning_rate": 9.975043438627836e-06,
-      "loss": 0.8416,
+      "epoch": 1.1585512113216598,
+      "grad_norm": 1.2538864728177044,
+      "learning_rate": 6.162306711004474e-06,
+      "loss": 0.4927,
       "step": 805
     },
     {
-      "epoch": 0.3885823938594387,
-      "grad_norm": 3.0978237393301735,
-      "learning_rate": 9.973630988825071e-06,
-      "loss": 0.9354,
+      "epoch": 1.165747181578316,
+      "grad_norm": 1.3142042016311857,
+      "learning_rate": 6.134008839347575e-06,
+      "loss": 0.4884,
       "step": 810
     },
     {
-      "epoch": 0.3909810506116575,
-      "grad_norm": 4.2752235499170705,
-      "learning_rate": 9.972179767435428e-06,
-      "loss": 0.9026,
+      "epoch": 1.1729431518349724,
+      "grad_norm": 1.232777503769632,
+      "learning_rate": 6.105560875578994e-06,
+      "loss": 0.5273,
       "step": 815
     },
     {
-      "epoch": 0.39337970736387623,
-      "grad_norm": 2.430423958859156,
-      "learning_rate": 9.970689785771798e-06,
-      "loss": 0.8654,
+      "epoch": 1.1801391220916286,
+      "grad_norm": 1.502479848796588,
+      "learning_rate": 6.076964820540937e-06,
+      "loss": 0.5086,
       "step": 820
     },
     {
-      "epoch": 0.39577836411609496,
-      "grad_norm": 2.9029155463869527,
-      "learning_rate": 9.96916105544923e-06,
-      "loss": 0.8403,
+      "epoch": 1.187335092348285,
+      "grad_norm": 1.5121283948236117,
+      "learning_rate": 6.048222685491374e-06,
+      "loss": 0.5374,
       "step": 825
     },
     {
-      "epoch": 0.39817702086831375,
-      "grad_norm": 2.7152107809882065,
-      "learning_rate": 9.967593588384832e-06,
-      "loss": 0.7184,
+      "epoch": 1.1945310626049412,
+      "grad_norm": 1.8544252686881426,
+      "learning_rate": 6.019336491962581e-06,
+      "loss": 0.5381,
       "step": 830
     },
     {
-      "epoch": 0.4005756776205325,
-      "grad_norm": 2.725981699021402,
-      "learning_rate": 9.965987396797678e-06,
-      "loss": 0.9184,
+      "epoch": 1.2017270328615974,
+      "grad_norm": 1.3559009500960952,
+      "learning_rate": 5.990308271618956e-06,
+      "loss": 0.4939,
       "step": 835
     },
     {
-      "epoch": 0.40297433437275126,
-      "grad_norm": 3.1490594445803626,
-      "learning_rate": 9.964342493208727e-06,
-      "loss": 0.9349,
+      "epoch": 1.2089230031182538,
+      "grad_norm": 1.6991974082062777,
+      "learning_rate": 5.961140066114128e-06,
+      "loss": 0.5429,
       "step": 840
     },
     {
-      "epoch": 0.40537299112497,
-      "grad_norm": 2.4002360503838847,
-      "learning_rate": 9.962658890440703e-06,
-      "loss": 0.8469,
+      "epoch": 1.21611897337491,
+      "grad_norm": 1.40613574888153,
+      "learning_rate": 5.931833926947358e-06,
+      "loss": 0.4778,
       "step": 845
     },
     {
-      "epoch": 0.4077716478771888,
-      "grad_norm": 3.3266215600879994,
-      "learning_rate": 9.960936601618014e-06,
-      "loss": 1.0367,
+      "epoch": 1.2233149436315662,
+      "grad_norm": 1.291827429944612,
+      "learning_rate": 5.902391915319252e-06,
+      "loss": 0.4604,
       "step": 850
     },
     {
-      "epoch": 0.4101703046294075,
-      "grad_norm": 2.965406316734529,
-      "learning_rate": 9.959175640166639e-06,
-      "loss": 0.8879,
+      "epoch": 1.2305109138882226,
+      "grad_norm": 1.5403421830718962,
+      "learning_rate": 5.872816101986789e-06,
+      "loss": 0.4993,
       "step": 855
     },
     {
-      "epoch": 0.4125689613816263,
-      "grad_norm": 2.5176873752286757,
-      "learning_rate": 9.957376019814028e-06,
-      "loss": 0.7977,
+      "epoch": 1.2377068841448788,
+      "grad_norm": 1.295984623620993,
+      "learning_rate": 5.843108567117678e-06,
+      "loss": 0.4972,
       "step": 860
     },
     {
-      "epoch": 0.41496761813384503,
-      "grad_norm": 3.221299896812408,
-      "learning_rate": 9.955537754588996e-06,
-      "loss": 0.8624,
+      "epoch": 1.244902854401535,
+      "grad_norm": 1.372473134492806,
+      "learning_rate": 5.813271400144051e-06,
+      "loss": 0.5199,
       "step": 865
     },
     {
-      "epoch": 0.4173662748860638,
-      "grad_norm": 3.4035598410054995,
-      "learning_rate": 9.953660858821604e-06,
-      "loss": 0.842,
+      "epoch": 1.2520988246581914,
+      "grad_norm": 1.3747684408273264,
+      "learning_rate": 5.783306699615512e-06,
+      "loss": 0.5136,
       "step": 870
     },
     {
-      "epoch": 0.41976493163828255,
-      "grad_norm": 2.759621830247372,
-      "learning_rate": 9.951745347143065e-06,
-      "loss": 0.831,
+      "epoch": 1.2592947949148476,
+      "grad_norm": 1.3792887268238399,
+      "learning_rate": 5.753216573051526e-06,
+      "loss": 0.5045,
       "step": 875
     },
     {
-      "epoch": 0.42216358839050133,
-      "grad_norm": 3.071767749542282,
-      "learning_rate": 9.94979123448561e-06,
-      "loss": 0.8394,
+      "epoch": 1.266490765171504,
+      "grad_norm": 1.6127625261536036,
+      "learning_rate": 5.723003136793208e-06,
+      "loss": 0.5003,
       "step": 880
     },
     {
-      "epoch": 0.42456224514272006,
-      "grad_norm": 2.728604857418505,
-      "learning_rate": 9.947798536082386e-06,
-      "loss": 0.9297,
+      "epoch": 1.2736867354281602,
+      "grad_norm": 1.4440630275399904,
+      "learning_rate": 5.692668515854457e-06,
+      "loss": 0.4521,
       "step": 885
     },
     {
-      "epoch": 0.42696090189493885,
-      "grad_norm": 3.2357927761859564,
-      "learning_rate": 9.945767267467335e-06,
-      "loss": 0.9248,
+      "epoch": 1.2808827056848164,
+      "grad_norm": 1.5683931030948375,
+      "learning_rate": 5.662214843772506e-06,
+      "loss": 0.5435,
       "step": 890
     },
     {
-      "epoch": 0.4293595586471576,
-      "grad_norm": 3.0566660546759326,
-      "learning_rate": 9.943697444475062e-06,
-      "loss": 0.9137,
+      "epoch": 1.2880786759414729,
+      "grad_norm": 1.4069551760135997,
+      "learning_rate": 5.631644262457861e-06,
+      "loss": 0.5326,
       "step": 895
     },
     {
-      "epoch": 0.43175821539937637,
-      "grad_norm": 3.281447736907799,
-      "learning_rate": 9.94158908324073e-06,
-      "loss": 0.8446,
-      "step": 900
-    },
-    {
-      "epoch": 0.43175821539937637,
-      "eval_loss": 0.891981303691864,
-      "eval_runtime": 739.0277,
-      "eval_samples_per_second": 10.027,
-      "eval_steps_per_second": 0.628,
+      "epoch": 1.295274646198129,
+      "grad_norm": 1.3933443999205188,
+      "learning_rate": 5.600958922043651e-06,
+      "loss": 0.4905,
       "step": 900
     },
     {
-      "epoch": 0.4341568721515951,
-      "grad_norm": 2.669716984027581,
-      "learning_rate": 9.939442200199917e-06,
-      "loss": 0.8603,
+      "epoch": 1.3024706164547855,
+      "grad_norm": 2.0908428594407344,
+      "learning_rate": 5.570160980734405e-06,
+      "loss": 0.4444,
       "step": 905
     },
     {
-      "epoch": 0.4365555289038139,
-      "grad_norm": 2.6752353630977854,
-      "learning_rate": 9.937256812088496e-06,
-      "loss": 0.8846,
+      "epoch": 1.3096665867114416,
+      "grad_norm": 1.6732069684734259,
+      "learning_rate": 5.539252604654256e-06,
+      "loss": 0.5535,
       "step": 910
     },
     {
-      "epoch": 0.4389541856560326,
-      "grad_norm": 4.071925103373792,
-      "learning_rate": 9.935032935942506e-06,
-      "loss": 0.9126,
+      "epoch": 1.3168625569680978,
+      "grad_norm": 1.4666496678005971,
+      "learning_rate": 5.50823596769459e-06,
+      "loss": 0.4977,
       "step": 915
     },
     {
-      "epoch": 0.4413528424082514,
-      "grad_norm": 2.9599856546225274,
-      "learning_rate": 9.932770589098015e-06,
-      "loss": 0.9073,
+      "epoch": 1.324058527224754,
+      "grad_norm": 1.3633392168469545,
+      "learning_rate": 5.477113251361149e-06,
+      "loss": 0.5118,
       "step": 920
     },
     {
-      "epoch": 0.44375149916047013,
-      "grad_norm": 2.4671508950529186,
-      "learning_rate": 9.930469789190988e-06,
-      "loss": 0.8806,
+      "epoch": 1.3312544974814104,
+      "grad_norm": 1.2762316788464472,
+      "learning_rate": 5.445886644620601e-06,
+      "loss": 0.5136,
       "step": 925
     },
     {
-      "epoch": 0.4461501559126889,
-      "grad_norm": 6.163188231431187,
-      "learning_rate": 9.928130554157144e-06,
-      "loss": 0.8668,
+      "epoch": 1.3384504677380666,
+      "grad_norm": 1.5424740001396617,
+      "learning_rate": 5.414558343746579e-06,
+      "loss": 0.4926,
       "step": 930
     },
     {
-      "epoch": 0.44854881266490765,
-      "grad_norm": 4.3889573037879535,
-      "learning_rate": 9.925752902231829e-06,
-      "loss": 0.9947,
+      "epoch": 1.345646437994723,
+      "grad_norm": 1.3847453947317292,
+      "learning_rate": 5.38313055216521e-06,
+      "loss": 0.5458,
       "step": 935
     },
     {
-      "epoch": 0.45094746941712643,
-      "grad_norm": 3.099558653114257,
-      "learning_rate": 9.923336851949855e-06,
-      "loss": 0.8214,
+      "epoch": 1.3528424082513792,
+      "grad_norm": 1.3877715039925074,
+      "learning_rate": 5.351605480300143e-06,
+      "loss": 0.4637,
       "step": 940
     },
     {
-      "epoch": 0.45334612616934516,
-      "grad_norm": 3.960300474080028,
-      "learning_rate": 9.920882422145372e-06,
-      "loss": 0.9271,
+      "epoch": 1.3600383785080354,
+      "grad_norm": 1.3969957173831602,
+      "learning_rate": 5.319985345417079e-06,
+      "loss": 0.4787,
       "step": 945
     },
     {
-      "epoch": 0.45574478292156395,
-      "grad_norm": 4.305333975019212,
-      "learning_rate": 9.918389631951712e-06,
-      "loss": 0.8557,
+      "epoch": 1.3672343487646919,
+      "grad_norm": 1.669129051283974,
+      "learning_rate": 5.288272371467827e-06,
+      "loss": 0.484,
       "step": 950
     },
     {
-      "epoch": 0.4581434396737827,
-      "grad_norm": 2.703221339883744,
-      "learning_rate": 9.915858500801245e-06,
-      "loss": 0.8476,
+      "epoch": 1.374430319021348,
+      "grad_norm": 1.3646255552767974,
+      "learning_rate": 5.256468788933881e-06,
+      "loss": 0.4782,
       "step": 955
     },
     {
-      "epoch": 0.4605420964260014,
-      "grad_norm": 2.0714764562806476,
-      "learning_rate": 9.913289048425225e-06,
-      "loss": 0.8258,
+      "epoch": 1.3816262892780042,
+      "grad_norm": 1.384029909257828,
+      "learning_rate": 5.2245768346695494e-06,
+      "loss": 0.5021,
       "step": 960
     },
     {
-      "epoch": 0.4629407531782202,
-      "grad_norm": 3.079398404907295,
-      "learning_rate": 9.910681294853632e-06,
-      "loss": 0.9569,
+      "epoch": 1.3888222595346607,
+      "grad_norm": 1.3642725031029292,
+      "learning_rate": 5.192598751744621e-06,
+      "loss": 0.476,
       "step": 965
     },
     {
-      "epoch": 0.46533940993043893,
-      "grad_norm": 3.1990029825784094,
-      "learning_rate": 9.908035260415028e-06,
-      "loss": 0.8552,
+      "epoch": 1.3960182297913168,
+      "grad_norm": 1.4324224845783868,
+      "learning_rate": 5.160536789286612e-06,
+      "loss": 0.4966,
       "step": 970
     },
     {
-      "epoch": 0.4677380666826577,
-      "grad_norm": 2.8699251494441738,
-      "learning_rate": 9.905350965736382e-06,
-      "loss": 0.8553,
+      "epoch": 1.403214200047973,
+      "grad_norm": 1.3426650482840705,
+      "learning_rate": 5.128393202322565e-06,
+      "loss": 0.5116,
       "step": 975
     },
     {
-      "epoch": 0.47013672343487645,
-      "grad_norm": 4.49023764637979,
-      "learning_rate": 9.902628431742924e-06,
-      "loss": 0.8762,
+      "epoch": 1.4104101703046295,
+      "grad_norm": 1.4860297463165402,
+      "learning_rate": 5.096170251620458e-06,
+      "loss": 0.512,
       "step": 980
     },
     {
-      "epoch": 0.47253538018709523,
-      "grad_norm": 3.0297852374638956,
-      "learning_rate": 9.899867679657974e-06,
-      "loss": 0.855,
+      "epoch": 1.4176061405612856,
+      "grad_norm": 1.3809734192523142,
+      "learning_rate": 5.063870203530188e-06,
+      "loss": 0.5128,
       "step": 985
     },
     {
-      "epoch": 0.47493403693931396,
-      "grad_norm": 3.4604086734874007,
-      "learning_rate": 9.89706873100278e-06,
-      "loss": 0.8564,
+      "epoch": 1.424802110817942,
+      "grad_norm": 1.61363323216821,
+      "learning_rate": 5.031495329824175e-06,
+      "loss": 0.5342,
       "step": 990
     },
     {
-      "epoch": 0.47733269369153275,
-      "grad_norm": 2.3167328747243365,
-      "learning_rate": 9.894231607596348e-06,
-      "loss": 0.8405,
+      "epoch": 1.4319980810745983,
+      "grad_norm": 2.088959286090713,
+      "learning_rate": 4.999047907537582e-06,
+      "loss": 0.489,
       "step": 995
     },
     {
-      "epoch": 0.4797313504437515,
-      "grad_norm": 2.334415977229892,
-      "learning_rate": 9.891356331555272e-06,
-      "loss": 0.8908,
-      "step": 1000
-    },
-    {
-      "epoch": 0.4797313504437515,
-      "eval_loss": 0.898030698299408,
-      "eval_runtime": 739.1272,
-      "eval_samples_per_second": 10.025,
-      "eval_steps_per_second": 0.628,
+      "epoch": 1.4391940513312544,
+      "grad_norm": 1.5116312432174273,
+      "learning_rate": 4.966530218808157e-06,
+      "loss": 0.4968,
       "step": 1000
     },
     {
-      "epoch": 0.48213000719597027,
-      "grad_norm": 3.3169042936391526,
-      "learning_rate": 9.888442925293555e-06,
-      "loss": 0.9384,
+      "epoch": 1.4463900215879106,
+      "grad_norm": 1.4453482163933629,
+      "learning_rate": 4.933944550715725e-06,
+      "loss": 0.5297,
       "step": 1005
     },
     {
-      "epoch": 0.484528663948189,
-      "grad_norm": 2.680090693916463,
-      "learning_rate": 9.885491411522457e-06,
-      "loss": 0.9831,
+      "epoch": 1.453585991844567,
+      "grad_norm": 1.5874368851505785,
+      "learning_rate": 4.901293195121338e-06,
+      "loss": 0.5005,
       "step": 1010
     },
     {
-      "epoch": 0.4869273207004078,
-      "grad_norm": 2.7070444583755195,
-      "learning_rate": 9.882501813250292e-06,
-      "loss": 0.8285,
+      "epoch": 1.4607819621012232,
+      "grad_norm": 1.5140612281307557,
+      "learning_rate": 4.868578448506067e-06,
+      "loss": 0.5425,
       "step": 1015
     },
     {
-      "epoch": 0.4893259774526265,
-      "grad_norm": 3.8337165580593204,
-      "learning_rate": 9.879474153782258e-06,
-      "loss": 0.9792,
+      "epoch": 1.4679779323578797,
+      "grad_norm": 1.5247613079347937,
+      "learning_rate": 4.835802611809492e-06,
+      "loss": 0.5246,
       "step": 1020
     },
     {
-      "epoch": 0.4917246342048453,
-      "grad_norm": 2.3683185044283706,
-      "learning_rate": 9.876408456720261e-06,
-      "loss": 0.8057,
+      "epoch": 1.4751739026145358,
+      "grad_norm": 1.4922253999621244,
+      "learning_rate": 4.802967990267867e-06,
+      "loss": 0.5129,
       "step": 1025
     },
     {
-      "epoch": 0.49412329095706403,
-      "grad_norm": 3.0270798858712586,
-      "learning_rate": 9.873304745962725e-06,
-      "loss": 0.8337,
+      "epoch": 1.482369872871192,
+      "grad_norm": 1.5098918292072203,
+      "learning_rate": 4.770076893251986e-06,
+      "loss": 0.5239,
       "step": 1030
     },
     {
-      "epoch": 0.4965219477092828,
-      "grad_norm": 3.6556525120504655,
-      "learning_rate": 9.870163045704403e-06,
-      "loss": 0.9186,
+      "epoch": 1.4895658431278485,
+      "grad_norm": 1.6715329520651785,
+      "learning_rate": 4.7371316341047484e-06,
+      "loss": 0.5659,
       "step": 1035
     },
     {
-      "epoch": 0.49892060446150155,
-      "grad_norm": 3.5057620622825887,
-      "learning_rate": 9.866983380436202e-06,
-      "loss": 0.8611,
+      "epoch": 1.4967618133845046,
+      "grad_norm": 1.4935236860112138,
+      "learning_rate": 4.704134529978471e-06,
+      "loss": 0.4914,
       "step": 1040
     },
     {
-      "epoch": 0.5013192612137203,
-      "grad_norm": 2.703061130697398,
-      "learning_rate": 9.863765774944973e-06,
-      "loss": 0.791,
+      "epoch": 1.503957783641161,
+      "grad_norm": 1.4122572036404084,
+      "learning_rate": 4.671087901671899e-06,
+      "loss": 0.4798,
       "step": 1045
     },
     {
-      "epoch": 0.5037179179659391,
-      "grad_norm": 2.58369559163552,
-      "learning_rate": 9.860510254313332e-06,
-      "loss": 0.9499,
+      "epoch": 1.5111537538978173,
+      "grad_norm": 1.4681688735493286,
+      "learning_rate": 4.637994073466981e-06,
+      "loss": 0.5051,
       "step": 1050
     },
     {
-      "epoch": 0.5061165747181579,
-      "grad_norm": 2.682160545465096,
-      "learning_rate": 9.857216843919457e-06,
-      "loss": 0.942,
+      "epoch": 1.5183497241544734,
+      "grad_norm": 1.44942902813713,
+      "learning_rate": 4.604855372965394e-06,
+      "loss": 0.539,
       "step": 1055
     },
     {
-      "epoch": 0.5085152314703766,
-      "grad_norm": 2.780862397297652,
-      "learning_rate": 9.853885569436896e-06,
-      "loss": 0.9212,
+      "epoch": 1.5255456944111296,
+      "grad_norm": 1.236269924531647,
+      "learning_rate": 4.5716741309248445e-06,
+      "loss": 0.5305,
       "step": 1060
     },
     {
-      "epoch": 0.5109138882225953,
-      "grad_norm": 2.022760314849931,
-      "learning_rate": 9.850516456834364e-06,
-      "loss": 0.8786,
+      "epoch": 1.532741664667786,
+      "grad_norm": 1.6240033228942266,
+      "learning_rate": 4.538452681095123e-06,
+      "loss": 0.5531,
       "step": 1065
     },
     {
-      "epoch": 0.5133125449748142,
-      "grad_norm": 2.604771721280209,
-      "learning_rate": 9.847109532375536e-06,
-      "loss": 0.9393,
+      "epoch": 1.5399376349244425,
+      "grad_norm": 1.5622316143038788,
+      "learning_rate": 4.5051933600539705e-06,
+      "loss": 0.494,
       "step": 1070
     },
     {
-      "epoch": 0.5157112017270329,
-      "grad_norm": 3.5331046029013553,
-      "learning_rate": 9.84366482261885e-06,
-      "loss": 0.8909,
+      "epoch": 1.5471336051810987,
+      "grad_norm": 1.5638109251408436,
+      "learning_rate": 4.471898507042745e-06,
+      "loss": 0.533,
       "step": 1075
     },
     {
-      "epoch": 0.5181098584792516,
-      "grad_norm": 2.640697846099872,
-      "learning_rate": 9.840182354417298e-06,
-      "loss": 0.9053,
+      "epoch": 1.5543295754377549,
+      "grad_norm": 1.5029077042303538,
+      "learning_rate": 4.438570463801884e-06,
+      "loss": 0.513,
       "step": 1080
     },
     {
-      "epoch": 0.5205085152314703,
-      "grad_norm": 2.5514635987479077,
-      "learning_rate": 9.836662154918212e-06,
-      "loss": 0.8935,
+      "epoch": 1.561525545694411,
+      "grad_norm": 1.372369046315647,
+      "learning_rate": 4.405211574406209e-06,
+      "loss": 0.4698,
       "step": 1085
     },
     {
-      "epoch": 0.5229071719836891,
-      "grad_norm": 2.7348046479736374,
-      "learning_rate": 9.833104251563058e-06,
-      "loss": 0.8916,
+      "epoch": 1.5687215159510672,
+      "grad_norm": 1.6322850083649267,
+      "learning_rate": 4.371824185100054e-06,
+      "loss": 0.4607,
       "step": 1090
     },
     {
-      "epoch": 0.5253058287359079,
-      "grad_norm": 2.8303414330543686,
-      "learning_rate": 9.829508672087216e-06,
-      "loss": 0.8965,
+      "epoch": 1.5759174862077237,
+      "grad_norm": 1.6038884982185644,
+      "learning_rate": 4.338410644132256e-06,
+      "loss": 0.4918,
       "step": 1095
     },
     {
-      "epoch": 0.5277044854881267,
-      "grad_norm": 2.9495795537053553,
-      "learning_rate": 9.825875444519772e-06,
-      "loss": 0.8806,
-      "step": 1100
-    },
-    {
-      "epoch": 0.5277044854881267,
-      "eval_loss": 0.8870354294776917,
-      "eval_runtime": 739.0135,
-      "eval_samples_per_second": 10.027,
-      "eval_steps_per_second": 0.628,
+      "epoch": 1.58311345646438,
+      "grad_norm": 1.510163759577701,
+      "learning_rate": 4.304973301590977e-06,
+      "loss": 0.5141,
       "step": 1100
     },
     {
-      "epoch": 0.5301031422403454,
-      "grad_norm": 2.6141874348758583,
-      "learning_rate": 9.822204597183295e-06,
-      "loss": 0.8272,
+      "epoch": 1.5903094267210363,
+      "grad_norm": 1.3447962919620353,
+      "learning_rate": 4.271514509238434e-06,
+      "loss": 0.5719,
       "step": 1105
     },
     {
-      "epoch": 0.5325017989925641,
-      "grad_norm": 2.2732654732109387,
-      "learning_rate": 9.818496158693611e-06,
-      "loss": 0.8833,
+      "epoch": 1.5975053969776924,
+      "grad_norm": 1.4095809553505452,
+      "learning_rate": 4.238036620345477e-06,
+      "loss": 0.5378,
       "step": 1110
     },
     {
-      "epoch": 0.534900455744783,
-      "grad_norm": 3.143024300051868,
-      "learning_rate": 9.814750157959591e-06,
-      "loss": 0.938,
+      "epoch": 1.6047013672343486,
+      "grad_norm": 1.5471600021180223,
+      "learning_rate": 4.204541989526083e-06,
+      "loss": 0.5159,
       "step": 1115
     },
     {
-      "epoch": 0.5372991124970017,
-      "grad_norm": 2.1585701131113737,
-      "learning_rate": 9.810966624182917e-06,
-      "loss": 0.7753,
+      "epoch": 1.611897337491005,
+      "grad_norm": 1.3407123732434036,
+      "learning_rate": 4.171032972571744e-06,
+      "loss": 0.514,
       "step": 1120
     },
     {
-      "epoch": 0.5396977692492204,
-      "grad_norm": 2.6267326211662914,
-      "learning_rate": 9.807145586857861e-06,
-      "loss": 0.8769,
+      "epoch": 1.6190933077476615,
+      "grad_norm": 1.5587421912806174,
+      "learning_rate": 4.137511926285779e-06,
+      "loss": 0.4943,
       "step": 1125
     },
     {
-      "epoch": 0.5420964260014391,
-      "grad_norm": 3.036194720896026,
-      "learning_rate": 9.80328707577104e-06,
-      "loss": 0.949,
+      "epoch": 1.6262892780043177,
+      "grad_norm": 1.5228893274275985,
+      "learning_rate": 4.103981208317571e-06,
+      "loss": 0.5161,
       "step": 1130
     },
     {
-      "epoch": 0.544495082753658,
-      "grad_norm": 2.9408794675718926,
-      "learning_rate": 9.79939112100121e-06,
-      "loss": 0.9483,
+      "epoch": 1.6334852482609739,
+      "grad_norm": 1.3816307683209126,
+      "learning_rate": 4.070443176996745e-06,
+      "loss": 0.5036,
       "step": 1135
     },
     {
-      "epoch": 0.5468937395058767,
-      "grad_norm": 3.0771625978930754,
-      "learning_rate": 9.795457752919007e-06,
-      "loss": 0.899,
+      "epoch": 1.64068121851763,
+      "grad_norm": 1.551360790563111,
+      "learning_rate": 4.036900191167301e-06,
+      "loss": 0.4973,
       "step": 1140
     },
     {
-      "epoch": 0.5492923962580955,
-      "grad_norm": 2.5540520239747444,
-      "learning_rate": 9.791487002186721e-06,
-      "loss": 1.0063,
+      "epoch": 1.6478771887742862,
+      "grad_norm": 1.3412086458854404,
+      "learning_rate": 4.003354610021701e-06,
+      "loss": 0.5029,
       "step": 1145
     },
     {
-      "epoch": 0.5516910530103142,
-      "grad_norm": 3.1679148544487155,
-      "learning_rate": 9.787478899758059e-06,
-      "loss": 0.8657,
+      "epoch": 1.6550731590309427,
+      "grad_norm": 1.4793955535305545,
+      "learning_rate": 3.96980879293495e-06,
+      "loss": 0.4925,
       "step": 1150
     },
     {
-      "epoch": 0.554089709762533,
-      "grad_norm": 2.976757653301214,
-      "learning_rate": 9.783433476877898e-06,
-      "loss": 0.8958,
+      "epoch": 1.662269129287599,
+      "grad_norm": 1.2902630164808577,
+      "learning_rate": 3.9362650992986465e-06,
+      "loss": 0.4906,
       "step": 1155
     },
     {
-      "epoch": 0.5564883665147518,
-      "grad_norm": 2.5443050711582513,
-      "learning_rate": 9.779350765082045e-06,
-      "loss": 0.857,
+      "epoch": 1.6694650995442553,
+      "grad_norm": 1.5152262911206174,
+      "learning_rate": 3.902725888355037e-06,
+      "loss": 0.5019,
       "step": 1160
     },
     {
-      "epoch": 0.5588870232669705,
-      "grad_norm": 2.86585544843531,
-      "learning_rate": 9.775230796196989e-06,
-      "loss": 0.9084,
+      "epoch": 1.6766610698009115,
+      "grad_norm": 1.577445286461562,
+      "learning_rate": 3.869193519031086e-06,
+      "loss": 0.49,
       "step": 1165
     },
     {
-      "epoch": 0.5612856800191892,
-      "grad_norm": 2.963216428022191,
-      "learning_rate": 9.771073602339655e-06,
-      "loss": 0.8169,
+      "epoch": 1.6838570400575676,
+      "grad_norm": 1.4532186104713023,
+      "learning_rate": 3.835670349772566e-06,
+      "loss": 0.47,
       "step": 1170
     },
     {
-      "epoch": 0.5636843367714081,
-      "grad_norm": 2.760026417999172,
-      "learning_rate": 9.766879215917152e-06,
-      "loss": 0.8466,
+      "epoch": 1.691053010314224,
+      "grad_norm": 1.2527595036522912,
+      "learning_rate": 3.802158738378176e-06,
+      "loss": 0.4508,
       "step": 1175
     },
     {
-      "epoch": 0.5660829935236268,
-      "grad_norm": 2.6454017962342125,
-      "learning_rate": 9.762647669626522e-06,
-      "loss": 0.8039,
+      "epoch": 1.6982489805708805,
+      "grad_norm": 1.4942224309571124,
+      "learning_rate": 3.7686610418337083e-06,
+      "loss": 0.5039,
       "step": 1180
     },
     {
-      "epoch": 0.5684816502758455,
-      "grad_norm": 2.4626436519353616,
-      "learning_rate": 9.758378996454482e-06,
-      "loss": 0.923,
+      "epoch": 1.7054449508275367,
+      "grad_norm": 1.72911326258422,
+      "learning_rate": 3.7351796161462796e-06,
+      "loss": 0.4808,
       "step": 1185
     },
     {
-      "epoch": 0.5708803070280642,
-      "grad_norm": 4.129179717625037,
-      "learning_rate": 9.754073229677171e-06,
-      "loss": 0.972,
+      "epoch": 1.7126409210841929,
+      "grad_norm": 1.3747290591542072,
+      "learning_rate": 3.7017168161786215e-06,
+      "loss": 0.4993,
       "step": 1190
     },
     {
-      "epoch": 0.5732789637802831,
-      "grad_norm": 2.2820077397958123,
-      "learning_rate": 9.749730402859887e-06,
-      "loss": 0.994,
+      "epoch": 1.719836891340849,
+      "grad_norm": 1.4764634140281585,
+      "learning_rate": 3.6682749954834548e-06,
+      "loss": 0.5115,
       "step": 1195
     },
     {
-      "epoch": 0.5756776205325018,
-      "grad_norm": 2.609944354559159,
-      "learning_rate": 9.745350549856831e-06,
-      "loss": 0.8549,
-      "step": 1200
-    },
-    {
-      "epoch": 0.5756776205325018,
-      "eval_loss": 0.8886787295341492,
-      "eval_runtime": 739.2649,
-      "eval_samples_per_second": 10.023,
-      "eval_steps_per_second": 0.628,
+      "epoch": 1.7270328615975052,
+      "grad_norm": 1.6701936544900278,
+      "learning_rate": 3.634856506137956e-06,
+      "loss": 0.5653,
       "step": 1200
     },
     {
-      "epoch": 0.5780762772847206,
-      "grad_norm": 2.8948575829375565,
-      "learning_rate": 9.740933704810832e-06,
-      "loss": 0.8167,
+      "epoch": 1.7342288318541617,
+      "grad_norm": 1.5206960936360632,
+      "learning_rate": 3.6014636985783287e-06,
+      "loss": 0.521,
       "step": 1205
     },
     {
-      "epoch": 0.5804749340369393,
-      "grad_norm": 2.3042933975353463,
-      "learning_rate": 9.736479902153093e-06,
-      "loss": 0.8773,
+      "epoch": 1.741424802110818,
+      "grad_norm": 1.5138010545314067,
+      "learning_rate": 3.568098921434488e-06,
+      "loss": 0.4856,
       "step": 1210
     },
     {
-      "epoch": 0.5828735907891581,
-      "grad_norm": 3.3042761579341153,
-      "learning_rate": 9.731989176602918e-06,
-      "loss": 0.8711,
+      "epoch": 1.7486207723674743,
+      "grad_norm": 1.5508467662920749,
+      "learning_rate": 3.534764521364879e-06,
+      "loss": 0.4846,
       "step": 1215
     },
     {
-      "epoch": 0.5852722475413769,
-      "grad_norm": 3.448353009463417,
-      "learning_rate": 9.727461563167436e-06,
-      "loss": 0.8695,
+      "epoch": 1.7558167426241305,
+      "grad_norm": 1.294214634658577,
+      "learning_rate": 3.501462842891418e-06,
+      "loss": 0.4876,
       "step": 1220
     },
     {
-      "epoch": 0.5876709042935956,
-      "grad_norm": 2.628400374964348,
-      "learning_rate": 9.722897097141336e-06,
-      "loss": 0.9043,
+      "epoch": 1.7630127128807866,
+      "grad_norm": 1.4066532330498682,
+      "learning_rate": 3.4681962282346023e-06,
+      "loss": 0.4644,
       "step": 1225
     },
     {
-      "epoch": 0.5900695610458143,
-      "grad_norm": 3.519964559295127,
-      "learning_rate": 9.718295814106589e-06,
-      "loss": 0.8924,
+      "epoch": 1.770208683137443,
+      "grad_norm": 1.4435875920097863,
+      "learning_rate": 3.4349670171487714e-06,
+      "loss": 0.5199,
       "step": 1230
     },
     {
-      "epoch": 0.592468217798033,
-      "grad_norm": 2.9319636027613742,
-      "learning_rate": 9.713657749932172e-06,
-      "loss": 0.931,
+      "epoch": 1.7774046533940993,
+      "grad_norm": 1.5007291535231901,
+      "learning_rate": 3.4017775467575446e-06,
+      "loss": 0.5224,
       "step": 1235
     },
     {
-      "epoch": 0.5948668745502519,
-      "grad_norm": 2.938976926867282,
-      "learning_rate": 9.70898294077378e-06,
-      "loss": 0.8561,
+      "epoch": 1.7846006236507557,
+      "grad_norm": 1.2699334841273793,
+      "learning_rate": 3.3686301513894416e-06,
+      "loss": 0.4914,
       "step": 1240
     },
     {
-      "epoch": 0.5972655313024706,
-      "grad_norm": 2.29795980845488,
-      "learning_rate": 9.704271423073562e-06,
-      "loss": 0.8199,
+      "epoch": 1.7917965939074119,
+      "grad_norm": 1.2195698381350315,
+      "learning_rate": 3.3355271624137037e-06,
+      "loss": 0.4719,
       "step": 1245
     },
     {
-      "epoch": 0.5996641880546894,
-      "grad_norm": 2.398089494212856,
-      "learning_rate": 9.699523233559813e-06,
-      "loss": 0.8939,
+      "epoch": 1.798992564164068,
+      "grad_norm": 1.3895500926839512,
+      "learning_rate": 3.3024709080763186e-06,
+      "loss": 0.5144,
       "step": 1250
     },
     {
-      "epoch": 0.6020628448069081,
-      "grad_norm": 2.380808540854109,
-      "learning_rate": 9.69473840924671e-06,
-      "loss": 0.8052,
+      "epoch": 1.8061885344207242,
+      "grad_norm": 1.48913348325029,
+      "learning_rate": 3.269463713336268e-06,
+      "loss": 0.5103,
       "step": 1255
     },
     {
-      "epoch": 0.6044615015591269,
-      "grad_norm": 2.8543322126129915,
-      "learning_rate": 9.689916987434012e-06,
-      "loss": 0.8934,
+      "epoch": 1.8133845046773807,
+      "grad_norm": 1.3658603273721261,
+      "learning_rate": 3.236507899702005e-06,
+      "loss": 0.473,
       "step": 1260
     },
     {
-      "epoch": 0.6068601583113457,
-      "grad_norm": 4.013675664035515,
-      "learning_rate": 9.685059005706767e-06,
-      "loss": 0.8592,
+      "epoch": 1.820580474934037,
+      "grad_norm": 1.3775281668274946,
+      "learning_rate": 3.2036057850681745e-06,
+      "loss": 0.514,
       "step": 1265
     },
     {
-      "epoch": 0.6092588150635644,
-      "grad_norm": 2.557169604456725,
-      "learning_rate": 9.680164501935027e-06,
-      "loss": 0.9077,
+      "epoch": 1.8277764451906933,
+      "grad_norm": 1.631774330401289,
+      "learning_rate": 3.170759683552586e-06,
+      "loss": 0.5163,
       "step": 1270
     },
     {
-      "epoch": 0.6116574718157831,
-      "grad_norm": 2.8906956541267745,
-      "learning_rate": 9.675233514273552e-06,
-      "loss": 0.9295,
+      "epoch": 1.8349724154473495,
+      "grad_norm": 1.384315211463836,
+      "learning_rate": 3.137971905333458e-06,
+      "loss": 0.4752,
       "step": 1275
     },
     {
-      "epoch": 0.614056128568002,
-      "grad_norm": 2.550036815779647,
-      "learning_rate": 9.670266081161501e-06,
-      "loss": 0.9126,
+      "epoch": 1.8421683857040057,
+      "grad_norm": 1.4207130788508293,
+      "learning_rate": 3.1052447564869343e-06,
+      "loss": 0.5018,
       "step": 1280
     },
     {
-      "epoch": 0.6164547853202207,
-      "grad_norm": 3.015435597841737,
-      "learning_rate": 9.665262241322148e-06,
-      "loss": 0.8321,
+      "epoch": 1.849364355960662,
+      "grad_norm": 1.5148685580490273,
+      "learning_rate": 3.0725805388248834e-06,
+      "loss": 0.5127,
       "step": 1285
     },
     {
-      "epoch": 0.6188534420724394,
-      "grad_norm": 2.437308403492592,
-      "learning_rate": 9.66022203376257e-06,
-      "loss": 0.8764,
+      "epoch": 1.8565603262173183,
+      "grad_norm": 1.4919953654248346,
+      "learning_rate": 3.039981549733014e-06,
+      "loss": 0.4971,
       "step": 1290
     },
     {
-      "epoch": 0.6212520988246581,
-      "grad_norm": 2.655419936930409,
-      "learning_rate": 9.655145497773348e-06,
-      "loss": 0.8112,
+      "epoch": 1.8637562964739747,
+      "grad_norm": 2.065773895317616,
+      "learning_rate": 3.007450082009283e-06,
+      "loss": 0.4843,
       "step": 1295
     },
     {
-      "epoch": 0.623650755576877,
-      "grad_norm": 2.856040503173194,
-      "learning_rate": 9.650032672928256e-06,
-      "loss": 0.9197,
-      "step": 1300
-    },
-    {
-      "epoch": 0.623650755576877,
-      "eval_loss": 0.891351580619812,
-      "eval_runtime": 739.9462,
-      "eval_samples_per_second": 10.014,
-      "eval_steps_per_second": 0.627,
+      "epoch": 1.8709522667306309,
+      "grad_norm": 1.4849907079900204,
+      "learning_rate": 2.9749884237026426e-06,
+      "loss": 0.5102,
       "step": 1300
     },
     {
-      "epoch": 0.6260494123290957,
-      "grad_norm": 2.7662185790099976,
-      "learning_rate": 9.644883599083959e-06,
-      "loss": 0.7763,
+      "epoch": 1.878148236987287,
+      "grad_norm": 1.7567340972637704,
+      "learning_rate": 2.9425988579521103e-06,
+      "loss": 0.4901,
       "step": 1305
     },
     {
-      "epoch": 0.6284480690813145,
-      "grad_norm": 2.9200612566129838,
-      "learning_rate": 9.639698316379692e-06,
-      "loss": 0.8978,
+      "epoch": 1.8853442072439432,
+      "grad_norm": 1.4271802301503538,
+      "learning_rate": 2.910283662826188e-06,
+      "loss": 0.4805,
       "step": 1310
     },
     {
-      "epoch": 0.6308467258335332,
-      "grad_norm": 2.0716196866252163,
-      "learning_rate": 9.634476865236964e-06,
-      "loss": 0.838,
+      "epoch": 1.8925401775005997,
+      "grad_norm": 1.479518679920681,
+      "learning_rate": 2.8780451111626384e-06,
+      "loss": 0.4908,
       "step": 1315
     },
     {
-      "epoch": 0.633245382585752,
-      "grad_norm": 2.437636233160616,
-      "learning_rate": 9.62921928635922e-06,
-      "loss": 0.8614,
+      "epoch": 1.899736147757256,
+      "grad_norm": 1.5062976034854971,
+      "learning_rate": 2.8458854704086275e-06,
+      "loss": 0.491,
       "step": 1320
     },
     {
-      "epoch": 0.6356440393379708,
-      "grad_norm": 2.8685025444969385,
-      "learning_rate": 9.623925620731546e-06,
-      "loss": 0.8602,
+      "epoch": 1.9069321180139123,
+      "grad_norm": 1.552858614788501,
+      "learning_rate": 2.8138070024612504e-06,
+      "loss": 0.4787,
       "step": 1325
     },
     {
-      "epoch": 0.6380426960901895,
-      "grad_norm": 2.7442682238420826,
-      "learning_rate": 9.618595909620335e-06,
-      "loss": 0.8944,
+      "epoch": 1.9141280882705685,
+      "grad_norm": 1.668691870366938,
+      "learning_rate": 2.7818119635084392e-06,
+      "loss": 0.536,
       "step": 1330
     },
     {
-      "epoch": 0.6404413528424082,
-      "grad_norm": 2.9361485822984124,
-      "learning_rate": 9.613230194572972e-06,
-      "loss": 0.8433,
+      "epoch": 1.9213240585272247,
+      "grad_norm": 1.4857517196291667,
+      "learning_rate": 2.749902603870283e-06,
+      "loss": 0.5047,
       "step": 1335
     },
     {
-      "epoch": 0.6428400095946271,
-      "grad_norm": 3.095568397022331,
-      "learning_rate": 9.607828517417507e-06,
-      "loss": 0.8038,
+      "epoch": 1.928520028783881,
+      "grad_norm": 2.9875786042021932,
+      "learning_rate": 2.7180811678407525e-06,
+      "loss": 0.504,
       "step": 1340
     },
     {
-      "epoch": 0.6452386663468458,
-      "grad_norm": 2.46360142167789,
-      "learning_rate": 9.60239092026233e-06,
-      "loss": 0.8382,
+      "epoch": 1.9357159990405373,
+      "grad_norm": 1.3595766267302944,
+      "learning_rate": 2.686349893529849e-06,
+      "loss": 0.4863,
       "step": 1345
     },
     {
-      "epoch": 0.6476373230990645,
-      "grad_norm": 2.7582607273277344,
-      "learning_rate": 9.596917445495843e-06,
-      "loss": 0.8037,
+      "epoch": 1.9429119692971937,
+      "grad_norm": 1.4889142920291538,
+      "learning_rate": 2.6547110127061975e-06,
+      "loss": 0.4926,
       "step": 1350
     },
     {
-      "epoch": 0.6500359798512833,
-      "grad_norm": 2.9601669174372693,
-      "learning_rate": 9.591408135786132e-06,
-      "loss": 0.8638,
+      "epoch": 1.9501079395538499,
+      "grad_norm": 1.451668900558599,
+      "learning_rate": 2.6231667506400706e-06,
+      "loss": 0.4984,
       "step": 1355
     },
     {
-      "epoch": 0.6524346366035021,
-      "grad_norm": 3.127369086710566,
-      "learning_rate": 9.585863034080624e-06,
-      "loss": 0.8225,
+      "epoch": 1.957303909810506,
+      "grad_norm": 1.47137694607456,
+      "learning_rate": 2.591719325946883e-06,
+      "loss": 0.5209,
       "step": 1360
     },
     {
-      "epoch": 0.6548332933557208,
-      "grad_norm": 3.3773196297676256,
-      "learning_rate": 9.58028218360577e-06,
-      "loss": 0.8978,
+      "epoch": 1.9644998800671623,
+      "grad_norm": 1.5284770765800149,
+      "learning_rate": 2.560370950431146e-06,
+      "loss": 0.4603,
       "step": 1365
     },
     {
-      "epoch": 0.6572319501079396,
-      "grad_norm": 3.7552856472244502,
-      "learning_rate": 9.574665627866692e-06,
-      "loss": 0.8582,
+      "epoch": 1.9716958503238187,
+      "grad_norm": 1.402896466872614,
+      "learning_rate": 2.5291238289309054e-06,
+      "loss": 0.5077,
       "step": 1370
     },
     {
-      "epoch": 0.6596306068601583,
-      "grad_norm": 2.958749010394765,
-      "learning_rate": 9.56901341064685e-06,
-      "loss": 0.8175,
+      "epoch": 1.978891820580475,
+      "grad_norm": 1.4817926638302614,
+      "learning_rate": 2.497980159162667e-06,
+      "loss": 0.4839,
       "step": 1375
     },
     {
-      "epoch": 0.662029263612377,
-      "grad_norm": 2.9983246522004054,
-      "learning_rate": 9.563325576007702e-06,
-      "loss": 0.8431,
+      "epoch": 1.9860877908371313,
+      "grad_norm": 1.5453436757112435,
+      "learning_rate": 2.466942131566824e-06,
+      "loss": 0.4888,
       "step": 1380
     },
     {
-      "epoch": 0.6644279203645959,
-      "grad_norm": 3.074984178917188,
-      "learning_rate": 9.557602168288357e-06,
-      "loss": 0.9192,
+      "epoch": 1.9932837610937875,
+      "grad_norm": 1.4335485637084342,
+      "learning_rate": 2.4360119291535955e-06,
+      "loss": 0.4917,
       "step": 1385
     },
     {
-      "epoch": 0.6668265771168146,
-      "grad_norm": 2.531045908501192,
-      "learning_rate": 9.551843232105239e-06,
-      "loss": 0.8562,
+      "epoch": 2.0004797313504437,
+      "grad_norm": 1.6143369616539034,
+      "learning_rate": 2.405191727349489e-06,
+      "loss": 0.4993,
       "step": 1390
     },
     {
-      "epoch": 0.6692252338690333,
-      "grad_norm": 3.4852818268873174,
-      "learning_rate": 9.54604881235172e-06,
-      "loss": 0.8687,
+      "epoch": 2.0076757016071,
+      "grad_norm": 1.2224443867251211,
+      "learning_rate": 2.3744836938442936e-06,
+      "loss": 0.2088,
       "step": 1395
     },
     {
-      "epoch": 0.671623890621252,
-      "grad_norm": 2.23016124603345,
-      "learning_rate": 9.540218954197789e-06,
-      "loss": 0.8864,
+      "epoch": 2.0148716718637565,
+      "grad_norm": 1.2602299678447657,
+      "learning_rate": 2.3438899884386185e-06,
+      "loss": 0.1941,
       "step": 1400
     },
     {
-      "epoch": 0.671623890621252,
-      "eval_loss": 0.8827491998672485,
-      "eval_runtime": 739.7102,
-      "eval_samples_per_second": 10.017,
+      "epoch": 2.0148716718637565,
+      "eval_loss": 0.9261869192123413,
+      "eval_runtime": 740.3886,
+      "eval_samples_per_second": 10.008,
       "eval_steps_per_second": 0.627,
       "step": 1400
     },
     {
-      "epoch": 0.6740225473734709,
-      "grad_norm": 3.018147763364397,
-      "learning_rate": 9.534353703089692e-06,
-      "loss": 0.9238,
+      "epoch": 2.0220676421204127,
+      "grad_norm": 1.323505095452205,
+      "learning_rate": 2.3134127628919927e-06,
+      "loss": 0.1915,
       "step": 1405
     },
     {
-      "epoch": 0.6764212041256896,
-      "grad_norm": 2.554891580216872,
-      "learning_rate": 9.528453104749575e-06,
-      "loss": 0.8289,
+      "epoch": 2.029263612377069,
+      "grad_norm": 1.2958926807268987,
+      "learning_rate": 2.2830541607715136e-06,
+      "loss": 0.1736,
       "step": 1410
     },
     {
-      "epoch": 0.6788198608779084,
-      "grad_norm": 2.258802758649112,
-      "learning_rate": 9.522517205175133e-06,
-      "loss": 0.8195,
+      "epoch": 2.036459582633725,
+      "grad_norm": 1.271545256175082,
+      "learning_rate": 2.2528163173010927e-06,
+      "loss": 0.1845,
       "step": 1415
     },
     {
-      "epoch": 0.6812185176301271,
-      "grad_norm": 3.1309363033906594,
-      "learning_rate": 9.516546050639253e-06,
-      "loss": 0.9182,
+      "epoch": 2.0436555528903813,
+      "grad_norm": 1.5029993532268295,
+      "learning_rate": 2.2227013592112757e-06,
+      "loss": 0.1893,
       "step": 1420
     },
     {
-      "epoch": 0.6836171743823459,
-      "grad_norm": 2.799964298869419,
-      "learning_rate": 9.510539687689641e-06,
-      "loss": 0.8926,
+      "epoch": 2.0508515231470374,
+      "grad_norm": 1.2921857666544403,
+      "learning_rate": 2.192711404589658e-06,
+      "loss": 0.1958,
       "step": 1425
     },
     {
-      "epoch": 0.6860158311345647,
-      "grad_norm": 7.855666861983768,
-      "learning_rate": 9.504498163148476e-06,
-      "loss": 0.8561,
+      "epoch": 2.058047493403694,
+      "grad_norm": 1.2460289218576504,
+      "learning_rate": 2.162848562731916e-06,
+      "loss": 0.1994,
       "step": 1430
     },
     {
-      "epoch": 0.6884144878867834,
-      "grad_norm": 3.0545825578502996,
-      "learning_rate": 9.498421524112032e-06,
-      "loss": 0.8985,
+      "epoch": 2.0652434636603503,
+      "grad_norm": 1.2881623243419067,
+      "learning_rate": 2.133114933993452e-06,
+      "loss": 0.1935,
       "step": 1435
     },
     {
-      "epoch": 0.6908131446390021,
-      "grad_norm": 3.2954044794247404,
-      "learning_rate": 9.492309817950315e-06,
-      "loss": 0.9906,
+      "epoch": 2.0724394339170065,
+      "grad_norm": 1.1792218418956621,
+      "learning_rate": 2.1035126096416704e-06,
+      "loss": 0.1951,
       "step": 1440
     },
     {
-      "epoch": 0.693211801391221,
-      "grad_norm": 2.5123813211532684,
-      "learning_rate": 9.486163092306699e-06,
-      "loss": 0.8587,
+      "epoch": 2.0796354041736627,
+      "grad_norm": 1.284870948942911,
+      "learning_rate": 2.07404367170889e-06,
+      "loss": 0.1948,
       "step": 1445
     },
     {
-      "epoch": 0.6956104581434397,
-      "grad_norm": 3.0011863719016527,
-      "learning_rate": 9.479981395097545e-06,
-      "loss": 0.8932,
+      "epoch": 2.086831374430319,
+      "grad_norm": 1.2222749381574636,
+      "learning_rate": 2.0447101928459083e-06,
+      "loss": 0.1927,
       "step": 1450
     },
     {
-      "epoch": 0.6980091148956584,
-      "grad_norm": 3.0839066691421775,
-      "learning_rate": 9.473764774511833e-06,
-      "loss": 0.8058,
+      "epoch": 2.0940273446869755,
+      "grad_norm": 1.4089391140981338,
+      "learning_rate": 2.0155142361762256e-06,
+      "loss": 0.1553,
       "step": 1455
     },
     {
-      "epoch": 0.7004077716478772,
-      "grad_norm": 2.2891983688771,
-      "learning_rate": 9.467513279010792e-06,
-      "loss": 0.7818,
+      "epoch": 2.1012233149436317,
+      "grad_norm": 1.1670069976157664,
+      "learning_rate": 1.986457855150937e-06,
+      "loss": 0.1882,
       "step": 1460
     },
     {
-      "epoch": 0.702806428400096,
-      "grad_norm": 2.725073147047602,
-      "learning_rate": 9.461226957327506e-06,
-      "loss": 0.8881,
+      "epoch": 2.108419285200288,
+      "grad_norm": 1.20035808468667,
+      "learning_rate": 1.957543093404309e-06,
+      "loss": 0.1723,
       "step": 1465
     },
     {
-      "epoch": 0.7052050851523147,
-      "grad_norm": 3.1730042623525083,
-      "learning_rate": 9.454905858466547e-06,
-      "loss": 0.8823,
+      "epoch": 2.115615255456944,
+      "grad_norm": 1.2485627377889825,
+      "learning_rate": 1.9287719846100366e-06,
+      "loss": 0.1841,
       "step": 1470
     },
     {
-      "epoch": 0.7076037419045335,
-      "grad_norm": 2.8737424156514706,
-      "learning_rate": 9.448550031703599e-06,
-      "loss": 0.8467,
+      "epoch": 2.1228112257136003,
+      "grad_norm": 1.4758861123400375,
+      "learning_rate": 1.900146552338222e-06,
+      "loss": 0.1989,
       "step": 1475
     },
     {
-      "epoch": 0.7100023986567522,
-      "grad_norm": 2.5220020914196914,
-      "learning_rate": 9.442159526585052e-06,
-      "loss": 0.8828,
+      "epoch": 2.1300071959702565,
+      "grad_norm": 1.415660377393989,
+      "learning_rate": 1.8716688099130336e-06,
+      "loss": 0.1792,
       "step": 1480
     },
     {
-      "epoch": 0.712401055408971,
-      "grad_norm": 3.219145860445945,
-      "learning_rate": 9.435734392927639e-06,
-      "loss": 0.8935,
+      "epoch": 2.137203166226913,
+      "grad_norm": 1.1398390415745234,
+      "learning_rate": 1.8433407602711122e-06,
+      "loss": 0.1828,
       "step": 1485
     },
     {
-      "epoch": 0.7147997121611898,
-      "grad_norm": 3.4596708965275673,
-      "learning_rate": 9.429274680818029e-06,
-      "loss": 0.8935,
+      "epoch": 2.1443991364835693,
+      "grad_norm": 1.436825768706905,
+      "learning_rate": 1.8151643958206963e-06,
+      "loss": 0.1873,
       "step": 1490
     },
     {
-      "epoch": 0.7171983689134085,
-      "grad_norm": 2.180515903835741,
-      "learning_rate": 9.422780440612449e-06,
-      "loss": 0.9376,
+      "epoch": 2.1515951067402255,
+      "grad_norm": 1.2111903866598819,
+      "learning_rate": 1.7871416983014864e-06,
+      "loss": 0.1747,
       "step": 1495
     },
     {
-      "epoch": 0.7195970256656272,
-      "grad_norm": 3.5672104188512,
-      "learning_rate": 9.416251722936289e-06,
-      "loss": 0.8231,
-      "step": 1500
-    },
-    {
-      "epoch": 0.7195970256656272,
-      "eval_loss": 0.8757531642913818,
-      "eval_runtime": 740.9577,
-      "eval_samples_per_second": 10.001,
-      "eval_steps_per_second": 0.626,
+      "epoch": 2.1587910769968817,
+      "grad_norm": 1.592322648486121,
+      "learning_rate": 1.7592746386452641e-06,
+      "loss": 0.1981,
       "step": 1500
     },
     {
-      "epoch": 0.721995682417846,
-      "grad_norm": 5.348376903765216,
-      "learning_rate": 9.409688578683702e-06,
-      "loss": 0.9555,
+      "epoch": 2.165987047253538,
+      "grad_norm": 1.3381476033081696,
+      "learning_rate": 1.7315651768372734e-06,
+      "loss": 0.1752,
       "step": 1505
     },
     {
-      "epoch": 0.7243943391700648,
-      "grad_norm": 3.0987915733129,
-      "learning_rate": 9.403091059017214e-06,
-      "loss": 0.894,
+      "epoch": 2.1731830175101945,
+      "grad_norm": 1.5243125399513529,
+      "learning_rate": 1.7040152617783607e-06,
+      "loss": 0.1797,
       "step": 1510
     },
     {
-      "epoch": 0.7267929959222835,
-      "grad_norm": 2.1853482972820113,
-      "learning_rate": 9.396459215367319e-06,
-      "loss": 0.7919,
+      "epoch": 2.1803789877668507,
+      "grad_norm": 1.5148343638714192,
+      "learning_rate": 1.6766268311479078e-06,
+      "loss": 0.193,
       "step": 1515
     },
     {
-      "epoch": 0.7291916526745023,
-      "grad_norm": 2.950634652802006,
-      "learning_rate": 9.389793099432083e-06,
-      "loss": 0.7889,
+      "epoch": 2.187574958023507,
+      "grad_norm": 1.315102374687142,
+      "learning_rate": 1.649401811267546e-06,
+      "loss": 0.1889,
       "step": 1520
     },
     {
-      "epoch": 0.731590309426721,
-      "grad_norm": 2.551117094087631,
-      "learning_rate": 9.38309276317674e-06,
-      "loss": 0.9133,
+      "epoch": 2.194770928280163,
+      "grad_norm": 1.5038597370043303,
+      "learning_rate": 1.622342116965672e-06,
+      "loss": 0.2193,
       "step": 1525
     },
     {
-      "epoch": 0.7339889661789398,
-      "grad_norm": 3.484732257795789,
-      "learning_rate": 9.376358258833283e-06,
-      "loss": 0.9706,
+      "epoch": 2.2019668985368193,
+      "grad_norm": 1.3456620640508148,
+      "learning_rate": 1.595449651442771e-06,
+      "loss": 0.1842,
       "step": 1530
     },
     {
-      "epoch": 0.7363876229311586,
-      "grad_norm": 2.526861763894699,
-      "learning_rate": 9.369589638900065e-06,
-      "loss": 0.9191,
+      "epoch": 2.2091628687934755,
+      "grad_norm": 1.3647300470767014,
+      "learning_rate": 1.5687263061375595e-06,
+      "loss": 0.1752,
       "step": 1535
     },
     {
-      "epoch": 0.7387862796833773,
-      "grad_norm": 3.206173841917802,
-      "learning_rate": 9.362786956141382e-06,
-      "loss": 0.8839,
+      "epoch": 2.216358839050132,
+      "grad_norm": 1.417987485184227,
+      "learning_rate": 1.5421739605939518e-06,
+      "loss": 0.1728,
       "step": 1540
     },
     {
-      "epoch": 0.741184936435596,
-      "grad_norm": 4.53746680005531,
-      "learning_rate": 9.35595026358706e-06,
-      "loss": 0.8787,
+      "epoch": 2.2235548093067883,
+      "grad_norm": 1.5887020304276804,
+      "learning_rate": 1.5157944823288672e-06,
+      "loss": 0.1637,
       "step": 1545
     },
     {
-      "epoch": 0.7435835931878149,
-      "grad_norm": 3.1177510145335425,
-      "learning_rate": 9.34907961453205e-06,
-      "loss": 0.9194,
+      "epoch": 2.2307507795634445,
+      "grad_norm": 1.3375708672110973,
+      "learning_rate": 1.4895897267008782e-06,
+      "loss": 0.1792,
       "step": 1550
     },
     {
-      "epoch": 0.7459822499400336,
-      "grad_norm": 2.3736625117965278,
-      "learning_rate": 9.342175062536012e-06,
-      "loss": 0.8542,
+      "epoch": 2.2379467498201007,
+      "grad_norm": 1.3565700525423485,
+      "learning_rate": 1.463561536779724e-06,
+      "loss": 0.1921,
       "step": 1555
     },
     {
-      "epoch": 0.7483809066922523,
-      "grad_norm": 3.443392708153097,
-      "learning_rate": 9.33523666142289e-06,
-      "loss": 0.8262,
+      "epoch": 2.245142720076757,
+      "grad_norm": 1.5551856772129453,
+      "learning_rate": 1.4377117432166718e-06,
+      "loss": 0.1618,
       "step": 1560
     },
     {
-      "epoch": 0.750779563444471,
-      "grad_norm": 2.4419951167114515,
-      "learning_rate": 9.328264465280494e-06,
-      "loss": 0.8791,
+      "epoch": 2.2523386903334135,
+      "grad_norm": 1.2100448164204372,
+      "learning_rate": 1.4120421641157662e-06,
+      "loss": 0.1928,
       "step": 1565
     },
     {
-      "epoch": 0.7531782201966899,
-      "grad_norm": 3.3955749482732402,
-      "learning_rate": 9.321258528460087e-06,
-      "loss": 0.8317,
+      "epoch": 2.2595346605900697,
+      "grad_norm": 1.438877153368831,
+      "learning_rate": 1.386554604905955e-06,
+      "loss": 0.1774,
       "step": 1570
     },
     {
-      "epoch": 0.7555768769489086,
-      "grad_norm": 2.7100494409509364,
-      "learning_rate": 9.314218905575947e-06,
-      "loss": 0.8661,
+      "epoch": 2.266730630846726,
+      "grad_norm": 1.2780217507242704,
+      "learning_rate": 1.3612508582141065e-06,
+      "loss": 0.1871,
       "step": 1575
     },
     {
-      "epoch": 0.7579755337011274,
-      "grad_norm": 2.8677532226727607,
-      "learning_rate": 9.307145651504959e-06,
-      "loss": 0.7924,
+      "epoch": 2.273926601103382,
+      "grad_norm": 1.3558845492725387,
+      "learning_rate": 1.3361327037389295e-06,
+      "loss": 0.2018,
       "step": 1580
     },
     {
-      "epoch": 0.7603741904533461,
-      "grad_norm": 2.8165977913128133,
-      "learning_rate": 9.300038821386167e-06,
-      "loss": 0.8922,
+      "epoch": 2.2811225713600383,
+      "grad_norm": 1.3490250928179355,
+      "learning_rate": 1.3112019081257986e-06,
+      "loss": 0.1731,
       "step": 1585
     },
     {
-      "epoch": 0.7627728472055649,
-      "grad_norm": 2.632477549615508,
-      "learning_rate": 9.292898470620364e-06,
-      "loss": 0.79,
+      "epoch": 2.2883185416166945,
+      "grad_norm": 1.2405141654870557,
+      "learning_rate": 1.2864602248425018e-06,
+      "loss": 0.1886,
       "step": 1590
     },
     {
-      "epoch": 0.7651715039577837,
-      "grad_norm": 3.2008058338812155,
-      "learning_rate": 9.285724654869646e-06,
-      "loss": 0.8729,
+      "epoch": 2.295514511873351,
+      "grad_norm": 1.2873724354006912,
+      "learning_rate": 1.2619093940559138e-06,
+      "loss": 0.1868,
       "step": 1595
     },
     {
-      "epoch": 0.7675701607100024,
-      "grad_norm": 2.8865776414732096,
-      "learning_rate": 9.278517430056983e-06,
-      "loss": 0.8658,
-      "step": 1600
-    },
-    {
-      "epoch": 0.7675701607100024,
-      "eval_loss": 0.8723194599151611,
-      "eval_runtime": 741.5116,
-      "eval_samples_per_second": 9.993,
-      "eval_steps_per_second": 0.626,
+      "epoch": 2.3027104821300073,
+      "grad_norm": 1.3107124153475105,
+      "learning_rate": 1.2375511425096013e-06,
+      "loss": 0.187,
       "step": 1600
     },
     {
-      "epoch": 0.7699688174622211,
-      "grad_norm": 3.215530709050795,
-      "learning_rate": 9.271276852365785e-06,
-      "loss": 0.8832,
+      "epoch": 2.3099064523866635,
+      "grad_norm": 1.3468010137925535,
+      "learning_rate": 1.213387183402378e-06,
+      "loss": 0.1771,
       "step": 1605
     },
     {
-      "epoch": 0.77236747421444,
-      "grad_norm": 2.0508885653823263,
-      "learning_rate": 9.264002978239459e-06,
-      "loss": 0.809,
+      "epoch": 2.3171024226433197,
+      "grad_norm": 1.4179240822671797,
+      "learning_rate": 1.1894192162678086e-06,
+      "loss": 0.1654,
       "step": 1610
     },
     {
-      "epoch": 0.7747661309666587,
-      "grad_norm": 2.752742925623021,
-      "learning_rate": 9.256695864380975e-06,
-      "loss": 0.8717,
+      "epoch": 2.324298392899976,
+      "grad_norm": 1.3848546480668056,
+      "learning_rate": 1.165648926854672e-06,
+      "loss": 0.1838,
       "step": 1615
     },
     {
-      "epoch": 0.7771647877188774,
-      "grad_norm": 2.6147386599664824,
-      "learning_rate": 9.24935556775242e-06,
-      "loss": 0.8645,
+      "epoch": 2.331494363156632,
+      "grad_norm": 1.5195023852589002,
+      "learning_rate": 1.1420779870084052e-06,
+      "loss": 0.1955,
       "step": 1620
     },
     {
-      "epoch": 0.7795634444710962,
-      "grad_norm": 2.7748844905168153,
-      "learning_rate": 9.241982145574555e-06,
-      "loss": 0.8061,
+      "epoch": 2.3386903334132887,
+      "grad_norm": 1.2410727385995408,
+      "learning_rate": 1.1187080545535064e-06,
+      "loss": 0.1685,
       "step": 1625
     },
     {
-      "epoch": 0.781962101223315,
-      "grad_norm": 2.8763749911858216,
-      "learning_rate": 9.234575655326369e-06,
-      "loss": 0.8543,
+      "epoch": 2.345886303669945,
+      "grad_norm": 1.1237477417805415,
+      "learning_rate": 1.09554077317694e-06,
+      "loss": 0.1824,
       "step": 1630
     },
     {
-      "epoch": 0.7843607579755337,
-      "grad_norm": 2.3381964392401318,
-      "learning_rate": 9.227136154744626e-06,
-      "loss": 0.7687,
+      "epoch": 2.353082273926601,
+      "grad_norm": 1.2937342096954545,
+      "learning_rate": 1.0725777723125301e-06,
+      "loss": 0.1943,
       "step": 1635
     },
     {
-      "epoch": 0.7867594147277525,
-      "grad_norm": 2.798456190922495,
-      "learning_rate": 9.219663701823427e-06,
-      "loss": 0.7947,
+      "epoch": 2.3602782441832573,
+      "grad_norm": 1.2828779926698606,
+      "learning_rate": 1.0498206670263567e-06,
+      "loss": 0.1832,
       "step": 1640
     },
     {
-      "epoch": 0.7891580714799712,
-      "grad_norm": 4.15300651724386,
-      "learning_rate": 9.21215835481375e-06,
-      "loss": 0.84,
+      "epoch": 2.3674742144399135,
+      "grad_norm": 1.253425033010922,
+      "learning_rate": 1.0272710579031616e-06,
+      "loss": 0.2044,
       "step": 1645
     },
     {
-      "epoch": 0.7915567282321899,
-      "grad_norm": 3.176612904560953,
-      "learning_rate": 9.20462017222299e-06,
-      "loss": 0.8498,
+      "epoch": 2.37467018469657,
+      "grad_norm": 1.3678742472737333,
+      "learning_rate": 1.0049305309337758e-06,
+      "loss": 0.1672,
       "step": 1650
     },
     {
-      "epoch": 0.7939553849844088,
-      "grad_norm": 2.6908797712355934,
-      "learning_rate": 9.197049212814518e-06,
-      "loss": 0.8092,
+      "epoch": 2.3818661549532263,
+      "grad_norm": 1.5542727998398753,
+      "learning_rate": 9.82800657403569e-07,
+      "loss": 0.1955,
       "step": 1655
     },
     {
-      "epoch": 0.7963540417366275,
-      "grad_norm": 3.5662583410447324,
-      "learning_rate": 9.189445535607207e-06,
-      "loss": 0.9158,
+      "epoch": 2.3890621252098825,
+      "grad_norm": 1.4017624513152087,
+      "learning_rate": 9.60882993781937e-07,
+      "loss": 0.1733,
       "step": 1660
     },
     {
-      "epoch": 0.7987526984888462,
-      "grad_norm": 2.882549375268096,
-      "learning_rate": 9.181809199874983e-06,
-      "loss": 0.8807,
+      "epoch": 2.3962580954665387,
+      "grad_norm": 1.199342554533447,
+      "learning_rate": 9.391790816128304e-07,
+      "loss": 0.1649,
       "step": 1665
     },
     {
-      "epoch": 0.801151355241065,
-      "grad_norm": 2.2541644302033452,
-      "learning_rate": 9.174140265146356e-06,
-      "loss": 0.8684,
+      "epoch": 2.403454065723195,
+      "grad_norm": 1.2335679341459465,
+      "learning_rate": 9.176904474063319e-07,
+      "loss": 0.198,
       "step": 1670
     },
     {
-      "epoch": 0.8035500119932838,
-      "grad_norm": 2.403498333088569,
-      "learning_rate": 9.166438791203967e-06,
-      "loss": 0.8122,
+      "epoch": 2.4106500359798515,
+      "grad_norm": 1.4585828923578052,
+      "learning_rate": 8.964186025312908e-07,
+      "loss": 0.1988,
       "step": 1675
     },
     {
-      "epoch": 0.8059486687455025,
-      "grad_norm": 2.9396774647398383,
-      "learning_rate": 9.158704838084102e-06,
-      "loss": 0.8704,
+      "epoch": 2.4178460062365077,
+      "grad_norm": 1.4518660782918198,
+      "learning_rate": 8.753650431090252e-07,
+      "loss": 0.1701,
       "step": 1680
     },
     {
-      "epoch": 0.8083473254977213,
-      "grad_norm": 2.291793929286392,
-      "learning_rate": 9.15093846607625e-06,
-      "loss": 0.8499,
+      "epoch": 2.425041976493164,
+      "grad_norm": 1.3322728405275928,
+      "learning_rate": 8.545312499080922e-07,
+      "loss": 0.1729,
       "step": 1685
     },
     {
-      "epoch": 0.81074598224994,
-      "grad_norm": 4.646259487690163,
-      "learning_rate": 9.143139735722607e-06,
-      "loss": 0.8904,
+      "epoch": 2.43223794674982,
+      "grad_norm": 1.3067316057050342,
+      "learning_rate": 8.339186882401445e-07,
+      "loss": 0.1874,
       "step": 1690
     },
     {
-      "epoch": 0.8131446390021588,
-      "grad_norm": 3.3102326837687173,
-      "learning_rate": 9.135308707817623e-06,
-      "loss": 0.79,
+      "epoch": 2.4394339170064763,
+      "grad_norm": 1.4177678336114292,
+      "learning_rate": 8.135288078568656e-07,
+      "loss": 0.2021,
       "step": 1695
     },
     {
-      "epoch": 0.8155432957543776,
-      "grad_norm": 6.864638773101543,
-      "learning_rate": 9.12744544340752e-06,
-      "loss": 0.8506,
-      "step": 1700
-    },
-    {
-      "epoch": 0.8155432957543776,
-      "eval_loss": 0.8721764087677002,
-      "eval_runtime": 740.915,
-      "eval_samples_per_second": 10.001,
-      "eval_steps_per_second": 0.626,
+      "epoch": 2.4466298872631325,
+      "grad_norm": 1.3121080863750958,
+      "learning_rate": 7.933630428480049e-07,
+      "loss": 0.1699,
       "step": 1700
     },
     {
-      "epoch": 0.8179419525065963,
-      "grad_norm": 2.9260118718517245,
-      "learning_rate": 9.119550003789815e-06,
-      "loss": 0.9302,
+      "epoch": 2.453825857519789,
+      "grad_norm": 1.3185780885959946,
+      "learning_rate": 7.734228115405161e-07,
+      "loss": 0.1624,
       "step": 1705
     },
     {
-      "epoch": 0.820340609258815,
-      "grad_norm": 2.7721431237181133,
-      "learning_rate": 9.111622450512846e-06,
-      "loss": 0.8453,
+      "epoch": 2.4610218277764453,
+      "grad_norm": 1.33019533604804,
+      "learning_rate": 7.537095163987972e-07,
+      "loss": 0.1784,
       "step": 1710
     },
     {
-      "epoch": 0.8227392660110339,
-      "grad_norm": 2.955453435688259,
-      "learning_rate": 9.103662845375294e-06,
-      "loss": 0.9435,
+      "epoch": 2.4682177980331015,
+      "grad_norm": 1.3853774517952444,
+      "learning_rate": 7.342245439260537e-07,
+      "loss": 0.1824,
       "step": 1715
     },
     {
-      "epoch": 0.8251379227632526,
-      "grad_norm": 2.658722267205169,
-      "learning_rate": 9.095671250425693e-06,
-      "loss": 0.8011,
+      "epoch": 2.4754137682897577,
+      "grad_norm": 1.1804687459435843,
+      "learning_rate": 7.149692645667804e-07,
+      "loss": 0.1693,
       "step": 1720
     },
     {
-      "epoch": 0.8275365795154713,
-      "grad_norm": 2.6804272332523036,
-      "learning_rate": 9.087647727961956e-06,
-      "loss": 0.8641,
+      "epoch": 2.482609738546414,
+      "grad_norm": 1.250231314429457,
+      "learning_rate": 6.959450326103722e-07,
+      "loss": 0.2067,
       "step": 1725
     },
     {
-      "epoch": 0.8299352362676901,
-      "grad_norm": 3.39051606914781,
-      "learning_rate": 9.079592340530879e-06,
-      "loss": 0.8148,
+      "epoch": 2.48980570880307,
+      "grad_norm": 1.3184620916504868,
+      "learning_rate": 6.771531860958726e-07,
+      "loss": 0.1557,
       "step": 1730
     },
     {
-      "epoch": 0.8323338930199089,
-      "grad_norm": 2.537680913646963,
-      "learning_rate": 9.071505150927663e-06,
-      "loss": 0.8964,
+      "epoch": 2.4970016790597267,
+      "grad_norm": 1.3996911523285738,
+      "learning_rate": 6.585950467178656e-07,
+      "loss": 0.1984,
       "step": 1735
     },
     {
-      "epoch": 0.8347325497721276,
-      "grad_norm": 5.12661162717378,
-      "learning_rate": 9.06338622219542e-06,
-      "loss": 0.8701,
+      "epoch": 2.504197649316383,
+      "grad_norm": 1.330732277956789,
+      "learning_rate": 6.402719197335181e-07,
+      "loss": 0.1656,
       "step": 1740
     },
     {
-      "epoch": 0.8371312065243464,
-      "grad_norm": 4.645974768815863,
-      "learning_rate": 9.055235617624682e-06,
-      "loss": 0.8514,
+      "epoch": 2.511393619573039,
+      "grad_norm": 1.3782406997114114,
+      "learning_rate": 6.22185093870772e-07,
+      "loss": 0.1669,
       "step": 1745
     },
     {
-      "epoch": 0.8395298632765651,
-      "grad_norm": 3.084982306927161,
-      "learning_rate": 9.047053400752907e-06,
-      "loss": 0.8877,
+      "epoch": 2.5185895898296953,
+      "grad_norm": 1.4431968846802443,
+      "learning_rate": 6.043358412377069e-07,
+      "loss": 0.1799,
       "step": 1750
     },
     {
-      "epoch": 0.8419285200287839,
-      "grad_norm": 2.576877326965564,
-      "learning_rate": 9.038839635363987e-06,
-      "loss": 0.8163,
+      "epoch": 2.5257855600863515,
+      "grad_norm": 1.1865288276002492,
+      "learning_rate": 5.867254172330689e-07,
+      "loss": 0.1614,
       "step": 1755
     },
     {
-      "epoch": 0.8443271767810027,
-      "grad_norm": 3.51019570508426,
-      "learning_rate": 9.030594385487745e-06,
-      "loss": 0.8915,
+      "epoch": 2.532981530343008,
+      "grad_norm": 1.3447844251083265,
+      "learning_rate": 5.693550604579722e-07,
+      "loss": 0.1761,
       "step": 1760
     },
     {
-      "epoch": 0.8467258335332214,
-      "grad_norm": 3.8600342540679025,
-      "learning_rate": 9.02231771539944e-06,
-      "loss": 0.9304,
+      "epoch": 2.5401775005996643,
+      "grad_norm": 1.312290863998097,
+      "learning_rate": 5.52225992628784e-07,
+      "loss": 0.175,
       "step": 1765
     },
     {
-      "epoch": 0.8491244902854401,
-      "grad_norm": 2.703374255650966,
-      "learning_rate": 9.014009689619267e-06,
-      "loss": 0.8352,
+      "epoch": 2.5473734708563205,
+      "grad_norm": 1.325480546799902,
+      "learning_rate": 5.353394184912012e-07,
+      "loss": 0.1893,
       "step": 1770
     },
     {
-      "epoch": 0.8515231470376589,
-      "grad_norm": 2.3896428975452553,
-      "learning_rate": 9.005670372911848e-06,
-      "loss": 0.8556,
+      "epoch": 2.5545694411129767,
+      "grad_norm": 1.211006197074522,
+      "learning_rate": 5.186965257355092e-07,
+      "loss": 0.1738,
       "step": 1775
     },
     {
-      "epoch": 0.8539218037898777,
-      "grad_norm": 3.008401281219767,
-      "learning_rate": 8.997299830285735e-06,
-      "loss": 0.8637,
+      "epoch": 2.561765411369633,
+      "grad_norm": 1.2613128106853304,
+      "learning_rate": 5.022984849130542e-07,
+      "loss": 0.1735,
       "step": 1780
     },
     {
-      "epoch": 0.8563204605420964,
-      "grad_norm": 3.3845621217286315,
-      "learning_rate": 8.9888981269929e-06,
-      "loss": 0.9349,
+      "epoch": 2.5689613816262895,
+      "grad_norm": 1.4240080375407917,
+      "learning_rate": 4.861464493539116e-07,
+      "loss": 0.209,
       "step": 1785
     },
     {
-      "epoch": 0.8587191172943152,
-      "grad_norm": 3.066551980706959,
-      "learning_rate": 8.98046532852822e-06,
-      "loss": 1.0207,
+      "epoch": 2.5761573518829457,
+      "grad_norm": 1.212642870699417,
+      "learning_rate": 4.702415550857668e-07,
+      "loss": 0.1661,
       "step": 1790
     },
     {
-      "epoch": 0.8611177740465339,
-      "grad_norm": 3.593558463375603,
-      "learning_rate": 8.972001500628978e-06,
-      "loss": 0.8669,
+      "epoch": 2.583353322139602,
+      "grad_norm": 1.19899124906289,
+      "learning_rate": 4.5458492075401845e-07,
+      "loss": 0.1871,
       "step": 1795
     },
     {
-      "epoch": 0.8635164307987527,
-      "grad_norm": 4.827873639267246,
-      "learning_rate": 8.963506709274344e-06,
-      "loss": 0.9533,
-      "step": 1800
-    },
-    {
-      "epoch": 0.8635164307987527,
-      "eval_loss": 0.8710312247276306,
-      "eval_runtime": 740.1915,
-      "eval_samples_per_second": 10.011,
-      "eval_steps_per_second": 0.627,
+      "epoch": 2.590549292396258,
+      "grad_norm": 1.2451776201467897,
+      "learning_rate": 4.391776475430964e-07,
+      "loss": 0.1736,
       "step": 1800
     },
     {
-      "epoch": 0.8659150875509715,
-      "grad_norm": 2.1671636510448913,
-      "learning_rate": 8.95498102068486e-06,
-      "loss": 0.8423,
+      "epoch": 2.5977452626529143,
+      "grad_norm": 1.4217111682942414,
+      "learning_rate": 4.240208190990149e-07,
+      "loss": 0.1656,
       "step": 1805
     },
     {
-      "epoch": 0.8683137443031902,
-      "grad_norm": 2.2993204223464954,
-      "learning_rate": 8.946424501321922e-06,
-      "loss": 0.9611,
+      "epoch": 2.604941232909571,
+      "grad_norm": 1.154023125578338,
+      "learning_rate": 4.0911550145315356e-07,
+      "loss": 0.176,
       "step": 1810
     },
     {
-      "epoch": 0.8707124010554089,
-      "grad_norm": 3.6181635510094323,
-      "learning_rate": 8.937837217887273e-06,
-      "loss": 0.9857,
+      "epoch": 2.6121372031662267,
+      "grad_norm": 1.2517982852871838,
+      "learning_rate": 3.944627429472809e-07,
+      "loss": 0.168,
       "step": 1815
     },
     {
-      "epoch": 0.8731110578076278,
-      "grad_norm": 2.3157461628034968,
-      "learning_rate": 8.929219237322468e-06,
-      "loss": 0.8967,
+      "epoch": 2.6193331734228833,
+      "grad_norm": 1.3001175217867729,
+      "learning_rate": 3.8006357415981947e-07,
+      "loss": 0.1582,
       "step": 1820
     },
     {
-      "epoch": 0.8755097145598465,
-      "grad_norm": 2.3766663633520593,
-      "learning_rate": 8.920570626808364e-06,
-      "loss": 0.8718,
+      "epoch": 2.6265291436795395,
+      "grad_norm": 1.4179539106113206,
+      "learning_rate": 3.659190078333667e-07,
+      "loss": 0.1901,
       "step": 1825
     },
     {
-      "epoch": 0.8779083713120652,
-      "grad_norm": 3.1542811825496617,
-      "learning_rate": 8.911891453764587e-06,
-      "loss": 0.8869,
+      "epoch": 2.6337251139361957,
+      "grad_norm": 1.2865481274768071,
+      "learning_rate": 3.5203003880345786e-07,
+      "loss": 0.1825,
       "step": 1830
     },
     {
-      "epoch": 0.880307028064284,
-      "grad_norm": 2.3205179227394765,
-      "learning_rate": 8.903181785849016e-06,
-      "loss": 0.9165,
+      "epoch": 2.640921084192852,
+      "grad_norm": 1.2107327771575902,
+      "learning_rate": 3.383976439286007e-07,
+      "loss": 0.178,
       "step": 1835
     },
     {
-      "epoch": 0.8827056848165028,
-      "grad_norm": 2.395604854053,
-      "learning_rate": 8.894441690957246e-06,
-      "loss": 0.8951,
+      "epoch": 2.648117054449508,
+      "grad_norm": 1.4930579520298934,
+      "learning_rate": 3.250227820215694e-07,
+      "loss": 0.1795,
       "step": 1840
     },
     {
-      "epoch": 0.8851043415687215,
-      "grad_norm": 2.4063566793229674,
-      "learning_rate": 8.88567123722207e-06,
-      "loss": 0.867,
+      "epoch": 2.6553130247061647,
+      "grad_norm": 1.7580144453795274,
+      "learning_rate": 3.119063937819666e-07,
+      "loss": 0.1988,
       "step": 1845
     },
     {
-      "epoch": 0.8875029983209403,
-      "grad_norm": 3.262096284773896,
-      "learning_rate": 8.876870493012931e-06,
-      "loss": 0.8811,
+      "epoch": 2.662508994962821,
+      "grad_norm": 1.389677232858989,
+      "learning_rate": 2.990494017300604e-07,
+      "loss": 0.189,
       "step": 1850
     },
     {
-      "epoch": 0.889901655073159,
-      "grad_norm": 4.069658033112594,
-      "learning_rate": 8.868039526935407e-06,
-      "loss": 0.8339,
+      "epoch": 2.669704965219477,
+      "grad_norm": 1.4778063736068945,
+      "learning_rate": 2.864527101419032e-07,
+      "loss": 0.2053,
       "step": 1855
     },
     {
-      "epoch": 0.8923003118253778,
-      "grad_norm": 3.192868682184835,
-      "learning_rate": 8.859178407830668e-06,
-      "loss": 0.7906,
+      "epoch": 2.6769009354761333,
+      "grad_norm": 1.2577420076989798,
+      "learning_rate": 2.7411720498572744e-07,
+      "loss": 0.1917,
       "step": 1860
     },
     {
-      "epoch": 0.8946989685775966,
-      "grad_norm": 2.478055612898803,
-      "learning_rate": 8.850287204774936e-06,
-      "loss": 0.9756,
+      "epoch": 2.6840969057327895,
+      "grad_norm": 1.5130645195940433,
+      "learning_rate": 2.6204375385963494e-07,
+      "loss": 0.161,
       "step": 1865
     },
     {
-      "epoch": 0.8970976253298153,
-      "grad_norm": 2.509818105526665,
-      "learning_rate": 8.841365987078955e-06,
-      "loss": 0.8408,
+      "epoch": 2.691292875989446,
+      "grad_norm": 1.080302530956707,
+      "learning_rate": 2.502332059305745e-07,
+      "loss": 0.1752,
       "step": 1870
     },
     {
-      "epoch": 0.899496282082034,
-      "grad_norm": 2.726184599074837,
-      "learning_rate": 8.832414824287441e-06,
-      "loss": 0.885,
+      "epoch": 2.6984888462461023,
+      "grad_norm": 1.306131392662643,
+      "learning_rate": 2.386863918746167e-07,
+      "loss": 0.1968,
       "step": 1875
     },
     {
-      "epoch": 0.9018949388342529,
-      "grad_norm": 3.068184221265625,
-      "learning_rate": 8.823433786178549e-06,
-      "loss": 0.8282,
+      "epoch": 2.7056848165027585,
+      "grad_norm": 1.3461515984684975,
+      "learning_rate": 2.2740412381853223e-07,
+      "loss": 0.183,
       "step": 1880
     },
     {
-      "epoch": 0.9042935955864716,
-      "grad_norm": 5.076920426744664,
-      "learning_rate": 8.814422942763322e-06,
-      "loss": 0.8414,
+      "epoch": 2.7128807867594147,
+      "grad_norm": 1.4486810753503954,
+      "learning_rate": 2.1638719528266835e-07,
+      "loss": 0.1938,
       "step": 1885
     },
     {
-      "epoch": 0.9066922523386903,
-      "grad_norm": 2.0951998113700445,
-      "learning_rate": 8.805382364285152e-06,
-      "loss": 0.9026,
+      "epoch": 2.720076757016071,
+      "grad_norm": 1.035281562927121,
+      "learning_rate": 2.0563638112514047e-07,
+      "loss": 0.1823,
       "step": 1890
     },
     {
-      "epoch": 0.9090909090909091,
-      "grad_norm": 2.5304664436244564,
-      "learning_rate": 8.79631212121922e-06,
-      "loss": 0.8819,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 1.4407581503306328,
+      "learning_rate": 1.9515243748733455e-07,
+      "loss": 0.1648,
       "step": 1895
     },
     {
-      "epoch": 0.9114895658431279,
-      "grad_norm": 2.910233568728739,
-      "learning_rate": 8.787212284271969e-06,
-      "loss": 0.7901,
-      "step": 1900
-    },
-    {
-      "epoch": 0.9114895658431279,
-      "eval_loss": 0.865532398223877,
-      "eval_runtime": 739.6594,
-      "eval_samples_per_second": 10.018,
-      "eval_steps_per_second": 0.627,
+      "epoch": 2.7344686975293837,
+      "grad_norm": 1.2289916037617492,
+      "learning_rate": 1.8493610174072248e-07,
+      "loss": 0.1716,
       "step": 1900
     },
     {
-      "epoch": 0.9138882225953466,
-      "grad_norm": 3.0968516736110385,
-      "learning_rate": 8.778082924380527e-06,
-      "loss": 0.8504,
+      "epoch": 2.74166466778604,
+      "grad_norm": 1.1641895371111006,
+      "learning_rate": 1.7498809243500133e-07,
+      "loss": 0.1659,
       "step": 1905
     },
     {
-      "epoch": 0.9162868793475654,
-      "grad_norm": 2.843647032000645,
-      "learning_rate": 8.76892411271217e-06,
-      "loss": 0.8945,
+      "epoch": 2.748860638042696,
+      "grad_norm": 1.2177853046541605,
+      "learning_rate": 1.6530910924755603e-07,
+      "loss": 0.1905,
       "step": 1910
     },
     {
-      "epoch": 0.9186855360997841,
-      "grad_norm": 2.7639703028981915,
-      "learning_rate": 8.759735920663764e-06,
-      "loss": 0.8549,
+      "epoch": 2.7560566082993523,
+      "grad_norm": 1.373432351937655,
+      "learning_rate": 1.5589983293424802e-07,
+      "loss": 0.1948,
       "step": 1915
     },
     {
-      "epoch": 0.9210841928520028,
-      "grad_norm": 5.035770193454392,
-      "learning_rate": 8.750518419861206e-06,
-      "loss": 0.9528,
+      "epoch": 2.7632525785560085,
+      "grad_norm": 1.2223263222194338,
+      "learning_rate": 1.4676092528153495e-07,
+      "loss": 0.1635,
       "step": 1920
     },
     {
-      "epoch": 0.9234828496042217,
-      "grad_norm": 2.2246289636977004,
-      "learning_rate": 8.74127168215887e-06,
-      "loss": 0.8214,
+      "epoch": 2.7704485488126647,
+      "grad_norm": 1.323423227538223,
+      "learning_rate": 1.378930290599265e-07,
+      "loss": 0.1941,
       "step": 1925
     },
     {
-      "epoch": 0.9258815063564404,
-      "grad_norm": 3.5156544905473828,
-      "learning_rate": 8.731995779639042e-06,
-      "loss": 0.8866,
+      "epoch": 2.7776445190693213,
+      "grad_norm": 1.1747342118230812,
+      "learning_rate": 1.29296767978774e-07,
+      "loss": 0.1556,
       "step": 1930
     },
     {
-      "epoch": 0.9282801631086591,
-      "grad_norm": 3.0156056794105752,
-      "learning_rate": 8.72269078461136e-06,
-      "loss": 0.9573,
+      "epoch": 2.7848404893259775,
+      "grad_norm": 1.4551159106122102,
+      "learning_rate": 1.2097274664240486e-07,
+      "loss": 0.1778,
       "step": 1935
     },
     {
-      "epoch": 0.9306788198608779,
-      "grad_norm": 3.063423244582553,
-      "learning_rate": 8.713356769612254e-06,
-      "loss": 0.8277,
+      "epoch": 2.7920364595826337,
+      "grad_norm": 1.2601624341009796,
+      "learning_rate": 1.1292155050759689e-07,
+      "loss": 0.183,
       "step": 1940
     },
     {
-      "epoch": 0.9330774766130967,
-      "grad_norm": 2.900116122741938,
-      "learning_rate": 8.703993807404371e-06,
-      "loss": 0.8748,
+      "epoch": 2.79923242983929,
+      "grad_norm": 1.2755552581393579,
+      "learning_rate": 1.0514374584240338e-07,
+      "loss": 0.1623,
       "step": 1945
     },
     {
-      "epoch": 0.9354761333653154,
-      "grad_norm": 2.8160457003362196,
-      "learning_rate": 8.694601970976022e-06,
-      "loss": 0.9839,
+      "epoch": 2.806428400095946,
+      "grad_norm": 1.2967219351574655,
+      "learning_rate": 9.763987968632293e-08,
+      "loss": 0.1895,
       "step": 1950
     },
     {
-      "epoch": 0.9378747901175342,
-      "grad_norm": 2.9528327363198343,
-      "learning_rate": 8.685181333540596e-06,
-      "loss": 0.9682,
+      "epoch": 2.8136243703526027,
+      "grad_norm": 1.5216756650995487,
+      "learning_rate": 9.04104798118257e-08,
+      "loss": 0.18,
       "step": 1955
     },
     {
-      "epoch": 0.9402734468697529,
-      "grad_norm": 2.416154745371056,
-      "learning_rate": 8.675731968536004e-06,
-      "loss": 0.8996,
+      "epoch": 2.820820340609259,
+      "grad_norm": 1.2463724400999108,
+      "learning_rate": 8.345605468723427e-08,
+      "loss": 0.1855,
       "step": 1960
     },
     {
-      "epoch": 0.9426721036219717,
-      "grad_norm": 2.6496183375696902,
-      "learning_rate": 8.666253949624097e-06,
-      "loss": 0.8659,
+      "epoch": 2.828016310865915,
+      "grad_norm": 1.3612982361619894,
+      "learning_rate": 7.677709344095883e-08,
+      "loss": 0.1971,
       "step": 1965
     },
     {
-      "epoch": 0.9450707603741905,
-      "grad_norm": 2.5880937245783033,
-      "learning_rate": 8.656747350690102e-06,
-      "loss": 0.8052,
+      "epoch": 2.8352122811225713,
+      "grad_norm": 1.1485240309542117,
+      "learning_rate": 7.037406582709815e-08,
+      "loss": 0.1673,
       "step": 1970
     },
     {
-      "epoch": 0.9474694171264092,
-      "grad_norm": 3.7751706043240105,
-      "learning_rate": 8.647212245842035e-06,
-      "loss": 0.9261,
+      "epoch": 2.8424082513792275,
+      "grad_norm": 1.129849173373603,
+      "learning_rate": 6.424742219239698e-08,
+      "loss": 0.1688,
       "step": 1975
     },
     {
-      "epoch": 0.9498680738786279,
-      "grad_norm": 2.886044266318896,
-      "learning_rate": 8.637648709410125e-06,
-      "loss": 0.8753,
+      "epoch": 2.849604221635884,
+      "grad_norm": 1.271858319859489,
+      "learning_rate": 5.839759344457462e-08,
+      "loss": 0.1864,
       "step": 1980
     },
     {
-      "epoch": 0.9522667306308468,
-      "grad_norm": 2.5010200818615584,
-      "learning_rate": 8.628056815946243e-06,
-      "loss": 0.7647,
+      "epoch": 2.8568001918925403,
+      "grad_norm": 1.3695271997638596,
+      "learning_rate": 5.282499102201532e-08,
+      "loss": 0.182,
       "step": 1985
     },
     {
-      "epoch": 0.9546653873830655,
-      "grad_norm": 2.4230783414425443,
-      "learning_rate": 8.618436640223311e-06,
-      "loss": 0.9169,
+      "epoch": 2.8639961621491965,
+      "grad_norm": 1.2817216551660306,
+      "learning_rate": 4.753000686483189e-08,
+      "loss": 0.191,
       "step": 1990
     },
     {
-      "epoch": 0.9570640441352842,
-      "grad_norm": 5.70711797643376,
-      "learning_rate": 8.608788257234726e-06,
-      "loss": 0.9718,
+      "epoch": 2.8711921324058527,
+      "grad_norm": 1.2963119208915668,
+      "learning_rate": 4.2513013387298846e-08,
+      "loss": 0.1877,
       "step": 1995
     },
     {
-      "epoch": 0.959462700887503,
-      "grad_norm": 3.2409263975929146,
-      "learning_rate": 8.599111742193772e-06,
-      "loss": 0.8306,
-      "step": 2000
-    },
-    {
-      "epoch": 0.959462700887503,
-      "eval_loss": 0.8639265894889832,
-      "eval_runtime": 738.2244,
-      "eval_samples_per_second": 10.038,
-      "eval_steps_per_second": 0.629,
+      "epoch": 2.878388102662509,
+      "grad_norm": 1.290101189771667,
+      "learning_rate": 3.7774363451658744e-08,
+      "loss": 0.1796,
       "step": 2000
     },
     {
-      "epoch": 0.9618613576397218,
-      "grad_norm": 2.60287555153669,
-      "learning_rate": 8.589407170533035e-06,
-      "loss": 0.9463,
+      "epoch": 2.8855840729191655,
+      "grad_norm": 1.4819813983719878,
+      "learning_rate": 3.331439034330552e-08,
+      "loss": 0.1763,
       "step": 2005
     },
     {
-      "epoch": 0.9642600143919405,
-      "grad_norm": 4.0419154556516235,
-      "learning_rate": 8.579674617903812e-06,
-      "loss": 0.8895,
+      "epoch": 2.8927800431758213,
+      "grad_norm": 1.4019557468116053,
+      "learning_rate": 2.913340774734152e-08,
+      "loss": 0.1708,
       "step": 2010
     },
     {
-      "epoch": 0.9666586711441593,
-      "grad_norm": 3.1939130510098934,
-      "learning_rate": 8.569914160175527e-06,
-      "loss": 0.9188,
+      "epoch": 2.899976013432478,
+      "grad_norm": 1.2339439577357936,
+      "learning_rate": 2.5231709726516005e-08,
+      "loss": 0.1789,
       "step": 2015
     },
     {
-      "epoch": 0.969057327896378,
-      "grad_norm": 3.1878190598190304,
-      "learning_rate": 8.560125873435132e-06,
-      "loss": 0.8027,
+      "epoch": 2.907171983689134,
+      "grad_norm": 1.2294457543053794,
+      "learning_rate": 2.1609570700543478e-08,
+      "loss": 0.1575,
       "step": 2020
     },
     {
-      "epoch": 0.9714559846485968,
-      "grad_norm": 2.1973702930768937,
-      "learning_rate": 8.550309833986525e-06,
-      "loss": 0.7583,
+      "epoch": 2.9143679539457903,
+      "grad_norm": 1.6910934029285385,
+      "learning_rate": 1.826724542680047e-08,
+      "loss": 0.1853,
       "step": 2025
     },
     {
-      "epoch": 0.9738546414008156,
-      "grad_norm": 3.416693480749312,
-      "learning_rate": 8.540466118349938e-06,
-      "loss": 0.873,
+      "epoch": 2.9215639242024465,
+      "grad_norm": 1.5951062337979125,
+      "learning_rate": 1.5204968982410527e-08,
+      "loss": 0.1994,
       "step": 2030
     },
     {
-      "epoch": 0.9762532981530343,
-      "grad_norm": 4.297709957092606,
-      "learning_rate": 8.530594803261362e-06,
-      "loss": 0.8679,
+      "epoch": 2.9287598944591027,
+      "grad_norm": 1.244634963059678,
+      "learning_rate": 1.2422956747708546e-08,
+      "loss": 0.1792,
       "step": 2035
     },
     {
-      "epoch": 0.978651954905253,
-      "grad_norm": 2.7317028436026343,
-      "learning_rate": 8.520695965671928e-06,
-      "loss": 0.8326,
+      "epoch": 2.9359558647157593,
+      "grad_norm": 1.3731755256138354,
+      "learning_rate": 9.92140439109157e-09,
+      "loss": 0.1855,
       "step": 2040
     },
     {
-      "epoch": 0.9810506116574718,
-      "grad_norm": 3.819976358159541,
-      "learning_rate": 8.510769682747324e-06,
-      "loss": 0.992,
+      "epoch": 2.9431518349724155,
+      "grad_norm": 1.487158498377204,
+      "learning_rate": 7.700487855260007e-09,
+      "loss": 0.1713,
       "step": 2045
     },
     {
-      "epoch": 0.9834492684096906,
-      "grad_norm": 2.4139195174663883,
-      "learning_rate": 8.50081603186718e-06,
-      "loss": 0.7586,
+      "epoch": 2.9503478052290717,
+      "grad_norm": 1.195702473189198,
+      "learning_rate": 5.760363344839536e-09,
+      "loss": 0.1756,
       "step": 2050
     },
     {
-      "epoch": 0.9858479251619093,
-      "grad_norm": 2.438887753565038,
-      "learning_rate": 8.490835090624476e-06,
-      "loss": 0.9304,
+      "epoch": 2.957543775485728,
+      "grad_norm": 1.2219249635030713,
+      "learning_rate": 4.101167315396559e-09,
+      "loss": 0.1705,
       "step": 2055
     },
     {
-      "epoch": 0.9882465819141281,
-      "grad_norm": 3.320035848122713,
-      "learning_rate": 8.480826936824933e-06,
-      "loss": 0.8927,
+      "epoch": 2.964739745742384,
+      "grad_norm": 1.1937579061988086,
+      "learning_rate": 2.7230164638401e-09,
+      "loss": 0.1669,
       "step": 2060
     },
     {
-      "epoch": 0.9906452386663468,
-      "grad_norm": 2.0982737537627165,
-      "learning_rate": 8.470791648486396e-06,
-      "loss": 0.9131,
+      "epoch": 2.9719357159990407,
+      "grad_norm": 1.0911166566357549,
+      "learning_rate": 1.626007720214595e-09,
+      "loss": 0.1556,
       "step": 2065
     },
     {
-      "epoch": 0.9930438954185656,
-      "grad_norm": 4.233832052942202,
-      "learning_rate": 8.460729303838246e-06,
-      "loss": 0.8421,
+      "epoch": 2.979131686255697,
+      "grad_norm": 1.155309788494306,
+      "learning_rate": 8.102182408822322e-10,
+      "loss": 0.1572,
       "step": 2070
     },
     {
-      "epoch": 0.9954425521707844,
-      "grad_norm": 3.0600316774502523,
-      "learning_rate": 8.450639981320777e-06,
-      "loss": 0.8101,
+      "epoch": 2.986327656512353,
+      "grad_norm": 1.1809384871247852,
+      "learning_rate": 2.7570540309618253e-10,
+      "loss": 0.1651,
       "step": 2075
     },
     {
-      "epoch": 0.9978412089230031,
-      "grad_norm": 3.229891184239876,
-      "learning_rate": 8.440523759584583e-06,
-      "loss": 0.8834,
+      "epoch": 2.9935236267690093,
+      "grad_norm": 2.41813981269854,
+      "learning_rate": 2.2506800965604867e-11,
+      "loss": 0.1452,
       "step": 2080
     },
     {
-      "epoch": 1.0002398656752218,
-      "grad_norm": 4.476281436495972,
-      "learning_rate": 8.43038071748995e-06,
-      "loss": 0.9229,
-      "step": 2085
-    },
-    {
-      "epoch": 1.0026385224274406,
-      "grad_norm": 2.5593143307820743,
-      "learning_rate": 8.420210934106245e-06,
-      "loss": 0.4714,
-      "step": 2090
-    },
-    {
-      "epoch": 1.0050371791796593,
-      "grad_norm": 2.213171577306985,
-      "learning_rate": 8.41001448871129e-06,
-      "loss": 0.5274,
-      "step": 2095
-    },
-    {
-      "epoch": 1.0074358359318782,
-      "grad_norm": 2.0985588840636282,
-      "learning_rate": 8.399791460790752e-06,
-      "loss": 0.4803,
-      "step": 2100
-    },
-    {
-      "epoch": 1.009834492684097,
-      "grad_norm": 2.4553936796867637,
-      "learning_rate": 8.389541930037516e-06,
-      "loss": 0.4868,
-      "step": 2105
-    },
-    {
-      "epoch": 1.0122331494363157,
-      "grad_norm": 2.228861585976993,
-      "learning_rate": 8.379265976351074e-06,
-      "loss": 0.5068,
-      "step": 2110
-    },
-    {
-      "epoch": 1.0146318061885344,
-      "grad_norm": 2.890711742500512,
-      "learning_rate": 8.36896367983689e-06,
-      "loss": 0.4482,
-      "step": 2115
-    },
-    {
-      "epoch": 1.0170304629407532,
-      "grad_norm": 2.227985338155966,
-      "learning_rate": 8.358635120805783e-06,
-      "loss": 0.4309,
-      "step": 2120
-    },
-    {
-      "epoch": 1.019429119692972,
-      "grad_norm": 1.7753949748634992,
-      "learning_rate": 8.348280379773303e-06,
-      "loss": 0.3702,
-      "step": 2125
-    },
-    {
-      "epoch": 1.0218277764451906,
-      "grad_norm": 2.39040145648445,
-      "learning_rate": 8.337899537459098e-06,
-      "loss": 0.4513,
-      "step": 2130
-    },
-    {
-      "epoch": 1.0242264331974094,
-      "grad_norm": 2.1792944421413925,
-      "learning_rate": 8.327492674786286e-06,
-      "loss": 0.4522,
-      "step": 2135
-    },
-    {
-      "epoch": 1.0266250899496283,
-      "grad_norm": 2.704876180547788,
-      "learning_rate": 8.317059872880823e-06,
-      "loss": 0.4446,
-      "step": 2140
-    },
-    {
-      "epoch": 1.029023746701847,
-      "grad_norm": 2.5695079674783337,
-      "learning_rate": 8.30660121307088e-06,
-      "loss": 0.4808,
-      "step": 2145
-    },
-    {
-      "epoch": 1.0314224034540658,
-      "grad_norm": 2.4711435136084465,
-      "learning_rate": 8.296116776886192e-06,
-      "loss": 0.3853,
-      "step": 2150
-    },
-    {
-      "epoch": 1.0338210602062845,
-      "grad_norm": 2.9044848081202126,
-      "learning_rate": 8.285606646057435e-06,
-      "loss": 0.443,
-      "step": 2155
-    },
-    {
-      "epoch": 1.0362197169585032,
-      "grad_norm": 2.1130488032920303,
-      "learning_rate": 8.275070902515593e-06,
-      "loss": 0.4353,
-      "step": 2160
-    },
-    {
-      "epoch": 1.038618373710722,
-      "grad_norm": 1.6913616556429052,
-      "learning_rate": 8.264509628391299e-06,
-      "loss": 0.4457,
-      "step": 2165
-    },
-    {
-      "epoch": 1.0410170304629407,
-      "grad_norm": 2.6512115817567614,
-      "learning_rate": 8.25392290601422e-06,
-      "loss": 0.5117,
-      "step": 2170
-    },
-    {
-      "epoch": 1.0434156872151594,
-      "grad_norm": 3.956549519902379,
-      "learning_rate": 8.243310817912395e-06,
-      "loss": 0.4705,
-      "step": 2175
-    },
-    {
-      "epoch": 1.0458143439673782,
-      "grad_norm": 2.7210983741481565,
-      "learning_rate": 8.232673446811602e-06,
-      "loss": 0.5065,
-      "step": 2180
-    },
-    {
-      "epoch": 1.048213000719597,
-      "grad_norm": 2.5059382822841187,
-      "learning_rate": 8.222010875634713e-06,
-      "loss": 0.5251,
-      "step": 2185
-    },
-    {
-      "epoch": 1.0506116574718158,
-      "grad_norm": 2.262650183186798,
-      "learning_rate": 8.211323187501046e-06,
-      "loss": 0.4201,
-      "step": 2190
-    },
-    {
-      "epoch": 1.0530103142240346,
-      "grad_norm": 2.1701091490205453,
-      "learning_rate": 8.200610465725709e-06,
-      "loss": 0.432,
-      "step": 2195
-    },
-    {
-      "epoch": 1.0554089709762533,
-      "grad_norm": 3.203732105955941,
-      "learning_rate": 8.189872793818966e-06,
-      "loss": 0.4875,
-      "step": 2200
-    },
-    {
-      "epoch": 1.057807627728472,
-      "grad_norm": 2.129191102623214,
-      "learning_rate": 8.179110255485576e-06,
-      "loss": 0.4272,
-      "step": 2205
-    },
-    {
-      "epoch": 1.0602062844806908,
-      "grad_norm": 2.0600885302520506,
-      "learning_rate": 8.16832293462414e-06,
-      "loss": 0.4211,
-      "step": 2210
-    },
-    {
-      "epoch": 1.0626049412329095,
-      "grad_norm": 1.7498086907025407,
-      "learning_rate": 8.157510915326453e-06,
-      "loss": 0.4825,
-      "step": 2215
-    },
-    {
-      "epoch": 1.0650035979851282,
-      "grad_norm": 2.7615254217770895,
-      "learning_rate": 8.14667428187684e-06,
-      "loss": 0.4528,
-      "step": 2220
-    },
-    {
-      "epoch": 1.0674022547373472,
-      "grad_norm": 2.6704428087666607,
-      "learning_rate": 8.135813118751508e-06,
-      "loss": 0.502,
-      "step": 2225
-    },
-    {
-      "epoch": 1.069800911489566,
-      "grad_norm": 2.5880979319847977,
-      "learning_rate": 8.124927510617886e-06,
-      "loss": 0.4126,
-      "step": 2230
-    },
-    {
-      "epoch": 1.0721995682417846,
-      "grad_norm": 2.62434654551889,
-      "learning_rate": 8.114017542333955e-06,
-      "loss": 0.485,
-      "step": 2235
-    },
-    {
-      "epoch": 1.0745982249940034,
-      "grad_norm": 1.777812099558463,
-      "learning_rate": 8.103083298947594e-06,
-      "loss": 0.4729,
-      "step": 2240
-    },
-    {
-      "epoch": 1.076996881746222,
-      "grad_norm": 2.09259412996849,
-      "learning_rate": 8.092124865695928e-06,
-      "loss": 0.4283,
-      "step": 2245
-    },
-    {
-      "epoch": 1.0793955384984408,
-      "grad_norm": 2.38765858433518,
-      "learning_rate": 8.081142328004638e-06,
-      "loss": 0.523,
-      "step": 2250
-    },
-    {
-      "epoch": 1.0817941952506596,
-      "grad_norm": 3.237139015501116,
-      "learning_rate": 8.070135771487317e-06,
-      "loss": 0.519,
-      "step": 2255
-    },
-    {
-      "epoch": 1.0841928520028783,
-      "grad_norm": 2.4130326528367334,
-      "learning_rate": 8.059105281944799e-06,
-      "loss": 0.5109,
-      "step": 2260
-    },
-    {
-      "epoch": 1.0865915087550972,
-      "grad_norm": 2.4038606293430758,
-      "learning_rate": 8.048050945364477e-06,
-      "loss": 0.3813,
-      "step": 2265
-    },
-    {
-      "epoch": 1.088990165507316,
-      "grad_norm": 2.6627926967455187,
-      "learning_rate": 8.036972847919645e-06,
-      "loss": 0.4671,
-      "step": 2270
-    },
-    {
-      "epoch": 1.0913888222595347,
-      "grad_norm": 2.6725216982707547,
-      "learning_rate": 8.025871075968828e-06,
-      "loss": 0.4293,
-      "step": 2275
-    },
-    {
-      "epoch": 1.0937874790117534,
-      "grad_norm": 1.8642256511396682,
-      "learning_rate": 8.014745716055095e-06,
-      "loss": 0.3925,
-      "step": 2280
-    },
-    {
-      "epoch": 1.0961861357639722,
-      "grad_norm": 2.6136199232940176,
-      "learning_rate": 8.003596854905405e-06,
-      "loss": 0.4031,
-      "step": 2285
-    },
-    {
-      "epoch": 1.098584792516191,
-      "grad_norm": 1.9487935903728233,
-      "learning_rate": 7.99242457942991e-06,
-      "loss": 0.5139,
-      "step": 2290
-    },
-    {
-      "epoch": 1.1009834492684096,
-      "grad_norm": 2.259840081609626,
-      "learning_rate": 7.981228976721287e-06,
-      "loss": 0.4398,
-      "step": 2295
-    },
-    {
-      "epoch": 1.1033821060206284,
-      "grad_norm": 3.5633206330617555,
-      "learning_rate": 7.970010134054064e-06,
-      "loss": 0.4823,
-      "step": 2300
-    },
-    {
-      "epoch": 1.1057807627728473,
-      "grad_norm": 2.6015299960981144,
-      "learning_rate": 7.95876813888393e-06,
-      "loss": 0.4304,
-      "step": 2305
-    },
-    {
-      "epoch": 1.108179419525066,
-      "grad_norm": 2.7206316510278694,
-      "learning_rate": 7.947503078847061e-06,
-      "loss": 0.4586,
-      "step": 2310
-    },
-    {
-      "epoch": 1.1105780762772848,
-      "grad_norm": 2.794098011567633,
-      "learning_rate": 7.93621504175943e-06,
-      "loss": 0.471,
-      "step": 2315
-    },
-    {
-      "epoch": 1.1129767330295035,
-      "grad_norm": 2.7240226313790226,
-      "learning_rate": 7.92490411561613e-06,
-      "loss": 0.497,
-      "step": 2320
-    },
-    {
-      "epoch": 1.1153753897817222,
-      "grad_norm": 3.5098947402019443,
-      "learning_rate": 7.91357038859068e-06,
-      "loss": 0.4853,
-      "step": 2325
-    },
-    {
-      "epoch": 1.117774046533941,
-      "grad_norm": 2.0290357336252582,
-      "learning_rate": 7.902213949034345e-06,
-      "loss": 0.421,
-      "step": 2330
-    },
-    {
-      "epoch": 1.1201727032861597,
-      "grad_norm": 3.0022701990136302,
-      "learning_rate": 7.89083488547544e-06,
-      "loss": 0.4909,
-      "step": 2335
-    },
-    {
-      "epoch": 1.1225713600383784,
-      "grad_norm": 2.065209472044516,
-      "learning_rate": 7.87943328661865e-06,
-      "loss": 0.4806,
-      "step": 2340
-    },
-    {
-      "epoch": 1.1249700167905972,
-      "grad_norm": 2.8034071820688142,
-      "learning_rate": 7.868009241344323e-06,
-      "loss": 0.4474,
-      "step": 2345
-    },
-    {
-      "epoch": 1.1273686735428161,
-      "grad_norm": 2.580536233115346,
-      "learning_rate": 7.856562838707799e-06,
-      "loss": 0.4731,
-      "step": 2350
-    },
-    {
-      "epoch": 1.1297673302950348,
-      "grad_norm": 2.0441292412793426,
-      "learning_rate": 7.845094167938689e-06,
-      "loss": 0.4212,
-      "step": 2355
-    },
-    {
-      "epoch": 1.1321659870472536,
-      "grad_norm": 1.9698046885266003,
-      "learning_rate": 7.833603318440207e-06,
-      "loss": 0.4269,
-      "step": 2360
-    },
-    {
-      "epoch": 1.1345646437994723,
-      "grad_norm": 2.4020632540305336,
-      "learning_rate": 7.822090379788449e-06,
-      "loss": 0.4117,
-      "step": 2365
-    },
-    {
-      "epoch": 1.136963300551691,
-      "grad_norm": 2.3063432967429662,
-      "learning_rate": 7.810555441731715e-06,
-      "loss": 0.4061,
-      "step": 2370
-    },
-    {
-      "epoch": 1.1393619573039098,
-      "grad_norm": 2.6738090741389997,
-      "learning_rate": 7.79899859418979e-06,
-      "loss": 0.4746,
-      "step": 2375
-    },
-    {
-      "epoch": 1.1417606140561285,
-      "grad_norm": 2.22066686785184,
-      "learning_rate": 7.787419927253259e-06,
-      "loss": 0.4648,
-      "step": 2380
-    },
-    {
-      "epoch": 1.1441592708083472,
-      "grad_norm": 2.1609157489317,
-      "learning_rate": 7.775819531182797e-06,
-      "loss": 0.4836,
-      "step": 2385
-    },
-    {
-      "epoch": 1.1465579275605662,
-      "grad_norm": 2.825274671556735,
-      "learning_rate": 7.764197496408468e-06,
-      "loss": 0.3937,
-      "step": 2390
-    },
-    {
-      "epoch": 1.148956584312785,
-      "grad_norm": 1.8923634073444822,
-      "learning_rate": 7.752553913529019e-06,
-      "loss": 0.4561,
-      "step": 2395
-    },
-    {
-      "epoch": 1.1513552410650036,
-      "grad_norm": 2.4561463478420964,
-      "learning_rate": 7.74088887331117e-06,
-      "loss": 0.3996,
-      "step": 2400
-    },
-    {
-      "epoch": 1.1537538978172224,
-      "grad_norm": 2.3237076372641643,
-      "learning_rate": 7.729202466688914e-06,
-      "loss": 0.5374,
-      "step": 2405
-    },
-    {
-      "epoch": 1.156152554569441,
-      "grad_norm": 1.7336149449293585,
-      "learning_rate": 7.717494784762804e-06,
-      "loss": 0.4204,
-      "step": 2410
-    },
-    {
-      "epoch": 1.1585512113216598,
-      "grad_norm": 2.6072676927620244,
-      "learning_rate": 7.705765918799244e-06,
-      "loss": 0.3674,
-      "step": 2415
-    },
-    {
-      "epoch": 1.1609498680738786,
-      "grad_norm": 2.031726825542831,
-      "learning_rate": 7.69401596022977e-06,
-      "loss": 0.4854,
-      "step": 2420
-    },
-    {
-      "epoch": 1.1633485248260973,
-      "grad_norm": 3.044907096929622,
-      "learning_rate": 7.682245000650354e-06,
-      "loss": 0.4676,
-      "step": 2425
-    },
-    {
-      "epoch": 1.165747181578316,
-      "grad_norm": 1.7787272779987482,
-      "learning_rate": 7.670453131820672e-06,
-      "loss": 0.4123,
-      "step": 2430
-    },
-    {
-      "epoch": 1.168145838330535,
-      "grad_norm": 2.1958069851919646,
-      "learning_rate": 7.6586404456634e-06,
-      "loss": 0.4779,
-      "step": 2435
-    },
-    {
-      "epoch": 1.1705444950827537,
-      "grad_norm": 2.2233956624705424,
-      "learning_rate": 7.646807034263493e-06,
-      "loss": 0.4659,
-      "step": 2440
-    },
-    {
-      "epoch": 1.1729431518349724,
-      "grad_norm": 1.8414644463809167,
-      "learning_rate": 7.63495298986747e-06,
-      "loss": 0.4409,
-      "step": 2445
-    },
-    {
-      "epoch": 1.1753418085871912,
-      "grad_norm": 3.3591717272248878,
-      "learning_rate": 7.623078404882689e-06,
-      "loss": 0.4125,
-      "step": 2450
-    },
-    {
-      "epoch": 1.17774046533941,
-      "grad_norm": 2.2891639443640366,
-      "learning_rate": 7.611183371876636e-06,
-      "loss": 0.4907,
-      "step": 2455
-    },
-    {
-      "epoch": 1.1801391220916286,
-      "grad_norm": 2.099380667162895,
-      "learning_rate": 7.599267983576193e-06,
-      "loss": 0.4529,
-      "step": 2460
-    },
-    {
-      "epoch": 1.1825377788438474,
-      "grad_norm": 2.542301285127525,
-      "learning_rate": 7.587332332866923e-06,
-      "loss": 0.5088,
-      "step": 2465
-    },
-    {
-      "epoch": 1.1849364355960663,
-      "grad_norm": 2.2714860789638567,
-      "learning_rate": 7.5753765127923475e-06,
-      "loss": 0.4983,
-      "step": 2470
-    },
-    {
-      "epoch": 1.187335092348285,
-      "grad_norm": 2.949953798544259,
-      "learning_rate": 7.563400616553205e-06,
-      "loss": 0.4625,
-      "step": 2475
-    },
-    {
-      "epoch": 1.1897337491005038,
-      "grad_norm": 3.3907825639923255,
-      "learning_rate": 7.551404737506747e-06,
-      "loss": 0.5057,
-      "step": 2480
-    },
-    {
-      "epoch": 1.1921324058527225,
-      "grad_norm": 3.379571230136842,
-      "learning_rate": 7.539388969165996e-06,
-      "loss": 0.4563,
-      "step": 2485
-    },
-    {
-      "epoch": 1.1945310626049412,
-      "grad_norm": 2.278909491259441,
-      "learning_rate": 7.52735340519902e-06,
-      "loss": 0.4486,
-      "step": 2490
-    },
-    {
-      "epoch": 1.19692971935716,
-      "grad_norm": 2.009431291635904,
-      "learning_rate": 7.515298139428203e-06,
-      "loss": 0.416,
-      "step": 2495
-    },
-    {
-      "epoch": 1.1993283761093787,
-      "grad_norm": 2.326722204481952,
-      "learning_rate": 7.503223265829511e-06,
-      "loss": 0.4517,
-      "step": 2500
-    },
-    {
-      "epoch": 1.2017270328615974,
-      "grad_norm": 3.1329815616704404,
-      "learning_rate": 7.491128878531764e-06,
-      "loss": 0.5385,
-      "step": 2505
-    },
-    {
-      "epoch": 1.2041256896138162,
-      "grad_norm": 3.013574468212731,
-      "learning_rate": 7.479015071815899e-06,
-      "loss": 0.4193,
-      "step": 2510
-    },
-    {
-      "epoch": 1.2065243463660351,
-      "grad_norm": 2.1700351101261965,
-      "learning_rate": 7.466881940114232e-06,
-      "loss": 0.4916,
-      "step": 2515
-    },
-    {
-      "epoch": 1.2089230031182538,
-      "grad_norm": 2.6970784604552,
-      "learning_rate": 7.4547295780097336e-06,
-      "loss": 0.5202,
-      "step": 2520
-    },
-    {
-      "epoch": 1.2113216598704726,
-      "grad_norm": 1.9662106419481995,
-      "learning_rate": 7.442558080235276e-06,
-      "loss": 0.393,
-      "step": 2525
-    },
-    {
-      "epoch": 1.2137203166226913,
-      "grad_norm": 2.1077726186526786,
-      "learning_rate": 7.4303675416729075e-06,
-      "loss": 0.3749,
-      "step": 2530
-    },
-    {
-      "epoch": 1.21611897337491,
-      "grad_norm": 2.320554500963109,
-      "learning_rate": 7.418158057353099e-06,
-      "loss": 0.4419,
-      "step": 2535
-    },
-    {
-      "epoch": 1.2185176301271288,
-      "grad_norm": 2.895651302536835,
-      "learning_rate": 7.405929722454026e-06,
-      "loss": 0.4118,
-      "step": 2540
-    },
-    {
-      "epoch": 1.2209162868793475,
-      "grad_norm": 1.8616421523355307,
-      "learning_rate": 7.393682632300798e-06,
-      "loss": 0.4295,
-      "step": 2545
-    },
-    {
-      "epoch": 1.2233149436315662,
-      "grad_norm": 3.737431743510741,
-      "learning_rate": 7.38141688236474e-06,
-      "loss": 0.441,
-      "step": 2550
-    },
-    {
-      "epoch": 1.2257136003837852,
-      "grad_norm": 1.6958443120456206,
-      "learning_rate": 7.36913256826263e-06,
-      "loss": 0.4359,
-      "step": 2555
-    },
-    {
-      "epoch": 1.228112257136004,
-      "grad_norm": 2.313565281454391,
-      "learning_rate": 7.356829785755967e-06,
-      "loss": 0.4512,
-      "step": 2560
-    },
-    {
-      "epoch": 1.2305109138882226,
-      "grad_norm": 4.040641972263277,
-      "learning_rate": 7.344508630750219e-06,
-      "loss": 0.4191,
-      "step": 2565
-    },
-    {
-      "epoch": 1.2329095706404414,
-      "grad_norm": 2.4002855017322333,
-      "learning_rate": 7.3321691992940725e-06,
-      "loss": 0.5318,
-      "step": 2570
-    },
-    {
-      "epoch": 1.23530822739266,
-      "grad_norm": 2.30125067592132,
-      "learning_rate": 7.319811587578689e-06,
-      "loss": 0.4533,
-      "step": 2575
-    },
-    {
-      "epoch": 1.2377068841448788,
-      "grad_norm": 2.0772341503543466,
-      "learning_rate": 7.3074358919369535e-06,
-      "loss": 0.3884,
-      "step": 2580
-    },
-    {
-      "epoch": 1.2401055408970976,
-      "grad_norm": 2.0581617857738297,
-      "learning_rate": 7.295042208842718e-06,
-      "loss": 0.4596,
-      "step": 2585
-    },
-    {
-      "epoch": 1.2425041976493163,
-      "grad_norm": 2.044753740126548,
-      "learning_rate": 7.282630634910062e-06,
-      "loss": 0.3761,
-      "step": 2590
-    },
-    {
-      "epoch": 1.244902854401535,
-      "grad_norm": 1.8350309498563098,
-      "learning_rate": 7.270201266892528e-06,
-      "loss": 0.5365,
-      "step": 2595
-    },
-    {
-      "epoch": 1.247301511153754,
-      "grad_norm": 2.861599377026653,
-      "learning_rate": 7.2577542016823685e-06,
-      "loss": 0.4203,
-      "step": 2600
-    },
-    {
-      "epoch": 1.2497001679059727,
-      "grad_norm": 2.265775605763178,
-      "learning_rate": 7.2452895363097964e-06,
-      "loss": 0.4422,
-      "step": 2605
-    },
-    {
-      "epoch": 1.2520988246581914,
-      "grad_norm": 1.8956362962363034,
-      "learning_rate": 7.232807367942225e-06,
-      "loss": 0.4711,
-      "step": 2610
-    },
-    {
-      "epoch": 1.2544974814104102,
-      "grad_norm": 2.3672251257694237,
-      "learning_rate": 7.220307793883512e-06,
-      "loss": 0.483,
-      "step": 2615
-    },
-    {
-      "epoch": 1.256896138162629,
-      "grad_norm": 2.2661312851841737,
-      "learning_rate": 7.207790911573198e-06,
-      "loss": 0.4678,
-      "step": 2620
-    },
-    {
-      "epoch": 1.2592947949148476,
-      "grad_norm": 1.7750731193059077,
-      "learning_rate": 7.195256818585747e-06,
-      "loss": 0.4545,
-      "step": 2625
-    },
-    {
-      "epoch": 1.2616934516670664,
-      "grad_norm": 3.7693900871996577,
-      "learning_rate": 7.182705612629793e-06,
-      "loss": 0.4417,
-      "step": 2630
-    },
-    {
-      "epoch": 1.2640921084192853,
-      "grad_norm": 2.4900331079659117,
-      "learning_rate": 7.17013739154737e-06,
-      "loss": 0.4084,
-      "step": 2635
-    },
-    {
-      "epoch": 1.266490765171504,
-      "grad_norm": 2.7575013032291817,
-      "learning_rate": 7.157552253313148e-06,
-      "loss": 0.4247,
-      "step": 2640
-    },
-    {
-      "epoch": 1.2688894219237228,
-      "grad_norm": 4.7542236375020375,
-      "learning_rate": 7.144950296033682e-06,
-      "loss": 0.4535,
-      "step": 2645
-    },
-    {
-      "epoch": 1.2712880786759415,
-      "grad_norm": 2.270681456582455,
-      "learning_rate": 7.132331617946634e-06,
-      "loss": 0.4005,
-      "step": 2650
-    },
-    {
-      "epoch": 1.2736867354281602,
-      "grad_norm": 2.5862028929190712,
-      "learning_rate": 7.1196963174200105e-06,
-      "loss": 0.4171,
-      "step": 2655
-    },
-    {
-      "epoch": 1.276085392180379,
-      "grad_norm": 2.1393183657416763,
-      "learning_rate": 7.107044492951398e-06,
-      "loss": 0.4313,
-      "step": 2660
-    },
-    {
-      "epoch": 1.2784840489325977,
-      "grad_norm": 1.9123346331650293,
-      "learning_rate": 7.094376243167197e-06,
-      "loss": 0.4364,
-      "step": 2665
-    },
-    {
-      "epoch": 1.2808827056848164,
-      "grad_norm": 2.8631827616417556,
-      "learning_rate": 7.0816916668218485e-06,
-      "loss": 0.5317,
-      "step": 2670
-    },
-    {
-      "epoch": 1.2832813624370352,
-      "grad_norm": 2.5356423623576805,
-      "learning_rate": 7.068990862797064e-06,
-      "loss": 0.4507,
-      "step": 2675
-    },
-    {
-      "epoch": 1.285680019189254,
-      "grad_norm": 2.052684995040051,
-      "learning_rate": 7.056273930101061e-06,
-      "loss": 0.4467,
-      "step": 2680
-    },
-    {
-      "epoch": 1.2880786759414729,
-      "grad_norm": 2.914920449781333,
-      "learning_rate": 7.043540967867782e-06,
-      "loss": 0.4526,
-      "step": 2685
-    },
-    {
-      "epoch": 1.2904773326936916,
-      "grad_norm": 2.2350378228168992,
-      "learning_rate": 7.03079207535613e-06,
-      "loss": 0.4565,
-      "step": 2690
-    },
-    {
-      "epoch": 1.2928759894459103,
-      "grad_norm": 2.9232644902019236,
-      "learning_rate": 7.01802735194919e-06,
-      "loss": 0.4327,
-      "step": 2695
-    },
-    {
-      "epoch": 1.295274646198129,
-      "grad_norm": 2.152229747789746,
-      "learning_rate": 7.0052468971534574e-06,
-      "loss": 0.475,
-      "step": 2700
-    },
-    {
-      "epoch": 1.2976733029503478,
-      "grad_norm": 1.576250476623198,
-      "learning_rate": 6.992450810598059e-06,
-      "loss": 0.3645,
-      "step": 2705
-    },
-    {
-      "epoch": 1.3000719597025665,
-      "grad_norm": 2.47081607319409,
-      "learning_rate": 6.979639192033977e-06,
-      "loss": 0.4351,
-      "step": 2710
-    },
-    {
-      "epoch": 1.3024706164547855,
-      "grad_norm": 2.9694947631286364,
-      "learning_rate": 6.966812141333273e-06,
-      "loss": 0.4233,
-      "step": 2715
-    },
-    {
-      "epoch": 1.3048692732070042,
-      "grad_norm": 2.657085372294027,
-      "learning_rate": 6.9539697584883116e-06,
-      "loss": 0.547,
-      "step": 2720
-    },
-    {
-      "epoch": 1.307267929959223,
-      "grad_norm": 1.9832533039303075,
-      "learning_rate": 6.941112143610973e-06,
-      "loss": 0.4364,
-      "step": 2725
-    },
-    {
-      "epoch": 1.3096665867114416,
-      "grad_norm": 2.523218142090448,
-      "learning_rate": 6.928239396931882e-06,
-      "loss": 0.4612,
-      "step": 2730
-    },
-    {
-      "epoch": 1.3120652434636604,
-      "grad_norm": 2.406726447436041,
-      "learning_rate": 6.915351618799618e-06,
-      "loss": 0.4071,
-      "step": 2735
-    },
-    {
-      "epoch": 1.314463900215879,
-      "grad_norm": 2.926082789168316,
-      "learning_rate": 6.902448909679942e-06,
-      "loss": 0.5146,
-      "step": 2740
-    },
-    {
-      "epoch": 1.3168625569680978,
-      "grad_norm": 1.8386902076079574,
-      "learning_rate": 6.889531370155004e-06,
-      "loss": 0.4658,
-      "step": 2745
-    },
-    {
-      "epoch": 1.3192612137203166,
-      "grad_norm": 2.0797479375790537,
-      "learning_rate": 6.876599100922566e-06,
-      "loss": 0.4361,
-      "step": 2750
-    },
-    {
-      "epoch": 1.3216598704725353,
-      "grad_norm": 3.061445264656408,
-      "learning_rate": 6.863652202795215e-06,
-      "loss": 0.4968,
-      "step": 2755
-    },
-    {
-      "epoch": 1.324058527224754,
-      "grad_norm": 2.687479706556386,
-      "learning_rate": 6.850690776699574e-06,
-      "loss": 0.4246,
-      "step": 2760
-    },
-    {
-      "epoch": 1.3264571839769728,
-      "grad_norm": 1.9168048123882346,
-      "learning_rate": 6.837714923675516e-06,
-      "loss": 0.4289,
-      "step": 2765
-    },
-    {
-      "epoch": 1.3288558407291917,
-      "grad_norm": 2.098881194418661,
-      "learning_rate": 6.8247247448753866e-06,
-      "loss": 0.4942,
-      "step": 2770
-    },
-    {
-      "epoch": 1.3312544974814104,
-      "grad_norm": 2.9646922853585793,
-      "learning_rate": 6.811720341563197e-06,
-      "loss": 0.4919,
-      "step": 2775
-    },
-    {
-      "epoch": 1.3336531542336292,
-      "grad_norm": 2.0101478360020604,
-      "learning_rate": 6.798701815113851e-06,
-      "loss": 0.435,
-      "step": 2780
-    },
-    {
-      "epoch": 1.336051810985848,
-      "grad_norm": 1.992177399616399,
-      "learning_rate": 6.785669267012346e-06,
-      "loss": 0.406,
-      "step": 2785
-    },
-    {
-      "epoch": 1.3384504677380666,
-      "grad_norm": 2.310067473610632,
-      "learning_rate": 6.7726227988529834e-06,
-      "loss": 0.427,
-      "step": 2790
-    },
-    {
-      "epoch": 1.3408491244902854,
-      "grad_norm": 2.1802412610225006,
-      "learning_rate": 6.759562512338581e-06,
-      "loss": 0.4949,
-      "step": 2795
-    },
-    {
-      "epoch": 1.3432477812425043,
-      "grad_norm": 2.1743296833228176,
-      "learning_rate": 6.74648850927967e-06,
-      "loss": 0.4758,
-      "step": 2800
-    },
-    {
-      "epoch": 1.345646437994723,
-      "grad_norm": 2.354140603930052,
-      "learning_rate": 6.733400891593717e-06,
-      "loss": 0.4515,
-      "step": 2805
-    },
-    {
-      "epoch": 1.3480450947469418,
-      "grad_norm": 2.2693267222576283,
-      "learning_rate": 6.720299761304312e-06,
-      "loss": 0.4229,
-      "step": 2810
-    },
-    {
-      "epoch": 1.3504437514991605,
-      "grad_norm": 2.7850031035864338,
-      "learning_rate": 6.707185220540385e-06,
-      "loss": 0.4096,
-      "step": 2815
-    },
-    {
-      "epoch": 1.3528424082513792,
-      "grad_norm": 2.0590276052665417,
-      "learning_rate": 6.6940573715354075e-06,
-      "loss": 0.3731,
-      "step": 2820
-    },
-    {
-      "epoch": 1.355241065003598,
-      "grad_norm": 2.455922368927128,
-      "learning_rate": 6.6809163166265945e-06,
-      "loss": 0.4694,
-      "step": 2825
-    },
-    {
-      "epoch": 1.3576397217558167,
-      "grad_norm": 3.2507513740552443,
-      "learning_rate": 6.667762158254104e-06,
-      "loss": 0.4457,
-      "step": 2830
-    },
-    {
-      "epoch": 1.3600383785080354,
-      "grad_norm": 2.2692392527198755,
-      "learning_rate": 6.654594998960242e-06,
-      "loss": 0.4451,
-      "step": 2835
-    },
-    {
-      "epoch": 1.3624370352602542,
-      "grad_norm": 2.03006411627659,
-      "learning_rate": 6.641414941388666e-06,
-      "loss": 0.4626,
-      "step": 2840
-    },
-    {
-      "epoch": 1.364835692012473,
-      "grad_norm": 2.3940134301391947,
-      "learning_rate": 6.628222088283574e-06,
-      "loss": 0.4573,
-      "step": 2845
-    },
-    {
-      "epoch": 1.3672343487646919,
-      "grad_norm": 2.966132014556057,
-      "learning_rate": 6.615016542488917e-06,
-      "loss": 0.4224,
-      "step": 2850
-    },
-    {
-      "epoch": 1.3696330055169106,
-      "grad_norm": 1.9966562170831368,
-      "learning_rate": 6.601798406947587e-06,
-      "loss": 0.4081,
-      "step": 2855
-    },
-    {
-      "epoch": 1.3720316622691293,
-      "grad_norm": 2.492563728485512,
-      "learning_rate": 6.588567784700623e-06,
-      "loss": 0.4479,
-      "step": 2860
-    },
-    {
-      "epoch": 1.374430319021348,
-      "grad_norm": 2.3512673785706606,
-      "learning_rate": 6.575324778886399e-06,
-      "loss": 0.4658,
-      "step": 2865
-    },
-    {
-      "epoch": 1.3768289757735668,
-      "grad_norm": 2.7381141478453253,
-      "learning_rate": 6.562069492739824e-06,
-      "loss": 0.3877,
-      "step": 2870
-    },
-    {
-      "epoch": 1.3792276325257855,
-      "grad_norm": 2.4848123819551904,
-      "learning_rate": 6.548802029591542e-06,
-      "loss": 0.4748,
-      "step": 2875
-    },
-    {
-      "epoch": 1.3816262892780042,
-      "grad_norm": 1.925224171020017,
-      "learning_rate": 6.535522492867118e-06,
-      "loss": 0.4298,
-      "step": 2880
-    },
-    {
-      "epoch": 1.3840249460302232,
-      "grad_norm": 2.589600737598477,
-      "learning_rate": 6.522230986086234e-06,
-      "loss": 0.3791,
-      "step": 2885
-    },
-    {
-      "epoch": 1.386423602782442,
-      "grad_norm": 2.276724697441552,
-      "learning_rate": 6.50892761286189e-06,
-      "loss": 0.4497,
-      "step": 2890
-    },
-    {
-      "epoch": 1.3888222595346607,
-      "grad_norm": 2.944103381058651,
-      "learning_rate": 6.4956124768995855e-06,
-      "loss": 0.4675,
-      "step": 2895
-    },
-    {
-      "epoch": 1.3912209162868794,
-      "grad_norm": 2.2595645772519055,
-      "learning_rate": 6.482285681996516e-06,
-      "loss": 0.4558,
-      "step": 2900
-    },
-    {
-      "epoch": 1.3936195730390981,
-      "grad_norm": 2.4610301955189735,
-      "learning_rate": 6.468947332040764e-06,
-      "loss": 0.4907,
-      "step": 2905
-    },
-    {
-      "epoch": 1.3960182297913168,
-      "grad_norm": 2.3168770293484857,
-      "learning_rate": 6.455597531010489e-06,
-      "loss": 0.4145,
-      "step": 2910
-    },
-    {
-      "epoch": 1.3984168865435356,
-      "grad_norm": 2.0741330602408166,
-      "learning_rate": 6.442236382973115e-06,
-      "loss": 0.423,
-      "step": 2915
-    },
-    {
-      "epoch": 1.4008155432957543,
-      "grad_norm": 2.9099644962421296,
-      "learning_rate": 6.428863992084523e-06,
-      "loss": 0.4216,
-      "step": 2920
-    },
-    {
-      "epoch": 1.403214200047973,
-      "grad_norm": 2.8760922784455523,
-      "learning_rate": 6.415480462588235e-06,
-      "loss": 0.4346,
-      "step": 2925
-    },
-    {
-      "epoch": 1.4056128568001918,
-      "grad_norm": 2.546333453485272,
-      "learning_rate": 6.402085898814605e-06,
-      "loss": 0.5081,
-      "step": 2930
-    },
-    {
-      "epoch": 1.4080115135524107,
-      "grad_norm": 2.4728709538253337,
-      "learning_rate": 6.388680405180003e-06,
-      "loss": 0.4637,
-      "step": 2935
-    },
-    {
-      "epoch": 1.4104101703046295,
-      "grad_norm": 1.9286327814952449,
-      "learning_rate": 6.375264086185999e-06,
-      "loss": 0.4532,
-      "step": 2940
-    },
-    {
-      "epoch": 1.4128088270568482,
-      "grad_norm": 2.1675714888403066,
-      "learning_rate": 6.361837046418558e-06,
-      "loss": 0.4243,
-      "step": 2945
-    },
-    {
-      "epoch": 1.415207483809067,
-      "grad_norm": 2.859235664034111,
-      "learning_rate": 6.348399390547213e-06,
-      "loss": 0.4288,
-      "step": 2950
-    },
-    {
-      "epoch": 1.4176061405612856,
-      "grad_norm": 1.9916645003102105,
-      "learning_rate": 6.334951223324254e-06,
-      "loss": 0.4518,
-      "step": 2955
-    },
-    {
-      "epoch": 1.4200047973135044,
-      "grad_norm": 3.334082881089859,
-      "learning_rate": 6.3214926495839135e-06,
-      "loss": 0.4372,
-      "step": 2960
-    },
-    {
-      "epoch": 1.4224034540657233,
-      "grad_norm": 2.051225494757854,
-      "learning_rate": 6.308023774241547e-06,
-      "loss": 0.4134,
-      "step": 2965
-    },
-    {
-      "epoch": 1.424802110817942,
-      "grad_norm": 2.2747949811500585,
-      "learning_rate": 6.294544702292814e-06,
-      "loss": 0.4694,
-      "step": 2970
-    },
-    {
-      "epoch": 1.4272007675701608,
-      "grad_norm": 2.195088989988513,
-      "learning_rate": 6.281055538812861e-06,
-      "loss": 0.4546,
-      "step": 2975
-    },
-    {
-      "epoch": 1.4295994243223795,
-      "grad_norm": 1.827059462779726,
-      "learning_rate": 6.267556388955508e-06,
-      "loss": 0.4284,
-      "step": 2980
-    },
-    {
-      "epoch": 1.4319980810745983,
-      "grad_norm": 1.7451018183946225,
-      "learning_rate": 6.254047357952413e-06,
-      "loss": 0.452,
-      "step": 2985
-    },
-    {
-      "epoch": 1.434396737826817,
-      "grad_norm": 2.4438753461794764,
-      "learning_rate": 6.24052855111227e-06,
-      "loss": 0.4416,
-      "step": 2990
-    },
-    {
-      "epoch": 1.4367953945790357,
-      "grad_norm": 1.927460965161698,
-      "learning_rate": 6.227000073819973e-06,
-      "loss": 0.3954,
-      "step": 2995
-    },
-    {
-      "epoch": 1.4391940513312544,
-      "grad_norm": 2.7660502397338047,
-      "learning_rate": 6.213462031535811e-06,
-      "loss": 0.4431,
-      "step": 3000
-    },
-    {
-      "epoch": 1.4391940513312544,
-      "eval_loss": 0.8874278664588928,
-      "eval_runtime": 746.2797,
-      "eval_samples_per_second": 9.929,
-      "eval_steps_per_second": 0.622,
-      "step": 3000
-    },
-    {
-      "epoch": 1.4415927080834732,
-      "grad_norm": 2.611336223360776,
-      "learning_rate": 6.199914529794627e-06,
-      "loss": 0.4706,
-      "step": 3005
-    },
-    {
-      "epoch": 1.443991364835692,
-      "grad_norm": 2.5667262714384003,
-      "learning_rate": 6.186357674205008e-06,
-      "loss": 0.5219,
-      "step": 3010
-    },
-    {
-      "epoch": 1.4463900215879106,
-      "grad_norm": 2.548110034722319,
-      "learning_rate": 6.172791570448458e-06,
-      "loss": 0.4481,
-      "step": 3015
-    },
-    {
-      "epoch": 1.4487886783401296,
-      "grad_norm": 2.597919097756615,
-      "learning_rate": 6.159216324278577e-06,
-      "loss": 0.4718,
-      "step": 3020
-    },
-    {
-      "epoch": 1.4511873350923483,
-      "grad_norm": 3.674189698827941,
-      "learning_rate": 6.145632041520229e-06,
-      "loss": 0.4611,
-      "step": 3025
-    },
-    {
-      "epoch": 1.453585991844567,
-      "grad_norm": 2.67480821934183,
-      "learning_rate": 6.132038828068725e-06,
-      "loss": 0.457,
-      "step": 3030
-    },
-    {
-      "epoch": 1.4559846485967858,
-      "grad_norm": 2.8863659126445764,
-      "learning_rate": 6.118436789888995e-06,
-      "loss": 0.4911,
-      "step": 3035
-    },
-    {
-      "epoch": 1.4583833053490045,
-      "grad_norm": 2.9651286406517574,
-      "learning_rate": 6.104826033014762e-06,
-      "loss": 0.4528,
-      "step": 3040
-    },
-    {
-      "epoch": 1.4607819621012232,
-      "grad_norm": 2.25092477044882,
-      "learning_rate": 6.091206663547713e-06,
-      "loss": 0.4568,
-      "step": 3045
-    },
-    {
-      "epoch": 1.4631806188534422,
-      "grad_norm": 2.5707376320364337,
-      "learning_rate": 6.077578787656673e-06,
-      "loss": 0.4713,
-      "step": 3050
-    },
-    {
-      "epoch": 1.465579275605661,
-      "grad_norm": 2.631364741417103,
-      "learning_rate": 6.063942511576782e-06,
-      "loss": 0.481,
-      "step": 3055
-    },
-    {
-      "epoch": 1.4679779323578797,
-      "grad_norm": 2.800004593675168,
-      "learning_rate": 6.050297941608661e-06,
-      "loss": 0.4427,
-      "step": 3060
-    },
-    {
-      "epoch": 1.4703765891100984,
-      "grad_norm": 2.95675973317759,
-      "learning_rate": 6.036645184117585e-06,
-      "loss": 0.4765,
-      "step": 3065
-    },
-    {
-      "epoch": 1.4727752458623171,
-      "grad_norm": 2.2729302837721557,
-      "learning_rate": 6.022984345532654e-06,
-      "loss": 0.4479,
-      "step": 3070
-    },
-    {
-      "epoch": 1.4751739026145358,
-      "grad_norm": 2.5945168985243305,
-      "learning_rate": 6.009315532345964e-06,
-      "loss": 0.4236,
-      "step": 3075
-    },
-    {
-      "epoch": 1.4775725593667546,
-      "grad_norm": 2.010953790417192,
-      "learning_rate": 5.995638851111778e-06,
-      "loss": 0.4607,
-      "step": 3080
-    },
-    {
-      "epoch": 1.4799712161189733,
-      "grad_norm": 2.4135304972011773,
-      "learning_rate": 5.981954408445691e-06,
-      "loss": 0.3831,
-      "step": 3085
-    },
-    {
-      "epoch": 1.482369872871192,
-      "grad_norm": 3.5469108380440106,
-      "learning_rate": 5.968262311023807e-06,
-      "loss": 0.4975,
-      "step": 3090
-    },
-    {
-      "epoch": 1.4847685296234108,
-      "grad_norm": 2.4819879566093763,
-      "learning_rate": 5.954562665581893e-06,
-      "loss": 0.4859,
-      "step": 3095
-    },
-    {
-      "epoch": 1.4871671863756297,
-      "grad_norm": 3.092020698716526,
-      "learning_rate": 5.940855578914564e-06,
-      "loss": 0.4222,
-      "step": 3100
-    },
-    {
-      "epoch": 1.4895658431278485,
-      "grad_norm": 2.603907807597384,
-      "learning_rate": 5.927141157874441e-06,
-      "loss": 0.5217,
-      "step": 3105
-    },
-    {
-      "epoch": 1.4919644998800672,
-      "grad_norm": 1.9127228659037592,
-      "learning_rate": 5.913419509371317e-06,
-      "loss": 0.4107,
-      "step": 3110
-    },
-    {
-      "epoch": 1.494363156632286,
-      "grad_norm": 1.9952834925260585,
-      "learning_rate": 5.899690740371327e-06,
-      "loss": 0.4634,
-      "step": 3115
-    },
-    {
-      "epoch": 1.4967618133845046,
-      "grad_norm": 2.3004395086577785,
-      "learning_rate": 5.885954957896115e-06,
-      "loss": 0.4243,
-      "step": 3120
-    },
-    {
-      "epoch": 1.4991604701367234,
-      "grad_norm": 2.712237010357561,
-      "learning_rate": 5.872212269021996e-06,
-      "loss": 0.4012,
-      "step": 3125
-    },
-    {
-      "epoch": 1.5015591268889423,
-      "grad_norm": 1.9707705697898028,
-      "learning_rate": 5.858462780879127e-06,
-      "loss": 0.3963,
-      "step": 3130
-    },
-    {
-      "epoch": 1.503957783641161,
-      "grad_norm": 2.1775128153767516,
-      "learning_rate": 5.844706600650662e-06,
-      "loss": 0.4788,
-      "step": 3135
-    },
-    {
-      "epoch": 1.5063564403933798,
-      "grad_norm": 2.2650398526849296,
-      "learning_rate": 5.83094383557193e-06,
-      "loss": 0.442,
-      "step": 3140
-    },
-    {
-      "epoch": 1.5087550971455985,
-      "grad_norm": 2.3021274108333176,
-      "learning_rate": 5.817174592929585e-06,
-      "loss": 0.4619,
-      "step": 3145
-    },
-    {
-      "epoch": 1.5111537538978173,
-      "grad_norm": 3.0408864818621266,
-      "learning_rate": 5.803398980060782e-06,
-      "loss": 0.4848,
-      "step": 3150
-    },
-    {
-      "epoch": 1.513552410650036,
-      "grad_norm": 2.662432672052024,
-      "learning_rate": 5.789617104352327e-06,
-      "loss": 0.4537,
-      "step": 3155
-    },
-    {
-      "epoch": 1.5159510674022547,
-      "grad_norm": 2.1199816412732826,
-      "learning_rate": 5.775829073239859e-06,
-      "loss": 0.4684,
-      "step": 3160
-    },
-    {
-      "epoch": 1.5183497241544734,
-      "grad_norm": 1.9990361255845992,
-      "learning_rate": 5.762034994206992e-06,
-      "loss": 0.4333,
-      "step": 3165
-    },
-    {
-      "epoch": 1.5207483809066922,
-      "grad_norm": 3.1609456768731614,
-      "learning_rate": 5.748234974784487e-06,
-      "loss": 0.4952,
-      "step": 3170
-    },
-    {
-      "epoch": 1.523147037658911,
-      "grad_norm": 2.4476985735005927,
-      "learning_rate": 5.734429122549415e-06,
-      "loss": 0.4948,
-      "step": 3175
-    },
-    {
-      "epoch": 1.5255456944111296,
-      "grad_norm": 2.148237529847839,
-      "learning_rate": 5.720617545124319e-06,
-      "loss": 0.451,
-      "step": 3180
-    },
-    {
-      "epoch": 1.5279443511633484,
-      "grad_norm": 2.3650819595385397,
-      "learning_rate": 5.706800350176366e-06,
-      "loss": 0.5046,
-      "step": 3185
-    },
-    {
-      "epoch": 1.5303430079155673,
-      "grad_norm": 2.4798863005701026,
-      "learning_rate": 5.692977645416519e-06,
-      "loss": 0.4512,
-      "step": 3190
-    },
-    {
-      "epoch": 1.532741664667786,
-      "grad_norm": 2.5313251041761466,
-      "learning_rate": 5.679149538598689e-06,
-      "loss": 0.4968,
-      "step": 3195
-    },
-    {
-      "epoch": 1.5351403214200048,
-      "grad_norm": 5.398569331104567,
-      "learning_rate": 5.6653161375189e-06,
-      "loss": 0.5006,
-      "step": 3200
-    },
-    {
-      "epoch": 1.5375389781722235,
-      "grad_norm": 2.7840966965275005,
-      "learning_rate": 5.651477550014447e-06,
-      "loss": 0.4485,
-      "step": 3205
-    },
-    {
-      "epoch": 1.5399376349244425,
-      "grad_norm": 3.262453782851082,
-      "learning_rate": 5.637633883963057e-06,
-      "loss": 0.3752,
-      "step": 3210
-    },
-    {
-      "epoch": 1.5423362916766612,
-      "grad_norm": 2.237442214349481,
-      "learning_rate": 5.623785247282044e-06,
-      "loss": 0.4064,
-      "step": 3215
-    },
-    {
-      "epoch": 1.54473494842888,
-      "grad_norm": 2.9173406947413616,
-      "learning_rate": 5.609931747927469e-06,
-      "loss": 0.4112,
-      "step": 3220
-    },
-    {
-      "epoch": 1.5471336051810987,
-      "grad_norm": 3.0776007451762024,
-      "learning_rate": 5.5960734938933055e-06,
-      "loss": 0.5295,
-      "step": 3225
-    },
-    {
-      "epoch": 1.5495322619333174,
-      "grad_norm": 2.4545633340960666,
-      "learning_rate": 5.582210593210584e-06,
-      "loss": 0.3961,
-      "step": 3230
-    },
-    {
-      "epoch": 1.5519309186855361,
-      "grad_norm": 2.009510951840261,
-      "learning_rate": 5.568343153946563e-06,
-      "loss": 0.4884,
-      "step": 3235
-    },
-    {
-      "epoch": 1.5543295754377549,
-      "grad_norm": 2.2007510609373195,
-      "learning_rate": 5.5544712842038795e-06,
-      "loss": 0.484,
-      "step": 3240
-    },
-    {
-      "epoch": 1.5567282321899736,
-      "grad_norm": 2.8520591253973073,
-      "learning_rate": 5.540595092119709e-06,
-      "loss": 0.4786,
-      "step": 3245
-    },
-    {
-      "epoch": 1.5591268889421923,
-      "grad_norm": 1.604583271616001,
-      "learning_rate": 5.526714685864921e-06,
-      "loss": 0.3909,
-      "step": 3250
-    },
-    {
-      "epoch": 1.561525545694411,
-      "grad_norm": 2.6979214589842333,
-      "learning_rate": 5.5128301736432335e-06,
-      "loss": 0.4753,
-      "step": 3255
-    },
-    {
-      "epoch": 1.5639242024466298,
-      "grad_norm": 2.1624025899920074,
-      "learning_rate": 5.498941663690376e-06,
-      "loss": 0.3832,
-      "step": 3260
-    },
-    {
-      "epoch": 1.5663228591988485,
-      "grad_norm": 2.2374635152759357,
-      "learning_rate": 5.485049264273241e-06,
-      "loss": 0.4296,
-      "step": 3265
-    },
-    {
-      "epoch": 1.5687215159510672,
-      "grad_norm": 1.973150872622609,
-      "learning_rate": 5.471153083689041e-06,
-      "loss": 0.3775,
-      "step": 3270
-    },
-    {
-      "epoch": 1.5711201727032862,
-      "grad_norm": 2.0848939395577633,
-      "learning_rate": 5.457253230264463e-06,
-      "loss": 0.4023,
-      "step": 3275
-    },
-    {
-      "epoch": 1.573518829455505,
-      "grad_norm": 2.12845976407844,
-      "learning_rate": 5.443349812354828e-06,
-      "loss": 0.426,
-      "step": 3280
-    },
-    {
-      "epoch": 1.5759174862077237,
-      "grad_norm": 2.2433674944507462,
-      "learning_rate": 5.429442938343241e-06,
-      "loss": 0.4825,
-      "step": 3285
-    },
-    {
-      "epoch": 1.5783161429599424,
-      "grad_norm": 1.8440111398002124,
-      "learning_rate": 5.41553271663975e-06,
-      "loss": 0.3807,
-      "step": 3290
-    },
-    {
-      "epoch": 1.5807147997121613,
-      "grad_norm": 2.8628882707095014,
-      "learning_rate": 5.401619255680501e-06,
-      "loss": 0.5037,
-      "step": 3295
-    },
-    {
-      "epoch": 1.58311345646438,
-      "grad_norm": 2.3477798507731045,
-      "learning_rate": 5.38770266392689e-06,
-      "loss": 0.431,
-      "step": 3300
-    },
-    {
-      "epoch": 1.5855121132165988,
-      "grad_norm": 1.8413235582299343,
-      "learning_rate": 5.373783049864714e-06,
-      "loss": 0.4954,
-      "step": 3305
-    },
-    {
-      "epoch": 1.5879107699688175,
-      "grad_norm": 2.608126509725837,
-      "learning_rate": 5.3598605220033406e-06,
-      "loss": 0.5038,
-      "step": 3310
-    },
-    {
-      "epoch": 1.5903094267210363,
-      "grad_norm": 2.14593361907128,
-      "learning_rate": 5.345935188874845e-06,
-      "loss": 0.4741,
-      "step": 3315
-    },
-    {
-      "epoch": 1.592708083473255,
-      "grad_norm": 2.6895229502835774,
-      "learning_rate": 5.3320071590331725e-06,
-      "loss": 0.4673,
-      "step": 3320
-    },
-    {
-      "epoch": 1.5951067402254737,
-      "grad_norm": 2.3486279127223058,
-      "learning_rate": 5.318076541053287e-06,
-      "loss": 0.4204,
-      "step": 3325
-    },
-    {
-      "epoch": 1.5975053969776924,
-      "grad_norm": 2.3779256132380753,
-      "learning_rate": 5.304143443530333e-06,
-      "loss": 0.479,
-      "step": 3330
-    },
-    {
-      "epoch": 1.5999040537299112,
-      "grad_norm": 2.382506024792259,
-      "learning_rate": 5.290207975078783e-06,
-      "loss": 0.4944,
-      "step": 3335
-    },
-    {
-      "epoch": 1.60230271048213,
-      "grad_norm": 2.324583019078476,
-      "learning_rate": 5.276270244331589e-06,
-      "loss": 0.4517,
-      "step": 3340
-    },
-    {
-      "epoch": 1.6047013672343486,
-      "grad_norm": 1.907971038729415,
-      "learning_rate": 5.2623303599393385e-06,
-      "loss": 0.3964,
-      "step": 3345
-    },
-    {
-      "epoch": 1.6071000239865674,
-      "grad_norm": 2.793523495104258,
-      "learning_rate": 5.248388430569415e-06,
-      "loss": 0.4268,
-      "step": 3350
-    },
-    {
-      "epoch": 1.6094986807387863,
-      "grad_norm": 2.0671079091787576,
-      "learning_rate": 5.234444564905132e-06,
-      "loss": 0.4985,
-      "step": 3355
-    },
-    {
-      "epoch": 1.611897337491005,
-      "grad_norm": 2.9592798078808102,
-      "learning_rate": 5.220498871644908e-06,
-      "loss": 0.4388,
-      "step": 3360
-    },
-    {
-      "epoch": 1.6142959942432238,
-      "grad_norm": 2.1426600987595656,
-      "learning_rate": 5.206551459501401e-06,
-      "loss": 0.3915,
-      "step": 3365
-    },
-    {
-      "epoch": 1.6166946509954425,
-      "grad_norm": 2.7506183229782395,
-      "learning_rate": 5.192602437200673e-06,
-      "loss": 0.4848,
-      "step": 3370
-    },
-    {
-      "epoch": 1.6190933077476615,
-      "grad_norm": 2.8796183674653615,
-      "learning_rate": 5.178651913481334e-06,
-      "loss": 0.4312,
-      "step": 3375
-    },
-    {
-      "epoch": 1.6214919644998802,
-      "grad_norm": 2.4712356902239265,
-      "learning_rate": 5.164699997093699e-06,
-      "loss": 0.4644,
-      "step": 3380
-    },
-    {
-      "epoch": 1.623890621252099,
-      "grad_norm": 2.4895946890030802,
-      "learning_rate": 5.150746796798946e-06,
-      "loss": 0.4597,
-      "step": 3385
-    },
-    {
-      "epoch": 1.6262892780043177,
-      "grad_norm": 2.7280161391548825,
-      "learning_rate": 5.136792421368251e-06,
-      "loss": 0.4394,
-      "step": 3390
-    },
-    {
-      "epoch": 1.6286879347565364,
-      "grad_norm": 2.0191439616934725,
-      "learning_rate": 5.122836979581958e-06,
-      "loss": 0.4216,
-      "step": 3395
-    },
-    {
-      "epoch": 1.6310865915087551,
-      "grad_norm": 2.710615920560287,
-      "learning_rate": 5.1088805802287215e-06,
-      "loss": 0.5304,
-      "step": 3400
-    },
-    {
-      "epoch": 1.6334852482609739,
-      "grad_norm": 2.0928161833734,
-      "learning_rate": 5.09492333210466e-06,
-      "loss": 0.4244,
-      "step": 3405
-    },
-    {
-      "epoch": 1.6358839050131926,
-      "grad_norm": 1.8191682812562566,
-      "learning_rate": 5.080965344012509e-06,
-      "loss": 0.426,
-      "step": 3410
-    },
-    {
-      "epoch": 1.6382825617654113,
-      "grad_norm": 1.915320356764598,
-      "learning_rate": 5.067006724760772e-06,
-      "loss": 0.3817,
-      "step": 3415
-    },
-    {
-      "epoch": 1.64068121851763,
-      "grad_norm": 3.0320685782151062,
-      "learning_rate": 5.053047583162877e-06,
-      "loss": 0.5218,
-      "step": 3420
-    },
-    {
-      "epoch": 1.6430798752698488,
-      "grad_norm": 2.140553134411092,
-      "learning_rate": 5.039088028036317e-06,
-      "loss": 0.4807,
-      "step": 3425
-    },
-    {
-      "epoch": 1.6454785320220675,
-      "grad_norm": 2.410914424743909,
-      "learning_rate": 5.0251281682018125e-06,
-      "loss": 0.4184,
-      "step": 3430
-    },
-    {
-      "epoch": 1.6478771887742862,
-      "grad_norm": 2.6739455394590745,
-      "learning_rate": 5.011168112482459e-06,
-      "loss": 0.4477,
-      "step": 3435
-    },
-    {
-      "epoch": 1.6502758455265052,
-      "grad_norm": 3.163665779437731,
-      "learning_rate": 4.997207969702879e-06,
-      "loss": 0.4681,
-      "step": 3440
-    },
-    {
-      "epoch": 1.652674502278724,
-      "grad_norm": 2.553283311318629,
-      "learning_rate": 4.98324784868837e-06,
-      "loss": 0.4042,
-      "step": 3445
-    },
-    {
-      "epoch": 1.6550731590309427,
-      "grad_norm": 1.677275594926634,
-      "learning_rate": 4.969287858264068e-06,
-      "loss": 0.4427,
-      "step": 3450
-    },
-    {
-      "epoch": 1.6574718157831614,
-      "grad_norm": 3.518235414569066,
-      "learning_rate": 4.955328107254084e-06,
-      "loss": 0.4576,
-      "step": 3455
-    },
-    {
-      "epoch": 1.6598704725353803,
-      "grad_norm": 2.560656886321086,
-      "learning_rate": 4.941368704480666e-06,
-      "loss": 0.4832,
-      "step": 3460
-    },
-    {
-      "epoch": 1.662269129287599,
-      "grad_norm": 1.8114516258560824,
-      "learning_rate": 4.927409758763343e-06,
-      "loss": 0.419,
-      "step": 3465
-    },
-    {
-      "epoch": 1.6646677860398178,
-      "grad_norm": 2.1836726473973416,
-      "learning_rate": 4.913451378918087e-06,
-      "loss": 0.5074,
-      "step": 3470
-    },
-    {
-      "epoch": 1.6670664427920365,
-      "grad_norm": 2.0729685092094905,
-      "learning_rate": 4.899493673756456e-06,
-      "loss": 0.3973,
-      "step": 3475
-    },
-    {
-      "epoch": 1.6694650995442553,
-      "grad_norm": 2.332454402343443,
-      "learning_rate": 4.885536752084748e-06,
-      "loss": 0.437,
-      "step": 3480
-    },
-    {
-      "epoch": 1.671863756296474,
-      "grad_norm": 1.9422102291303713,
-      "learning_rate": 4.871580722703153e-06,
-      "loss": 0.4099,
-      "step": 3485
-    },
-    {
-      "epoch": 1.6742624130486927,
-      "grad_norm": 2.45511686162037,
-      "learning_rate": 4.857625694404907e-06,
-      "loss": 0.4417,
-      "step": 3490
-    },
-    {
-      "epoch": 1.6766610698009115,
-      "grad_norm": 2.477532923554625,
-      "learning_rate": 4.843671775975441e-06,
-      "loss": 0.4183,
-      "step": 3495
-    },
-    {
-      "epoch": 1.6790597265531302,
-      "grad_norm": 2.0134977559736424,
-      "learning_rate": 4.8297190761915335e-06,
-      "loss": 0.4511,
-      "step": 3500
-    },
-    {
-      "epoch": 1.681458383305349,
-      "grad_norm": 2.7722220766364476,
-      "learning_rate": 4.815767703820466e-06,
-      "loss": 0.4241,
-      "step": 3505
-    },
-    {
-      "epoch": 1.6838570400575676,
-      "grad_norm": 2.4222195684237007,
-      "learning_rate": 4.80181776761917e-06,
-      "loss": 0.4166,
-      "step": 3510
-    },
-    {
-      "epoch": 1.6862556968097864,
-      "grad_norm": 1.7497888557935886,
-      "learning_rate": 4.787869376333381e-06,
-      "loss": 0.384,
-      "step": 3515
-    },
-    {
-      "epoch": 1.6886543535620053,
-      "grad_norm": 2.252998156949262,
-      "learning_rate": 4.773922638696792e-06,
-      "loss": 0.4858,
-      "step": 3520
-    },
-    {
-      "epoch": 1.691053010314224,
-      "grad_norm": 2.156538837955339,
-      "learning_rate": 4.759977663430204e-06,
-      "loss": 0.3671,
-      "step": 3525
-    },
-    {
-      "epoch": 1.6934516670664428,
-      "grad_norm": 2.8590215367905345,
-      "learning_rate": 4.746034559240682e-06,
-      "loss": 0.455,
-      "step": 3530
-    },
-    {
-      "epoch": 1.6958503238186615,
-      "grad_norm": 2.774007584511685,
-      "learning_rate": 4.732093434820703e-06,
-      "loss": 0.4291,
-      "step": 3535
-    },
-    {
-      "epoch": 1.6982489805708805,
-      "grad_norm": 2.634909995207892,
-      "learning_rate": 4.7181543988473116e-06,
-      "loss": 0.4459,
-      "step": 3540
-    },
-    {
-      "epoch": 1.7006476373230992,
-      "grad_norm": 2.133691456687859,
-      "learning_rate": 4.704217559981273e-06,
-      "loss": 0.3831,
-      "step": 3545
-    },
-    {
-      "epoch": 1.703046294075318,
-      "grad_norm": 2.2922171922400163,
-      "learning_rate": 4.690283026866221e-06,
-      "loss": 0.4505,
-      "step": 3550
-    },
-    {
-      "epoch": 1.7054449508275367,
-      "grad_norm": 3.7940433061602357,
-      "learning_rate": 4.6763509081278215e-06,
-      "loss": 0.4534,
-      "step": 3555
-    },
-    {
-      "epoch": 1.7078436075797554,
-      "grad_norm": 2.208035399901243,
-      "learning_rate": 4.6624213123729146e-06,
-      "loss": 0.4009,
-      "step": 3560
-    },
-    {
-      "epoch": 1.7102422643319741,
-      "grad_norm": 2.356109324316714,
-      "learning_rate": 4.648494348188675e-06,
-      "loss": 0.5101,
-      "step": 3565
-    },
-    {
-      "epoch": 1.7126409210841929,
-      "grad_norm": 2.5764709432365316,
-      "learning_rate": 4.634570124141761e-06,
-      "loss": 0.4425,
-      "step": 3570
-    },
-    {
-      "epoch": 1.7150395778364116,
-      "grad_norm": 1.9758783243784745,
-      "learning_rate": 4.620648748777472e-06,
-      "loss": 0.445,
-      "step": 3575
-    },
-    {
-      "epoch": 1.7174382345886303,
-      "grad_norm": 2.2527649874877334,
-      "learning_rate": 4.606730330618899e-06,
-      "loss": 0.4632,
-      "step": 3580
-    },
-    {
-      "epoch": 1.719836891340849,
-      "grad_norm": 2.132159506220732,
-      "learning_rate": 4.592814978166084e-06,
-      "loss": 0.4718,
-      "step": 3585
-    },
-    {
-      "epoch": 1.7222355480930678,
-      "grad_norm": 2.8446174142457,
-      "learning_rate": 4.578902799895165e-06,
-      "loss": 0.4792,
-      "step": 3590
-    },
-    {
-      "epoch": 1.7246342048452865,
-      "grad_norm": 2.230569296862582,
-      "learning_rate": 4.5649939042575415e-06,
-      "loss": 0.4598,
-      "step": 3595
-    },
-    {
-      "epoch": 1.7270328615975052,
-      "grad_norm": 3.444809582847871,
-      "learning_rate": 4.5510883996790175e-06,
-      "loss": 0.6123,
-      "step": 3600
-    },
-    {
-      "epoch": 1.7294315183497242,
-      "grad_norm": 1.9684733506440177,
-      "learning_rate": 4.53718639455897e-06,
-      "loss": 0.3768,
-      "step": 3605
-    },
-    {
-      "epoch": 1.731830175101943,
-      "grad_norm": 2.368594527234669,
-      "learning_rate": 4.523287997269489e-06,
-      "loss": 0.4619,
-      "step": 3610
-    },
-    {
-      "epoch": 1.7342288318541617,
-      "grad_norm": 2.104619044987494,
-      "learning_rate": 4.509393316154545e-06,
-      "loss": 0.4856,
-      "step": 3615
-    },
-    {
-      "epoch": 1.7366274886063804,
-      "grad_norm": 2.824474105567847,
-      "learning_rate": 4.495502459529135e-06,
-      "loss": 0.4595,
-      "step": 3620
-    },
-    {
-      "epoch": 1.7390261453585993,
-      "grad_norm": 2.0922330865937235,
-      "learning_rate": 4.481615535678444e-06,
-      "loss": 0.4544,
-      "step": 3625
-    },
-    {
-      "epoch": 1.741424802110818,
-      "grad_norm": 3.699659997300747,
-      "learning_rate": 4.467732652857003e-06,
-      "loss": 0.4202,
-      "step": 3630
-    },
-    {
-      "epoch": 1.7438234588630368,
-      "grad_norm": 2.9983464058160516,
-      "learning_rate": 4.453853919287836e-06,
-      "loss": 0.423,
-      "step": 3635
-    },
-    {
-      "epoch": 1.7462221156152555,
-      "grad_norm": 2.0846769162351366,
-      "learning_rate": 4.439979443161624e-06,
-      "loss": 0.388,
-      "step": 3640
-    },
-    {
-      "epoch": 1.7486207723674743,
-      "grad_norm": 2.477314432741006,
-      "learning_rate": 4.42610933263586e-06,
-      "loss": 0.4506,
-      "step": 3645
-    },
-    {
-      "epoch": 1.751019429119693,
-      "grad_norm": 2.589674357755263,
-      "learning_rate": 4.4122436958340035e-06,
-      "loss": 0.3728,
-      "step": 3650
-    },
-    {
-      "epoch": 1.7534180858719117,
-      "grad_norm": 2.6366312614017122,
-      "learning_rate": 4.398382640844644e-06,
-      "loss": 0.479,
-      "step": 3655
-    },
-    {
-      "epoch": 1.7558167426241305,
-      "grad_norm": 2.014560929851438,
-      "learning_rate": 4.3845262757206495e-06,
-      "loss": 0.4408,
-      "step": 3660
-    },
-    {
-      "epoch": 1.7582153993763492,
-      "grad_norm": 1.7221380583217487,
-      "learning_rate": 4.370674708478326e-06,
-      "loss": 0.4317,
-      "step": 3665
-    },
-    {
-      "epoch": 1.760614056128568,
-      "grad_norm": 2.1638693438484995,
-      "learning_rate": 4.3568280470965865e-06,
-      "loss": 0.3856,
-      "step": 3670
-    },
-    {
-      "epoch": 1.7630127128807866,
-      "grad_norm": 2.6456939593036184,
-      "learning_rate": 4.342986399516092e-06,
-      "loss": 0.4169,
-      "step": 3675
-    },
-    {
-      "epoch": 1.7654113696330054,
-      "grad_norm": 2.585012432731737,
-      "learning_rate": 4.329149873638423e-06,
-      "loss": 0.4074,
-      "step": 3680
-    },
-    {
-      "epoch": 1.767810026385224,
-      "grad_norm": 2.192226239922893,
-      "learning_rate": 4.315318577325235e-06,
-      "loss": 0.4805,
-      "step": 3685
-    },
-    {
-      "epoch": 1.770208683137443,
-      "grad_norm": 2.3384135122928997,
-      "learning_rate": 4.3014926183974155e-06,
-      "loss": 0.4866,
-      "step": 3690
-    },
-    {
-      "epoch": 1.7726073398896618,
-      "grad_norm": 2.365443373467843,
-      "learning_rate": 4.287672104634241e-06,
-      "loss": 0.4481,
-      "step": 3695
-    },
-    {
-      "epoch": 1.7750059966418805,
-      "grad_norm": 2.542894793017788,
-      "learning_rate": 4.27385714377255e-06,
-      "loss": 0.4389,
-      "step": 3700
-    },
-    {
-      "epoch": 1.7774046533940993,
-      "grad_norm": 2.1437152802123096,
-      "learning_rate": 4.260047843505883e-06,
-      "loss": 0.4747,
-      "step": 3705
-    },
-    {
-      "epoch": 1.7798033101463182,
-      "grad_norm": 1.9219033326157453,
-      "learning_rate": 4.246244311483661e-06,
-      "loss": 0.4277,
-      "step": 3710
-    },
-    {
-      "epoch": 1.782201966898537,
-      "grad_norm": 2.5391812417621518,
-      "learning_rate": 4.232446655310337e-06,
-      "loss": 0.4594,
-      "step": 3715
-    },
-    {
-      "epoch": 1.7846006236507557,
-      "grad_norm": 2.6367801196236345,
-      "learning_rate": 4.218654982544559e-06,
-      "loss": 0.451,
-      "step": 3720
-    },
-    {
-      "epoch": 1.7869992804029744,
-      "grad_norm": 1.998666444994321,
-      "learning_rate": 4.204869400698331e-06,
-      "loss": 0.4611,
-      "step": 3725
-    },
-    {
-      "epoch": 1.7893979371551931,
-      "grad_norm": 2.396249588875812,
-      "learning_rate": 4.191090017236177e-06,
-      "loss": 0.4102,
-      "step": 3730
-    },
-    {
-      "epoch": 1.7917965939074119,
-      "grad_norm": 2.0438518913068906,
-      "learning_rate": 4.177316939574302e-06,
-      "loss": 0.3945,
-      "step": 3735
-    },
-    {
-      "epoch": 1.7941952506596306,
-      "grad_norm": 3.3647778164933895,
-      "learning_rate": 4.163550275079753e-06,
-      "loss": 0.4793,
-      "step": 3740
-    },
-    {
-      "epoch": 1.7965939074118493,
-      "grad_norm": 2.0510954320612833,
-      "learning_rate": 4.149790131069584e-06,
-      "loss": 0.4312,
-      "step": 3745
-    },
-    {
-      "epoch": 1.798992564164068,
-      "grad_norm": 2.1133217662374197,
-      "learning_rate": 4.136036614810022e-06,
-      "loss": 0.4127,
-      "step": 3750
-    },
-    {
-      "epoch": 1.8013912209162868,
-      "grad_norm": 2.110615768818984,
-      "learning_rate": 4.122289833515621e-06,
-      "loss": 0.4555,
-      "step": 3755
-    },
-    {
-      "epoch": 1.8037898776685055,
-      "grad_norm": 2.490483001412966,
-      "learning_rate": 4.1085498943484444e-06,
-      "loss": 0.4967,
-      "step": 3760
-    },
-    {
-      "epoch": 1.8061885344207242,
-      "grad_norm": 3.002844293787043,
-      "learning_rate": 4.094816904417207e-06,
-      "loss": 0.4706,
-      "step": 3765
-    },
-    {
-      "epoch": 1.8085871911729432,
-      "grad_norm": 2.183976582582103,
-      "learning_rate": 4.0810909707764555e-06,
-      "loss": 0.4638,
-      "step": 3770
-    },
-    {
-      "epoch": 1.810985847925162,
-      "grad_norm": 2.8167634413211466,
-      "learning_rate": 4.067372200425732e-06,
-      "loss": 0.4355,
-      "step": 3775
-    },
-    {
-      "epoch": 1.8133845046773807,
-      "grad_norm": 3.1466044507478825,
-      "learning_rate": 4.053660700308734e-06,
-      "loss": 0.4076,
-      "step": 3780
-    },
-    {
-      "epoch": 1.8157831614295994,
-      "grad_norm": 2.1976932894705,
-      "learning_rate": 4.039956577312486e-06,
-      "loss": 0.4186,
-      "step": 3785
-    },
-    {
-      "epoch": 1.8181818181818183,
-      "grad_norm": 2.6344969999958625,
-      "learning_rate": 4.026259938266508e-06,
-      "loss": 0.4368,
-      "step": 3790
-    },
-    {
-      "epoch": 1.820580474934037,
-      "grad_norm": 2.463495815471494,
-      "learning_rate": 4.012570889941973e-06,
-      "loss": 0.4712,
-      "step": 3795
-    },
-    {
-      "epoch": 1.8229791316862558,
-      "grad_norm": 2.3535074134581255,
-      "learning_rate": 3.998889539050884e-06,
-      "loss": 0.4478,
-      "step": 3800
-    },
-    {
-      "epoch": 1.8253777884384745,
-      "grad_norm": 1.8539164713669094,
-      "learning_rate": 3.98521599224524e-06,
-      "loss": 0.4817,
-      "step": 3805
-    },
-    {
-      "epoch": 1.8277764451906933,
-      "grad_norm": 2.4488933427042405,
-      "learning_rate": 3.971550356116205e-06,
-      "loss": 0.4517,
-      "step": 3810
-    },
-    {
-      "epoch": 1.830175101942912,
-      "grad_norm": 2.544480168164569,
-      "learning_rate": 3.957892737193274e-06,
-      "loss": 0.4154,
-      "step": 3815
-    },
-    {
-      "epoch": 1.8325737586951307,
-      "grad_norm": 2.3012603822017144,
-      "learning_rate": 3.944243241943442e-06,
-      "loss": 0.3977,
-      "step": 3820
-    },
-    {
-      "epoch": 1.8349724154473495,
-      "grad_norm": 2.5924353524670045,
-      "learning_rate": 3.930601976770383e-06,
-      "loss": 0.4526,
-      "step": 3825
-    },
-    {
-      "epoch": 1.8373710721995682,
-      "grad_norm": 2.0573535751534275,
-      "learning_rate": 3.916969048013607e-06,
-      "loss": 0.4115,
-      "step": 3830
-    },
-    {
-      "epoch": 1.839769728951787,
-      "grad_norm": 2.6580914984639237,
-      "learning_rate": 3.9033445619476415e-06,
-      "loss": 0.4362,
-      "step": 3835
-    },
-    {
-      "epoch": 1.8421683857040057,
-      "grad_norm": 2.1023106414276973,
-      "learning_rate": 3.8897286247812006e-06,
-      "loss": 0.4124,
-      "step": 3840
-    },
-    {
-      "epoch": 1.8445670424562244,
-      "grad_norm": 2.1445435272357445,
-      "learning_rate": 3.8761213426563546e-06,
-      "loss": 0.5047,
-      "step": 3845
-    },
-    {
-      "epoch": 1.8469656992084431,
-      "grad_norm": 2.2160537149648523,
-      "learning_rate": 3.862522821647702e-06,
-      "loss": 0.4063,
-      "step": 3850
-    },
-    {
-      "epoch": 1.849364355960662,
-      "grad_norm": 2.582471907598274,
-      "learning_rate": 3.848933167761549e-06,
-      "loss": 0.39,
-      "step": 3855
-    },
-    {
-      "epoch": 1.8517630127128808,
-      "grad_norm": 2.4913859988764173,
-      "learning_rate": 3.835352486935076e-06,
-      "loss": 0.4045,
-      "step": 3860
-    },
-    {
-      "epoch": 1.8541616694650995,
-      "grad_norm": 2.4347035193394366,
-      "learning_rate": 3.821780885035516e-06,
-      "loss": 0.4174,
-      "step": 3865
-    },
-    {
-      "epoch": 1.8565603262173183,
-      "grad_norm": 2.557816266559845,
-      "learning_rate": 3.808218467859326e-06,
-      "loss": 0.4846,
-      "step": 3870
-    },
-    {
-      "epoch": 1.8589589829695372,
-      "grad_norm": 2.871791500574788,
-      "learning_rate": 3.794665341131365e-06,
-      "loss": 0.3993,
-      "step": 3875
-    },
-    {
-      "epoch": 1.861357639721756,
-      "grad_norm": 4.453945141895481,
-      "learning_rate": 3.7811216105040705e-06,
-      "loss": 0.3666,
-      "step": 3880
-    },
-    {
-      "epoch": 1.8637562964739747,
-      "grad_norm": 2.7539613985240443,
-      "learning_rate": 3.7675873815566307e-06,
-      "loss": 0.4947,
-      "step": 3885
-    },
-    {
-      "epoch": 1.8661549532261934,
-      "grad_norm": 2.757681777789113,
-      "learning_rate": 3.754062759794165e-06,
-      "loss": 0.4364,
-      "step": 3890
-    },
-    {
-      "epoch": 1.8685536099784121,
-      "grad_norm": 1.8453147641515586,
-      "learning_rate": 3.740547850646902e-06,
-      "loss": 0.3905,
-      "step": 3895
-    },
-    {
-      "epoch": 1.8709522667306309,
-      "grad_norm": 4.16401248546863,
-      "learning_rate": 3.7270427594693565e-06,
-      "loss": 0.5428,
-      "step": 3900
-    },
-    {
-      "epoch": 1.8733509234828496,
-      "grad_norm": 2.7452054734057816,
-      "learning_rate": 3.713547591539504e-06,
-      "loss": 0.4806,
-      "step": 3905
-    },
-    {
-      "epoch": 1.8757495802350683,
-      "grad_norm": 2.188982964972483,
-      "learning_rate": 3.7000624520579677e-06,
-      "loss": 0.4278,
-      "step": 3910
-    },
-    {
-      "epoch": 1.878148236987287,
-      "grad_norm": 2.306468705586506,
-      "learning_rate": 3.6865874461471974e-06,
-      "loss": 0.4893,
-      "step": 3915
-    },
-    {
-      "epoch": 1.8805468937395058,
-      "grad_norm": 2.63525840513707,
-      "learning_rate": 3.6731226788506412e-06,
-      "loss": 0.4033,
-      "step": 3920
-    },
-    {
-      "epoch": 1.8829455504917245,
-      "grad_norm": 2.6391854567866573,
-      "learning_rate": 3.6596682551319347e-06,
-      "loss": 0.4328,
-      "step": 3925
-    },
-    {
-      "epoch": 1.8853442072439432,
-      "grad_norm": 2.3123066585387697,
-      "learning_rate": 3.6462242798740842e-06,
-      "loss": 0.4837,
-      "step": 3930
-    },
-    {
-      "epoch": 1.8877428639961622,
-      "grad_norm": 2.5028797266020484,
-      "learning_rate": 3.632790857878642e-06,
-      "loss": 0.3742,
-      "step": 3935
-    },
-    {
-      "epoch": 1.890141520748381,
-      "grad_norm": 2.0753848490152085,
-      "learning_rate": 3.6193680938648933e-06,
-      "loss": 0.3936,
-      "step": 3940
-    },
-    {
-      "epoch": 1.8925401775005997,
-      "grad_norm": 1.5576844963836263,
-      "learning_rate": 3.605956092469043e-06,
-      "loss": 0.47,
-      "step": 3945
-    },
-    {
-      "epoch": 1.8949388342528184,
-      "grad_norm": 1.7508643803207475,
-      "learning_rate": 3.5925549582433937e-06,
-      "loss": 0.3943,
-      "step": 3950
-    },
-    {
-      "epoch": 1.8973374910050371,
-      "grad_norm": 2.504931518208198,
-      "learning_rate": 3.5791647956555347e-06,
-      "loss": 0.4571,
-      "step": 3955
-    },
-    {
-      "epoch": 1.899736147757256,
-      "grad_norm": 1.372851159785718,
-      "learning_rate": 3.565785709087527e-06,
-      "loss": 0.3731,
-      "step": 3960
-    },
-    {
-      "epoch": 1.9021348045094748,
-      "grad_norm": 2.550831447418303,
-      "learning_rate": 3.5524178028350886e-06,
-      "loss": 0.4265,
-      "step": 3965
-    },
-    {
-      "epoch": 1.9045334612616935,
-      "grad_norm": 1.9108344497032907,
-      "learning_rate": 3.5390611811067844e-06,
-      "loss": 0.3443,
-      "step": 3970
-    },
-    {
-      "epoch": 1.9069321180139123,
-      "grad_norm": 2.112501662712423,
-      "learning_rate": 3.5257159480232085e-06,
-      "loss": 0.4774,
-      "step": 3975
-    },
-    {
-      "epoch": 1.909330774766131,
-      "grad_norm": 2.3842938517301517,
-      "learning_rate": 3.512382207616176e-06,
-      "loss": 0.5172,
-      "step": 3980
-    },
-    {
-      "epoch": 1.9117294315183497,
-      "grad_norm": 2.173550770298736,
-      "learning_rate": 3.4990600638279143e-06,
-      "loss": 0.4375,
-      "step": 3985
-    },
-    {
-      "epoch": 1.9141280882705685,
-      "grad_norm": 2.533900150370561,
-      "learning_rate": 3.4857496205102475e-06,
-      "loss": 0.4809,
-      "step": 3990
-    },
-    {
-      "epoch": 1.9165267450227872,
-      "grad_norm": 2.3636952079374325,
-      "learning_rate": 3.4724509814237894e-06,
-      "loss": 0.4117,
-      "step": 3995
-    },
-    {
-      "epoch": 1.918925401775006,
-      "grad_norm": 2.838190159310154,
-      "learning_rate": 3.4591642502371383e-06,
-      "loss": 0.4682,
-      "step": 4000
-    },
-    {
-      "epoch": 1.918925401775006,
-      "eval_loss": 0.8692654967308044,
-      "eval_runtime": 740.4824,
-      "eval_samples_per_second": 10.007,
-      "eval_steps_per_second": 0.627,
-      "step": 4000
-    },
-    {
-      "epoch": 1.9213240585272247,
-      "grad_norm": 2.1605145377053927,
-      "learning_rate": 3.445889530526061e-06,
-      "loss": 0.4229,
-      "step": 4005
-    },
-    {
-      "epoch": 1.9237227152794434,
-      "grad_norm": 2.2539865850184113,
-      "learning_rate": 3.4326269257726907e-06,
-      "loss": 0.4552,
-      "step": 4010
-    },
-    {
-      "epoch": 1.9261213720316621,
-      "grad_norm": 2.4596434765648127,
-      "learning_rate": 3.4193765393647217e-06,
-      "loss": 0.4016,
-      "step": 4015
-    },
-    {
-      "epoch": 1.928520028783881,
-      "grad_norm": 2.4021095095636045,
-      "learning_rate": 3.406138474594601e-06,
-      "loss": 0.4841,
-      "step": 4020
-    },
-    {
-      "epoch": 1.9309186855360998,
-      "grad_norm": 3.030531595041229,
-      "learning_rate": 3.3929128346587192e-06,
-      "loss": 0.4347,
-      "step": 4025
-    },
-    {
-      "epoch": 1.9333173422883185,
-      "grad_norm": 1.7489998010970578,
-      "learning_rate": 3.379699722656612e-06,
-      "loss": 0.4193,
-      "step": 4030
-    },
-    {
-      "epoch": 1.9357159990405373,
-      "grad_norm": 2.6910202006360446,
-      "learning_rate": 3.3664992415901575e-06,
-      "loss": 0.4323,
-      "step": 4035
-    },
-    {
-      "epoch": 1.9381146557927562,
-      "grad_norm": 3.0994878154800998,
-      "learning_rate": 3.3533114943627666e-06,
-      "loss": 0.4332,
-      "step": 4040
-    },
-    {
-      "epoch": 1.940513312544975,
-      "grad_norm": 2.31042002303757,
-      "learning_rate": 3.3401365837785828e-06,
-      "loss": 0.3942,
-      "step": 4045
-    },
-    {
-      "epoch": 1.9429119692971937,
-      "grad_norm": 2.233640476594205,
-      "learning_rate": 3.326974612541689e-06,
-      "loss": 0.4187,
-      "step": 4050
-    },
-    {
-      "epoch": 1.9453106260494124,
-      "grad_norm": 2.315235530807908,
-      "learning_rate": 3.313825683255295e-06,
-      "loss": 0.4322,
-      "step": 4055
-    },
-    {
-      "epoch": 1.9477092828016311,
-      "grad_norm": 2.453655450256884,
-      "learning_rate": 3.3006898984209436e-06,
-      "loss": 0.44,
-      "step": 4060
-    },
-    {
-      "epoch": 1.9501079395538499,
-      "grad_norm": 1.6854114253515358,
-      "learning_rate": 3.287567360437712e-06,
-      "loss": 0.439,
-      "step": 4065
-    },
-    {
-      "epoch": 1.9525065963060686,
-      "grad_norm": 2.0995266216210258,
-      "learning_rate": 3.274458171601414e-06,
-      "loss": 0.4386,
-      "step": 4070
-    },
-    {
-      "epoch": 1.9549052530582873,
-      "grad_norm": 2.947024685833257,
-      "learning_rate": 3.2613624341037994e-06,
-      "loss": 0.4262,
-      "step": 4075
-    },
-    {
-      "epoch": 1.957303909810506,
-      "grad_norm": 2.055447214443006,
-      "learning_rate": 3.248280250031758e-06,
-      "loss": 0.4234,
-      "step": 4080
-    },
-    {
-      "epoch": 1.9597025665627248,
-      "grad_norm": 2.404771167899297,
-      "learning_rate": 3.2352117213665247e-06,
-      "loss": 0.4254,
-      "step": 4085
-    },
-    {
-      "epoch": 1.9621012233149435,
-      "grad_norm": 2.5779058142571465,
-      "learning_rate": 3.222156949982889e-06,
-      "loss": 0.3543,
-      "step": 4090
-    },
-    {
-      "epoch": 1.9644998800671623,
-      "grad_norm": 2.713609918807594,
-      "learning_rate": 3.209116037648391e-06,
-      "loss": 0.4684,
-      "step": 4095
-    },
-    {
-      "epoch": 1.966898536819381,
-      "grad_norm": 2.479178615500214,
-      "learning_rate": 3.1960890860225364e-06,
-      "loss": 0.4405,
-      "step": 4100
-    },
-    {
-      "epoch": 1.9692971935716,
-      "grad_norm": 2.731958193705481,
-      "learning_rate": 3.1830761966560013e-06,
-      "loss": 0.4894,
-      "step": 4105
-    },
-    {
-      "epoch": 1.9716958503238187,
-      "grad_norm": 1.7562274667575501,
-      "learning_rate": 3.1700774709898403e-06,
-      "loss": 0.4007,
-      "step": 4110
-    },
-    {
-      "epoch": 1.9740945070760374,
-      "grad_norm": 2.3200173678085525,
-      "learning_rate": 3.157093010354692e-06,
-      "loss": 0.3679,
-      "step": 4115
-    },
-    {
-      "epoch": 1.9764931638282561,
-      "grad_norm": 2.4945622359580533,
-      "learning_rate": 3.1441229159700004e-06,
-      "loss": 0.4625,
-      "step": 4120
-    },
-    {
-      "epoch": 1.978891820580475,
-      "grad_norm": 2.488700609342745,
-      "learning_rate": 3.1311672889432145e-06,
-      "loss": 0.4021,
-      "step": 4125
-    },
-    {
-      "epoch": 1.9812904773326938,
-      "grad_norm": 1.3886751642557622,
-      "learning_rate": 3.1182262302690025e-06,
-      "loss": 0.3989,
-      "step": 4130
-    },
-    {
-      "epoch": 1.9836891340849125,
-      "grad_norm": 2.9028776554358866,
-      "learning_rate": 3.1052998408284664e-06,
-      "loss": 0.4477,
-      "step": 4135
-    },
-    {
-      "epoch": 1.9860877908371313,
-      "grad_norm": 2.4419026307903224,
-      "learning_rate": 3.0923882213883593e-06,
-      "loss": 0.4037,
-      "step": 4140
-    },
-    {
-      "epoch": 1.98848644758935,
-      "grad_norm": 2.8596964505015614,
-      "learning_rate": 3.0794914726002904e-06,
-      "loss": 0.4253,
-      "step": 4145
-    },
-    {
-      "epoch": 1.9908851043415687,
-      "grad_norm": 2.3119885246996614,
-      "learning_rate": 3.0666096949999493e-06,
-      "loss": 0.4408,
-      "step": 4150
-    },
-    {
-      "epoch": 1.9932837610937875,
-      "grad_norm": 1.603854256806702,
-      "learning_rate": 3.0537429890063177e-06,
-      "loss": 0.4348,
-      "step": 4155
-    },
-    {
-      "epoch": 1.9956824178460062,
-      "grad_norm": 2.247485386854169,
-      "learning_rate": 3.0408914549208867e-06,
-      "loss": 0.4137,
-      "step": 4160
-    },
-    {
-      "epoch": 1.998081074598225,
-      "grad_norm": 2.323713545683984,
-      "learning_rate": 3.0280551929268763e-06,
-      "loss": 0.429,
-      "step": 4165
-    },
-    {
-      "epoch": 2.0004797313504437,
-      "grad_norm": 1.7492359777642301,
-      "learning_rate": 3.0152343030884586e-06,
-      "loss": 0.4368,
-      "step": 4170
-    },
-    {
-      "epoch": 2.0028783881026624,
-      "grad_norm": 1.2104549778941618,
-      "learning_rate": 3.002428885349965e-06,
-      "loss": 0.1384,
-      "step": 4175
-    },
-    {
-      "epoch": 2.005277044854881,
-      "grad_norm": 1.2533431726821433,
-      "learning_rate": 2.9896390395351215e-06,
-      "loss": 0.1307,
-      "step": 4180
-    },
-    {
-      "epoch": 2.0076757016071,
-      "grad_norm": 1.6153822677232479,
-      "learning_rate": 2.9768648653462617e-06,
-      "loss": 0.1509,
-      "step": 4185
-    },
-    {
-      "epoch": 2.0100743583593186,
-      "grad_norm": 1.9706051369619293,
-      "learning_rate": 2.964106462363551e-06,
-      "loss": 0.1407,
-      "step": 4190
-    },
-    {
-      "epoch": 2.0124730151115378,
-      "grad_norm": 1.5835329547161818,
-      "learning_rate": 2.9513639300442156e-06,
-      "loss": 0.1423,
-      "step": 4195
-    },
-    {
-      "epoch": 2.0148716718637565,
-      "grad_norm": 1.4932494900950153,
-      "learning_rate": 2.9386373677217583e-06,
-      "loss": 0.1328,
-      "step": 4200
-    },
-    {
-      "epoch": 2.017270328615975,
-      "grad_norm": 1.373976549075157,
-      "learning_rate": 2.925926874605192e-06,
-      "loss": 0.1239,
-      "step": 4205
-    },
-    {
-      "epoch": 2.019668985368194,
-      "grad_norm": 2.6678759030139245,
-      "learning_rate": 2.9132325497782637e-06,
-      "loss": 0.1576,
-      "step": 4210
-    },
-    {
-      "epoch": 2.0220676421204127,
-      "grad_norm": 1.8908086442023389,
-      "learning_rate": 2.9005544921986774e-06,
-      "loss": 0.1438,
-      "step": 4215
-    },
-    {
-      "epoch": 2.0244662988726314,
-      "grad_norm": 1.4406396259902392,
-      "learning_rate": 2.8878928006973326e-06,
-      "loss": 0.1172,
-      "step": 4220
-    },
-    {
-      "epoch": 2.02686495562485,
-      "grad_norm": 3.144383893347305,
-      "learning_rate": 2.8752475739775456e-06,
-      "loss": 0.1248,
-      "step": 4225
-    },
-    {
-      "epoch": 2.029263612377069,
-      "grad_norm": 1.5254881527613178,
-      "learning_rate": 2.862618910614281e-06,
-      "loss": 0.1166,
-      "step": 4230
-    },
-    {
-      "epoch": 2.0316622691292876,
-      "grad_norm": 1.5380395388583785,
-      "learning_rate": 2.8500069090533845e-06,
-      "loss": 0.122,
-      "step": 4235
-    },
-    {
-      "epoch": 2.0340609258815063,
-      "grad_norm": 1.8517578832693309,
-      "learning_rate": 2.837411667610823e-06,
-      "loss": 0.1264,
-      "step": 4240
-    },
-    {
-      "epoch": 2.036459582633725,
-      "grad_norm": 1.5611087867626268,
-      "learning_rate": 2.8248332844719033e-06,
-      "loss": 0.1273,
-      "step": 4245
-    },
-    {
-      "epoch": 2.038858239385944,
-      "grad_norm": 1.9829636176688326,
-      "learning_rate": 2.8122718576905174e-06,
-      "loss": 0.1211,
-      "step": 4250
-    },
-    {
-      "epoch": 2.0412568961381625,
-      "grad_norm": 2.336890942831196,
-      "learning_rate": 2.7997274851883748e-06,
-      "loss": 0.1329,
-      "step": 4255
-    },
-    {
-      "epoch": 2.0436555528903813,
-      "grad_norm": 2.123217631063763,
-      "learning_rate": 2.7872002647542395e-06,
-      "loss": 0.1294,
-      "step": 4260
-    },
-    {
-      "epoch": 2.0460542096426,
-      "grad_norm": 1.3603389345629862,
-      "learning_rate": 2.7746902940431665e-06,
-      "loss": 0.1183,
-      "step": 4265
-    },
-    {
-      "epoch": 2.0484528663948187,
-      "grad_norm": 2.9582290350986686,
-      "learning_rate": 2.7621976705757446e-06,
-      "loss": 0.1573,
-      "step": 4270
-    },
-    {
-      "epoch": 2.0508515231470374,
-      "grad_norm": 1.6318991512169887,
-      "learning_rate": 2.7497224917373355e-06,
-      "loss": 0.1487,
-      "step": 4275
-    },
-    {
-      "epoch": 2.0532501798992566,
-      "grad_norm": 1.317567144216617,
-      "learning_rate": 2.7372648547773063e-06,
-      "loss": 0.1214,
-      "step": 4280
-    },
-    {
-      "epoch": 2.0556488366514754,
-      "grad_norm": 1.4968986861944702,
-      "learning_rate": 2.7248248568082796e-06,
-      "loss": 0.1311,
-      "step": 4285
-    },
-    {
-      "epoch": 2.058047493403694,
-      "grad_norm": 2.066601497926396,
-      "learning_rate": 2.712402594805377e-06,
-      "loss": 0.132,
-      "step": 4290
-    },
-    {
-      "epoch": 2.060446150155913,
-      "grad_norm": 2.0874309705799505,
-      "learning_rate": 2.6999981656054587e-06,
-      "loss": 0.1406,
-      "step": 4295
-    },
-    {
-      "epoch": 2.0628448069081315,
-      "grad_norm": 1.2870381784002116,
-      "learning_rate": 2.687611665906369e-06,
-      "loss": 0.1385,
-      "step": 4300
-    },
-    {
-      "epoch": 2.0652434636603503,
-      "grad_norm": 1.5033982222031699,
-      "learning_rate": 2.675243192266189e-06,
-      "loss": 0.1082,
-      "step": 4305
-    },
-    {
-      "epoch": 2.067642120412569,
-      "grad_norm": 1.4861250410016393,
-      "learning_rate": 2.662892841102474e-06,
-      "loss": 0.1354,
-      "step": 4310
-    },
-    {
-      "epoch": 2.0700407771647877,
-      "grad_norm": 1.3891683331641131,
-      "learning_rate": 2.65056070869151e-06,
-      "loss": 0.1365,
-      "step": 4315
-    },
-    {
-      "epoch": 2.0724394339170065,
-      "grad_norm": 1.6768904681198002,
-      "learning_rate": 2.6382468911675553e-06,
-      "loss": 0.1319,
-      "step": 4320
-    },
-    {
-      "epoch": 2.074838090669225,
-      "grad_norm": 2.125719533478666,
-      "learning_rate": 2.6259514845221045e-06,
-      "loss": 0.1524,
-      "step": 4325
-    },
-    {
-      "epoch": 2.077236747421444,
-      "grad_norm": 1.724095586385624,
-      "learning_rate": 2.6136745846031232e-06,
-      "loss": 0.1382,
-      "step": 4330
-    },
-    {
-      "epoch": 2.0796354041736627,
-      "grad_norm": 1.5007407801751065,
-      "learning_rate": 2.6014162871143123e-06,
-      "loss": 0.1319,
-      "step": 4335
-    },
-    {
-      "epoch": 2.0820340609258814,
-      "grad_norm": 1.7606564742094755,
-      "learning_rate": 2.589176687614357e-06,
-      "loss": 0.125,
-      "step": 4340
-    },
-    {
-      "epoch": 2.0844327176781,
-      "grad_norm": 1.8469687804386385,
-      "learning_rate": 2.5769558815161887e-06,
-      "loss": 0.1351,
-      "step": 4345
-    },
-    {
-      "epoch": 2.086831374430319,
-      "grad_norm": 1.8472855123668117,
-      "learning_rate": 2.5647539640862316e-06,
-      "loss": 0.1246,
-      "step": 4350
-    },
-    {
-      "epoch": 2.0892300311825376,
-      "grad_norm": 3.765807007677004,
-      "learning_rate": 2.5525710304436658e-06,
-      "loss": 0.1063,
-      "step": 4355
-    },
-    {
-      "epoch": 2.0916286879347563,
-      "grad_norm": 1.589551356385384,
-      "learning_rate": 2.540407175559684e-06,
-      "loss": 0.1129,
-      "step": 4360
-    },
-    {
-      "epoch": 2.0940273446869755,
-      "grad_norm": 1.7923356652831175,
-      "learning_rate": 2.5282624942567523e-06,
-      "loss": 0.119,
-      "step": 4365
-    },
-    {
-      "epoch": 2.096426001439194,
-      "grad_norm": 2.4656699199229273,
-      "learning_rate": 2.5161370812078687e-06,
-      "loss": 0.1119,
-      "step": 4370
-    },
-    {
-      "epoch": 2.098824658191413,
-      "grad_norm": 3.4413079643746376,
-      "learning_rate": 2.504031030935832e-06,
-      "loss": 0.1495,
-      "step": 4375
-    },
-    {
-      "epoch": 2.1012233149436317,
-      "grad_norm": 1.5898426640404113,
-      "learning_rate": 2.491944437812497e-06,
-      "loss": 0.1322,
-      "step": 4380
-    },
-    {
-      "epoch": 2.1036219716958504,
-      "grad_norm": 1.305705593587179,
-      "learning_rate": 2.4798773960580403e-06,
-      "loss": 0.1462,
-      "step": 4385
-    },
-    {
-      "epoch": 2.106020628448069,
-      "grad_norm": 1.54107319224433,
-      "learning_rate": 2.4678299997402276e-06,
-      "loss": 0.1191,
-      "step": 4390
-    },
-    {
-      "epoch": 2.108419285200288,
-      "grad_norm": 1.4056141128475004,
-      "learning_rate": 2.4558023427736817e-06,
-      "loss": 0.0998,
-      "step": 4395
-    },
-    {
-      "epoch": 2.1108179419525066,
-      "grad_norm": 1.5206415072257935,
-      "learning_rate": 2.4437945189191464e-06,
-      "loss": 0.1625,
-      "step": 4400
-    },
-    {
-      "epoch": 2.1132165987047253,
-      "grad_norm": 1.7861408291312562,
-      "learning_rate": 2.431806621782758e-06,
-      "loss": 0.1334,
-      "step": 4405
-    },
-    {
-      "epoch": 2.115615255456944,
-      "grad_norm": 1.4757372455135827,
-      "learning_rate": 2.4198387448153205e-06,
-      "loss": 0.1038,
-      "step": 4410
-    },
-    {
-      "epoch": 2.118013912209163,
-      "grad_norm": 1.9894165818208251,
-      "learning_rate": 2.407890981311566e-06,
-      "loss": 0.1199,
-      "step": 4415
-    },
-    {
-      "epoch": 2.1204125689613815,
-      "grad_norm": 2.0026704399459336,
-      "learning_rate": 2.3959634244094336e-06,
-      "loss": 0.1275,
-      "step": 4420
-    },
-    {
-      "epoch": 2.1228112257136003,
-      "grad_norm": 1.5674286742154901,
-      "learning_rate": 2.38405616708935e-06,
-      "loss": 0.1255,
-      "step": 4425
-    },
-    {
-      "epoch": 2.125209882465819,
-      "grad_norm": 1.5064556826612898,
-      "learning_rate": 2.37216930217349e-06,
-      "loss": 0.1347,
-      "step": 4430
-    },
-    {
-      "epoch": 2.1276085392180377,
-      "grad_norm": 1.8166945572467463,
-      "learning_rate": 2.3603029223250657e-06,
-      "loss": 0.1168,
-      "step": 4435
-    },
-    {
-      "epoch": 2.1300071959702565,
-      "grad_norm": 1.4920164777188665,
-      "learning_rate": 2.348457120047597e-06,
-      "loss": 0.1164,
-      "step": 4440
-    },
-    {
-      "epoch": 2.132405852722475,
-      "grad_norm": 1.8683865713478673,
-      "learning_rate": 2.336631987684192e-06,
-      "loss": 0.1335,
-      "step": 4445
-    },
-    {
-      "epoch": 2.1348045094746944,
-      "grad_norm": 1.5965976402737696,
-      "learning_rate": 2.3248276174168333e-06,
-      "loss": 0.1333,
-      "step": 4450
-    },
-    {
-      "epoch": 2.137203166226913,
-      "grad_norm": 2.0761993267672527,
-      "learning_rate": 2.3130441012656485e-06,
-      "loss": 0.117,
-      "step": 4455
-    },
-    {
-      "epoch": 2.139601822979132,
-      "grad_norm": 1.9907693084636011,
-      "learning_rate": 2.3012815310882e-06,
-      "loss": 0.1529,
-      "step": 4460
-    },
-    {
-      "epoch": 2.1420004797313505,
-      "grad_norm": 1.427001377831548,
-      "learning_rate": 2.289539998578768e-06,
-      "loss": 0.1144,
-      "step": 4465
-    },
-    {
-      "epoch": 2.1443991364835693,
-      "grad_norm": 1.9701633935879637,
-      "learning_rate": 2.2778195952676345e-06,
-      "loss": 0.1346,
-      "step": 4470
-    },
-    {
-      "epoch": 2.146797793235788,
-      "grad_norm": 1.8539619958760687,
-      "learning_rate": 2.266120412520369e-06,
-      "loss": 0.1171,
-      "step": 4475
-    },
-    {
-      "epoch": 2.1491964499880067,
-      "grad_norm": 1.3067217228012888,
-      "learning_rate": 2.25444254153712e-06,
-      "loss": 0.1147,
-      "step": 4480
-    },
-    {
-      "epoch": 2.1515951067402255,
-      "grad_norm": 2.077001334839875,
-      "learning_rate": 2.2427860733519034e-06,
-      "loss": 0.1338,
-      "step": 4485
-    },
-    {
-      "epoch": 2.153993763492444,
-      "grad_norm": 2.696463098963866,
-      "learning_rate": 2.2311510988318865e-06,
-      "loss": 0.1376,
-      "step": 4490
-    },
-    {
-      "epoch": 2.156392420244663,
-      "grad_norm": 1.4341789258623368,
-      "learning_rate": 2.2195377086766862e-06,
-      "loss": 0.1102,
-      "step": 4495
-    },
-    {
-      "epoch": 2.1587910769968817,
-      "grad_norm": 2.0711628473207466,
-      "learning_rate": 2.2079459934176596e-06,
-      "loss": 0.1632,
-      "step": 4500
-    },
-    {
-      "epoch": 2.1611897337491004,
-      "grad_norm": 1.6336280241127499,
-      "learning_rate": 2.1963760434172e-06,
-      "loss": 0.1233,
-      "step": 4505
-    },
-    {
-      "epoch": 2.163588390501319,
-      "grad_norm": 1.8495280260216118,
-      "learning_rate": 2.184827948868028e-06,
-      "loss": 0.1315,
-      "step": 4510
-    },
-    {
-      "epoch": 2.165987047253538,
-      "grad_norm": 1.9526925374165924,
-      "learning_rate": 2.173301799792497e-06,
-      "loss": 0.116,
-      "step": 4515
-    },
-    {
-      "epoch": 2.1683857040057566,
-      "grad_norm": 3.3827621253228974,
-      "learning_rate": 2.161797686041882e-06,
-      "loss": 0.1279,
-      "step": 4520
-    },
-    {
-      "epoch": 2.1707843607579758,
-      "grad_norm": 1.5109194055996138,
-      "learning_rate": 2.1503156972956823e-06,
-      "loss": 0.1154,
-      "step": 4525
-    },
-    {
-      "epoch": 2.1731830175101945,
-      "grad_norm": 2.2477580547961473,
-      "learning_rate": 2.1388559230609278e-06,
-      "loss": 0.1401,
-      "step": 4530
-    },
-    {
-      "epoch": 2.1755816742624132,
-      "grad_norm": 2.0474896431675096,
-      "learning_rate": 2.1274184526714733e-06,
-      "loss": 0.1051,
-      "step": 4535
-    },
-    {
-      "epoch": 2.177980331014632,
-      "grad_norm": 1.5451984571822854,
-      "learning_rate": 2.1160033752873053e-06,
-      "loss": 0.1326,
-      "step": 4540
-    },
-    {
-      "epoch": 2.1803789877668507,
-      "grad_norm": 2.126613331135665,
-      "learning_rate": 2.1046107798938493e-06,
-      "loss": 0.1549,
-      "step": 4545
-    },
-    {
-      "epoch": 2.1827776445190694,
-      "grad_norm": 1.4317126044877562,
-      "learning_rate": 2.09324075530127e-06,
-      "loss": 0.1164,
-      "step": 4550
-    },
-    {
-      "epoch": 2.185176301271288,
-      "grad_norm": 2.1177926714585587,
-      "learning_rate": 2.0818933901437905e-06,
-      "loss": 0.1415,
-      "step": 4555
-    },
-    {
-      "epoch": 2.187574958023507,
-      "grad_norm": 1.283216594231968,
-      "learning_rate": 2.0705687728789876e-06,
-      "loss": 0.1271,
-      "step": 4560
-    },
-    {
-      "epoch": 2.1899736147757256,
-      "grad_norm": 1.2736682129552483,
-      "learning_rate": 2.0592669917871093e-06,
-      "loss": 0.1424,
-      "step": 4565
-    },
-    {
-      "epoch": 2.1923722715279443,
-      "grad_norm": 1.829115525077864,
-      "learning_rate": 2.0479881349703885e-06,
-      "loss": 0.1298,
-      "step": 4570
-    },
-    {
-      "epoch": 2.194770928280163,
-      "grad_norm": 1.795266905694527,
-      "learning_rate": 2.0367322903523483e-06,
-      "loss": 0.1233,
-      "step": 4575
-    },
-    {
-      "epoch": 2.197169585032382,
-      "grad_norm": 1.2760353352996483,
-      "learning_rate": 2.0254995456771303e-06,
-      "loss": 0.1279,
-      "step": 4580
-    },
-    {
-      "epoch": 2.1995682417846005,
-      "grad_norm": 1.7111968554436063,
-      "learning_rate": 2.0142899885087937e-06,
-      "loss": 0.1275,
-      "step": 4585
-    },
-    {
-      "epoch": 2.2019668985368193,
-      "grad_norm": 1.4094962158008097,
-      "learning_rate": 2.0031037062306467e-06,
-      "loss": 0.1219,
-      "step": 4590
-    },
-    {
-      "epoch": 2.204365555289038,
-      "grad_norm": 1.4066646174683728,
-      "learning_rate": 1.9919407860445585e-06,
-      "loss": 0.1231,
-      "step": 4595
-    },
-    {
-      "epoch": 2.2067642120412567,
-      "grad_norm": 1.6071130525622066,
-      "learning_rate": 1.9808013149702776e-06,
-      "loss": 0.1323,
-      "step": 4600
-    },
-    {
-      "epoch": 2.2091628687934755,
-      "grad_norm": 2.547322876356862,
-      "learning_rate": 1.9696853798447603e-06,
-      "loss": 0.127,
-      "step": 4605
-    },
-    {
-      "epoch": 2.2115615255456946,
-      "grad_norm": 1.5483008900838984,
-      "learning_rate": 1.9585930673214885e-06,
-      "loss": 0.1074,
-      "step": 4610
-    },
-    {
-      "epoch": 2.2139601822979134,
-      "grad_norm": 1.5382419648678334,
-      "learning_rate": 1.9475244638697943e-06,
-      "loss": 0.1526,
-      "step": 4615
-    },
-    {
-      "epoch": 2.216358839050132,
-      "grad_norm": 2.8665464265307534,
-      "learning_rate": 1.936479655774193e-06,
-      "loss": 0.1432,
-      "step": 4620
-    },
-    {
-      "epoch": 2.218757495802351,
-      "grad_norm": 1.3472673683557654,
-      "learning_rate": 1.9254587291336975e-06,
-      "loss": 0.1111,
-      "step": 4625
-    },
-    {
-      "epoch": 2.2211561525545696,
-      "grad_norm": 3.2736752179779467,
-      "learning_rate": 1.9144617698611616e-06,
-      "loss": 0.1251,
-      "step": 4630
-    },
-    {
-      "epoch": 2.2235548093067883,
-      "grad_norm": 1.5309415711503425,
-      "learning_rate": 1.9034888636825976e-06,
-      "loss": 0.107,
-      "step": 4635
-    },
-    {
-      "epoch": 2.225953466059007,
-      "grad_norm": 1.6722342536437953,
-      "learning_rate": 1.8925400961365164e-06,
-      "loss": 0.1271,
-      "step": 4640
-    },
-    {
-      "epoch": 2.2283521228112257,
-      "grad_norm": 1.2072514830382435,
-      "learning_rate": 1.8816155525732566e-06,
-      "loss": 0.1161,
-      "step": 4645
-    },
-    {
-      "epoch": 2.2307507795634445,
-      "grad_norm": 1.8683460607192535,
-      "learning_rate": 1.870715318154322e-06,
-      "loss": 0.1153,
-      "step": 4650
-    },
-    {
-      "epoch": 2.233149436315663,
-      "grad_norm": 1.8005183568624918,
-      "learning_rate": 1.8598394778517125e-06,
-      "loss": 0.1312,
-      "step": 4655
-    },
-    {
-      "epoch": 2.235548093067882,
-      "grad_norm": 1.803493471243067,
-      "learning_rate": 1.8489881164472722e-06,
-      "loss": 0.144,
-      "step": 4660
-    },
-    {
-      "epoch": 2.2379467498201007,
-      "grad_norm": 1.2566663432508494,
-      "learning_rate": 1.8381613185320158e-06,
-      "loss": 0.1136,
-      "step": 4665
-    },
-    {
-      "epoch": 2.2403454065723194,
-      "grad_norm": 1.7624206247680096,
-      "learning_rate": 1.827359168505477e-06,
-      "loss": 0.1045,
-      "step": 4670
-    },
-    {
-      "epoch": 2.242744063324538,
-      "grad_norm": 1.5566145746764684,
-      "learning_rate": 1.8165817505750483e-06,
-      "loss": 0.0988,
-      "step": 4675
-    },
-    {
-      "epoch": 2.245142720076757,
-      "grad_norm": 1.5186784492397856,
-      "learning_rate": 1.8058291487553243e-06,
-      "loss": 0.1304,
-      "step": 4680
-    },
-    {
-      "epoch": 2.2475413768289756,
-      "grad_norm": 1.825383805349096,
-      "learning_rate": 1.795101446867451e-06,
-      "loss": 0.1324,
-      "step": 4685
-    },
-    {
-      "epoch": 2.2499400335811943,
-      "grad_norm": 1.8077386836692138,
-      "learning_rate": 1.784398728538463e-06,
-      "loss": 0.1119,
-      "step": 4690
-    },
-    {
-      "epoch": 2.2523386903334135,
-      "grad_norm": 1.3139307927703525,
-      "learning_rate": 1.773721077200643e-06,
-      "loss": 0.1284,
-      "step": 4695
-    },
-    {
-      "epoch": 2.2547373470856322,
-      "grad_norm": 1.9359776479828592,
-      "learning_rate": 1.7630685760908623e-06,
-      "loss": 0.1128,
-      "step": 4700
-    },
-    {
-      "epoch": 2.257136003837851,
-      "grad_norm": 1.5512067156614446,
-      "learning_rate": 1.7524413082499354e-06,
-      "loss": 0.1199,
-      "step": 4705
-    },
-    {
-      "epoch": 2.2595346605900697,
-      "grad_norm": 2.06959679504393,
-      "learning_rate": 1.7418393565219737e-06,
-      "loss": 0.1325,
-      "step": 4710
-    },
-    {
-      "epoch": 2.2619333173422884,
-      "grad_norm": 2.2066122371220804,
-      "learning_rate": 1.7312628035537388e-06,
-      "loss": 0.1191,
-      "step": 4715
-    },
-    {
-      "epoch": 2.264331974094507,
-      "grad_norm": 2.1618555873923513,
-      "learning_rate": 1.720711731793996e-06,
-      "loss": 0.141,
-      "step": 4720
-    },
-    {
-      "epoch": 2.266730630846726,
-      "grad_norm": 2.2477809764108643,
-      "learning_rate": 1.710186223492878e-06,
-      "loss": 0.1607,
-      "step": 4725
-    },
-    {
-      "epoch": 2.2691292875989446,
-      "grad_norm": 2.445594160005071,
-      "learning_rate": 1.6996863607012337e-06,
-      "loss": 0.1241,
-      "step": 4730
-    },
-    {
-      "epoch": 2.2715279443511633,
-      "grad_norm": 2.2636372479823024,
-      "learning_rate": 1.6892122252700005e-06,
-      "loss": 0.1471,
-      "step": 4735
-    },
-    {
-      "epoch": 2.273926601103382,
-      "grad_norm": 2.165613571335606,
-      "learning_rate": 1.6787638988495548e-06,
-      "loss": 0.1171,
-      "step": 4740
-    },
-    {
-      "epoch": 2.276325257855601,
-      "grad_norm": 1.5501086091890792,
-      "learning_rate": 1.668341462889083e-06,
-      "loss": 0.1152,
-      "step": 4745
-    },
-    {
-      "epoch": 2.2787239146078195,
-      "grad_norm": 2.324272542162646,
-      "learning_rate": 1.657944998635944e-06,
-      "loss": 0.1413,
-      "step": 4750
-    },
-    {
-      "epoch": 2.2811225713600383,
-      "grad_norm": 1.3705751476924688,
-      "learning_rate": 1.6475745871350356e-06,
-      "loss": 0.1255,
-      "step": 4755
-    },
-    {
-      "epoch": 2.283521228112257,
-      "grad_norm": 1.6009570905341144,
-      "learning_rate": 1.6372303092281616e-06,
-      "loss": 0.1013,
-      "step": 4760
-    },
-    {
-      "epoch": 2.2859198848644757,
-      "grad_norm": 2.0897856937356054,
-      "learning_rate": 1.6269122455534097e-06,
-      "loss": 0.1121,
-      "step": 4765
-    },
-    {
-      "epoch": 2.2883185416166945,
-      "grad_norm": 2.0733573382322046,
-      "learning_rate": 1.6166204765445093e-06,
-      "loss": 0.1211,
-      "step": 4770
-    },
-    {
-      "epoch": 2.290717198368913,
-      "grad_norm": 1.6760313776999054,
-      "learning_rate": 1.6063550824302143e-06,
-      "loss": 0.147,
-      "step": 4775
-    },
-    {
-      "epoch": 2.2931158551211324,
-      "grad_norm": 1.4628986967241502,
-      "learning_rate": 1.5961161432336736e-06,
-      "loss": 0.1017,
-      "step": 4780
-    },
-    {
-      "epoch": 2.295514511873351,
-      "grad_norm": 1.5119823167350506,
-      "learning_rate": 1.585903738771813e-06,
-      "loss": 0.1186,
-      "step": 4785
-    },
-    {
-      "epoch": 2.29791316862557,
-      "grad_norm": 2.114617503567997,
-      "learning_rate": 1.5757179486547043e-06,
-      "loss": 0.1107,
-      "step": 4790
-    },
-    {
-      "epoch": 2.3003118253777886,
-      "grad_norm": 2.1299147651150254,
-      "learning_rate": 1.5655588522849486e-06,
-      "loss": 0.1228,
-      "step": 4795
-    },
-    {
-      "epoch": 2.3027104821300073,
-      "grad_norm": 2.0091940656270784,
-      "learning_rate": 1.555426528857063e-06,
-      "loss": 0.1328,
-      "step": 4800
-    },
-    {
-      "epoch": 2.305109138882226,
-      "grad_norm": 2.278463048269149,
-      "learning_rate": 1.5453210573568523e-06,
-      "loss": 0.0995,
-      "step": 4805
-    },
-    {
-      "epoch": 2.3075077956344447,
-      "grad_norm": 2.0911503889315672,
-      "learning_rate": 1.5352425165608025e-06,
-      "loss": 0.1281,
-      "step": 4810
-    },
-    {
-      "epoch": 2.3099064523866635,
-      "grad_norm": 2.2008586407114072,
-      "learning_rate": 1.5251909850354613e-06,
-      "loss": 0.1319,
-      "step": 4815
-    },
-    {
-      "epoch": 2.312305109138882,
-      "grad_norm": 1.7793237602426986,
-      "learning_rate": 1.5151665411368288e-06,
-      "loss": 0.1015,
-      "step": 4820
-    },
-    {
-      "epoch": 2.314703765891101,
-      "grad_norm": 1.892162960562313,
-      "learning_rate": 1.5051692630097426e-06,
-      "loss": 0.1213,
-      "step": 4825
-    },
-    {
-      "epoch": 2.3171024226433197,
-      "grad_norm": 1.454131556874955,
-      "learning_rate": 1.4951992285872779e-06,
-      "loss": 0.1135,
-      "step": 4830
-    },
-    {
-      "epoch": 2.3195010793955384,
-      "grad_norm": 1.9721817974769122,
-      "learning_rate": 1.4852565155901254e-06,
-      "loss": 0.1032,
-      "step": 4835
-    },
-    {
-      "epoch": 2.321899736147757,
-      "grad_norm": 1.5362146521829505,
-      "learning_rate": 1.4753412015260017e-06,
-      "loss": 0.15,
-      "step": 4840
-    },
-    {
-      "epoch": 2.324298392899976,
-      "grad_norm": 1.883056468435468,
-      "learning_rate": 1.46545336368903e-06,
-      "loss": 0.1357,
-      "step": 4845
-    },
-    {
-      "epoch": 2.3266970496521946,
-      "grad_norm": 1.5230339355008213,
-      "learning_rate": 1.4555930791591483e-06,
-      "loss": 0.1174,
-      "step": 4850
-    },
-    {
-      "epoch": 2.3290957064044138,
-      "grad_norm": 2.0634402334235427,
-      "learning_rate": 1.4457604248015039e-06,
-      "loss": 0.1594,
-      "step": 4855
-    },
-    {
-      "epoch": 2.331494363156632,
-      "grad_norm": 2.5978427352784346,
-      "learning_rate": 1.4359554772658551e-06,
-      "loss": 0.1453,
-      "step": 4860
-    },
-    {
-      "epoch": 2.3338930199088512,
-      "grad_norm": 1.4828036272318572,
-      "learning_rate": 1.4261783129859712e-06,
-      "loss": 0.1135,
-      "step": 4865
-    },
-    {
-      "epoch": 2.33629167666107,
-      "grad_norm": 1.6377747205147426,
-      "learning_rate": 1.4164290081790449e-06,
-      "loss": 0.131,
-      "step": 4870
-    },
-    {
-      "epoch": 2.3386903334132887,
-      "grad_norm": 1.3431874234243162,
-      "learning_rate": 1.406707638845088e-06,
-      "loss": 0.0933,
-      "step": 4875
-    },
-    {
-      "epoch": 2.3410889901655074,
-      "grad_norm": 1.7396553876263976,
-      "learning_rate": 1.3970142807663434e-06,
-      "loss": 0.1035,
-      "step": 4880
-    },
-    {
-      "epoch": 2.343487646917726,
-      "grad_norm": 1.038152016301309,
-      "learning_rate": 1.3873490095066933e-06,
-      "loss": 0.1312,
-      "step": 4885
-    },
-    {
-      "epoch": 2.345886303669945,
-      "grad_norm": 1.9332345560468491,
-      "learning_rate": 1.3777119004110773e-06,
-      "loss": 0.119,
-      "step": 4890
-    },
-    {
-      "epoch": 2.3482849604221636,
-      "grad_norm": 1.2020052938888175,
-      "learning_rate": 1.3681030286048913e-06,
-      "loss": 0.1247,
-      "step": 4895
-    },
-    {
-      "epoch": 2.3506836171743823,
-      "grad_norm": 1.3034758430102855,
-      "learning_rate": 1.3585224689934123e-06,
-      "loss": 0.1417,
-      "step": 4900
-    },
-    {
-      "epoch": 2.353082273926601,
-      "grad_norm": 1.6110679516673296,
-      "learning_rate": 1.348970296261214e-06,
-      "loss": 0.1232,
-      "step": 4905
-    },
-    {
-      "epoch": 2.35548093067882,
-      "grad_norm": 1.3235836550393674,
-      "learning_rate": 1.3394465848715787e-06,
-      "loss": 0.1143,
-      "step": 4910
-    },
-    {
-      "epoch": 2.3578795874310385,
-      "grad_norm": 1.7551178365926876,
-      "learning_rate": 1.3299514090659216e-06,
-      "loss": 0.1253,
-      "step": 4915
-    },
-    {
-      "epoch": 2.3602782441832573,
-      "grad_norm": 1.8650654499857218,
-      "learning_rate": 1.320484842863211e-06,
-      "loss": 0.1238,
-      "step": 4920
-    },
-    {
-      "epoch": 2.362676900935476,
-      "grad_norm": 1.8525194939207048,
-      "learning_rate": 1.311046960059391e-06,
-      "loss": 0.1579,
-      "step": 4925
-    },
-    {
-      "epoch": 2.3650755576876947,
-      "grad_norm": 1.5852491550316363,
-      "learning_rate": 1.301637834226806e-06,
-      "loss": 0.1429,
-      "step": 4930
-    },
-    {
-      "epoch": 2.3674742144399135,
-      "grad_norm": 1.429254229854141,
-      "learning_rate": 1.2922575387136266e-06,
-      "loss": 0.1468,
-      "step": 4935
-    },
-    {
-      "epoch": 2.3698728711921326,
-      "grad_norm": 1.4466523780422678,
-      "learning_rate": 1.2829061466432807e-06,
-      "loss": 0.1008,
-      "step": 4940
-    },
-    {
-      "epoch": 2.372271527944351,
-      "grad_norm": 1.5252773310736942,
-      "learning_rate": 1.2735837309138827e-06,
-      "loss": 0.1144,
-      "step": 4945
-    },
-    {
-      "epoch": 2.37467018469657,
-      "grad_norm": 1.5855398691958067,
-      "learning_rate": 1.2642903641976595e-06,
-      "loss": 0.1279,
-      "step": 4950
-    },
-    {
-      "epoch": 2.377068841448789,
-      "grad_norm": 1.9995841417171394,
-      "learning_rate": 1.2550261189403912e-06,
-      "loss": 0.1472,
-      "step": 4955
-    },
-    {
-      "epoch": 2.3794674982010076,
-      "grad_norm": 1.8707045496556725,
-      "learning_rate": 1.2457910673608431e-06,
-      "loss": 0.122,
-      "step": 4960
-    },
-    {
-      "epoch": 2.3818661549532263,
-      "grad_norm": 1.7079603628945372,
-      "learning_rate": 1.2365852814502044e-06,
-      "loss": 0.1363,
-      "step": 4965
-    },
-    {
-      "epoch": 2.384264811705445,
-      "grad_norm": 1.739198799924586,
-      "learning_rate": 1.2274088329715218e-06,
-      "loss": 0.114,
-      "step": 4970
-    },
-    {
-      "epoch": 2.3866634684576638,
-      "grad_norm": 1.7400466578261822,
-      "learning_rate": 1.2182617934591518e-06,
-      "loss": 0.1263,
-      "step": 4975
-    },
-    {
-      "epoch": 2.3890621252098825,
-      "grad_norm": 1.7442631450776291,
-      "learning_rate": 1.209144234218188e-06,
-      "loss": 0.1183,
-      "step": 4980
-    },
-    {
-      "epoch": 2.391460781962101,
-      "grad_norm": 1.453171635685411,
-      "learning_rate": 1.2000562263239147e-06,
-      "loss": 0.1229,
-      "step": 4985
-    },
-    {
-      "epoch": 2.39385943871432,
-      "grad_norm": 1.774275653754628,
-      "learning_rate": 1.190997840621254e-06,
-      "loss": 0.1053,
-      "step": 4990
-    },
-    {
-      "epoch": 2.3962580954665387,
-      "grad_norm": 1.9739289764276322,
-      "learning_rate": 1.1819691477242069e-06,
-      "loss": 0.1166,
-      "step": 4995
-    },
-    {
-      "epoch": 2.3986567522187574,
-      "grad_norm": 1.5591416089972125,
-      "learning_rate": 1.172970218015307e-06,
-      "loss": 0.139,
-      "step": 5000
-    },
-    {
-      "epoch": 2.3986567522187574,
-      "eval_loss": 0.9828721284866333,
-      "eval_runtime": 739.7974,
-      "eval_samples_per_second": 10.016,
-      "eval_steps_per_second": 0.627,
-      "step": 5000
-    },
-    {
-      "epoch": 2.401055408970976,
-      "grad_norm": 1.6293250795813292,
-      "learning_rate": 1.164001121645069e-06,
-      "loss": 0.1341,
-      "step": 5005
-    },
-    {
-      "epoch": 2.403454065723195,
-      "grad_norm": 1.3132218762272834,
-      "learning_rate": 1.1550619285314506e-06,
-      "loss": 0.1555,
-      "step": 5010
-    },
-    {
-      "epoch": 2.4058527224754136,
-      "grad_norm": 1.9570731793037472,
-      "learning_rate": 1.146152708359294e-06,
-      "loss": 0.1604,
-      "step": 5015
-    },
-    {
-      "epoch": 2.4082513792276323,
-      "grad_norm": 1.5540385866208424,
-      "learning_rate": 1.1372735305797915e-06,
-      "loss": 0.13,
-      "step": 5020
-    },
-    {
-      "epoch": 2.4106500359798515,
-      "grad_norm": 1.5740122868688697,
-      "learning_rate": 1.1284244644099429e-06,
-      "loss": 0.1385,
-      "step": 5025
-    },
-    {
-      "epoch": 2.4130486927320702,
-      "grad_norm": 1.2901353951281462,
-      "learning_rate": 1.1196055788320148e-06,
-      "loss": 0.1055,
-      "step": 5030
-    },
-    {
-      "epoch": 2.415447349484289,
-      "grad_norm": 2.307961658041504,
-      "learning_rate": 1.1108169425930028e-06,
-      "loss": 0.1226,
-      "step": 5035
-    },
-    {
-      "epoch": 2.4178460062365077,
-      "grad_norm": 1.7431624927432852,
-      "learning_rate": 1.1020586242040943e-06,
-      "loss": 0.1233,
-      "step": 5040
-    },
-    {
-      "epoch": 2.4202446629887264,
-      "grad_norm": 1.7139513380658922,
-      "learning_rate": 1.09333069194014e-06,
-      "loss": 0.1123,
-      "step": 5045
-    },
-    {
-      "epoch": 2.422643319740945,
-      "grad_norm": 1.3214268846749446,
-      "learning_rate": 1.0846332138391158e-06,
-      "loss": 0.1282,
-      "step": 5050
-    },
-    {
-      "epoch": 2.425041976493164,
-      "grad_norm": 1.3545971596216881,
-      "learning_rate": 1.0759662577015934e-06,
-      "loss": 0.1234,
-      "step": 5055
-    },
-    {
-      "epoch": 2.4274406332453826,
-      "grad_norm": 1.5840901905123996,
-      "learning_rate": 1.067329891090213e-06,
-      "loss": 0.1085,
-      "step": 5060
-    },
-    {
-      "epoch": 2.4298392899976013,
-      "grad_norm": 1.8510314040612086,
-      "learning_rate": 1.0587241813291577e-06,
-      "loss": 0.1241,
-      "step": 5065
-    },
-    {
-      "epoch": 2.43223794674982,
-      "grad_norm": 1.6443534535540818,
-      "learning_rate": 1.0501491955036248e-06,
-      "loss": 0.1283,
-      "step": 5070
-    },
-    {
-      "epoch": 2.434636603502039,
-      "grad_norm": 1.630490467331264,
-      "learning_rate": 1.0416050004593064e-06,
-      "loss": 0.1309,
-      "step": 5075
-    },
-    {
-      "epoch": 2.4370352602542575,
-      "grad_norm": 1.546896056501047,
-      "learning_rate": 1.0330916628018706e-06,
-      "loss": 0.1251,
-      "step": 5080
-    },
-    {
-      "epoch": 2.4394339170064763,
-      "grad_norm": 2.444462757677636,
-      "learning_rate": 1.0246092488964338e-06,
-      "loss": 0.1567,
-      "step": 5085
-    },
-    {
-      "epoch": 2.441832573758695,
-      "grad_norm": 2.0106589140381788,
-      "learning_rate": 1.01615782486705e-06,
-      "loss": 0.1176,
-      "step": 5090
-    },
-    {
-      "epoch": 2.4442312305109137,
-      "grad_norm": 1.277389727131116,
-      "learning_rate": 1.0077374565961977e-06,
-      "loss": 0.1023,
-      "step": 5095
-    },
-    {
-      "epoch": 2.4466298872631325,
-      "grad_norm": 1.330955589156616,
-      "learning_rate": 9.993482097242568e-07,
-      "loss": 0.1302,
-      "step": 5100
-    },
-    {
-      "epoch": 2.449028544015351,
-      "grad_norm": 1.421052118495812,
-      "learning_rate": 9.90990149649006e-07,
-      "loss": 0.1065,
-      "step": 5105
-    },
-    {
-      "epoch": 2.4514272007675704,
-      "grad_norm": 2.470850062777196,
-      "learning_rate": 9.826633415251063e-07,
-      "loss": 0.0987,
-      "step": 5110
-    },
-    {
-      "epoch": 2.453825857519789,
-      "grad_norm": 1.9901504584119645,
-      "learning_rate": 9.743678502636028e-07,
-      "loss": 0.1376,
-      "step": 5115
-    },
-    {
-      "epoch": 2.456224514272008,
-      "grad_norm": 1.8474285543280216,
-      "learning_rate": 9.66103740531405e-07,
-      "loss": 0.143,
-      "step": 5120
-    },
-    {
-      "epoch": 2.4586231710242266,
-      "grad_norm": 1.7017419659624342,
-      "learning_rate": 9.578710767507938e-07,
-      "loss": 0.1094,
-      "step": 5125
-    },
-    {
-      "epoch": 2.4610218277764453,
-      "grad_norm": 1.3695825075670094,
-      "learning_rate": 9.496699230989149e-07,
-      "loss": 0.1131,
-      "step": 5130
-    },
-    {
-      "epoch": 2.463420484528664,
-      "grad_norm": 2.2153426080080645,
-      "learning_rate": 9.415003435072778e-07,
-      "loss": 0.1146,
-      "step": 5135
-    },
-    {
-      "epoch": 2.4658191412808828,
-      "grad_norm": 3.5120243652823238,
-      "learning_rate": 9.333624016612586e-07,
-      "loss": 0.1525,
-      "step": 5140
-    },
-    {
-      "epoch": 2.4682177980331015,
-      "grad_norm": 1.817355800367923,
-      "learning_rate": 9.252561609996075e-07,
-      "loss": 0.1159,
-      "step": 5145
-    },
-    {
-      "epoch": 2.47061645478532,
-      "grad_norm": 1.9821073225380732,
-      "learning_rate": 9.171816847139447e-07,
-      "loss": 0.1041,
-      "step": 5150
-    },
-    {
-      "epoch": 2.473015111537539,
-      "grad_norm": 1.459100052526109,
-      "learning_rate": 9.091390357482793e-07,
-      "loss": 0.1181,
-      "step": 5155
-    },
-    {
-      "epoch": 2.4754137682897577,
-      "grad_norm": 1.952875712371669,
-      "learning_rate": 9.011282767985069e-07,
-      "loss": 0.1098,
-      "step": 5160
-    },
-    {
-      "epoch": 2.4778124250419764,
-      "grad_norm": 1.534591950431475,
-      "learning_rate": 8.931494703119309e-07,
-      "loss": 0.1408,
-      "step": 5165
-    },
-    {
-      "epoch": 2.480211081794195,
-      "grad_norm": 1.7689562715329574,
-      "learning_rate": 8.852026784867701e-07,
-      "loss": 0.1313,
-      "step": 5170
-    },
-    {
-      "epoch": 2.482609738546414,
-      "grad_norm": 1.7339149389894069,
-      "learning_rate": 8.772879632716746e-07,
-      "loss": 0.1092,
-      "step": 5175
-    },
-    {
-      "epoch": 2.4850083952986326,
-      "grad_norm": 1.6464606444231056,
-      "learning_rate": 8.694053863652435e-07,
-      "loss": 0.1196,
-      "step": 5180
-    },
-    {
-      "epoch": 2.4874070520508513,
-      "grad_norm": 1.6648519910066735,
-      "learning_rate": 8.615550092155478e-07,
-      "loss": 0.0914,
-      "step": 5185
-    },
-    {
-      "epoch": 2.48980570880307,
-      "grad_norm": 1.7874580399443132,
-      "learning_rate": 8.53736893019641e-07,
-      "loss": 0.0952,
-      "step": 5190
-    },
-    {
-      "epoch": 2.4922043655552892,
-      "grad_norm": 1.4683803781623317,
-      "learning_rate": 8.459510987230957e-07,
-      "loss": 0.1495,
-      "step": 5195
-    },
-    {
-      "epoch": 2.494603022307508,
-      "grad_norm": 1.5568299629889224,
-      "learning_rate": 8.381976870195169e-07,
-      "loss": 0.1236,
-      "step": 5200
-    },
-    {
-      "epoch": 2.4970016790597267,
-      "grad_norm": 2.4873939051158263,
-      "learning_rate": 8.304767183500734e-07,
-      "loss": 0.1356,
-      "step": 5205
-    },
-    {
-      "epoch": 2.4994003358119454,
-      "grad_norm": 1.5690964117556039,
-      "learning_rate": 8.227882529030284e-07,
-      "loss": 0.1091,
-      "step": 5210
-    },
-    {
-      "epoch": 2.501798992564164,
-      "grad_norm": 1.6954049205320723,
-      "learning_rate": 8.151323506132663e-07,
-      "loss": 0.113,
-      "step": 5215
-    },
-    {
-      "epoch": 2.504197649316383,
-      "grad_norm": 1.5579032194602551,
-      "learning_rate": 8.075090711618322e-07,
-      "loss": 0.1085,
-      "step": 5220
-    },
-    {
-      "epoch": 2.5065963060686016,
-      "grad_norm": 1.764174191146571,
-      "learning_rate": 7.999184739754578e-07,
-      "loss": 0.1096,
-      "step": 5225
-    },
-    {
-      "epoch": 2.5089949628208204,
-      "grad_norm": 1.951667226426689,
-      "learning_rate": 7.92360618226104e-07,
-      "loss": 0.1288,
-      "step": 5230
-    },
-    {
-      "epoch": 2.511393619573039,
-      "grad_norm": 1.3975084205150126,
-      "learning_rate": 7.848355628304976e-07,
-      "loss": 0.1473,
-      "step": 5235
-    },
-    {
-      "epoch": 2.513792276325258,
-      "grad_norm": 1.4583105685332018,
-      "learning_rate": 7.773433664496738e-07,
-      "loss": 0.098,
-      "step": 5240
-    },
-    {
-      "epoch": 2.5161909330774765,
-      "grad_norm": 2.052173821531935,
-      "learning_rate": 7.698840874885161e-07,
-      "loss": 0.1119,
-      "step": 5245
-    },
-    {
-      "epoch": 2.5185895898296953,
-      "grad_norm": 2.1552887506028506,
-      "learning_rate": 7.624577840953046e-07,
-      "loss": 0.1497,
-      "step": 5250
-    },
-    {
-      "epoch": 2.520988246581914,
-      "grad_norm": 1.6928562717383377,
-      "learning_rate": 7.550645141612583e-07,
-      "loss": 0.1153,
-      "step": 5255
-    },
-    {
-      "epoch": 2.5233869033341327,
-      "grad_norm": 1.532142293479699,
-      "learning_rate": 7.477043353200891e-07,
-      "loss": 0.1266,
-      "step": 5260
-    },
-    {
-      "epoch": 2.5257855600863515,
-      "grad_norm": 1.5075112493817633,
-      "learning_rate": 7.403773049475477e-07,
-      "loss": 0.1195,
-      "step": 5265
-    },
-    {
-      "epoch": 2.5281842168385706,
-      "grad_norm": 1.4691450646957636,
-      "learning_rate": 7.330834801609782e-07,
-      "loss": 0.1091,
-      "step": 5270
-    },
-    {
-      "epoch": 2.530582873590789,
-      "grad_norm": 1.5131542211003923,
-      "learning_rate": 7.25822917818873e-07,
-      "loss": 0.1177,
-      "step": 5275
-    },
-    {
-      "epoch": 2.532981530343008,
-      "grad_norm": 1.3665626938316375,
-      "learning_rate": 7.185956745204298e-07,
-      "loss": 0.1369,
-      "step": 5280
-    },
-    {
-      "epoch": 2.535380187095227,
-      "grad_norm": 1.7874005964681707,
-      "learning_rate": 7.114018066051098e-07,
-      "loss": 0.127,
-      "step": 5285
-    },
-    {
-      "epoch": 2.5377788438474456,
-      "grad_norm": 2.660178675102634,
-      "learning_rate": 7.042413701522005e-07,
-      "loss": 0.1121,
-      "step": 5290
-    },
-    {
-      "epoch": 2.5401775005996643,
-      "grad_norm": 1.8021482036576246,
-      "learning_rate": 6.971144209803738e-07,
-      "loss": 0.1203,
-      "step": 5295
-    },
-    {
-      "epoch": 2.542576157351883,
-      "grad_norm": 3.5023632685823576,
-      "learning_rate": 6.900210146472564e-07,
-      "loss": 0.121,
-      "step": 5300
-    },
-    {
-      "epoch": 2.5449748141041018,
-      "grad_norm": 1.4117798013702134,
-      "learning_rate": 6.829612064489933e-07,
-      "loss": 0.1132,
-      "step": 5305
-    },
-    {
-      "epoch": 2.5473734708563205,
-      "grad_norm": 2.182251500140064,
-      "learning_rate": 6.759350514198171e-07,
-      "loss": 0.1195,
-      "step": 5310
-    },
-    {
-      "epoch": 2.549772127608539,
-      "grad_norm": 1.7117310660076108,
-      "learning_rate": 6.6894260433162e-07,
-      "loss": 0.1111,
-      "step": 5315
-    },
-    {
-      "epoch": 2.552170784360758,
-      "grad_norm": 1.5233014936758835,
-      "learning_rate": 6.619839196935251e-07,
-      "loss": 0.1361,
-      "step": 5320
-    },
-    {
-      "epoch": 2.5545694411129767,
-      "grad_norm": 1.3687130333207045,
-      "learning_rate": 6.550590517514666e-07,
-      "loss": 0.109,
-      "step": 5325
-    },
-    {
-      "epoch": 2.5569680978651954,
-      "grad_norm": 1.8037859341746436,
-      "learning_rate": 6.481680544877584e-07,
-      "loss": 0.1316,
-      "step": 5330
-    },
-    {
-      "epoch": 2.559366754617414,
-      "grad_norm": 2.2640573897300205,
-      "learning_rate": 6.413109816206803e-07,
-      "loss": 0.1333,
-      "step": 5335
-    },
-    {
-      "epoch": 2.561765411369633,
-      "grad_norm": 1.592815519460586,
-      "learning_rate": 6.344878866040571e-07,
-      "loss": 0.1116,
-      "step": 5340
-    },
-    {
-      "epoch": 2.5641640681218516,
-      "grad_norm": 1.9130650669785958,
-      "learning_rate": 6.276988226268388e-07,
-      "loss": 0.1327,
-      "step": 5345
-    },
-    {
-      "epoch": 2.5665627248740703,
-      "grad_norm": 1.7208032444949648,
-      "learning_rate": 6.209438426126946e-07,
-      "loss": 0.1693,
-      "step": 5350
-    },
-    {
-      "epoch": 2.5689613816262895,
-      "grad_norm": 1.5896417871178385,
-      "learning_rate": 6.142229992195886e-07,
-      "loss": 0.1156,
-      "step": 5355
-    },
-    {
-      "epoch": 2.571360038378508,
-      "grad_norm": 1.3231503619333465,
-      "learning_rate": 6.075363448393778e-07,
-      "loss": 0.0997,
-      "step": 5360
-    },
-    {
-      "epoch": 2.573758695130727,
-      "grad_norm": 2.3180058541025224,
-      "learning_rate": 6.00883931597403e-07,
-      "loss": 0.1207,
-      "step": 5365
-    },
-    {
-      "epoch": 2.5761573518829457,
-      "grad_norm": 1.8598922709868437,
-      "learning_rate": 5.942658113520778e-07,
-      "loss": 0.1187,
-      "step": 5370
-    },
-    {
-      "epoch": 2.5785560086351644,
-      "grad_norm": 1.741716946232789,
-      "learning_rate": 5.876820356944879e-07,
-      "loss": 0.1782,
-      "step": 5375
-    },
-    {
-      "epoch": 2.580954665387383,
-      "grad_norm": 2.0613835405768346,
-      "learning_rate": 5.811326559479885e-07,
-      "loss": 0.1276,
-      "step": 5380
-    },
-    {
-      "epoch": 2.583353322139602,
-      "grad_norm": 1.2306263180991488,
-      "learning_rate": 5.746177231678046e-07,
-      "loss": 0.1134,
-      "step": 5385
-    },
-    {
-      "epoch": 2.5857519788918206,
-      "grad_norm": 1.5217893010086523,
-      "learning_rate": 5.681372881406295e-07,
-      "loss": 0.1151,
-      "step": 5390
-    },
-    {
-      "epoch": 2.5881506356440394,
-      "grad_norm": 1.8109437720576,
-      "learning_rate": 5.616914013842362e-07,
-      "loss": 0.0977,
-      "step": 5395
-    },
-    {
-      "epoch": 2.590549292396258,
-      "grad_norm": 1.2622629020176752,
-      "learning_rate": 5.55280113147077e-07,
-      "loss": 0.1271,
-      "step": 5400
-    },
-    {
-      "epoch": 2.592947949148477,
-      "grad_norm": 1.571274143304735,
-      "learning_rate": 5.489034734078924e-07,
-      "loss": 0.1192,
-      "step": 5405
-    },
-    {
-      "epoch": 2.5953466059006955,
-      "grad_norm": 1.9698287341511391,
-      "learning_rate": 5.425615318753252e-07,
-      "loss": 0.1057,
-      "step": 5410
-    },
-    {
-      "epoch": 2.5977452626529143,
-      "grad_norm": 1.9701923754542536,
-      "learning_rate": 5.36254337987529e-07,
-      "loss": 0.1282,
-      "step": 5415
-    },
-    {
-      "epoch": 2.600143919405133,
-      "grad_norm": 2.2686032744150015,
-      "learning_rate": 5.299819409117857e-07,
-      "loss": 0.1429,
-      "step": 5420
-    },
-    {
-      "epoch": 2.6025425761573517,
-      "grad_norm": 1.2611154676332545,
-      "learning_rate": 5.237443895441213e-07,
-      "loss": 0.1188,
-      "step": 5425
-    },
-    {
-      "epoch": 2.604941232909571,
-      "grad_norm": 1.5048543165518988,
-      "learning_rate": 5.175417325089227e-07,
-      "loss": 0.1052,
-      "step": 5430
-    },
-    {
-      "epoch": 2.607339889661789,
-      "grad_norm": 1.5118763112028835,
-      "learning_rate": 5.113740181585646e-07,
-      "loss": 0.1258,
-      "step": 5435
-    },
-    {
-      "epoch": 2.6097385464140084,
-      "grad_norm": 2.3811131252232056,
-      "learning_rate": 5.05241294573024e-07,
-      "loss": 0.1081,
-      "step": 5440
-    },
-    {
-      "epoch": 2.6121372031662267,
-      "grad_norm": 2.454395400405562,
-      "learning_rate": 4.99143609559512e-07,
-      "loss": 0.1461,
-      "step": 5445
-    },
-    {
-      "epoch": 2.614535859918446,
-      "grad_norm": 1.5225164499950974,
-      "learning_rate": 4.930810106520983e-07,
-      "loss": 0.126,
-      "step": 5450
-    },
-    {
-      "epoch": 2.6169345166706646,
-      "grad_norm": 2.9727690647821166,
-      "learning_rate": 4.870535451113434e-07,
-      "loss": 0.0866,
-      "step": 5455
-    },
-    {
-      "epoch": 2.6193331734228833,
-      "grad_norm": 1.8766682156530599,
-      "learning_rate": 4.81061259923925e-07,
-      "loss": 0.1161,
-      "step": 5460
-    },
-    {
-      "epoch": 2.621731830175102,
-      "grad_norm": 1.3747965137374512,
-      "learning_rate": 4.751042018022761e-07,
-      "loss": 0.1177,
-      "step": 5465
-    },
-    {
-      "epoch": 2.6241304869273208,
-      "grad_norm": 2.4902927641849764,
-      "learning_rate": 4.691824171842219e-07,
-      "loss": 0.1545,
-      "step": 5470
-    },
-    {
-      "epoch": 2.6265291436795395,
-      "grad_norm": 2.20678634424907,
-      "learning_rate": 4.6329595223261193e-07,
-      "loss": 0.1212,
-      "step": 5475
-    },
-    {
-      "epoch": 2.628927800431758,
-      "grad_norm": 2.36358294096013,
-      "learning_rate": 4.574448528349662e-07,
-      "loss": 0.1337,
-      "step": 5480
-    },
-    {
-      "epoch": 2.631326457183977,
-      "grad_norm": 2.1209790191903153,
-      "learning_rate": 4.5162916460311346e-07,
-      "loss": 0.1327,
-      "step": 5485
-    },
-    {
-      "epoch": 2.6337251139361957,
-      "grad_norm": 2.1750708889962285,
-      "learning_rate": 4.4584893287283773e-07,
-      "loss": 0.1115,
-      "step": 5490
-    },
-    {
-      "epoch": 2.6361237706884144,
-      "grad_norm": 1.7181125564121318,
-      "learning_rate": 4.4010420270352463e-07,
-      "loss": 0.1152,
-      "step": 5495
-    },
-    {
-      "epoch": 2.638522427440633,
-      "grad_norm": 1.458326231995438,
-      "learning_rate": 4.3439501887781064e-07,
-      "loss": 0.1257,
-      "step": 5500
-    },
-    {
-      "epoch": 2.640921084192852,
-      "grad_norm": 1.349194936754368,
-      "learning_rate": 4.287214259012329e-07,
-      "loss": 0.1094,
-      "step": 5505
-    },
-    {
-      "epoch": 2.6433197409450706,
-      "grad_norm": 1.3226300725693332,
-      "learning_rate": 4.2308346800188193e-07,
-      "loss": 0.1287,
-      "step": 5510
-    },
-    {
-      "epoch": 2.64571839769729,
-      "grad_norm": 1.8303296939894655,
-      "learning_rate": 4.1748118913005675e-07,
-      "loss": 0.1119,
-      "step": 5515
-    },
-    {
-      "epoch": 2.648117054449508,
-      "grad_norm": 2.1148807551463658,
-      "learning_rate": 4.1191463295792545e-07,
-      "loss": 0.1229,
-      "step": 5520
-    },
-    {
-      "epoch": 2.6505157112017272,
-      "grad_norm": 1.8398900807422445,
-      "learning_rate": 4.063838428791805e-07,
-      "loss": 0.1479,
-      "step": 5525
-    },
-    {
-      "epoch": 2.6529143679539455,
-      "grad_norm": 1.9081399715986063,
-      "learning_rate": 4.0088886200870256e-07,
-      "loss": 0.1518,
-      "step": 5530
-    },
-    {
-      "epoch": 2.6553130247061647,
-      "grad_norm": 2.5071133379640123,
-      "learning_rate": 3.9542973318222376e-07,
-      "loss": 0.1453,
-      "step": 5535
-    },
-    {
-      "epoch": 2.6577116814583834,
-      "grad_norm": 1.3617983583001947,
-      "learning_rate": 3.900064989559965e-07,
-      "loss": 0.1369,
-      "step": 5540
-    },
-    {
-      "epoch": 2.660110338210602,
-      "grad_norm": 1.7350007966826908,
-      "learning_rate": 3.846192016064576e-07,
-      "loss": 0.1131,
-      "step": 5545
-    },
-    {
-      "epoch": 2.662508994962821,
-      "grad_norm": 1.4222833594841213,
-      "learning_rate": 3.7926788312989904e-07,
-      "loss": 0.1273,
-      "step": 5550
-    },
-    {
-      "epoch": 2.6649076517150396,
-      "grad_norm": 1.4888034133853167,
-      "learning_rate": 3.739525852421455e-07,
-      "loss": 0.1325,
-      "step": 5555
-    },
-    {
-      "epoch": 2.6673063084672584,
-      "grad_norm": 1.4598038003756197,
-      "learning_rate": 3.68673349378223e-07,
-      "loss": 0.1403,
-      "step": 5560
-    },
-    {
-      "epoch": 2.669704965219477,
-      "grad_norm": 1.595545300083971,
-      "learning_rate": 3.634302166920395e-07,
-      "loss": 0.1385,
-      "step": 5565
-    },
-    {
-      "epoch": 2.672103621971696,
-      "grad_norm": 1.7762121350952307,
-      "learning_rate": 3.5822322805606267e-07,
-      "loss": 0.1374,
-      "step": 5570
-    },
-    {
-      "epoch": 2.6745022787239146,
-      "grad_norm": 1.703923143430848,
-      "learning_rate": 3.5305242406100395e-07,
-      "loss": 0.1331,
-      "step": 5575
-    },
-    {
-      "epoch": 2.6769009354761333,
-      "grad_norm": 1.4304866394510962,
-      "learning_rate": 3.479178450154974e-07,
-      "loss": 0.1373,
-      "step": 5580
-    },
-    {
-      "epoch": 2.679299592228352,
-      "grad_norm": 1.7486370046393058,
-      "learning_rate": 3.4281953094578877e-07,
-      "loss": 0.1376,
-      "step": 5585
-    },
-    {
-      "epoch": 2.6816982489805707,
-      "grad_norm": 1.533651476967988,
-      "learning_rate": 3.3775752159542375e-07,
-      "loss": 0.1115,
-      "step": 5590
-    },
-    {
-      "epoch": 2.6840969057327895,
-      "grad_norm": 1.9305760137068588,
-      "learning_rate": 3.32731856424936e-07,
-      "loss": 0.1105,
-      "step": 5595
-    },
-    {
-      "epoch": 2.6864955624850086,
-      "grad_norm": 1.8129731172858077,
-      "learning_rate": 3.277425746115398e-07,
-      "loss": 0.112,
-      "step": 5600
-    },
-    {
-      "epoch": 2.688894219237227,
-      "grad_norm": 1.7960561922883571,
-      "learning_rate": 3.227897150488285e-07,
-      "loss": 0.1174,
-      "step": 5605
-    },
-    {
-      "epoch": 2.691292875989446,
-      "grad_norm": 2.0259280213774176,
-      "learning_rate": 3.178733163464676e-07,
-      "loss": 0.118,
-      "step": 5610
-    },
-    {
-      "epoch": 2.6936915327416644,
-      "grad_norm": 1.6681393489978669,
-      "learning_rate": 3.1299341682989313e-07,
-      "loss": 0.1258,
-      "step": 5615
-    },
-    {
-      "epoch": 2.6960901894938836,
-      "grad_norm": 1.5698047263231931,
-      "learning_rate": 3.08150054540014e-07,
-      "loss": 0.1089,
-      "step": 5620
-    },
-    {
-      "epoch": 2.6984888462461023,
-      "grad_norm": 1.8806087643705454,
-      "learning_rate": 3.033432672329173e-07,
-      "loss": 0.1584,
-      "step": 5625
-    },
-    {
-      "epoch": 2.700887502998321,
-      "grad_norm": 1.6181003470798663,
-      "learning_rate": 2.985730923795721e-07,
-      "loss": 0.1197,
-      "step": 5630
-    },
-    {
-      "epoch": 2.7032861597505398,
-      "grad_norm": 1.903755595617439,
-      "learning_rate": 2.93839567165537e-07,
-      "loss": 0.1082,
-      "step": 5635
-    },
-    {
-      "epoch": 2.7056848165027585,
-      "grad_norm": 1.4481162983792952,
-      "learning_rate": 2.891427284906706e-07,
-      "loss": 0.1235,
-      "step": 5640
-    },
-    {
-      "epoch": 2.7080834732549772,
-      "grad_norm": 1.3940387482587855,
-      "learning_rate": 2.8448261296884715e-07,
-      "loss": 0.1135,
-      "step": 5645
-    },
-    {
-      "epoch": 2.710482130007196,
-      "grad_norm": 1.601511394808099,
-      "learning_rate": 2.7985925692766426e-07,
-      "loss": 0.1317,
-      "step": 5650
-    },
-    {
-      "epoch": 2.7128807867594147,
-      "grad_norm": 1.3269174628464722,
-      "learning_rate": 2.7527269640816537e-07,
-      "loss": 0.1294,
-      "step": 5655
-    },
-    {
-      "epoch": 2.7152794435116334,
-      "grad_norm": 1.4993241080528643,
-      "learning_rate": 2.707229671645578e-07,
-      "loss": 0.1289,
-      "step": 5660
-    },
-    {
-      "epoch": 2.717678100263852,
-      "grad_norm": 1.779375793335884,
-      "learning_rate": 2.662101046639326e-07,
-      "loss": 0.1239,
-      "step": 5665
-    },
-    {
-      "epoch": 2.720076757016071,
-      "grad_norm": 1.9455875856444815,
-      "learning_rate": 2.617341440859883e-07,
-      "loss": 0.1219,
-      "step": 5670
-    },
-    {
-      "epoch": 2.7224754137682896,
-      "grad_norm": 1.672660061147154,
-      "learning_rate": 2.572951203227564e-07,
-      "loss": 0.1306,
-      "step": 5675
-    },
-    {
-      "epoch": 2.7248740705205083,
-      "grad_norm": 1.6430355652699298,
-      "learning_rate": 2.528930679783331e-07,
-      "loss": 0.0924,
-      "step": 5680
-    },
-    {
-      "epoch": 2.7272727272727275,
-      "grad_norm": 1.52811198720444,
-      "learning_rate": 2.485280213686031e-07,
-      "loss": 0.1083,
-      "step": 5685
-    },
-    {
-      "epoch": 2.729671384024946,
-      "grad_norm": 2.0031601863687545,
-      "learning_rate": 2.442000145209783e-07,
-      "loss": 0.1123,
-      "step": 5690
-    },
-    {
-      "epoch": 2.732070040777165,
-      "grad_norm": 2.183846855918397,
-      "learning_rate": 2.3990908117412725e-07,
-      "loss": 0.128,
-      "step": 5695
-    },
-    {
-      "epoch": 2.7344686975293837,
-      "grad_norm": 1.3334592284309938,
-      "learning_rate": 2.3565525477771635e-07,
-      "loss": 0.1044,
-      "step": 5700
-    },
-    {
-      "epoch": 2.7368673542816024,
-      "grad_norm": 1.7121458933661435,
-      "learning_rate": 2.3143856849214608e-07,
-      "loss": 0.1149,
-      "step": 5705
-    },
-    {
-      "epoch": 2.739266011033821,
-      "grad_norm": 1.6207522141153845,
-      "learning_rate": 2.2725905518829582e-07,
-      "loss": 0.1167,
-      "step": 5710
-    },
-    {
-      "epoch": 2.74166466778604,
-      "grad_norm": 1.2168116030947638,
-      "learning_rate": 2.231167474472651e-07,
-      "loss": 0.127,
-      "step": 5715
-    },
-    {
-      "epoch": 2.7440633245382586,
-      "grad_norm": 1.90993157648795,
-      "learning_rate": 2.190116775601181e-07,
-      "loss": 0.1255,
-      "step": 5720
-    },
-    {
-      "epoch": 2.7464619812904774,
-      "grad_norm": 1.6206812916963553,
-      "learning_rate": 2.149438775276358e-07,
-      "loss": 0.1203,
-      "step": 5725
-    },
-    {
-      "epoch": 2.748860638042696,
-      "grad_norm": 2.2235187423218092,
-      "learning_rate": 2.109133790600648e-07,
-      "loss": 0.1578,
-      "step": 5730
-    },
-    {
-      "epoch": 2.751259294794915,
-      "grad_norm": 1.8139066015122434,
-      "learning_rate": 2.0692021357686886e-07,
-      "loss": 0.1352,
-      "step": 5735
-    },
-    {
-      "epoch": 2.7536579515471336,
-      "grad_norm": 2.031694397609124,
-      "learning_rate": 2.0296441220648554e-07,
-      "loss": 0.1373,
-      "step": 5740
-    },
-    {
-      "epoch": 2.7560566082993523,
-      "grad_norm": 2.2355401023781516,
-      "learning_rate": 1.990460057860827e-07,
-      "loss": 0.1212,
-      "step": 5745
-    },
-    {
-      "epoch": 2.758455265051571,
-      "grad_norm": 1.7755393370077912,
-      "learning_rate": 1.9516502486132072e-07,
-      "loss": 0.1027,
-      "step": 5750
-    },
-    {
-      "epoch": 2.7608539218037897,
-      "grad_norm": 1.6871643573081383,
-      "learning_rate": 1.9132149968610903e-07,
-      "loss": 0.1231,
-      "step": 5755
-    },
-    {
-      "epoch": 2.7632525785560085,
-      "grad_norm": 1.7906723105033933,
-      "learning_rate": 1.875154602223761e-07,
-      "loss": 0.1286,
-      "step": 5760
-    },
-    {
-      "epoch": 2.765651235308227,
-      "grad_norm": 1.4557489056351423,
-      "learning_rate": 1.8374693613983197e-07,
-      "loss": 0.1195,
-      "step": 5765
-    },
-    {
-      "epoch": 2.7680498920604464,
-      "grad_norm": 1.2501535667579526,
-      "learning_rate": 1.800159568157378e-07,
-      "loss": 0.122,
-      "step": 5770
-    },
-    {
-      "epoch": 2.7704485488126647,
-      "grad_norm": 1.9933839036208927,
-      "learning_rate": 1.7632255133467836e-07,
-      "loss": 0.1157,
-      "step": 5775
-    },
-    {
-      "epoch": 2.772847205564884,
-      "grad_norm": 1.8434279839218284,
-      "learning_rate": 1.7266674848833155e-07,
-      "loss": 0.0927,
-      "step": 5780
-    },
-    {
-      "epoch": 2.7752458623171026,
-      "grad_norm": 1.5358444407642786,
-      "learning_rate": 1.6904857677525145e-07,
-      "loss": 0.1041,
-      "step": 5785
-    },
-    {
-      "epoch": 2.7776445190693213,
-      "grad_norm": 1.3531209392466852,
-      "learning_rate": 1.654680644006368e-07,
-      "loss": 0.1018,
-      "step": 5790
-    },
-    {
-      "epoch": 2.78004317582154,
-      "grad_norm": 2.1858517794960783,
-      "learning_rate": 1.6192523927611835e-07,
-      "loss": 0.119,
-      "step": 5795
-    },
-    {
-      "epoch": 2.7824418325737588,
-      "grad_norm": 1.5246185793951936,
-      "learning_rate": 1.584201290195364e-07,
-      "loss": 0.1271,
-      "step": 5800
-    },
-    {
-      "epoch": 2.7848404893259775,
-      "grad_norm": 1.6607987943748377,
-      "learning_rate": 1.5495276095472967e-07,
-      "loss": 0.1175,
-      "step": 5805
-    },
-    {
-      "epoch": 2.7872391460781962,
-      "grad_norm": 1.5511169416761914,
-      "learning_rate": 1.5152316211131835e-07,
-      "loss": 0.1189,
-      "step": 5810
-    },
-    {
-      "epoch": 2.789637802830415,
-      "grad_norm": 1.3636834104396771,
-      "learning_rate": 1.481313592244976e-07,
-      "loss": 0.0974,
-      "step": 5815
-    },
-    {
-      "epoch": 2.7920364595826337,
-      "grad_norm": 1.7780659172726045,
-      "learning_rate": 1.4477737873482556e-07,
-      "loss": 0.116,
-      "step": 5820
-    },
-    {
-      "epoch": 2.7944351163348524,
-      "grad_norm": 1.9201443415453343,
-      "learning_rate": 1.4146124678801886e-07,
-      "loss": 0.1048,
-      "step": 5825
-    },
-    {
-      "epoch": 2.796833773087071,
-      "grad_norm": 1.6060426355740423,
-      "learning_rate": 1.381829892347486e-07,
-      "loss": 0.1134,
-      "step": 5830
-    },
-    {
-      "epoch": 2.79923242983929,
-      "grad_norm": 1.8753892090672297,
-      "learning_rate": 1.3494263163043752e-07,
-      "loss": 0.1022,
-      "step": 5835
-    },
-    {
-      "epoch": 2.8016310865915086,
-      "grad_norm": 1.7909391408691406,
-      "learning_rate": 1.3174019923506365e-07,
-      "loss": 0.1083,
-      "step": 5840
-    },
-    {
-      "epoch": 2.804029743343728,
-      "grad_norm": 1.8615512629565283,
-      "learning_rate": 1.2857571701296146e-07,
-      "loss": 0.1248,
-      "step": 5845
-    },
-    {
-      "epoch": 2.806428400095946,
-      "grad_norm": 1.4977699048787447,
-      "learning_rate": 1.2544920963262653e-07,
-      "loss": 0.1261,
-      "step": 5850
-    },
-    {
-      "epoch": 2.8088270568481652,
-      "grad_norm": 1.8717329377806204,
-      "learning_rate": 1.2236070146652512e-07,
-      "loss": 0.1123,
-      "step": 5855
-    },
-    {
-      "epoch": 2.8112257136003835,
-      "grad_norm": 1.8892706724274941,
-      "learning_rate": 1.1931021659090325e-07,
-      "loss": 0.1276,
-      "step": 5860
-    },
-    {
-      "epoch": 2.8136243703526027,
-      "grad_norm": 1.691896819156689,
-      "learning_rate": 1.1629777878559956e-07,
-      "loss": 0.1244,
-      "step": 5865
-    },
-    {
-      "epoch": 2.8160230271048214,
-      "grad_norm": 1.8682571614765742,
-      "learning_rate": 1.1332341153385773e-07,
-      "loss": 0.1355,
-      "step": 5870
-    },
-    {
-      "epoch": 2.81842168385704,
-      "grad_norm": 1.2268638639315688,
-      "learning_rate": 1.1038713802214718e-07,
-      "loss": 0.1225,
-      "step": 5875
-    },
-    {
-      "epoch": 2.820820340609259,
-      "grad_norm": 2.0069927297948555,
-      "learning_rate": 1.0748898113997875e-07,
-      "loss": 0.1195,
-      "step": 5880
-    },
-    {
-      "epoch": 2.8232189973614776,
-      "grad_norm": 2.308460157379254,
-      "learning_rate": 1.0462896347972818e-07,
-      "loss": 0.1174,
-      "step": 5885
-    },
-    {
-      "epoch": 2.8256176541136964,
-      "grad_norm": 2.0122573238619075,
-      "learning_rate": 1.0180710733646127e-07,
-      "loss": 0.1556,
-      "step": 5890
-    },
-    {
-      "epoch": 2.828016310865915,
-      "grad_norm": 1.3824343924619586,
-      "learning_rate": 9.902343470775566e-08,
-      "loss": 0.1403,
-      "step": 5895
-    },
-    {
-      "epoch": 2.830414967618134,
-      "grad_norm": 2.967278007524201,
-      "learning_rate": 9.627796729353434e-08,
-      "loss": 0.1106,
-      "step": 5900
-    },
-    {
-      "epoch": 2.8328136243703526,
-      "grad_norm": 1.6475498343063206,
-      "learning_rate": 9.357072649589238e-08,
-      "loss": 0.1169,
-      "step": 5905
-    },
-    {
-      "epoch": 2.8352122811225713,
-      "grad_norm": 1.2585469032506666,
-      "learning_rate": 9.090173341893383e-08,
-      "loss": 0.1277,
-      "step": 5910
-    },
-    {
-      "epoch": 2.83761093787479,
-      "grad_norm": 1.1238112527321122,
-      "learning_rate": 8.827100886860451e-08,
-      "loss": 0.1043,
-      "step": 5915
-    },
-    {
-      "epoch": 2.8400095946270087,
-      "grad_norm": 1.488550388006429,
-      "learning_rate": 8.567857335253005e-08,
-      "loss": 0.13,
-      "step": 5920
-    },
-    {
-      "epoch": 2.8424082513792275,
-      "grad_norm": 1.3312873362378537,
-      "learning_rate": 8.312444707985811e-08,
-      "loss": 0.1245,
-      "step": 5925
-    },
-    {
-      "epoch": 2.8448069081314467,
-      "grad_norm": 2.290161387016882,
-      "learning_rate": 8.060864996109807e-08,
-      "loss": 0.1307,
-      "step": 5930
-    },
-    {
-      "epoch": 2.847205564883665,
-      "grad_norm": 1.9033120811547106,
-      "learning_rate": 7.813120160796772e-08,
-      "loss": 0.1371,
-      "step": 5935
-    },
-    {
-      "epoch": 2.849604221635884,
-      "grad_norm": 2.0871629668196676,
-      "learning_rate": 7.569212133323956e-08,
-      "loss": 0.1239,
-      "step": 5940
-    },
-    {
-      "epoch": 2.8520028783881024,
-      "grad_norm": 1.4650768736158892,
-      "learning_rate": 7.329142815059032e-08,
-      "loss": 0.119,
-      "step": 5945
-    },
-    {
-      "epoch": 2.8544015351403216,
-      "grad_norm": 1.1743412799238815,
-      "learning_rate": 7.092914077445335e-08,
-      "loss": 0.1018,
-      "step": 5950
-    },
-    {
-      "epoch": 2.8568001918925403,
-      "grad_norm": 1.5498522487672175,
-      "learning_rate": 6.860527761987145e-08,
-      "loss": 0.1197,
-      "step": 5955
-    },
-    {
-      "epoch": 2.859198848644759,
-      "grad_norm": 1.6056923353016053,
-      "learning_rate": 6.631985680235487e-08,
-      "loss": 0.1115,
-      "step": 5960
-    },
-    {
-      "epoch": 2.8615975053969778,
-      "grad_norm": 1.791424778629989,
-      "learning_rate": 6.407289613774014e-08,
-      "loss": 0.1414,
-      "step": 5965
-    },
-    {
-      "epoch": 2.8639961621491965,
-      "grad_norm": 1.4221448461917148,
-      "learning_rate": 6.186441314204816e-08,
-      "loss": 0.1289,
-      "step": 5970
-    },
-    {
-      "epoch": 2.8663948189014152,
-      "grad_norm": 1.495279144021361,
-      "learning_rate": 5.969442503135192e-08,
-      "loss": 0.112,
-      "step": 5975
-    },
-    {
-      "epoch": 2.868793475653634,
-      "grad_norm": 2.545350173896225,
-      "learning_rate": 5.756294872163948e-08,
-      "loss": 0.131,
-      "step": 5980
-    },
-    {
-      "epoch": 2.8711921324058527,
-      "grad_norm": 1.2008746098462784,
-      "learning_rate": 5.547000082868237e-08,
-      "loss": 0.1219,
-      "step": 5985
-    },
-    {
-      "epoch": 2.8735907891580714,
-      "grad_norm": 1.487751985051309,
-      "learning_rate": 5.341559766790738e-08,
-      "loss": 0.1323,
-      "step": 5990
-    },
-    {
-      "epoch": 2.87598944591029,
-      "grad_norm": 1.3980272709675226,
-      "learning_rate": 5.139975525426833e-08,
-      "loss": 0.1467,
-      "step": 5995
-    },
-    {
-      "epoch": 2.878388102662509,
-      "grad_norm": 2.10880272054709,
-      "learning_rate": 4.942248930212224e-08,
-      "loss": 0.146,
-      "step": 6000
-    },
-    {
-      "epoch": 2.878388102662509,
-      "eval_loss": 0.9810638427734375,
-      "eval_runtime": 739.6233,
-      "eval_samples_per_second": 10.019,
-      "eval_steps_per_second": 0.627,
-      "step": 6000
-    },
-    {
-      "epoch": 2.8807867594147276,
-      "grad_norm": 1.483009143164286,
-      "learning_rate": 4.748381522510392e-08,
-      "loss": 0.1158,
-      "step": 6005
-    },
-    {
-      "epoch": 2.8831854161669463,
-      "grad_norm": 1.539223427422332,
-      "learning_rate": 4.558374813601052e-08,
-      "loss": 0.1336,
-      "step": 6010
-    },
-    {
-      "epoch": 2.8855840729191655,
-      "grad_norm": 1.622424329571053,
-      "learning_rate": 4.372230284667878e-08,
-      "loss": 0.1108,
-      "step": 6015
-    },
-    {
-      "epoch": 2.887982729671384,
-      "grad_norm": 1.7328410726207495,
-      "learning_rate": 4.189949386787462e-08,
-      "loss": 0.1207,
-      "step": 6020
-    },
-    {
-      "epoch": 2.890381386423603,
-      "grad_norm": 1.954059556795084,
-      "learning_rate": 4.011533540917489e-08,
-      "loss": 0.1056,
-      "step": 6025
-    },
-    {
-      "epoch": 2.8927800431758213,
-      "grad_norm": 2.137802294017243,
-      "learning_rate": 3.836984137886024e-08,
-      "loss": 0.1277,
-      "step": 6030
-    },
-    {
-      "epoch": 2.8951786999280404,
-      "grad_norm": 1.6698284932650442,
-      "learning_rate": 3.6663025383805174e-08,
-      "loss": 0.1172,
-      "step": 6035
-    },
-    {
-      "epoch": 2.897577356680259,
-      "grad_norm": 1.6220510726257216,
-      "learning_rate": 3.4994900729373173e-08,
-      "loss": 0.1232,
-      "step": 6040
-    },
-    {
-      "epoch": 2.899976013432478,
-      "grad_norm": 1.9213808029719575,
-      "learning_rate": 3.3365480419310646e-08,
-      "loss": 0.1151,
-      "step": 6045
-    },
-    {
-      "epoch": 2.9023746701846966,
-      "grad_norm": 1.6501341986627394,
-      "learning_rate": 3.177477715564869e-08,
-      "loss": 0.1006,
-      "step": 6050
-    },
-    {
-      "epoch": 2.9047733269369154,
-      "grad_norm": 1.9659614646400356,
-      "learning_rate": 3.0222803338600394e-08,
-      "loss": 0.1014,
-      "step": 6055
-    },
-    {
-      "epoch": 2.907171983689134,
-      "grad_norm": 2.060843304606416,
-      "learning_rate": 2.87095710664681e-08,
-      "loss": 0.1278,
-      "step": 6060
-    },
-    {
-      "epoch": 2.909570640441353,
-      "grad_norm": 1.4858740242674862,
-      "learning_rate": 2.723509213554576e-08,
-      "loss": 0.117,
-      "step": 6065
-    },
-    {
-      "epoch": 2.9119692971935716,
-      "grad_norm": 1.6542843067890352,
-      "learning_rate": 2.5799378040030075e-08,
-      "loss": 0.1243,
-      "step": 6070
-    },
-    {
-      "epoch": 2.9143679539457903,
-      "grad_norm": 1.2106121391404152,
-      "learning_rate": 2.4402439971927816e-08,
-      "loss": 0.1129,
-      "step": 6075
-    },
-    {
-      "epoch": 2.916766610698009,
-      "grad_norm": 1.637222521799144,
-      "learning_rate": 2.3044288820969764e-08,
-      "loss": 0.1147,
-      "step": 6080
-    },
-    {
-      "epoch": 2.9191652674502278,
-      "grad_norm": 1.4227992719038707,
-      "learning_rate": 2.1724935174528006e-08,
-      "loss": 0.1277,
-      "step": 6085
-    },
-    {
-      "epoch": 2.9215639242024465,
-      "grad_norm": 1.6446009283719198,
-      "learning_rate": 2.044438931752879e-08,
-      "loss": 0.1555,
-      "step": 6090
-    },
-    {
-      "epoch": 2.923962580954665,
-      "grad_norm": 1.8090094335987503,
-      "learning_rate": 1.9202661232377018e-08,
-      "loss": 0.1245,
-      "step": 6095
-    },
-    {
-      "epoch": 2.9263612377068844,
-      "grad_norm": 1.3319652375080722,
-      "learning_rate": 1.799976059887576e-08,
-      "loss": 0.1353,
-      "step": 6100
-    },
-    {
-      "epoch": 2.9287598944591027,
-      "grad_norm": 1.5406269983982313,
-      "learning_rate": 1.6835696794151312e-08,
-      "loss": 0.1212,
-      "step": 6105
-    },
-    {
-      "epoch": 2.931158551211322,
-      "grad_norm": 1.690521455881083,
-      "learning_rate": 1.5710478892579927e-08,
-      "loss": 0.1586,
-      "step": 6110
-    },
-    {
-      "epoch": 2.9335572079635406,
-      "grad_norm": 1.705355483123506,
-      "learning_rate": 1.4624115665717865e-08,
-      "loss": 0.1324,
-      "step": 6115
-    },
-    {
-      "epoch": 2.9359558647157593,
-      "grad_norm": 1.7101043349695386,
-      "learning_rate": 1.3576615582233666e-08,
-      "loss": 0.0989,
-      "step": 6120
-    },
-    {
-      "epoch": 2.938354521467978,
-      "grad_norm": 1.6013100594407221,
-      "learning_rate": 1.2567986807838217e-08,
-      "loss": 0.095,
-      "step": 6125
-    },
-    {
-      "epoch": 2.9407531782201968,
-      "grad_norm": 1.906186669455512,
-      "learning_rate": 1.1598237205225904e-08,
-      "loss": 0.1555,
-      "step": 6130
-    },
-    {
-      "epoch": 2.9431518349724155,
-      "grad_norm": 1.60057625401919,
-      "learning_rate": 1.0667374334011327e-08,
-      "loss": 0.1195,
-      "step": 6135
-    },
-    {
-      "epoch": 2.9455504917246342,
-      "grad_norm": 1.854263325199233,
-      "learning_rate": 9.77540545066935e-09,
-      "loss": 0.1251,
-      "step": 6140
-    },
-    {
-      "epoch": 2.947949148476853,
-      "grad_norm": 1.453650090583012,
-      "learning_rate": 8.92233750848015e-09,
-      "loss": 0.1021,
-      "step": 6145
-    },
-    {
-      "epoch": 2.9503478052290717,
-      "grad_norm": 1.4057914251921044,
-      "learning_rate": 8.10817715747425e-09,
-      "loss": 0.1061,
-      "step": 6150
-    },
-    {
-      "epoch": 2.9527464619812904,
-      "grad_norm": 2.005590745242888,
-      "learning_rate": 7.332930744380906e-09,
-      "loss": 0.1232,
-      "step": 6155
-    },
-    {
-      "epoch": 2.955145118733509,
-      "grad_norm": 1.6052684537774102,
-      "learning_rate": 6.596604312578136e-09,
-      "loss": 0.0927,
-      "step": 6160
-    },
-    {
-      "epoch": 2.957543775485728,
-      "grad_norm": 1.4712133769137725,
-      "learning_rate": 5.899203602046655e-09,
-      "loss": 0.1218,
-      "step": 6165
-    },
-    {
-      "epoch": 2.9599424322379466,
-      "grad_norm": 1.6115906553260235,
-      "learning_rate": 5.240734049323237e-09,
-      "loss": 0.1501,
-      "step": 6170
-    },
-    {
-      "epoch": 2.9623410889901653,
-      "grad_norm": 1.5596040657708192,
-      "learning_rate": 4.621200787461866e-09,
-      "loss": 0.1269,
-      "step": 6175
-    },
-    {
-      "epoch": 2.964739745742384,
-      "grad_norm": 1.2433232808545898,
-      "learning_rate": 4.0406086459893194e-09,
-      "loss": 0.1087,
-      "step": 6180
-    },
-    {
-      "epoch": 2.9671384024946033,
-      "grad_norm": 1.520664803555638,
-      "learning_rate": 3.4989621508707548e-09,
-      "loss": 0.1001,
-      "step": 6185
-    },
-    {
-      "epoch": 2.9695370592468215,
-      "grad_norm": 2.213337556588271,
-      "learning_rate": 2.996265524472519e-09,
-      "loss": 0.1257,
-      "step": 6190
-    },
-    {
-      "epoch": 2.9719357159990407,
-      "grad_norm": 1.551517099633644,
-      "learning_rate": 2.532522685529948e-09,
-      "loss": 0.1055,
-      "step": 6195
-    },
-    {
-      "epoch": 2.9743343727512594,
-      "grad_norm": 1.5250172751119109,
-      "learning_rate": 2.107737249118502e-09,
-      "loss": 0.1058,
-      "step": 6200
-    },
-    {
-      "epoch": 2.976733029503478,
-      "grad_norm": 2.028832683383426,
-      "learning_rate": 1.7219125266221271e-09,
-      "loss": 0.1005,
-      "step": 6205
-    },
-    {
-      "epoch": 2.979131686255697,
-      "grad_norm": 3.659500022563047,
-      "learning_rate": 1.3750515257104913e-09,
-      "loss": 0.1084,
-      "step": 6210
-    },
-    {
-      "epoch": 2.9815303430079156,
-      "grad_norm": 1.8413767130375305,
-      "learning_rate": 1.0671569503134528e-09,
-      "loss": 0.1046,
-      "step": 6215
-    },
-    {
-      "epoch": 2.9839289997601344,
-      "grad_norm": 1.3540342868737936,
-      "learning_rate": 7.982312006010739e-10,
-      "loss": 0.1176,
-      "step": 6220
-    },
-    {
-      "epoch": 2.986327656512353,
-      "grad_norm": 1.0218623424780813,
-      "learning_rate": 5.68276372965304e-10,
-      "loss": 0.0957,
-      "step": 6225
-    },
-    {
-      "epoch": 2.988726313264572,
-      "grad_norm": 1.4590475891432568,
-      "learning_rate": 3.772942600033247e-10,
-      "loss": 0.1145,
-      "step": 6230
-    },
-    {
-      "epoch": 2.9911249700167906,
-      "grad_norm": 1.639227053529018,
-      "learning_rate": 2.2528635050145243e-10,
-      "loss": 0.1,
-      "step": 6235
-    },
-    {
-      "epoch": 2.9935236267690093,
-      "grad_norm": 1.3287877021744292,
-      "learning_rate": 1.1225382942736673e-10,
-      "loss": 0.0781,
-      "step": 6240
-    },
-    {
-      "epoch": 2.995922283521228,
-      "grad_norm": 1.86336909717487,
-      "learning_rate": 3.819757791734269e-11,
-      "loss": 0.1027,
-      "step": 6245
-    },
-    {
-      "epoch": 2.9983209402734468,
-      "grad_norm": 1.1404093519785887,
-      "learning_rate": 3.1181732718099656e-12,
-      "loss": 0.1106,
-      "step": 6250
-    },
-    {
-      "epoch": 2.9992804029743345,
-      "step": 6252,
-      "total_flos": 3.055078882714583e+17,
-      "train_loss": 0.20201521048409316,
-      "train_runtime": 48283.8245,
-      "train_samples_per_second": 4.144,
-      "train_steps_per_second": 0.129
+      "epoch": 2.9964020148716717,
+      "step": 2082,
+      "total_flos": 5.073775214995702e+17,
+      "train_loss": 0.5172660127031643,
+      "train_runtime": 62603.1223,
+      "train_samples_per_second": 3.196,
+      "train_steps_per_second": 0.033
     }
   ],
   "logging_steps": 5,
-  "max_steps": 6252,
+  "max_steps": 2082,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
-  "save_steps": 1000,
+  "save_steps": 700,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -8977,7 +2955,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.055078882714583e+17,
+  "total_flos": 5.073775214995702e+17,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null