diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8134 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5779,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 5.7683433317951084e-05,
+      "grad_norm": 0.805375337600708,
+      "learning_rate": 1.1534025374855825e-07,
+      "loss": 1.1819,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002884171665897554,
+      "grad_norm": 0.6597965359687805,
+      "learning_rate": 5.767012687427913e-07,
+      "loss": 1.0888,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005768343331795108,
+      "grad_norm": 0.8368421792984009,
+      "learning_rate": 1.1534025374855826e-06,
+      "loss": 1.2132,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008652514997692663,
+      "grad_norm": 0.9829678535461426,
+      "learning_rate": 1.7301038062283738e-06,
+      "loss": 1.1889,
+      "step": 15
+    },
+    {
+      "epoch": 0.0011536686663590216,
+      "grad_norm": 0.7449169158935547,
+      "learning_rate": 2.3068050749711653e-06,
+      "loss": 1.2103,
+      "step": 20
+    },
+    {
+      "epoch": 0.001442085832948777,
+      "grad_norm": 0.7239571809768677,
+      "learning_rate": 2.8835063437139563e-06,
+      "loss": 1.1712,
+      "step": 25
+    },
+    {
+      "epoch": 0.0017305029995385325,
+      "grad_norm": 0.6156097650527954,
+      "learning_rate": 3.4602076124567477e-06,
+      "loss": 1.0438,
+      "step": 30
+    },
+    {
+      "epoch": 0.0020189201661282878,
+      "grad_norm": 0.7826522588729858,
+      "learning_rate": 4.036908881199539e-06,
+      "loss": 1.212,
+      "step": 35
+    },
+    {
+      "epoch": 0.0023073373327180432,
+      "grad_norm": 0.6530287861824036,
+      "learning_rate": 4.6136101499423305e-06,
+      "loss": 1.1553,
+      "step": 40
+    },
+    {
+      "epoch": 0.0025957544993077987,
+      "grad_norm": 0.7508683204650879,
+      "learning_rate": 5.190311418685121e-06,
+      "loss": 1.1354,
+      "step": 45
+    },
+    {
+      "epoch": 0.002884171665897554,
+      "grad_norm": 0.6567290425300598,
+      "learning_rate": 5.7670126874279126e-06,
+      "loss": 1.1375,
+      "step": 50
+    },
+    {
+      "epoch": 0.0031725888324873096,
+      "grad_norm": 0.5964232683181763,
+      "learning_rate": 6.3437139561707036e-06,
+      "loss": 1.1193,
+      "step": 55
+    },
+    {
+      "epoch": 0.003461005999077065,
+      "grad_norm": 0.7864587306976318,
+      "learning_rate": 6.920415224913495e-06,
+      "loss": 1.1051,
+      "step": 60
+    },
+    {
+      "epoch": 0.0037494231656668205,
+      "grad_norm": 0.643945574760437,
+      "learning_rate": 7.497116493656286e-06,
+      "loss": 1.107,
+      "step": 65
+    },
+    {
+      "epoch": 0.0040378403322565756,
+      "grad_norm": 0.6281377077102661,
+      "learning_rate": 8.073817762399077e-06,
+      "loss": 1.0673,
+      "step": 70
+    },
+    {
+      "epoch": 0.0043262574988463314,
+      "grad_norm": 0.6308133006095886,
+      "learning_rate": 8.650519031141868e-06,
+      "loss": 1.0958,
+      "step": 75
+    },
+    {
+      "epoch": 0.0046146746654360865,
+      "grad_norm": 0.566056489944458,
+      "learning_rate": 9.227220299884661e-06,
+      "loss": 1.0433,
+      "step": 80
+    },
+    {
+      "epoch": 0.004903091832025842,
+      "grad_norm": 0.6622409820556641,
+      "learning_rate": 9.803921568627451e-06,
+      "loss": 0.9888,
+      "step": 85
+    },
+    {
+      "epoch": 0.005191508998615597,
+      "grad_norm": 0.6094119548797607,
+      "learning_rate": 1.0380622837370241e-05,
+      "loss": 1.1145,
+      "step": 90
+    },
+    {
+      "epoch": 0.005479926165205353,
+      "grad_norm": 0.6860048770904541,
+      "learning_rate": 1.0957324106113035e-05,
+      "loss": 1.0594,
+      "step": 95
+    },
+    {
+      "epoch": 0.005768343331795108,
+      "grad_norm": 0.6527470946311951,
+      "learning_rate": 1.1534025374855825e-05,
+      "loss": 0.9886,
+      "step": 100
+    },
+    {
+      "epoch": 0.006056760498384864,
+      "grad_norm": 0.6624901294708252,
+      "learning_rate": 1.2110726643598615e-05,
+      "loss": 1.0142,
+      "step": 105
+    },
+    {
+      "epoch": 0.006345177664974619,
+      "grad_norm": 0.680112361907959,
+      "learning_rate": 1.2687427912341407e-05,
+      "loss": 0.9999,
+      "step": 110
+    },
+    {
+      "epoch": 0.006633594831564375,
+      "grad_norm": 0.6242687702178955,
+      "learning_rate": 1.3264129181084197e-05,
+      "loss": 0.9571,
+      "step": 115
+    },
+    {
+      "epoch": 0.00692201199815413,
+      "grad_norm": 0.6047985553741455,
+      "learning_rate": 1.384083044982699e-05,
+      "loss": 1.0342,
+      "step": 120
+    },
+    {
+      "epoch": 0.007210429164743885,
+      "grad_norm": 0.650623083114624,
+      "learning_rate": 1.4417531718569783e-05,
+      "loss": 1.0183,
+      "step": 125
+    },
+    {
+      "epoch": 0.007498846331333641,
+      "grad_norm": 0.5915358066558838,
+      "learning_rate": 1.4994232987312573e-05,
+      "loss": 1.0469,
+      "step": 130
+    },
+    {
+      "epoch": 0.007787263497923396,
+      "grad_norm": 0.653743326663971,
+      "learning_rate": 1.5570934256055363e-05,
+      "loss": 1.0459,
+      "step": 135
+    },
+    {
+      "epoch": 0.008075680664513151,
+      "grad_norm": 0.6202878355979919,
+      "learning_rate": 1.6147635524798155e-05,
+      "loss": 1.047,
+      "step": 140
+    },
+    {
+      "epoch": 0.008364097831102908,
+      "grad_norm": 0.6978910565376282,
+      "learning_rate": 1.6724336793540947e-05,
+      "loss": 1.0783,
+      "step": 145
+    },
+    {
+      "epoch": 0.008652514997692663,
+      "grad_norm": 0.6709557771682739,
+      "learning_rate": 1.7301038062283735e-05,
+      "loss": 1.0576,
+      "step": 150
+    },
+    {
+      "epoch": 0.008940932164282418,
+      "grad_norm": 0.6336321830749512,
+      "learning_rate": 1.787773933102653e-05,
+      "loss": 1.0142,
+      "step": 155
+    },
+    {
+      "epoch": 0.009229349330872173,
+      "grad_norm": 0.6070622801780701,
+      "learning_rate": 1.8454440599769322e-05,
+      "loss": 0.9873,
+      "step": 160
+    },
+    {
+      "epoch": 0.00951776649746193,
+      "grad_norm": 0.6129010915756226,
+      "learning_rate": 1.903114186851211e-05,
+      "loss": 0.9731,
+      "step": 165
+    },
+    {
+      "epoch": 0.009806183664051685,
+      "grad_norm": 0.7349147200584412,
+      "learning_rate": 1.9607843137254903e-05,
+      "loss": 1.0242,
+      "step": 170
+    },
+    {
+      "epoch": 0.01009460083064144,
+      "grad_norm": 0.6842703223228455,
+      "learning_rate": 2.0184544405997694e-05,
+      "loss": 1.1214,
+      "step": 175
+    },
+    {
+      "epoch": 0.010383017997231195,
+      "grad_norm": 0.7253619432449341,
+      "learning_rate": 2.0761245674740483e-05,
+      "loss": 1.0322,
+      "step": 180
+    },
+    {
+      "epoch": 0.01067143516382095,
+      "grad_norm": 0.7551470994949341,
+      "learning_rate": 2.1337946943483278e-05,
+      "loss": 0.9613,
+      "step": 185
+    },
+    {
+      "epoch": 0.010959852330410707,
+      "grad_norm": 0.6953349113464355,
+      "learning_rate": 2.191464821222607e-05,
+      "loss": 1.075,
+      "step": 190
+    },
+    {
+      "epoch": 0.011248269497000462,
+      "grad_norm": 0.6868691444396973,
+      "learning_rate": 2.249134948096886e-05,
+      "loss": 1.0639,
+      "step": 195
+    },
+    {
+      "epoch": 0.011536686663590217,
+      "grad_norm": 0.7638917565345764,
+      "learning_rate": 2.306805074971165e-05,
+      "loss": 1.0104,
+      "step": 200
+    },
+    {
+      "epoch": 0.011825103830179972,
+      "grad_norm": 0.6889998316764832,
+      "learning_rate": 2.3644752018454442e-05,
+      "loss": 1.0104,
+      "step": 205
+    },
+    {
+      "epoch": 0.012113520996769728,
+      "grad_norm": 0.6709694862365723,
+      "learning_rate": 2.422145328719723e-05,
+      "loss": 1.0671,
+      "step": 210
+    },
+    {
+      "epoch": 0.012401938163359483,
+      "grad_norm": 0.7222777605056763,
+      "learning_rate": 2.4798154555940022e-05,
+      "loss": 0.9234,
+      "step": 215
+    },
+    {
+      "epoch": 0.012690355329949238,
+      "grad_norm": 0.8851092457771301,
+      "learning_rate": 2.5374855824682814e-05,
+      "loss": 1.0463,
+      "step": 220
+    },
+    {
+      "epoch": 0.012978772496538993,
+      "grad_norm": 0.7736419439315796,
+      "learning_rate": 2.5951557093425606e-05,
+      "loss": 1.0129,
+      "step": 225
+    },
+    {
+      "epoch": 0.01326718966312875,
+      "grad_norm": 0.853778064250946,
+      "learning_rate": 2.6528258362168395e-05,
+      "loss": 1.0253,
+      "step": 230
+    },
+    {
+      "epoch": 0.013555606829718505,
+      "grad_norm": 0.7991467714309692,
+      "learning_rate": 2.7104959630911193e-05,
+      "loss": 0.9774,
+      "step": 235
+    },
+    {
+      "epoch": 0.01384402399630826,
+      "grad_norm": 0.718079149723053,
+      "learning_rate": 2.768166089965398e-05,
+      "loss": 1.0044,
+      "step": 240
+    },
+    {
+      "epoch": 0.014132441162898015,
+      "grad_norm": 0.7406652569770813,
+      "learning_rate": 2.8258362168396773e-05,
+      "loss": 1.0082,
+      "step": 245
+    },
+    {
+      "epoch": 0.01442085832948777,
+      "grad_norm": 0.7190337181091309,
+      "learning_rate": 2.8835063437139565e-05,
+      "loss": 0.9858,
+      "step": 250
+    },
+    {
+      "epoch": 0.014709275496077527,
+      "grad_norm": 0.7842580676078796,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 1.0378,
+      "step": 255
+    },
+    {
+      "epoch": 0.014997692662667282,
+      "grad_norm": 0.7193072438240051,
+      "learning_rate": 2.9988465974625146e-05,
+      "loss": 1.0055,
+      "step": 260
+    },
+    {
+      "epoch": 0.015286109829257037,
+      "grad_norm": 0.7311102151870728,
+      "learning_rate": 3.0565167243367934e-05,
+      "loss": 0.9497,
+      "step": 265
+    },
+    {
+      "epoch": 0.015574526995846792,
+      "grad_norm": 0.7522133588790894,
+      "learning_rate": 3.1141868512110726e-05,
+      "loss": 1.0437,
+      "step": 270
+    },
+    {
+      "epoch": 0.015862944162436547,
+      "grad_norm": 0.7482451796531677,
+      "learning_rate": 3.171856978085352e-05,
+      "loss": 0.9474,
+      "step": 275
+    },
+    {
+      "epoch": 0.016151361329026302,
+      "grad_norm": 0.7732366919517517,
+      "learning_rate": 3.229527104959631e-05,
+      "loss": 0.9688,
+      "step": 280
+    },
+    {
+      "epoch": 0.01643977849561606,
+      "grad_norm": 0.746648907661438,
+      "learning_rate": 3.28719723183391e-05,
+      "loss": 0.914,
+      "step": 285
+    },
+    {
+      "epoch": 0.016728195662205816,
+      "grad_norm": 0.6986602544784546,
+      "learning_rate": 3.344867358708189e-05,
+      "loss": 0.9613,
+      "step": 290
+    },
+    {
+      "epoch": 0.01701661282879557,
+      "grad_norm": 0.7043907642364502,
+      "learning_rate": 3.4025374855824685e-05,
+      "loss": 0.9693,
+      "step": 295
+    },
+    {
+      "epoch": 0.017305029995385326,
+      "grad_norm": 0.7637396454811096,
+      "learning_rate": 3.460207612456747e-05,
+      "loss": 0.9629,
+      "step": 300
+    },
+    {
+      "epoch": 0.01759344716197508,
+      "grad_norm": 0.7434666156768799,
+      "learning_rate": 3.517877739331027e-05,
+      "loss": 1.0924,
+      "step": 305
+    },
+    {
+      "epoch": 0.017881864328564836,
+      "grad_norm": 0.7435088157653809,
+      "learning_rate": 3.575547866205306e-05,
+      "loss": 1.0479,
+      "step": 310
+    },
+    {
+      "epoch": 0.01817028149515459,
+      "grad_norm": 0.6840139031410217,
+      "learning_rate": 3.633217993079585e-05,
+      "loss": 1.0394,
+      "step": 315
+    },
+    {
+      "epoch": 0.018458698661744346,
+      "grad_norm": 0.6994503736495972,
+      "learning_rate": 3.6908881199538644e-05,
+      "loss": 1.0005,
+      "step": 320
+    },
+    {
+      "epoch": 0.0187471158283341,
+      "grad_norm": 0.6722626686096191,
+      "learning_rate": 3.748558246828143e-05,
+      "loss": 0.9682,
+      "step": 325
+    },
+    {
+      "epoch": 0.01903553299492386,
+      "grad_norm": 0.6701967716217041,
+      "learning_rate": 3.806228373702422e-05,
+      "loss": 0.9973,
+      "step": 330
+    },
+    {
+      "epoch": 0.019323950161513614,
+      "grad_norm": 0.658715546131134,
+      "learning_rate": 3.863898500576701e-05,
+      "loss": 0.9848,
+      "step": 335
+    },
+    {
+      "epoch": 0.01961236732810337,
+      "grad_norm": 0.7802160382270813,
+      "learning_rate": 3.9215686274509805e-05,
+      "loss": 0.9661,
+      "step": 340
+    },
+    {
+      "epoch": 0.019900784494693124,
+      "grad_norm": 0.8040672540664673,
+      "learning_rate": 3.97923875432526e-05,
+      "loss": 0.9959,
+      "step": 345
+    },
+    {
+      "epoch": 0.02018920166128288,
+      "grad_norm": 0.687023937702179,
+      "learning_rate": 4.036908881199539e-05,
+      "loss": 0.9384,
+      "step": 350
+    },
+    {
+      "epoch": 0.020477618827872635,
+      "grad_norm": 0.6872786283493042,
+      "learning_rate": 4.094579008073818e-05,
+      "loss": 0.9794,
+      "step": 355
+    },
+    {
+      "epoch": 0.02076603599446239,
+      "grad_norm": 0.6797603368759155,
+      "learning_rate": 4.1522491349480966e-05,
+      "loss": 1.013,
+      "step": 360
+    },
+    {
+      "epoch": 0.021054453161052145,
+      "grad_norm": 0.6620325446128845,
+      "learning_rate": 4.209919261822376e-05,
+      "loss": 0.9782,
+      "step": 365
+    },
+    {
+      "epoch": 0.0213428703276419,
+      "grad_norm": 0.7120839357376099,
+      "learning_rate": 4.2675893886966556e-05,
+      "loss": 1.001,
+      "step": 370
+    },
+    {
+      "epoch": 0.021631287494231658,
+      "grad_norm": 0.7110047936439514,
+      "learning_rate": 4.325259515570935e-05,
+      "loss": 1.0208,
+      "step": 375
+    },
+    {
+      "epoch": 0.021919704660821413,
+      "grad_norm": 0.674694299697876,
+      "learning_rate": 4.382929642445214e-05,
+      "loss": 0.976,
+      "step": 380
+    },
+    {
+      "epoch": 0.022208121827411168,
+      "grad_norm": 0.6173574924468994,
+      "learning_rate": 4.440599769319493e-05,
+      "loss": 0.8755,
+      "step": 385
+    },
+    {
+      "epoch": 0.022496538994000923,
+      "grad_norm": 0.6821759343147278,
+      "learning_rate": 4.498269896193772e-05,
+      "loss": 1.0689,
+      "step": 390
+    },
+    {
+      "epoch": 0.022784956160590678,
+      "grad_norm": 0.709820568561554,
+      "learning_rate": 4.555940023068051e-05,
+      "loss": 1.0277,
+      "step": 395
+    },
+    {
+      "epoch": 0.023073373327180433,
+      "grad_norm": 0.6610621809959412,
+      "learning_rate": 4.61361014994233e-05,
+      "loss": 1.0061,
+      "step": 400
+    },
+    {
+      "epoch": 0.023361790493770188,
+      "grad_norm": 0.7234196662902832,
+      "learning_rate": 4.671280276816609e-05,
+      "loss": 1.0004,
+      "step": 405
+    },
+    {
+      "epoch": 0.023650207660359943,
+      "grad_norm": 0.7094035744667053,
+      "learning_rate": 4.7289504036908884e-05,
+      "loss": 0.9914,
+      "step": 410
+    },
+    {
+      "epoch": 0.0239386248269497,
+      "grad_norm": 0.6732367873191833,
+      "learning_rate": 4.7866205305651676e-05,
+      "loss": 0.9613,
+      "step": 415
+    },
+    {
+      "epoch": 0.024227041993539457,
+      "grad_norm": 0.6192376613616943,
+      "learning_rate": 4.844290657439446e-05,
+      "loss": 1.0058,
+      "step": 420
+    },
+    {
+      "epoch": 0.024515459160129212,
+      "grad_norm": 0.6171934604644775,
+      "learning_rate": 4.901960784313725e-05,
+      "loss": 1.027,
+      "step": 425
+    },
+    {
+      "epoch": 0.024803876326718967,
+      "grad_norm": 0.6176871657371521,
+      "learning_rate": 4.9596309111880045e-05,
+      "loss": 0.9524,
+      "step": 430
+    },
+    {
+      "epoch": 0.025092293493308722,
+      "grad_norm": 0.6524646878242493,
+      "learning_rate": 5.017301038062284e-05,
+      "loss": 1.0202,
+      "step": 435
+    },
+    {
+      "epoch": 0.025380710659898477,
+      "grad_norm": 0.6370823979377747,
+      "learning_rate": 5.074971164936563e-05,
+      "loss": 0.9185,
+      "step": 440
+    },
+    {
+      "epoch": 0.025669127826488232,
+      "grad_norm": 0.6290232539176941,
+      "learning_rate": 5.132641291810843e-05,
+      "loss": 1.0323,
+      "step": 445
+    },
+    {
+      "epoch": 0.025957544993077987,
+      "grad_norm": 0.6379473805427551,
+      "learning_rate": 5.190311418685121e-05,
+      "loss": 0.9921,
+      "step": 450
+    },
+    {
+      "epoch": 0.026245962159667742,
+      "grad_norm": 0.6444252729415894,
+      "learning_rate": 5.2479815455594004e-05,
+      "loss": 1.0232,
+      "step": 455
+    },
+    {
+      "epoch": 0.0265343793262575,
+      "grad_norm": 0.6765018105506897,
+      "learning_rate": 5.305651672433679e-05,
+      "loss": 0.9987,
+      "step": 460
+    },
+    {
+      "epoch": 0.026822796492847256,
+      "grad_norm": 0.6089096069335938,
+      "learning_rate": 5.363321799307959e-05,
+      "loss": 0.9996,
+      "step": 465
+    },
+    {
+      "epoch": 0.02711121365943701,
+      "grad_norm": 0.6402391195297241,
+      "learning_rate": 5.4209919261822386e-05,
+      "loss": 0.9029,
+      "step": 470
+    },
+    {
+      "epoch": 0.027399630826026766,
+      "grad_norm": 0.616694450378418,
+      "learning_rate": 5.478662053056517e-05,
+      "loss": 0.9625,
+      "step": 475
+    },
+    {
+      "epoch": 0.02768804799261652,
+      "grad_norm": 0.6394984126091003,
+      "learning_rate": 5.536332179930796e-05,
+      "loss": 0.984,
+      "step": 480
+    },
+    {
+      "epoch": 0.027976465159206276,
+      "grad_norm": 0.6072252988815308,
+      "learning_rate": 5.594002306805075e-05,
+      "loss": 0.9278,
+      "step": 485
+    },
+    {
+      "epoch": 0.02826488232579603,
+      "grad_norm": 0.6287209987640381,
+      "learning_rate": 5.651672433679355e-05,
+      "loss": 1.011,
+      "step": 490
+    },
+    {
+      "epoch": 0.028553299492385786,
+      "grad_norm": 0.610227644443512,
+      "learning_rate": 5.709342560553633e-05,
+      "loss": 0.9999,
+      "step": 495
+    },
+    {
+      "epoch": 0.02884171665897554,
+      "grad_norm": 0.5862972736358643,
+      "learning_rate": 5.767012687427913e-05,
+      "loss": 0.9724,
+      "step": 500
+    },
+    {
+      "epoch": 0.08738536078906385,
+      "grad_norm": 0.5334975719451904,
+      "learning_rate": 0.00017474048442906573,
+      "loss": 0.9612,
+      "step": 505
+    },
+    {
+      "epoch": 0.08825056238103478,
+      "grad_norm": 0.4906371831893921,
+      "learning_rate": 0.00017647058823529413,
+      "loss": 1.0106,
+      "step": 510
+    },
+    {
+      "epoch": 0.0891157639730057,
+      "grad_norm": 0.6553336977958679,
+      "learning_rate": 0.0001782006920415225,
+      "loss": 1.0372,
+      "step": 515
+    },
+    {
+      "epoch": 0.08998096556497664,
+      "grad_norm": 0.4532735347747803,
+      "learning_rate": 0.00017993079584775087,
+      "loss": 1.0166,
+      "step": 520
+    },
+    {
+      "epoch": 0.09084616715694757,
+      "grad_norm": 0.474738210439682,
+      "learning_rate": 0.00018166089965397926,
+      "loss": 0.9979,
+      "step": 525
+    },
+    {
+      "epoch": 0.0917113687489185,
+      "grad_norm": 0.4620087146759033,
+      "learning_rate": 0.0001833910034602076,
+      "loss": 1.0254,
+      "step": 530
+    },
+    {
+      "epoch": 0.09257657034088942,
+      "grad_norm": 0.4435741901397705,
+      "learning_rate": 0.000185121107266436,
+      "loss": 0.9888,
+      "step": 535
+    },
+    {
+      "epoch": 0.09344177193286035,
+      "grad_norm": 0.43698352575302124,
+      "learning_rate": 0.00018685121107266437,
+      "loss": 0.9884,
+      "step": 540
+    },
+    {
+      "epoch": 0.09430697352483129,
+      "grad_norm": 0.41308143734931946,
+      "learning_rate": 0.00018858131487889274,
+      "loss": 1.0003,
+      "step": 545
+    },
+    {
+      "epoch": 0.09517217511680222,
+      "grad_norm": 0.4729122817516327,
+      "learning_rate": 0.00019031141868512113,
+      "loss": 0.9812,
+      "step": 550
+    },
+    {
+      "epoch": 0.09603737670877315,
+      "grad_norm": 0.42877742648124695,
+      "learning_rate": 0.00019204152249134948,
+      "loss": 0.9832,
+      "step": 555
+    },
+    {
+      "epoch": 0.09690257830074407,
+      "grad_norm": 0.4206876754760742,
+      "learning_rate": 0.00019377162629757784,
+      "loss": 1.009,
+      "step": 560
+    },
+    {
+      "epoch": 0.097767779892715,
+      "grad_norm": 0.4647689461708069,
+      "learning_rate": 0.00019550173010380624,
+      "loss": 0.945,
+      "step": 565
+    },
+    {
+      "epoch": 0.09863298148468594,
+      "grad_norm": 0.41559451818466187,
+      "learning_rate": 0.0001972318339100346,
+      "loss": 1.0342,
+      "step": 570
+    },
+    {
+      "epoch": 0.09949818307665687,
+      "grad_norm": 0.44591063261032104,
+      "learning_rate": 0.000198961937716263,
+      "loss": 0.9927,
+      "step": 575
+    },
+    {
+      "epoch": 0.10036338466862779,
+      "grad_norm": 0.41210636496543884,
+      "learning_rate": 0.00019999992702804517,
+      "loss": 0.9835,
+      "step": 580
+    },
+    {
+      "epoch": 0.10122858626059872,
+      "grad_norm": 0.4488976001739502,
+      "learning_rate": 0.0001999991060947763,
+      "loss": 0.9783,
+      "step": 585
+    },
+    {
+      "epoch": 0.10209378785256965,
+      "grad_norm": 0.49996423721313477,
+      "learning_rate": 0.0001999973730208081,
+      "loss": 1.0076,
+      "step": 590
+    },
+    {
+      "epoch": 0.10295898944454057,
+      "grad_norm": 0.3953074812889099,
+      "learning_rate": 0.00019999472782194876,
+      "loss": 0.9747,
+      "step": 595
+    },
+    {
+      "epoch": 0.10382419103651151,
+      "grad_norm": 0.42277634143829346,
+      "learning_rate": 0.0001999911705223265,
+      "loss": 0.9959,
+      "step": 600
+    },
+    {
+      "epoch": 0.10468939262848244,
+      "grad_norm": 0.4215725064277649,
+      "learning_rate": 0.00019998670115438909,
+      "loss": 1.0212,
+      "step": 605
+    },
+    {
+      "epoch": 0.10555459422045337,
+      "grad_norm": 0.4542864263057709,
+      "learning_rate": 0.0001999813197589039,
+      "loss": 0.9818,
+      "step": 610
+    },
+    {
+      "epoch": 0.10641979581242429,
+      "grad_norm": 0.4049575626850128,
+      "learning_rate": 0.00019997502638495723,
+      "loss": 0.9868,
+      "step": 615
+    },
+    {
+      "epoch": 0.10728499740439522,
+      "grad_norm": 0.4267943799495697,
+      "learning_rate": 0.00019996782108995404,
+      "loss": 1.0023,
+      "step": 620
+    },
+    {
+      "epoch": 0.10815019899636616,
+      "grad_norm": 0.3867630660533905,
+      "learning_rate": 0.00019995970393961733,
+      "loss": 0.9584,
+      "step": 625
+    },
+    {
+      "epoch": 0.10901540058833709,
+      "grad_norm": 0.4118204116821289,
+      "learning_rate": 0.00019995067500798758,
+      "loss": 1.0091,
+      "step": 630
+    },
+    {
+      "epoch": 0.10988060218030801,
+      "grad_norm": 0.4244009852409363,
+      "learning_rate": 0.00019994073437742214,
+      "loss": 0.9887,
+      "step": 635
+    },
+    {
+      "epoch": 0.11074580377227894,
+      "grad_norm": 0.37798789143562317,
+      "learning_rate": 0.0001999298821385943,
+      "loss": 0.9533,
+      "step": 640
+    },
+    {
+      "epoch": 0.11161100536424987,
+      "grad_norm": 0.4218810796737671,
+      "learning_rate": 0.00019991811839049263,
+      "loss": 0.9813,
+      "step": 645
+    },
+    {
+      "epoch": 0.1124762069562208,
+      "grad_norm": 0.419813871383667,
+      "learning_rate": 0.00019990544324042007,
+      "loss": 0.9898,
+      "step": 650
+    },
+    {
+      "epoch": 0.11334140854819173,
+      "grad_norm": 0.4155246615409851,
+      "learning_rate": 0.00019989185680399283,
+      "loss": 1.0018,
+      "step": 655
+    },
+    {
+      "epoch": 0.11420661014016266,
+      "grad_norm": 0.45885682106018066,
+      "learning_rate": 0.00019987735920513943,
+      "loss": 1.0023,
+      "step": 660
+    },
+    {
+      "epoch": 0.11507181173213359,
+      "grad_norm": 0.45299044251441956,
+      "learning_rate": 0.00019986195057609957,
+      "loss": 1.0061,
+      "step": 665
+    },
+    {
+      "epoch": 0.11593701332410451,
+      "grad_norm": 0.4102267920970917,
+      "learning_rate": 0.00019984563105742285,
+      "loss": 1.0284,
+      "step": 670
+    },
+    {
+      "epoch": 0.11680221491607544,
+      "grad_norm": 0.41383153200149536,
+      "learning_rate": 0.0001998284007979676,
+      "loss": 0.9999,
+      "step": 675
+    },
+    {
+      "epoch": 0.11766741650804638,
+      "grad_norm": 0.4205871522426605,
+      "learning_rate": 0.00019981025995489943,
+      "loss": 0.9574,
+      "step": 680
+    },
+    {
+      "epoch": 0.1185326181000173,
+      "grad_norm": 0.43072691559791565,
+      "learning_rate": 0.00019979120869368982,
+      "loss": 0.9703,
+      "step": 685
+    },
+    {
+      "epoch": 0.11939781969198823,
+      "grad_norm": 0.4149424135684967,
+      "learning_rate": 0.00019977124718811463,
+      "loss": 1.0093,
+      "step": 690
+    },
+    {
+      "epoch": 0.12026302128395916,
+      "grad_norm": 0.46796900033950806,
+      "learning_rate": 0.00019975037562025254,
+      "loss": 0.9802,
+      "step": 695
+    },
+    {
+      "epoch": 0.12112822287593009,
+      "grad_norm": 0.3972123861312866,
+      "learning_rate": 0.00019972859418048328,
+      "loss": 1.0109,
+      "step": 700
+    },
+    {
+      "epoch": 0.12199342446790103,
+      "grad_norm": 0.4356140196323395,
+      "learning_rate": 0.00019970590306748603,
+      "loss": 1.0295,
+      "step": 705
+    },
+    {
+      "epoch": 0.12285862605987195,
+      "grad_norm": 0.43844765424728394,
+      "learning_rate": 0.00019968230248823746,
+      "loss": 0.9474,
+      "step": 710
+    },
+    {
+      "epoch": 0.12372382765184288,
+      "grad_norm": 0.42332398891448975,
+      "learning_rate": 0.00019965779265801006,
+      "loss": 1.0008,
+      "step": 715
+    },
+    {
+      "epoch": 0.1245890292438138,
+      "grad_norm": 0.4387184679508209,
+      "learning_rate": 0.00019963237380036993,
+      "loss": 0.9893,
+      "step": 720
+    },
+    {
+      "epoch": 0.12545423083578475,
+      "grad_norm": 0.42322424054145813,
+      "learning_rate": 0.00019960604614717485,
+      "loss": 0.9831,
+      "step": 725
+    },
+    {
+      "epoch": 0.12631943242775567,
+      "grad_norm": 0.41627830266952515,
+      "learning_rate": 0.00019957880993857227,
+      "loss": 1.006,
+      "step": 730
+    },
+    {
+      "epoch": 0.1271846340197266,
+      "grad_norm": 0.4279492199420929,
+      "learning_rate": 0.00019955066542299695,
+      "loss": 0.9847,
+      "step": 735
+    },
+    {
+      "epoch": 0.12804983561169753,
+      "grad_norm": 0.43340563774108887,
+      "learning_rate": 0.00019952161285716872,
+      "loss": 0.9798,
+      "step": 740
+    },
+    {
+      "epoch": 0.12891503720366845,
+      "grad_norm": 0.4125729501247406,
+      "learning_rate": 0.00019949165250609022,
+      "loss": 0.976,
+      "step": 745
+    },
+    {
+      "epoch": 0.12978023879563938,
+      "grad_norm": 0.43918147683143616,
+      "learning_rate": 0.0001994607846430445,
+      "loss": 0.9614,
+      "step": 750
+    },
+    {
+      "epoch": 0.1306454403876103,
+      "grad_norm": 0.43273496627807617,
+      "learning_rate": 0.00019942900954959244,
+      "loss": 1.0095,
+      "step": 755
+    },
+    {
+      "epoch": 0.13151064197958123,
+      "grad_norm": 0.43806779384613037,
+      "learning_rate": 0.0001993963275155701,
+      "loss": 0.9866,
+      "step": 760
+    },
+    {
+      "epoch": 0.13237584357155216,
+      "grad_norm": 0.43065160512924194,
+      "learning_rate": 0.00019936273883908637,
+      "loss": 0.9964,
+      "step": 765
+    },
+    {
+      "epoch": 0.13324104516352311,
+      "grad_norm": 0.41607436537742615,
+      "learning_rate": 0.0001993282438265199,
+      "loss": 0.9926,
+      "step": 770
+    },
+    {
+      "epoch": 0.13410624675549404,
+      "grad_norm": 0.419233113527298,
+      "learning_rate": 0.00019929284279251658,
+      "loss": 0.995,
+      "step": 775
+    },
+    {
+      "epoch": 0.13497144834746497,
+      "grad_norm": 0.41398414969444275,
+      "learning_rate": 0.00019925653605998655,
+      "loss": 0.983,
+      "step": 780
+    },
+    {
+      "epoch": 0.1358366499394359,
+      "grad_norm": 0.43536847829818726,
+      "learning_rate": 0.00019921932396010122,
+      "loss": 0.9669,
+      "step": 785
+    },
+    {
+      "epoch": 0.13670185153140682,
+      "grad_norm": 0.41906699538230896,
+      "learning_rate": 0.00019918120683229043,
+      "loss": 0.9839,
+      "step": 790
+    },
+    {
+      "epoch": 0.13756705312337775,
+      "grad_norm": 0.47571635246276855,
+      "learning_rate": 0.00019914218502423898,
+      "loss": 1.0062,
+      "step": 795
+    },
+    {
+      "epoch": 0.13843225471534867,
+      "grad_norm": 0.4490084648132324,
+      "learning_rate": 0.00019910225889188397,
+      "loss": 1.0186,
+      "step": 800
+    },
+    {
+      "epoch": 0.1392974563073196,
+      "grad_norm": 0.43234339356422424,
+      "learning_rate": 0.00019906142879941107,
+      "loss": 0.973,
+      "step": 805
+    },
+    {
+      "epoch": 0.14016265789929053,
+      "grad_norm": 0.42632317543029785,
+      "learning_rate": 0.00019901969511925153,
+      "loss": 0.9917,
+      "step": 810
+    },
+    {
+      "epoch": 0.14102785949126145,
+      "grad_norm": 0.4437479078769684,
+      "learning_rate": 0.00019897705823207867,
+      "loss": 0.9794,
+      "step": 815
+    },
+    {
+      "epoch": 0.14189306108323238,
+      "grad_norm": 0.416469931602478,
+      "learning_rate": 0.00019893351852680433,
+      "loss": 0.9611,
+      "step": 820
+    },
+    {
+      "epoch": 0.14275826267520333,
+      "grad_norm": 0.4988054037094116,
+      "learning_rate": 0.00019888907640057543,
+      "loss": 0.9644,
+      "step": 825
+    },
+    {
+      "epoch": 0.14362346426717426,
+      "grad_norm": 0.44064489006996155,
+      "learning_rate": 0.0001988437322587703,
+      "loss": 0.9768,
+      "step": 830
+    },
+    {
+      "epoch": 0.1444886658591452,
+      "grad_norm": 0.43101176619529724,
+      "learning_rate": 0.00019879748651499508,
+      "loss": 1.0167,
+      "step": 835
+    },
+    {
+      "epoch": 0.14535386745111611,
+      "grad_norm": 0.4331214725971222,
+      "learning_rate": 0.0001987503395910797,
+      "loss": 0.9585,
+      "step": 840
+    },
+    {
+      "epoch": 0.14621906904308704,
+      "grad_norm": 0.4513319730758667,
+      "learning_rate": 0.00019870229191707437,
+      "loss": 0.9678,
+      "step": 845
+    },
+    {
+      "epoch": 0.14708427063505797,
+      "grad_norm": 0.43121349811553955,
+      "learning_rate": 0.00019865334393124535,
+      "loss": 0.9701,
+      "step": 850
+    },
+    {
+      "epoch": 0.1479494722270289,
+      "grad_norm": 0.4441729485988617,
+      "learning_rate": 0.00019860349608007111,
+      "loss": 0.9771,
+      "step": 855
+    },
+    {
+      "epoch": 0.14881467381899982,
+      "grad_norm": 0.447622686624527,
+      "learning_rate": 0.00019855274881823833,
+      "loss": 0.9707,
+      "step": 860
+    },
+    {
+      "epoch": 0.14967987541097075,
+      "grad_norm": 0.439892441034317,
+      "learning_rate": 0.0001985011026086375,
+      "loss": 0.945,
+      "step": 865
+    },
+    {
+      "epoch": 0.15054507700294167,
+      "grad_norm": 0.45804309844970703,
+      "learning_rate": 0.00019844855792235897,
+      "loss": 1.0298,
+      "step": 870
+    },
+    {
+      "epoch": 0.15141027859491263,
+      "grad_norm": 0.427182674407959,
+      "learning_rate": 0.00019839511523868847,
+      "loss": 0.9811,
+      "step": 875
+    },
+    {
+      "epoch": 0.15227548018688356,
+      "grad_norm": 0.397521436214447,
+      "learning_rate": 0.00019834077504510283,
+      "loss": 0.9782,
+      "step": 880
+    },
+    {
+      "epoch": 0.15314068177885448,
+      "grad_norm": 0.44817712903022766,
+      "learning_rate": 0.00019828553783726553,
+      "loss": 1.0242,
+      "step": 885
+    },
+    {
+      "epoch": 0.1540058833708254,
+      "grad_norm": 0.4743613302707672,
+      "learning_rate": 0.00019822940411902212,
+      "loss": 0.9646,
+      "step": 890
+    },
+    {
+      "epoch": 0.15487108496279633,
+      "grad_norm": 0.4440392255783081,
+      "learning_rate": 0.0001981723744023956,
+      "loss": 1.013,
+      "step": 895
+    },
+    {
+      "epoch": 0.15573628655476726,
+      "grad_norm": 0.4056445062160492,
+      "learning_rate": 0.000198114449207582,
+      "loss": 0.9629,
+      "step": 900
+    },
+    {
+      "epoch": 0.1566014881467382,
+      "grad_norm": 0.44350746273994446,
+      "learning_rate": 0.0001980556290629452,
+      "loss": 0.944,
+      "step": 905
+    },
+    {
+      "epoch": 0.15746668973870911,
+      "grad_norm": 0.4516960382461548,
+      "learning_rate": 0.00019799591450501253,
+      "loss": 0.9772,
+      "step": 910
+    },
+    {
+      "epoch": 0.15833189133068004,
+      "grad_norm": 0.4214520752429962,
+      "learning_rate": 0.00019793530607846967,
+      "loss": 0.959,
+      "step": 915
+    },
+    {
+      "epoch": 0.15919709292265097,
+      "grad_norm": 0.4131547212600708,
+      "learning_rate": 0.00019787380433615562,
+      "loss": 0.9922,
+      "step": 920
+    },
+    {
+      "epoch": 0.1600622945146219,
+      "grad_norm": 0.43274828791618347,
+      "learning_rate": 0.0001978114098390579,
+      "loss": 0.9896,
+      "step": 925
+    },
+    {
+      "epoch": 0.16092749610659285,
+      "grad_norm": 0.41580265760421753,
+      "learning_rate": 0.0001977481231563071,
+      "loss": 0.9829,
+      "step": 930
+    },
+    {
+      "epoch": 0.16179269769856378,
+      "grad_norm": 0.4266924560070038,
+      "learning_rate": 0.0001976839448651721,
+      "loss": 1.0039,
+      "step": 935
+    },
+    {
+      "epoch": 0.1626578992905347,
+      "grad_norm": 0.4301856756210327,
+      "learning_rate": 0.00019761887555105428,
+      "loss": 0.9754,
+      "step": 940
+    },
+    {
+      "epoch": 0.16352310088250563,
+      "grad_norm": 0.4580918848514557,
+      "learning_rate": 0.00019755291580748278,
+      "loss": 0.9878,
+      "step": 945
+    },
+    {
+      "epoch": 0.16438830247447656,
+      "grad_norm": 0.42018818855285645,
+      "learning_rate": 0.0001974860662361086,
+      "loss": 1.0021,
+      "step": 950
+    },
+    {
+      "epoch": 0.16525350406644748,
+      "grad_norm": 0.44303232431411743,
+      "learning_rate": 0.00019741832744669938,
+      "loss": 0.9844,
+      "step": 955
+    },
+    {
+      "epoch": 0.1661187056584184,
+      "grad_norm": 0.4127476215362549,
+      "learning_rate": 0.0001973497000571336,
+      "loss": 0.9969,
+      "step": 960
+    },
+    {
+      "epoch": 0.16698390725038934,
+      "grad_norm": 0.413120299577713,
+      "learning_rate": 0.00019728018469339532,
+      "loss": 1.0029,
+      "step": 965
+    },
+    {
+      "epoch": 0.16784910884236026,
+      "grad_norm": 0.41281285881996155,
+      "learning_rate": 0.00019720978198956807,
+      "loss": 0.9744,
+      "step": 970
+    },
+    {
+      "epoch": 0.1687143104343312,
+      "grad_norm": 0.42970991134643555,
+      "learning_rate": 0.00019713849258782933,
+      "loss": 0.9722,
+      "step": 975
+    },
+    {
+      "epoch": 0.16957951202630211,
+      "grad_norm": 0.4328010678291321,
+      "learning_rate": 0.00019706631713844455,
+      "loss": 0.9985,
+      "step": 980
+    },
+    {
+      "epoch": 0.17044471361827307,
+      "grad_norm": 0.42618030309677124,
+      "learning_rate": 0.00019699325629976127,
+      "loss": 0.9983,
+      "step": 985
+    },
+    {
+      "epoch": 0.171309915210244,
+      "grad_norm": 0.44706591963768005,
+      "learning_rate": 0.00019691931073820312,
+      "loss": 1.0262,
+      "step": 990
+    },
+    {
+      "epoch": 0.17217511680221492,
+      "grad_norm": 0.41572490334510803,
+      "learning_rate": 0.00019684448112826361,
+      "loss": 0.9898,
+      "step": 995
+    },
+    {
+      "epoch": 0.17304031839418585,
+      "grad_norm": 0.4546913206577301,
+      "learning_rate": 0.00019676876815250024,
+      "loss": 0.9845,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17390551998615678,
+      "grad_norm": 0.41583436727523804,
+      "learning_rate": 0.00019669217250152806,
+      "loss": 0.9786,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1747707215781277,
+      "grad_norm": 0.4279019236564636,
+      "learning_rate": 0.00019661469487401335,
+      "loss": 0.9573,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17563592317009863,
+      "grad_norm": 0.4380638003349304,
+      "learning_rate": 0.00019653633597666744,
+      "loss": 0.9666,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17650112476206956,
+      "grad_norm": 0.4483678638935089,
+      "learning_rate": 0.00019645709652424016,
+      "loss": 1.0032,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17736632635404048,
+      "grad_norm": 0.4582922160625458,
+      "learning_rate": 0.0001963769772395132,
+      "loss": 1.0026,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1782315279460114,
+      "grad_norm": 0.39407840371131897,
+      "learning_rate": 0.00019629597885329373,
+      "loss": 0.9871,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17909672953798236,
+      "grad_norm": 0.4242671728134155,
+      "learning_rate": 0.00019621410210440759,
+      "loss": 1.006,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1799619311299533,
+      "grad_norm": 0.4255340099334717,
+      "learning_rate": 0.00019613134773969256,
+      "loss": 1.036,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18082713272192422,
+      "grad_norm": 0.4595358967781067,
+      "learning_rate": 0.00019604771651399175,
+      "loss": 0.9635,
+      "step": 1045
+    },
+    {
+      "epoch": 0.18169233431389514,
+      "grad_norm": 0.4306567907333374,
+      "learning_rate": 0.0001959632091901463,
+      "loss": 0.9313,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18255753590586607,
+      "grad_norm": 0.4518164396286011,
+      "learning_rate": 0.00019587782653898884,
+      "loss": 1.0311,
+      "step": 1055
+    },
+    {
+      "epoch": 0.183422737497837,
+      "grad_norm": 0.41974586248397827,
+      "learning_rate": 0.00019579156933933632,
+      "loss": 0.9976,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18428793908980792,
+      "grad_norm": 0.4630170464515686,
+      "learning_rate": 0.00019570443837798265,
+      "loss": 0.9915,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18515314068177885,
+      "grad_norm": 0.4572630524635315,
+      "learning_rate": 0.00019561643444969203,
+      "loss": 1.0228,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18601834227374978,
+      "grad_norm": 0.46339964866638184,
+      "learning_rate": 0.00019552755835719116,
+      "loss": 0.9616,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1868835438657207,
+      "grad_norm": 0.43660199642181396,
+      "learning_rate": 0.00019543781091116243,
+      "loss": 0.9879,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18774874545769163,
+      "grad_norm": 0.4161589741706848,
+      "learning_rate": 0.00019534719293023604,
+      "loss": 0.9961,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18861394704966258,
+      "grad_norm": 0.4922807216644287,
+      "learning_rate": 0.00019525570524098288,
+      "loss": 1.0022,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1894791486416335,
+      "grad_norm": 0.4476989805698395,
+      "learning_rate": 0.0001951633486779069,
+      "loss": 1.007,
+      "step": 1095
+    },
+    {
+      "epoch": 0.19034435023360444,
+      "grad_norm": 0.4585910141468048,
+      "learning_rate": 0.0001950701240834374,
+      "loss": 0.9368,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19120955182557536,
+      "grad_norm": 0.440684050321579,
+      "learning_rate": 0.00019497603230792145,
+      "loss": 1.0081,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1920747534175463,
+      "grad_norm": 0.41014137864112854,
+      "learning_rate": 0.00019488107420961612,
+      "loss": 0.971,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19293995500951722,
+      "grad_norm": 0.4705316126346588,
+      "learning_rate": 0.00019478525065468058,
+      "loss": 0.9501,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19380515660148814,
+      "grad_norm": 0.4233025014400482,
+      "learning_rate": 0.00019468856251716833,
+      "loss": 1.0,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19467035819345907,
+      "grad_norm": 0.47275301814079285,
+      "learning_rate": 0.00019459101067901906,
+      "loss": 1.0192,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19553555978543,
+      "grad_norm": 0.47677910327911377,
+      "learning_rate": 0.0001944925960300508,
+      "loss": 0.9638,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19640076137740092,
+      "grad_norm": 0.4414370656013489,
+      "learning_rate": 0.00019439331946795158,
+      "loss": 0.9603,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19726596296937188,
+      "grad_norm": 0.43756040930747986,
+      "learning_rate": 0.0001942931818982715,
+      "loss": 1.0019,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1981311645613428,
+      "grad_norm": 0.434892863035202,
+      "learning_rate": 0.00019419218423441414,
+      "loss": 0.9573,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19899636615331373,
+      "grad_norm": 0.43879738450050354,
+      "learning_rate": 0.00019409032739762863,
+      "loss": 0.9556,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19986156774528466,
+      "grad_norm": 0.4677106738090515,
+      "learning_rate": 0.00019398761231700088,
+      "loss": 1.026,
+      "step": 1155
+    },
+    {
+      "epoch": 0.20072676933725558,
+      "grad_norm": 0.44820311665534973,
+      "learning_rate": 0.0001938840399294454,
+      "loss": 0.9921,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2015919709292265,
+      "grad_norm": 0.4272112250328064,
+      "learning_rate": 0.00019377961117969644,
+      "loss": 0.9864,
+      "step": 1165
+    },
+    {
+      "epoch": 0.20245717252119744,
+      "grad_norm": 0.44302093982696533,
+      "learning_rate": 0.0001936743270202997,
+      "loss": 0.9945,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20332237411316836,
+      "grad_norm": 0.43296101689338684,
+      "learning_rate": 0.0001935681884116034,
+      "loss": 0.9394,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2041875757051393,
+      "grad_norm": 0.4531594514846802,
+      "learning_rate": 0.00019346119632174968,
+      "loss": 1.0248,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20505277729711022,
+      "grad_norm": 0.4667482078075409,
+      "learning_rate": 0.00019335335172666565,
+      "loss": 0.944,
+      "step": 1185
+    },
+    {
+      "epoch": 0.20591797888908114,
+      "grad_norm": 0.41053980588912964,
+      "learning_rate": 0.00019324465561005452,
+      "loss": 0.9948,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2067831804810521,
+      "grad_norm": 0.4748595952987671,
+      "learning_rate": 0.0001931351089633867,
+      "loss": 0.9561,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20764838207302302,
+      "grad_norm": 0.45077425241470337,
+      "learning_rate": 0.00019302471278589061,
+      "loss": 0.9711,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20851358366499395,
+      "grad_norm": 0.4596273899078369,
+      "learning_rate": 0.00019291346808454382,
+      "loss": 0.9931,
+      "step": 1205
+    },
+    {
+      "epoch": 0.20937878525696488,
+      "grad_norm": 0.4171757102012634,
+      "learning_rate": 0.00019280137587406352,
+      "loss": 0.9529,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2102439868489358,
+      "grad_norm": 0.43460047245025635,
+      "learning_rate": 0.00019268843717689754,
+      "loss": 0.9599,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21110918844090673,
+      "grad_norm": 0.44364187121391296,
+      "learning_rate": 0.00019257465302321495,
+      "loss": 0.973,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21197439003287766,
+      "grad_norm": 0.4260512590408325,
+      "learning_rate": 0.00019246002445089656,
+      "loss": 0.9951,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21283959162484858,
+      "grad_norm": 0.47502508759498596,
+      "learning_rate": 0.00019234455250552554,
+      "loss": 1.0132,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2137047932168195,
+      "grad_norm": 0.43591979146003723,
+      "learning_rate": 0.00019222823824037804,
+      "loss": 0.9455,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21456999480879044,
+      "grad_norm": 0.42019540071487427,
+      "learning_rate": 0.00019211108271641316,
+      "loss": 1.009,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21543519640076136,
+      "grad_norm": 0.4755167067050934,
+      "learning_rate": 0.00019199308700226374,
+      "loss": 1.0004,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21630039799273232,
+      "grad_norm": 0.4237242639064789,
+      "learning_rate": 0.00019187425217422632,
+      "loss": 0.9806,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21716559958470324,
+      "grad_norm": 0.4744568169116974,
+      "learning_rate": 0.0001917545793162514,
+      "loss": 0.9841,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21803080117667417,
+      "grad_norm": 0.4334425628185272,
+      "learning_rate": 0.00019163406951993363,
+      "loss": 0.9809,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2188960027686451,
+      "grad_norm": 0.419150710105896,
+      "learning_rate": 0.00019151272388450173,
+      "loss": 0.9872,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21976120436061602,
+      "grad_norm": 0.4238300621509552,
+      "learning_rate": 0.00019139054351680846,
+      "loss": 0.9613,
+      "step": 1270
+    },
+    {
+      "epoch": 0.22062640595258695,
+      "grad_norm": 0.4214048385620117,
+      "learning_rate": 0.00019126752953132069,
+      "loss": 0.9791,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22149160754455788,
+      "grad_norm": 0.4584942162036896,
+      "learning_rate": 0.00019114368305010905,
+      "loss": 1.0075,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2223568091365288,
+      "grad_norm": 0.435167133808136,
+      "learning_rate": 0.00019101900520283784,
+      "loss": 0.9987,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22322201072849973,
+      "grad_norm": 0.43369898200035095,
+      "learning_rate": 0.0001908934971267546,
+      "loss": 0.9703,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22408721232047066,
+      "grad_norm": 0.452718585729599,
+      "learning_rate": 0.00019076715996667977,
+      "loss": 0.9758,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2249524139124416,
+      "grad_norm": 0.43349212408065796,
+      "learning_rate": 0.00019063999487499637,
+      "loss": 0.9425,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22581761550441254,
+      "grad_norm": 0.5821581482887268,
+      "learning_rate": 0.00019051200301163922,
+      "loss": 1.0269,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22668281709638347,
+      "grad_norm": 0.4367370307445526,
+      "learning_rate": 0.00019038318554408479,
+      "loss": 0.981,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2275480186883544,
+      "grad_norm": 0.43077537417411804,
+      "learning_rate": 0.00019025354364734,
+      "loss": 0.9795,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22841322028032532,
+      "grad_norm": 0.43558841943740845,
+      "learning_rate": 0.00019012307850393197,
+      "loss": 0.9835,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22927842187229625,
+      "grad_norm": 0.4512316882610321,
+      "learning_rate": 0.000189991791303897,
+      "loss": 1.0004,
+      "step": 1325
+    },
+    {
+      "epoch": 0.23014362346426717,
+      "grad_norm": 0.4050248861312866,
+      "learning_rate": 0.0001898596832447698,
+      "loss": 0.9768,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2310088250562381,
+      "grad_norm": 0.45898815989494324,
+      "learning_rate": 0.00018972675553157253,
+      "loss": 1.0172,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23187402664820903,
+      "grad_norm": 0.4515649676322937,
+      "learning_rate": 0.00018959300937680372,
+      "loss": 1.0098,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23273922824017995,
+      "grad_norm": 0.42254117131233215,
+      "learning_rate": 0.00018945844600042747,
+      "loss": 0.9975,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23360442983215088,
+      "grad_norm": 0.4825042486190796,
+      "learning_rate": 0.00018932306662986209,
+      "loss": 1.008,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23446963142412183,
+      "grad_norm": 0.4267342686653137,
+      "learning_rate": 0.00018918687249996888,
+      "loss": 1.0035,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23533483301609276,
+      "grad_norm": 0.4438179135322571,
+      "learning_rate": 0.00018904986485304104,
+      "loss": 0.9971,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23620003460806369,
+      "grad_norm": 0.4820108413696289,
+      "learning_rate": 0.0001889120449387923,
+      "loss": 0.9482,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2370652362000346,
+      "grad_norm": 0.5275493264198303,
+      "learning_rate": 0.00018877341401434542,
+      "loss": 0.978,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23793043779200554,
+      "grad_norm": 0.40757259726524353,
+      "learning_rate": 0.00018863397334422074,
+      "loss": 1.0092,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23879563938397647,
+      "grad_norm": 0.5136415362358093,
+      "learning_rate": 0.00018849372420032482,
+      "loss": 0.9226,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2396608409759474,
+      "grad_norm": 0.41990041732788086,
+      "learning_rate": 0.0001883526678619385,
+      "loss": 1.0081,
+      "step": 1385
+    },
+    {
+      "epoch": 0.24052604256791832,
+      "grad_norm": 0.42340949177742004,
+      "learning_rate": 0.00018821080561570564,
+      "loss": 0.9498,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24139124415988925,
+      "grad_norm": 0.4153204560279846,
+      "learning_rate": 0.00018806813875562106,
+      "loss": 0.9773,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24225644575186017,
+      "grad_norm": 0.4485369622707367,
+      "learning_rate": 0.0001879246685830189,
+      "loss": 0.9816,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24312164734383113,
+      "grad_norm": 0.41648054122924805,
+      "learning_rate": 0.0001877803964065607,
+      "loss": 0.9895,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24398684893580205,
+      "grad_norm": 0.45885545015335083,
+      "learning_rate": 0.0001876353235422234,
+      "loss": 0.9631,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24485205052777298,
+      "grad_norm": 0.4401138722896576,
+      "learning_rate": 0.0001874894513132876,
+      "loss": 0.9606,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2457172521197439,
+      "grad_norm": 0.46568748354911804,
+      "learning_rate": 0.00018734278105032504,
+      "loss": 0.962,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24658245371171483,
+      "grad_norm": 0.4546853005886078,
+      "learning_rate": 0.0001871953140911869,
+      "loss": 0.9967,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24744765530368576,
+      "grad_norm": 0.5038758516311646,
+      "learning_rate": 0.0001870470517809913,
+      "loss": 0.9848,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2483128568956567,
+      "grad_norm": 0.44099846482276917,
+      "learning_rate": 0.0001868979954721113,
+      "loss": 1.0159,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2491780584876276,
+      "grad_norm": 0.4379419684410095,
+      "learning_rate": 0.0001867481465241622,
+      "loss": 0.9912,
+      "step": 1440
+    },
+    {
+      "epoch": 0.25004326007959854,
+      "grad_norm": 0.4447334408760071,
+      "learning_rate": 0.00018659750630398953,
+      "loss": 0.9705,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2509084616715695,
+      "grad_norm": 0.4537925720214844,
+      "learning_rate": 0.00018644607618565636,
+      "loss": 1.0205,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2517736632635404,
+      "grad_norm": 0.4481840133666992,
+      "learning_rate": 0.00018629385755043068,
+      "loss": 0.9782,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25263886485551135,
+      "grad_norm": 0.467529833316803,
+      "learning_rate": 0.0001861408517867731,
+      "loss": 0.9552,
+      "step": 1460
+    },
+    {
+      "epoch": 0.25350406644748225,
+      "grad_norm": 0.449776291847229,
+      "learning_rate": 0.0001859870602903239,
+      "loss": 0.9672,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2543692680394532,
+      "grad_norm": 0.6294044256210327,
+      "learning_rate": 0.0001858324844638905,
+      "loss": 0.927,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2552344696314241,
+      "grad_norm": 0.5393489599227905,
+      "learning_rate": 0.0001856771257174345,
+      "loss": 0.9396,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25609967122339505,
+      "grad_norm": 0.44971466064453125,
+      "learning_rate": 0.00018552098546805888,
+      "loss": 0.9652,
+      "step": 1480
+    },
+    {
+      "epoch": 0.256964872815366,
+      "grad_norm": 0.44472429156303406,
+      "learning_rate": 0.0001853640651399952,
+      "loss": 1.0073,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2578300744073369,
+      "grad_norm": 0.47125008702278137,
+      "learning_rate": 0.00018520636616459036,
+      "loss": 1.0136,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25869527599930786,
+      "grad_norm": 0.45552554726600647,
+      "learning_rate": 0.00018504788998029387,
+      "loss": 0.9856,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25956047759127876,
+      "grad_norm": 0.4563223421573639,
+      "learning_rate": 0.00018488863803264432,
+      "loss": 0.9882,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2604256791832497,
+      "grad_norm": 0.48147350549697876,
+      "learning_rate": 0.00018472861177425655,
+      "loss": 0.9911,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2612908807752206,
+      "grad_norm": 0.43628785014152527,
+      "learning_rate": 0.0001845678126648083,
+      "loss": 0.9816,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26215608236719157,
+      "grad_norm": 0.4450732171535492,
+      "learning_rate": 0.00018440624217102674,
+      "loss": 0.9672,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26302128395916247,
+      "grad_norm": 0.45051246881484985,
+      "learning_rate": 0.00018424390176667528,
+      "loss": 0.9984,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2638864855511334,
+      "grad_norm": 0.4232747554779053,
+      "learning_rate": 0.00018408079293254006,
+      "loss": 0.9948,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2647516871431043,
+      "grad_norm": 0.4707663059234619,
+      "learning_rate": 0.0001839169171564165,
+      "loss": 0.9916,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2656168887350753,
+      "grad_norm": 0.46704453229904175,
+      "learning_rate": 0.00018375227593309546,
+      "loss": 1.0485,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26648209032704623,
+      "grad_norm": 0.4697518050670624,
+      "learning_rate": 0.00018358687076435015,
+      "loss": 1.0106,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2673472919190171,
+      "grad_norm": 0.44449952244758606,
+      "learning_rate": 0.00018342070315892182,
+      "loss": 1.0133,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2682124935109881,
+      "grad_norm": 0.43841657042503357,
+      "learning_rate": 0.0001832537746325064,
+      "loss": 1.003,
+      "step": 1550
+    },
+    {
+      "epoch": 0.269077695102959,
+      "grad_norm": 0.45256274938583374,
+      "learning_rate": 0.00018308608670774054,
+      "loss": 0.991,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26994289669492993,
+      "grad_norm": 0.4906154274940491,
+      "learning_rate": 0.00018291764091418769,
+      "loss": 1.0044,
+      "step": 1560
+    },
+    {
+      "epoch": 0.27080809828690083,
+      "grad_norm": 0.42110729217529297,
+      "learning_rate": 0.00018274843878832427,
+      "loss": 0.9705,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2716732998788718,
+      "grad_norm": 0.4667767584323883,
+      "learning_rate": 0.00018257848187352543,
+      "loss": 1.0161,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2725385014708427,
+      "grad_norm": 0.46111616492271423,
+      "learning_rate": 0.00018240777172005138,
+      "loss": 0.9946,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27340370306281364,
+      "grad_norm": 0.420553594827652,
+      "learning_rate": 0.0001822363098850327,
+      "loss": 0.9672,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27426890465478454,
+      "grad_norm": 0.4221336543560028,
+      "learning_rate": 0.0001820640979324566,
+      "loss": 0.9421,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2751341062467555,
+      "grad_norm": 0.42993614077568054,
+      "learning_rate": 0.0001818911374331524,
+      "loss": 0.9645,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27599930783872645,
+      "grad_norm": 0.42793458700180054,
+      "learning_rate": 0.00018171742996477737,
+      "loss": 0.9735,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27686450943069735,
+      "grad_norm": 0.49229153990745544,
+      "learning_rate": 0.00018154297711180212,
+      "loss": 0.97,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2777297110226683,
+      "grad_norm": 0.4865579307079315,
+      "learning_rate": 0.00018136778046549642,
+      "loss": 0.9487,
+      "step": 1605
+    },
+    {
+      "epoch": 0.2785949126146392,
+      "grad_norm": 0.442254900932312,
+      "learning_rate": 0.0001811918416239144,
+      "loss": 0.9725,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27946011420661016,
+      "grad_norm": 0.45241814851760864,
+      "learning_rate": 0.0001810151621918802,
+      "loss": 0.9681,
+      "step": 1615
+    },
+    {
+      "epoch": 0.28032531579858105,
+      "grad_norm": 0.4351949095726013,
+      "learning_rate": 0.00018083774378097327,
+      "loss": 0.9728,
+      "step": 1620
+    },
+    {
+      "epoch": 0.281190517390552,
+      "grad_norm": 0.46627506613731384,
+      "learning_rate": 0.0001806595880095136,
+      "loss": 0.9751,
+      "step": 1625
+    },
+    {
+      "epoch": 0.2820557189825229,
+      "grad_norm": 0.4739546775817871,
+      "learning_rate": 0.000180480696502547,
+      "loss": 0.986,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28292092057449386,
+      "grad_norm": 0.4473097622394562,
+      "learning_rate": 0.00018030107089183028,
+      "loss": 0.9424,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28378612216646476,
+      "grad_norm": 0.4566546678543091,
+      "learning_rate": 0.00018012071281581644,
+      "loss": 0.9743,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2846513237584357,
+      "grad_norm": 0.49578657746315,
+      "learning_rate": 0.00017993962391963953,
+      "loss": 0.982,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28551652535040667,
+      "grad_norm": 0.43043839931488037,
+      "learning_rate": 0.00017975780585509994,
+      "loss": 0.9775,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28638172694237757,
+      "grad_norm": 0.4603051543235779,
+      "learning_rate": 0.000179575260280649,
+      "loss": 1.019,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2872469285343485,
+      "grad_norm": 0.4643576443195343,
+      "learning_rate": 0.00017939198886137401,
+      "loss": 1.0245,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2881121301263194,
+      "grad_norm": 0.4360422194004059,
+      "learning_rate": 0.00017920799326898328,
+      "loss": 0.9784,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2889773317182904,
+      "grad_norm": 0.44874873757362366,
+      "learning_rate": 0.00017902327518179033,
+      "loss": 0.9807,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2898425333102613,
+      "grad_norm": 0.4646460711956024,
+      "learning_rate": 0.0001788378362846992,
+      "loss": 1.009,
+      "step": 1675
+    },
+    {
+      "epoch": 0.29070773490223223,
+      "grad_norm": 0.44585877656936646,
+      "learning_rate": 0.00017865167826918856,
+      "loss": 0.9996,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2915729364942031,
+      "grad_norm": 0.43145236372947693,
+      "learning_rate": 0.0001784648028332967,
+      "loss": 0.9971,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2924381380861741,
+      "grad_norm": 0.4801592230796814,
+      "learning_rate": 0.0001782772116816057,
+      "loss": 0.9416,
+      "step": 1690
+    },
+    {
+      "epoch": 0.293303339678145,
+      "grad_norm": 0.4545969069004059,
+      "learning_rate": 0.00017808890652522613,
+      "loss": 0.9463,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29416854127011594,
+      "grad_norm": 0.4462297856807709,
+      "learning_rate": 0.0001778998890817813,
+      "loss": 0.9647,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2950337428620869,
+      "grad_norm": 0.480363667011261,
+      "learning_rate": 0.0001777101610753917,
+      "loss": 0.9527,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2958989444540578,
+      "grad_norm": 0.4686351418495178,
+      "learning_rate": 0.00017751972423665912,
+      "loss": 0.9553,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29676414604602874,
+      "grad_norm": 0.4903172552585602,
+      "learning_rate": 0.000177328580302651,
+      "loss": 0.9789,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29762934763799964,
+      "grad_norm": 0.4888453185558319,
+      "learning_rate": 0.00017713673101688452,
+      "loss": 1.002,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2984945492299706,
+      "grad_norm": 0.49155911803245544,
+      "learning_rate": 0.0001769441781293108,
+      "loss": 0.9818,
+      "step": 1725
+    },
+    {
+      "epoch": 0.2993597508219415,
+      "grad_norm": 0.4467802047729492,
+      "learning_rate": 0.00017675092339629876,
+      "loss": 1.0178,
+      "step": 1730
+    },
+    {
+      "epoch": 0.30022495241391245,
+      "grad_norm": 0.4385581612586975,
+      "learning_rate": 0.00017655696858061924,
+      "loss": 0.9938,
+      "step": 1735
+    },
+    {
+      "epoch": 0.30109015400588335,
+      "grad_norm": 0.4440990090370178,
+      "learning_rate": 0.00017636231545142884,
+      "loss": 1.0216,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3019553555978543,
+      "grad_norm": 0.4298366606235504,
+      "learning_rate": 0.0001761669657842539,
+      "loss": 1.0444,
+      "step": 1745
+    },
+    {
+      "epoch": 0.30282055718982526,
+      "grad_norm": 0.47890421748161316,
+      "learning_rate": 0.0001759709213609741,
+      "loss": 0.9882,
+      "step": 1750
+    },
+    {
+      "epoch": 0.30368575878179616,
+      "grad_norm": 0.4364605247974396,
+      "learning_rate": 0.00017577418396980647,
+      "loss": 1.0003,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3045509603737671,
+      "grad_norm": 0.42261895537376404,
+      "learning_rate": 0.0001755767554052888,
+      "loss": 0.9954,
+      "step": 1760
+    },
+    {
+      "epoch": 0.305416161965738,
+      "grad_norm": 0.448160320520401,
+      "learning_rate": 0.00017537863746826352,
+      "loss": 0.9995,
+      "step": 1765
+    },
+    {
+      "epoch": 0.30628136355770896,
+      "grad_norm": 0.46218201518058777,
+      "learning_rate": 0.0001751798319658611,
+      "loss": 0.9356,
+      "step": 1770
+    },
+    {
+      "epoch": 0.30714656514967986,
+      "grad_norm": 0.45802685618400574,
+      "learning_rate": 0.00017498034071148367,
+      "loss": 0.9978,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3080117667416508,
+      "grad_norm": 0.4749560058116913,
+      "learning_rate": 0.0001747801655247884,
+      "loss": 0.9887,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3088769683336217,
+      "grad_norm": 0.44203734397888184,
+      "learning_rate": 0.000174579308231671,
+      "loss": 0.9844,
+      "step": 1785
+    },
+    {
+      "epoch": 0.30974216992559267,
+      "grad_norm": 0.509231686592102,
+      "learning_rate": 0.0001743777706642489,
+      "loss": 1.0078,
+      "step": 1790
+    },
+    {
+      "epoch": 0.31060737151756357,
+      "grad_norm": 0.44051945209503174,
+      "learning_rate": 0.00017417555466084475,
+      "loss": 1.0002,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3114725731095345,
+      "grad_norm": 0.4557575285434723,
+      "learning_rate": 0.0001739726620659695,
+      "loss": 0.9577,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3123377747015055,
+      "grad_norm": 0.4637974798679352,
+      "learning_rate": 0.0001737690947303056,
+      "loss": 0.9786,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3132029762934764,
+      "grad_norm": 0.44868960976600647,
+      "learning_rate": 0.00017356485451069023,
+      "loss": 0.9733,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31406817788544733,
+      "grad_norm": 0.43835899233818054,
+      "learning_rate": 0.00017335994327009814,
+      "loss": 1.0112,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31493337947741823,
+      "grad_norm": 0.4374648928642273,
+      "learning_rate": 0.0001731543628776249,
+      "loss": 1.0003,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3157985810693892,
+      "grad_norm": 0.4368119239807129,
+      "learning_rate": 0.00017294811520846972,
+      "loss": 0.9886,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3166637826613601,
+      "grad_norm": 0.47979751229286194,
+      "learning_rate": 0.00017274120214391832,
+      "loss": 0.9984,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31752898425333104,
+      "grad_norm": 0.4430343210697174,
+      "learning_rate": 0.00017253362557132583,
+      "loss": 1.0221,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31839418584530194,
+      "grad_norm": 0.47513678669929504,
+      "learning_rate": 0.00017232538738409963,
+      "loss": 0.9883,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3192593874372729,
+      "grad_norm": 0.41813385486602783,
+      "learning_rate": 0.00017211648948168187,
+      "loss": 1.0066,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3201245890292438,
+      "grad_norm": 0.463078111410141,
+      "learning_rate": 0.0001719069337695325,
+      "loss": 0.9792,
+      "step": 1850
+    },
+    {
+      "epoch": 0.32098979062121474,
+      "grad_norm": 0.4492539167404175,
+      "learning_rate": 0.00017169672215911142,
+      "loss": 1.0032,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3218549922131857,
+      "grad_norm": 0.47429776191711426,
+      "learning_rate": 0.00017148585656786146,
+      "loss": 1.0013,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3227201938051566,
+      "grad_norm": 0.474584698677063,
+      "learning_rate": 0.00017127433891919074,
+      "loss": 0.9642,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32358539539712755,
+      "grad_norm": 0.4662615954875946,
+      "learning_rate": 0.000171062171142455,
+      "loss": 0.9688,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32445059698909845,
+      "grad_norm": 0.43062394857406616,
+      "learning_rate": 0.00017084935517294023,
+      "loss": 0.9717,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3253157985810694,
+      "grad_norm": 0.478098064661026,
+      "learning_rate": 0.00017063589295184483,
+      "loss": 0.9646,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3261810001730403,
+      "grad_norm": 0.4593997895717621,
+      "learning_rate": 0.000170421786426262,
+      "loss": 1.0208,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32704620176501126,
+      "grad_norm": 0.44811689853668213,
+      "learning_rate": 0.000170207037549162,
+      "loss": 0.982,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32791140335698216,
+      "grad_norm": 0.5170145034790039,
+      "learning_rate": 0.0001699916482793742,
+      "loss": 0.9832,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3287766049489531,
+      "grad_norm": 0.4444349706172943,
+      "learning_rate": 0.00016977562058156936,
+      "loss": 1.0137,
+      "step": 1900
+    },
+    {
+      "epoch": 0.329641806540924,
+      "grad_norm": 0.44498804211616516,
+      "learning_rate": 0.00016955895642624166,
+      "loss": 1.0198,
+      "step": 1905
+    },
+    {
+      "epoch": 0.33050700813289496,
+      "grad_norm": 0.44884684681892395,
+      "learning_rate": 0.00016934165778969074,
+      "loss": 0.9827,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3313722097248659,
+      "grad_norm": 0.48218467831611633,
+      "learning_rate": 0.00016912372665400354,
+      "loss": 0.9728,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3322374113168368,
+      "grad_norm": 0.45547670125961304,
+      "learning_rate": 0.00016890516500703645,
+      "loss": 0.9389,
+      "step": 1920
+    },
+    {
+      "epoch": 0.33310261290880777,
+      "grad_norm": 0.43529343605041504,
+      "learning_rate": 0.0001686859748423971,
+      "loss": 0.9806,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33396781450077867,
+      "grad_norm": 0.4809509813785553,
+      "learning_rate": 0.00016846615815942597,
+      "loss": 0.9655,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3348330160927496,
+      "grad_norm": 0.4193096160888672,
+      "learning_rate": 0.00016824571696317844,
+      "loss": 1.0132,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3356982176847205,
+      "grad_norm": 0.4560771882534027,
+      "learning_rate": 0.00016802465326440634,
+      "loss": 1.0416,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3365634192766915,
+      "grad_norm": 0.44306620955467224,
+      "learning_rate": 0.00016780296907953972,
+      "loss": 0.9739,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3374286208686624,
+      "grad_norm": 0.44249188899993896,
+      "learning_rate": 0.00016758066643066826,
+      "loss": 0.9777,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33829382246063333,
+      "grad_norm": 0.44801339507102966,
+      "learning_rate": 0.00016735774734552306,
+      "loss": 0.9981,
+      "step": 1955
+    },
+    {
+      "epoch": 0.33915902405260423,
+      "grad_norm": 0.4256415069103241,
+      "learning_rate": 0.000167134213857458,
+      "loss": 0.9594,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3400242256445752,
+      "grad_norm": 0.44402891397476196,
+      "learning_rate": 0.0001669100680054312,
+      "loss": 0.9862,
+      "step": 1965
+    },
+    {
+      "epoch": 0.34088942723654614,
+      "grad_norm": 0.5019104480743408,
+      "learning_rate": 0.0001666853118339865,
+      "loss": 0.975,
+      "step": 1970
+    },
+    {
+      "epoch": 0.34175462882851704,
+      "grad_norm": 0.47364917397499084,
+      "learning_rate": 0.00016645994739323473,
+      "loss": 0.969,
+      "step": 1975
+    },
+    {
+      "epoch": 0.342619830420488,
+      "grad_norm": 0.4283023774623871,
+      "learning_rate": 0.00016623397673883508,
+      "loss": 0.9786,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3434850320124589,
+      "grad_norm": 0.47522231936454773,
+      "learning_rate": 0.00016600740193197623,
+      "loss": 0.9779,
+      "step": 1985
+    },
+    {
+      "epoch": 0.34435023360442985,
+      "grad_norm": 0.46930992603302,
+      "learning_rate": 0.00016578022503935773,
+      "loss": 1.0124,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34521543519640074,
+      "grad_norm": 0.5017907619476318,
+      "learning_rate": 0.000165552448133171,
+      "loss": 0.9931,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3460806367883717,
+      "grad_norm": 0.5087099075317383,
+      "learning_rate": 0.0001653240732910804,
+      "loss": 0.94,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3469458383803426,
+      "grad_norm": 0.44298025965690613,
+      "learning_rate": 0.00016509510259620457,
+      "loss": 0.9793,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34781103997231355,
+      "grad_norm": 0.43045955896377563,
+      "learning_rate": 0.00016486553813709696,
+      "loss": 0.9673,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3486762415642845,
+      "grad_norm": 0.5183268189430237,
+      "learning_rate": 0.00016463538200772718,
+      "loss": 1.0128,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3495414431562554,
+      "grad_norm": 0.47518181800842285,
+      "learning_rate": 0.00016440463630746174,
+      "loss": 0.9851,
+      "step": 2020
+    },
+    {
+      "epoch": 0.35040664474822636,
+      "grad_norm": 0.4647828936576843,
+      "learning_rate": 0.00016417330314104491,
+      "loss": 0.9923,
+      "step": 2025
+    },
+    {
+      "epoch": 0.35127184634019726,
+      "grad_norm": 0.484781414270401,
+      "learning_rate": 0.0001639413846185795,
+      "loss": 0.9818,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3521370479321682,
+      "grad_norm": 0.4849836230278015,
+      "learning_rate": 0.00016370888285550763,
+      "loss": 0.9607,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3530022495241391,
+      "grad_norm": 0.48198217153549194,
+      "learning_rate": 0.0001634757999725915,
+      "loss": 0.9953,
+      "step": 2040
+    },
+    {
+      "epoch": 0.35386745111611007,
+      "grad_norm": 0.4720988869667053,
+      "learning_rate": 0.00016324213809589393,
+      "loss": 0.9718,
+      "step": 2045
+    },
+    {
+      "epoch": 0.35473265270808096,
+      "grad_norm": 0.48839956521987915,
+      "learning_rate": 0.00016300789935675908,
+      "loss": 0.9521,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3555978543000519,
+      "grad_norm": 0.47152647376060486,
+      "learning_rate": 0.00016277308589179287,
+      "loss": 0.9766,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3564630558920228,
+      "grad_norm": 0.4669671654701233,
+      "learning_rate": 0.00016253769984284365,
+      "loss": 1.0222,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35732825748399377,
+      "grad_norm": 0.4763648509979248,
+      "learning_rate": 0.00016230174335698258,
+      "loss": 1.0098,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3581934590759647,
+      "grad_norm": 0.45041629672050476,
+      "learning_rate": 0.00016206521858648406,
+      "loss": 0.9917,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3590586606679356,
+      "grad_norm": 0.4782506823539734,
+      "learning_rate": 0.00016182812768880602,
+      "loss": 0.935,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3599238622599066,
+      "grad_norm": 0.4275178015232086,
+      "learning_rate": 0.00016159047282657043,
+      "loss": 0.9905,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3607890638518775,
+      "grad_norm": 0.4617001414299011,
+      "learning_rate": 0.0001613522561675433,
+      "loss": 0.9867,
+      "step": 2085
+    },
+    {
+      "epoch": 0.36165426544384843,
+      "grad_norm": 0.5393168330192566,
+      "learning_rate": 0.00016111347988461523,
+      "loss": 0.9924,
+      "step": 2090
+    },
+    {
+      "epoch": 0.36251946703581933,
+      "grad_norm": 0.4483303427696228,
+      "learning_rate": 0.0001608741461557813,
+      "loss": 0.9911,
+      "step": 2095
+    },
+    {
+      "epoch": 0.3633846686277903,
+      "grad_norm": 0.4839470088481903,
+      "learning_rate": 0.0001606342571641214,
+      "loss": 0.9432,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3642498702197612,
+      "grad_norm": 0.4631226062774658,
+      "learning_rate": 0.00016039381509778017,
+      "loss": 0.9823,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36511507181173214,
+      "grad_norm": 0.4838786721229553,
+      "learning_rate": 0.00016015282214994724,
+      "loss": 0.9399,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36598027340370304,
+      "grad_norm": 0.458999365568161,
+      "learning_rate": 0.00015991128051883697,
+      "loss": 1.0042,
+      "step": 2115
+    },
+    {
+      "epoch": 0.366845474995674,
+      "grad_norm": 0.45957833528518677,
+      "learning_rate": 0.00015966919240766858,
+      "loss": 0.982,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36771067658764495,
+      "grad_norm": 0.4562719464302063,
+      "learning_rate": 0.000159426560024646,
+      "loss": 0.9766,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36857587817961585,
+      "grad_norm": 0.4500807821750641,
+      "learning_rate": 0.00015918338558293773,
+      "loss": 0.9552,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3694410797715868,
+      "grad_norm": 0.42124584317207336,
+      "learning_rate": 0.00015893967130065667,
+      "loss": 0.9275,
+      "step": 2135
+    },
+    {
+      "epoch": 0.3703062813635577,
+      "grad_norm": 0.4581054747104645,
+      "learning_rate": 0.00015869541940083978,
+      "loss": 1.0353,
+      "step": 2140
+    },
+    {
+      "epoch": 0.37117148295552865,
+      "grad_norm": 0.47650617361068726,
+      "learning_rate": 0.0001584506321114281,
+      "loss": 1.0077,
+      "step": 2145
+    },
+    {
+      "epoch": 0.37203668454749955,
+      "grad_norm": 0.45496097207069397,
+      "learning_rate": 0.00015820531166524593,
+      "loss": 1.0202,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3729018861394705,
+      "grad_norm": 0.46953004598617554,
+      "learning_rate": 0.000157959460299981,
+      "loss": 0.9695,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3737670877314414,
+      "grad_norm": 0.4414077401161194,
+      "learning_rate": 0.00015771308025816372,
+      "loss": 0.967,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37463228932341236,
+      "grad_norm": 0.4325014650821686,
+      "learning_rate": 0.00015746617378714674,
+      "loss": 0.9621,
+      "step": 2165
+    },
+    {
+      "epoch": 0.37549749091538326,
+      "grad_norm": 0.45029759407043457,
+      "learning_rate": 0.00015721874313908468,
+      "loss": 0.9789,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3763626925073542,
+      "grad_norm": 0.4882911443710327,
+      "learning_rate": 0.00015697079057091332,
+      "loss": 0.9836,
+      "step": 2175
+    },
+    {
+      "epoch": 0.37722789409932517,
+      "grad_norm": 0.4586533308029175,
+      "learning_rate": 0.0001567223183443292,
+      "loss": 0.9221,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37809309569129607,
+      "grad_norm": 0.4566982686519623,
+      "learning_rate": 0.00015647332872576887,
+      "loss": 0.9981,
+      "step": 2185
+    },
+    {
+      "epoch": 0.378958297283267,
+      "grad_norm": 0.4582918584346771,
+      "learning_rate": 0.00015622382398638825,
+      "loss": 0.959,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3798234988752379,
+      "grad_norm": 0.455666720867157,
+      "learning_rate": 0.000155973806402042,
+      "loss": 0.9668,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3806887004672089,
+      "grad_norm": 0.4562900960445404,
+      "learning_rate": 0.00015572327825326263,
+      "loss": 0.9914,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3815539020591798,
+      "grad_norm": 0.4523032307624817,
+      "learning_rate": 0.00015547224182523977,
+      "loss": 0.9311,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3824191036511507,
+      "grad_norm": 0.4746415913105011,
+      "learning_rate": 0.0001552206994077993,
+      "loss": 1.0036,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3832843052431216,
+      "grad_norm": 0.46040624380111694,
+      "learning_rate": 0.00015496865329538254,
+      "loss": 1.0197,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3841495068350926,
+      "grad_norm": 0.4492608308792114,
+      "learning_rate": 0.00015471610578702522,
+      "loss": 0.9614,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3850147084270635,
+      "grad_norm": 0.4558756649494171,
+      "learning_rate": 0.00015446305918633647,
+      "loss": 1.0048,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38587991001903443,
+      "grad_norm": 0.46587228775024414,
+      "learning_rate": 0.00015420951580147807,
+      "loss": 0.9887,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3867451116110054,
+      "grad_norm": 0.46439608931541443,
+      "learning_rate": 0.00015395547794514296,
+      "loss": 0.9568,
+      "step": 2235
+    },
+    {
+      "epoch": 0.3876103132029763,
+      "grad_norm": 0.5386447310447693,
+      "learning_rate": 0.00015370094793453466,
+      "loss": 1.0201,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38847551479494724,
+      "grad_norm": 0.46466854214668274,
+      "learning_rate": 0.0001534459280913456,
+      "loss": 0.9553,
+      "step": 2245
+    },
+    {
+      "epoch": 0.38934071638691814,
+      "grad_norm": 0.4588780999183655,
+      "learning_rate": 0.0001531904207417365,
+      "loss": 1.002,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3902059179788891,
+      "grad_norm": 0.4447367191314697,
+      "learning_rate": 0.00015293442821631467,
+      "loss": 0.9832,
+      "step": 2255
+    },
+    {
+      "epoch": 0.39107111957086,
+      "grad_norm": 0.4650314450263977,
+      "learning_rate": 0.00015267795285011296,
+      "loss": 0.9939,
+      "step": 2260
+    },
+    {
+      "epoch": 0.39193632116283095,
+      "grad_norm": 0.44617509841918945,
+      "learning_rate": 0.0001524209969825685,
+      "loss": 0.9858,
+      "step": 2265
+    },
+    {
+      "epoch": 0.39280152275480185,
+      "grad_norm": 0.469292551279068,
+      "learning_rate": 0.00015216356295750128,
+      "loss": 0.9522,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3936667243467728,
+      "grad_norm": 0.5017248392105103,
+      "learning_rate": 0.00015190565312309286,
+      "loss": 0.9806,
+      "step": 2275
+    },
+    {
+      "epoch": 0.39453192593874375,
+      "grad_norm": 0.4717511236667633,
+      "learning_rate": 0.00015164726983186476,
+      "loss": 0.9885,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39539712753071465,
+      "grad_norm": 0.4415047764778137,
+      "learning_rate": 0.00015138841544065724,
+      "loss": 0.977,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3962623291226856,
+      "grad_norm": 0.46711215376853943,
+      "learning_rate": 0.00015112909231060768,
+      "loss": 1.0132,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3971275307146565,
+      "grad_norm": 0.44584599137306213,
+      "learning_rate": 0.00015086930280712904,
+      "loss": 1.0131,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39799273230662746,
+      "grad_norm": 0.46857744455337524,
+      "learning_rate": 0.00015060904929988824,
+      "loss": 0.9522,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39885793389859836,
+      "grad_norm": 0.5087977647781372,
+      "learning_rate": 0.00015034833416278473,
+      "loss": 0.9703,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3997231354905693,
+      "grad_norm": 0.45563820004463196,
+      "learning_rate": 0.0001500871597739286,
+      "loss": 0.9969,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4005883370825402,
+      "grad_norm": 0.4566476345062256,
+      "learning_rate": 0.00014982552851561904,
+      "loss": 1.0003,
+      "step": 2315
+    },
+    {
+      "epoch": 0.40145353867451117,
+      "grad_norm": 0.5005662441253662,
+      "learning_rate": 0.00014956344277432257,
+      "loss": 0.9812,
+      "step": 2320
+    },
+    {
+      "epoch": 0.40231874026648207,
+      "grad_norm": 0.48637065291404724,
+      "learning_rate": 0.00014930090494065133,
+      "loss": 1.0326,
+      "step": 2325
+    },
+    {
+      "epoch": 0.403183941858453,
+      "grad_norm": 0.47612977027893066,
+      "learning_rate": 0.0001490379174093411,
+      "loss": 0.9604,
+      "step": 2330
+    },
+    {
+      "epoch": 0.404049143450424,
+      "grad_norm": 0.4837099015712738,
+      "learning_rate": 0.00014877448257922965,
+      "loss": 0.9985,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4049143450423949,
+      "grad_norm": 0.5050654411315918,
+      "learning_rate": 0.00014851060285323478,
+      "loss": 0.9943,
+      "step": 2340
+    },
+    {
+      "epoch": 0.40577954663436583,
+      "grad_norm": 0.4585602581501007,
+      "learning_rate": 0.00014824628063833233,
+      "loss": 0.9702,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4066447482263367,
+      "grad_norm": 0.4651777446269989,
+      "learning_rate": 0.0001479815183455344,
+      "loss": 0.9967,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4075099498183077,
+      "grad_norm": 0.46856093406677246,
+      "learning_rate": 0.00014771631838986717,
+      "loss": 0.9894,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4083751514102786,
+      "grad_norm": 0.4769989848136902,
+      "learning_rate": 0.00014745068319034905,
+      "loss": 0.9982,
+      "step": 2360
+    },
+    {
+      "epoch": 0.40924035300224954,
+      "grad_norm": 0.4633113145828247,
+      "learning_rate": 0.00014718461516996842,
+      "loss": 1.0114,
+      "step": 2365
+    },
+    {
+      "epoch": 0.41010555459422043,
+      "grad_norm": 0.46894222497940063,
+      "learning_rate": 0.0001469181167556617,
+      "loss": 1.001,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4109707561861914,
+      "grad_norm": 0.4495028853416443,
+      "learning_rate": 0.0001466511903782911,
+      "loss": 0.9475,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4118359577781623,
+      "grad_norm": 0.4745693504810333,
+      "learning_rate": 0.0001463838384726225,
+      "loss": 0.9911,
+      "step": 2380
+    },
+    {
+      "epoch": 0.41270115937013324,
+      "grad_norm": 0.4932793974876404,
+      "learning_rate": 0.00014611606347730326,
+      "loss": 0.9857,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4135663609621042,
+      "grad_norm": 0.45724374055862427,
+      "learning_rate": 0.00014584786783483996,
+      "loss": 0.9506,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4144315625540751,
+      "grad_norm": 0.48630237579345703,
+      "learning_rate": 0.000145579253991576,
+      "loss": 1.0016,
+      "step": 2395
+    },
+    {
+      "epoch": 0.41529676414604605,
+      "grad_norm": 0.48039719462394714,
+      "learning_rate": 0.00014531022439766956,
+      "loss": 1.0147,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41616196573801695,
+      "grad_norm": 0.4719230830669403,
+      "learning_rate": 0.00014504078150707092,
+      "loss": 0.9935,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4170271673299879,
+      "grad_norm": 0.4671541154384613,
+      "learning_rate": 0.00014477092777750037,
+      "loss": 0.9885,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4178923689219588,
+      "grad_norm": 0.4346173107624054,
+      "learning_rate": 0.00014450066567042556,
+      "loss": 0.9948,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41875757051392976,
+      "grad_norm": 0.44263723492622375,
+      "learning_rate": 0.00014422999765103923,
+      "loss": 0.9637,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41962277210590065,
+      "grad_norm": 0.45122030377388,
+      "learning_rate": 0.00014395892618823664,
+      "loss": 0.969,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4204879736978716,
+      "grad_norm": 0.526986837387085,
+      "learning_rate": 0.00014368745375459296,
+      "loss": 0.9778,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4213531752898425,
+      "grad_norm": 0.502812385559082,
+      "learning_rate": 0.00014341558282634096,
+      "loss": 0.9846,
+      "step": 2435
+    },
+    {
+      "epoch": 0.42221837688181346,
+      "grad_norm": 0.4661869406700134,
+      "learning_rate": 0.00014314331588334813,
+      "loss": 0.9688,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4230835784737844,
+      "grad_norm": 0.4835832715034485,
+      "learning_rate": 0.00014287065540909429,
+      "loss": 0.976,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4239487800657553,
+      "grad_norm": 0.46521812677383423,
+      "learning_rate": 0.00014259760389064884,
+      "loss": 1.0407,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42481398165772627,
+      "grad_norm": 0.4323074519634247,
+      "learning_rate": 0.00014232416381864803,
+      "loss": 0.9383,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42567918324969717,
+      "grad_norm": 0.4658597409725189,
+      "learning_rate": 0.00014205033768727242,
+      "loss": 1.011,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4265443848416681,
+      "grad_norm": 0.46207860112190247,
+      "learning_rate": 0.00014177612799422384,
+      "loss": 0.9617,
+      "step": 2465
+    },
+    {
+      "epoch": 0.427409586433639,
+      "grad_norm": 0.48865434527397156,
+      "learning_rate": 0.00014150153724070292,
+      "loss": 0.9585,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42827478802561,
+      "grad_norm": 0.44298145174980164,
+      "learning_rate": 0.00014122656793138605,
+      "loss": 0.9844,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4291399896175809,
+      "grad_norm": 0.4769092798233032,
+      "learning_rate": 0.00014095122257440265,
+      "loss": 1.0026,
+      "step": 2480
+    },
+    {
+      "epoch": 0.43000519120955183,
+      "grad_norm": 0.4551126956939697,
+      "learning_rate": 0.0001406755036813122,
+      "loss": 0.9775,
+      "step": 2485
+    },
+    {
+      "epoch": 0.43087039280152273,
+      "grad_norm": 0.45167794823646545,
+      "learning_rate": 0.00014039941376708142,
+      "loss": 0.9684,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4317355943934937,
+      "grad_norm": 0.5096551179885864,
+      "learning_rate": 0.00014012295535006136,
+      "loss": 1.0283,
+      "step": 2495
+    },
+    {
+      "epoch": 0.43260079598546464,
+      "grad_norm": 0.45402616262435913,
+      "learning_rate": 0.0001398461309519642,
+      "loss": 0.9989,
+      "step": 2500
+    },
+    {
+      "epoch": 0.43346599757743554,
+      "grad_norm": 0.4679848849773407,
+      "learning_rate": 0.0001395689430978406,
+      "loss": 1.0361,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4343311991694065,
+      "grad_norm": 0.4805293381214142,
+      "learning_rate": 0.00013929139431605633,
+      "loss": 0.979,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4351964007613774,
+      "grad_norm": 0.46778836846351624,
+      "learning_rate": 0.00013901348713826948,
+      "loss": 1.0005,
+      "step": 2515
+    },
+    {
+      "epoch": 0.43606160235334834,
+      "grad_norm": 0.4670213758945465,
+      "learning_rate": 0.00013873522409940713,
+      "loss": 0.9671,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43692680394531924,
+      "grad_norm": 0.44049155712127686,
+      "learning_rate": 0.00013845660773764243,
+      "loss": 0.9851,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4377920055372902,
+      "grad_norm": 0.4682471752166748,
+      "learning_rate": 0.00013817764059437132,
+      "loss": 0.9386,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4386572071292611,
+      "grad_norm": 0.4299015700817108,
+      "learning_rate": 0.00013789832521418941,
+      "loss": 0.9519,
+      "step": 2535
+    },
+    {
+      "epoch": 0.43952240872123205,
+      "grad_norm": 0.4857511520385742,
+      "learning_rate": 0.00013761866414486874,
+      "loss": 0.9507,
+      "step": 2540
+    },
+    {
+      "epoch": 0.440387610313203,
+      "grad_norm": 0.4522685408592224,
+      "learning_rate": 0.00013733865993733447,
+      "loss": 0.9712,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4412528119051739,
+      "grad_norm": 0.4425513446331024,
+      "learning_rate": 0.00013705831514564186,
+      "loss": 0.9616,
+      "step": 2550
+    },
+    {
+      "epoch": 0.44211801349714486,
+      "grad_norm": 0.4239989221096039,
+      "learning_rate": 0.0001367776323269527,
+      "loss": 0.9598,
+      "step": 2555
+    },
+    {
+      "epoch": 0.44298321508911576,
+      "grad_norm": 0.4723816215991974,
+      "learning_rate": 0.00013649661404151203,
+      "loss": 0.9917,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4438484166810867,
+      "grad_norm": 0.44909706711769104,
+      "learning_rate": 0.00013621526285262502,
+      "loss": 1.0044,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4447136182730576,
+      "grad_norm": 0.48389700055122375,
+      "learning_rate": 0.00013593358132663324,
+      "loss": 1.0128,
+      "step": 2570
+    },
+    {
+      "epoch": 0.44557881986502856,
+      "grad_norm": 0.45943713188171387,
+      "learning_rate": 0.0001356515720328915,
+      "loss": 0.9519,
+      "step": 2575
+    },
+    {
+      "epoch": 0.44644402145699946,
+      "grad_norm": 0.5405260324478149,
+      "learning_rate": 0.00013536923754374437,
+      "loss": 0.9431,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4473092230489704,
+      "grad_norm": 0.4995253086090088,
+      "learning_rate": 0.00013508658043450265,
+      "loss": 0.9976,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4481744246409413,
+      "grad_norm": 0.4669324457645416,
+      "learning_rate": 0.00013480360328341993,
+      "loss": 1.0101,
+      "step": 2590
+    },
+    {
+      "epoch": 0.44903962623291227,
+      "grad_norm": 0.44764748215675354,
+      "learning_rate": 0.00013452030867166901,
+      "loss": 0.9538,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4499048278248832,
+      "grad_norm": 0.46463918685913086,
+      "learning_rate": 0.00013423669918331848,
+      "loss": 0.9601,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4507700294168541,
+      "grad_norm": 0.45471301674842834,
+      "learning_rate": 0.000133952777405309,
+      "loss": 0.9727,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4516352310088251,
+      "grad_norm": 0.4745343029499054,
+      "learning_rate": 0.00013366854592742984,
+      "loss": 0.9931,
+      "step": 2610
+    },
+    {
+      "epoch": 0.452500432600796,
+      "grad_norm": 0.43110033869743347,
+      "learning_rate": 0.00013338400734229516,
+      "loss": 0.9657,
+      "step": 2615
+    },
+    {
+      "epoch": 0.45336563419276693,
+      "grad_norm": 0.48442110419273376,
+      "learning_rate": 0.00013309916424532035,
+      "loss": 0.9869,
+      "step": 2620
+    },
+    {
+      "epoch": 0.45423083578473783,
+      "grad_norm": 0.4661529064178467,
+      "learning_rate": 0.00013281401923469856,
+      "loss": 0.9899,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4550960373767088,
+      "grad_norm": 0.4315500557422638,
+      "learning_rate": 0.00013252857491137665,
+      "loss": 0.9737,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4559612389686797,
+      "grad_norm": 0.43988022208213806,
+      "learning_rate": 0.00013224283387903177,
+      "loss": 0.9847,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45682644056065064,
+      "grad_norm": 0.49706265330314636,
+      "learning_rate": 0.00013195679874404746,
+      "loss": 0.9938,
+      "step": 2640
+    },
+    {
+      "epoch": 0.45769164215262154,
+      "grad_norm": 0.48079583048820496,
+      "learning_rate": 0.0001316704721154899,
+      "loss": 0.9714,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4585568437445925,
+      "grad_norm": 0.4445064067840576,
+      "learning_rate": 0.0001313838566050842,
+      "loss": 0.9566,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45942204533656344,
+      "grad_norm": 0.46092909574508667,
+      "learning_rate": 0.0001310969548271904,
+      "loss": 0.9506,
+      "step": 2655
+    },
+    {
+      "epoch": 0.46028724692853434,
+      "grad_norm": 0.4312794804573059,
+      "learning_rate": 0.00013080976939877984,
+      "loss": 0.957,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4611524485205053,
+      "grad_norm": 0.43655863404273987,
+      "learning_rate": 0.00013052230293941107,
+      "loss": 1.0075,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4620176501124762,
+      "grad_norm": 0.483980655670166,
+      "learning_rate": 0.0001302345580712061,
+      "loss": 0.9979,
+      "step": 2670
+    },
+    {
+      "epoch": 0.46288285170444715,
+      "grad_norm": 0.4380267858505249,
+      "learning_rate": 0.00012994653741882646,
+      "loss": 0.9632,
+      "step": 2675
+    },
+    {
+      "epoch": 0.46374805329641805,
+      "grad_norm": 0.4459783434867859,
+      "learning_rate": 0.00012965824360944915,
+      "loss": 0.9637,
+      "step": 2680
+    },
+    {
+      "epoch": 0.464613254888389,
+      "grad_norm": 0.46052780747413635,
+      "learning_rate": 0.00012936967927274298,
+      "loss": 0.9944,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4654784564803599,
+      "grad_norm": 0.43334734439849854,
+      "learning_rate": 0.00012908084704084415,
+      "loss": 1.0045,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46634365807233086,
+      "grad_norm": 0.46884188055992126,
+      "learning_rate": 0.00012879174954833256,
+      "loss": 0.9334,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46720885966430176,
+      "grad_norm": 0.46058720350265503,
+      "learning_rate": 0.00012850238943220774,
+      "loss": 1.0021,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4680740612562727,
+      "grad_norm": 0.46159979701042175,
+      "learning_rate": 0.00012821276933186457,
+      "loss": 0.9597,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46893926284824367,
+      "grad_norm": 0.4704788029193878,
+      "learning_rate": 0.00012792289188906953,
+      "loss": 1.0088,
+      "step": 2710
+    },
+    {
+      "epoch": 0.46980446444021456,
+      "grad_norm": 0.4625367820262909,
+      "learning_rate": 0.00012763275974793636,
+      "loss": 0.966,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4706696660321855,
+      "grad_norm": 0.5280613899230957,
+      "learning_rate": 0.0001273423755549021,
+      "loss": 0.9922,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4715348676241564,
+      "grad_norm": 0.4669839143753052,
+      "learning_rate": 0.00012705174195870282,
+      "loss": 1.0138,
+      "step": 2725
+    },
+    {
+      "epoch": 0.47240006921612737,
+      "grad_norm": 0.4880688488483429,
+      "learning_rate": 0.00012676086161034951,
+      "loss": 0.9749,
+      "step": 2730
+    },
+    {
+      "epoch": 0.47326527080809827,
+      "grad_norm": 0.46175360679626465,
+      "learning_rate": 0.00012646973716310393,
+      "loss": 0.9597,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4741304724000692,
+      "grad_norm": 0.4644413888454437,
+      "learning_rate": 0.00012617837127245441,
+      "loss": 0.9941,
+      "step": 2740
+    },
+    {
+      "epoch": 0.4749956739920401,
+      "grad_norm": 0.41886743903160095,
+      "learning_rate": 0.00012588676659609159,
+      "loss": 0.9234,
+      "step": 2745
+    },
+    {
+      "epoch": 0.4758608755840111,
+      "grad_norm": 0.4897218346595764,
+      "learning_rate": 0.00012559492579388416,
+      "loss": 1.0088,
+      "step": 2750
+    },
+    {
+      "epoch": 0.476726077175982,
+      "grad_norm": 0.47003859281539917,
+      "learning_rate": 0.00012530285152785467,
+      "loss": 1.0095,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47759127876795293,
+      "grad_norm": 0.42755237221717834,
+      "learning_rate": 0.00012501054646215515,
+      "loss": 0.9769,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4784564803599239,
+      "grad_norm": 0.5013285279273987,
+      "learning_rate": 0.00012471801326304293,
+      "loss": 0.9739,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4793216819518948,
+      "grad_norm": 0.4810338318347931,
+      "learning_rate": 0.00012442525459885624,
+      "loss": 1.0089,
+      "step": 2770
+    },
+    {
+      "epoch": 0.48018688354386574,
+      "grad_norm": 0.42311739921569824,
+      "learning_rate": 0.00012413227313998984,
+      "loss": 0.9531,
+      "step": 2775
+    },
+    {
+      "epoch": 0.48105208513583664,
+      "grad_norm": 0.46264195442199707,
+      "learning_rate": 0.00012383907155887083,
+      "loss": 0.9483,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4819172867278076,
+      "grad_norm": 0.4477481245994568,
+      "learning_rate": 0.00012354565252993397,
+      "loss": 0.9452,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4827824883197785,
+      "grad_norm": 0.5055541396141052,
+      "learning_rate": 0.0001232520187295976,
+      "loss": 0.9698,
+      "step": 2790
+    },
+    {
+      "epoch": 0.48364768991174945,
+      "grad_norm": 0.4845668375492096,
+      "learning_rate": 0.00012295817283623908,
+      "loss": 0.9613,
+      "step": 2795
+    },
+    {
+      "epoch": 0.48451289150372034,
+      "grad_norm": 0.47771093249320984,
+      "learning_rate": 0.00012266411753017036,
+      "loss": 0.9475,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4853780930956913,
+      "grad_norm": 0.43290895223617554,
+      "learning_rate": 0.0001223698554936135,
+      "loss": 0.9655,
+      "step": 2805
+    },
+    {
+      "epoch": 0.48624329468766225,
+      "grad_norm": 0.4561658799648285,
+      "learning_rate": 0.00012207538941067637,
+      "loss": 0.9628,
+      "step": 2810
+    },
+    {
+      "epoch": 0.48710849627963315,
+      "grad_norm": 0.442929208278656,
+      "learning_rate": 0.00012178072196732788,
+      "loss": 1.0197,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4879736978716041,
+      "grad_norm": 0.4500119686126709,
+      "learning_rate": 0.00012148585585137377,
+      "loss": 1.0237,
+      "step": 2820
+    },
+    {
+      "epoch": 0.488838899463575,
+      "grad_norm": 0.44526350498199463,
+      "learning_rate": 0.00012119079375243194,
+      "loss": 0.9766,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48970410105554596,
+      "grad_norm": 0.48116928339004517,
+      "learning_rate": 0.00012089553836190795,
+      "loss": 0.994,
+      "step": 2830
+    },
+    {
+      "epoch": 0.49056930264751686,
+      "grad_norm": 0.5110150575637817,
+      "learning_rate": 0.00012060009237297046,
+      "loss": 0.9642,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4914345042394878,
+      "grad_norm": 0.4595610201358795,
+      "learning_rate": 0.00012030445848052665,
+      "loss": 0.9828,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4922997058314587,
+      "grad_norm": 0.4527970552444458,
+      "learning_rate": 0.00012000863938119768,
+      "loss": 0.9682,
+      "step": 2845
+    },
+    {
+      "epoch": 0.49316490742342967,
+      "grad_norm": 0.4377381503582001,
+      "learning_rate": 0.00011971263777329406,
+      "loss": 0.9387,
+      "step": 2850
+    },
+    {
+      "epoch": 0.49403010901540056,
+      "grad_norm": 0.44428756833076477,
+      "learning_rate": 0.00011941645635679111,
+      "loss": 0.8939,
+      "step": 2855
+    },
+    {
+      "epoch": 0.4948953106073715,
+      "grad_norm": 0.5483399629592896,
+      "learning_rate": 0.00011912009783330417,
+      "loss": 0.9681,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4957605121993425,
+      "grad_norm": 0.5041677951812744,
+      "learning_rate": 0.00011882356490606414,
+      "loss": 0.9968,
+      "step": 2865
+    },
+    {
+      "epoch": 0.4966257137913134,
+      "grad_norm": 0.4788263440132141,
+      "learning_rate": 0.00011852686027989273,
+      "loss": 0.9272,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4974909153832843,
+      "grad_norm": 0.44364604353904724,
+      "learning_rate": 0.00011822998666117775,
+      "loss": 0.9746,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4983561169752552,
+      "grad_norm": 0.44387176632881165,
+      "learning_rate": 0.00011793294675784853,
+      "loss": 0.9396,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4992213185672262,
+      "grad_norm": 0.4744800627231598,
+      "learning_rate": 0.00011763574327935113,
+      "loss": 0.9756,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5000865201591971,
+      "grad_norm": 0.420992910861969,
+      "learning_rate": 0.00011733837893662369,
+      "loss": 0.9588,
+      "step": 2890
+    },
+    {
+      "epoch": 0.500951721751168,
+      "grad_norm": 0.47278472781181335,
+      "learning_rate": 0.00011704085644207163,
+      "loss": 0.9438,
+      "step": 2895
+    },
+    {
+      "epoch": 0.501816923343139,
+      "grad_norm": 0.4406570494174957,
+      "learning_rate": 0.00011674317850954297,
+      "loss": 0.9592,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5026821249351099,
+      "grad_norm": 0.4493499994277954,
+      "learning_rate": 0.00011644534785430359,
+      "loss": 0.9935,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5035473265270808,
+      "grad_norm": 0.5532475709915161,
+      "learning_rate": 0.00011614736719301236,
+      "loss": 0.9957,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5044125281190517,
+      "grad_norm": 0.44797247648239136,
+      "learning_rate": 0.00011584923924369646,
+      "loss": 0.936,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5052777297110227,
+      "grad_norm": 0.4524749219417572,
+      "learning_rate": 0.00011555096672572659,
+      "loss": 1.0383,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5061429313029936,
+      "grad_norm": 0.477090984582901,
+      "learning_rate": 0.00011525255235979209,
+      "loss": 0.9769,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5070081328949645,
+      "grad_norm": 0.42532771825790405,
+      "learning_rate": 0.00011495399886787618,
+      "loss": 0.963,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5078733344869355,
+      "grad_norm": 0.4601391553878784,
+      "learning_rate": 0.0001146553089732311,
+      "loss": 0.9738,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5087385360789064,
+      "grad_norm": 0.4493047595024109,
+      "learning_rate": 0.00011435648540035328,
+      "loss": 0.9874,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5096037376708773,
+      "grad_norm": 0.5158471465110779,
+      "learning_rate": 0.00011405753087495859,
+      "loss": 0.9945,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5104689392628482,
+      "grad_norm": 0.4900023937225342,
+      "learning_rate": 0.00011375844812395722,
+      "loss": 0.9789,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5113341408548192,
+      "grad_norm": 0.4824017286300659,
+      "learning_rate": 0.00011345923987542913,
+      "loss": 0.9529,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5121993424467901,
+      "grad_norm": 0.4537661075592041,
+      "learning_rate": 0.00011315990885859885,
+      "loss": 0.9865,
+      "step": 2960
+    },
+    {
+      "epoch": 0.513064544038761,
+      "grad_norm": 0.48735764622688293,
+      "learning_rate": 0.00011286045780381085,
+      "loss": 1.007,
+      "step": 2965
+    },
+    {
+      "epoch": 0.513929745630732,
+      "grad_norm": 0.494494765996933,
+      "learning_rate": 0.00011256088944250446,
+      "loss": 0.9732,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5147949472227029,
+      "grad_norm": 0.4806153476238251,
+      "learning_rate": 0.00011226120650718912,
+      "loss": 0.983,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5156601488146738,
+      "grad_norm": 0.44597506523132324,
+      "learning_rate": 0.00011196141173141926,
+      "loss": 1.0055,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5165253504066447,
+      "grad_norm": 0.45618417859077454,
+      "learning_rate": 0.00011166150784976945,
+      "loss": 0.9788,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5173905519986157,
+      "grad_norm": 0.4606229066848755,
+      "learning_rate": 0.00011136149759780956,
+      "loss": 0.95,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5182557535905866,
+      "grad_norm": 0.42977404594421387,
+      "learning_rate": 0.00011106138371207963,
+      "loss": 0.9855,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5191209551825575,
+      "grad_norm": 0.45749431848526,
+      "learning_rate": 0.00011076116893006505,
+      "loss": 0.9748,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5199861567745284,
+      "grad_norm": 0.46687498688697815,
+      "learning_rate": 0.00011046085599017147,
+      "loss": 0.9932,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5208513583664994,
+      "grad_norm": 0.4712628424167633,
+      "learning_rate": 0.00011016044763169999,
+      "loss": 0.9233,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5217165599584703,
+      "grad_norm": 0.43779537081718445,
+      "learning_rate": 0.00010985994659482196,
+      "loss": 0.9391,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5225817615504412,
+      "grad_norm": 0.4388592839241028,
+      "learning_rate": 0.00010955935562055415,
+      "loss": 0.9942,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5234469631424122,
+      "grad_norm": 0.4575163424015045,
+      "learning_rate": 0.00010925867745073368,
+      "loss": 0.9802,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5243121647343831,
+      "grad_norm": 0.45447036623954773,
+      "learning_rate": 0.00010895791482799303,
+      "loss": 0.991,
+      "step": 3030
+    },
+    {
+      "epoch": 0.525177366326354,
+      "grad_norm": 0.4621734917163849,
+      "learning_rate": 0.00010865707049573501,
+      "loss": 0.9547,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5260425679183249,
+      "grad_norm": 0.4414147436618805,
+      "learning_rate": 0.00010835614719810776,
+      "loss": 0.9802,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5269077695102959,
+      "grad_norm": 0.447683721780777,
+      "learning_rate": 0.00010805514767997966,
+      "loss": 0.9481,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5277729711022668,
+      "grad_norm": 0.43929868936538696,
+      "learning_rate": 0.00010775407468691441,
+      "loss": 0.9902,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5286381726942377,
+      "grad_norm": 0.4730943441390991,
+      "learning_rate": 0.00010745293096514583,
+      "loss": 0.9888,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5295033742862086,
+      "grad_norm": 0.46589067578315735,
+      "learning_rate": 0.00010715171926155294,
+      "loss": 0.9888,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5303685758781796,
+      "grad_norm": 0.4755696952342987,
+      "learning_rate": 0.00010685044232363481,
+      "loss": 0.9891,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5312337774701505,
+      "grad_norm": 0.46226081252098083,
+      "learning_rate": 0.00010654910289948563,
+      "loss": 0.9916,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5320989790621214,
+      "grad_norm": 0.42151981592178345,
+      "learning_rate": 0.00010624770373776945,
+      "loss": 0.9415,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5329641806540925,
+      "grad_norm": 0.46804869174957275,
+      "learning_rate": 0.00010594624758769526,
+      "loss": 0.9871,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5338293822460634,
+      "grad_norm": 0.4632691442966461,
+      "learning_rate": 0.00010564473719899192,
+      "loss": 0.9903,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5346945838380343,
+      "grad_norm": 0.5181126594543457,
+      "learning_rate": 0.00010534317532188296,
+      "loss": 1.0247,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5355597854300052,
+      "grad_norm": 0.46891406178474426,
+      "learning_rate": 0.00010504156470706157,
+      "loss": 0.9627,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5364249870219762,
+      "grad_norm": 0.44084757566452026,
+      "learning_rate": 0.00010473990810566554,
+      "loss": 0.9644,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5372901886139471,
+      "grad_norm": 0.4899301528930664,
+      "learning_rate": 0.00010443820826925208,
+      "loss": 0.9962,
+      "step": 3105
+    },
+    {
+      "epoch": 0.538155390205918,
+      "grad_norm": 0.5310002565383911,
+      "learning_rate": 0.00010413646794977278,
+      "loss": 0.9352,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5390205917978889,
+      "grad_norm": 0.45059874653816223,
+      "learning_rate": 0.00010383468989954852,
+      "loss": 1.0001,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5398857933898599,
+      "grad_norm": 0.45391032099723816,
+      "learning_rate": 0.00010353287687124435,
+      "loss": 0.9482,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5407509949818308,
+      "grad_norm": 0.44090577960014343,
+      "learning_rate": 0.0001032310316178443,
+      "loss": 0.9858,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5416161965738017,
+      "grad_norm": 0.4861569404602051,
+      "learning_rate": 0.00010292915689262642,
+      "loss": 0.9731,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5424813981657727,
+      "grad_norm": 0.44434854388237,
+      "learning_rate": 0.00010262725544913756,
+      "loss": 0.9719,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5433465997577436,
+      "grad_norm": 0.44749364256858826,
+      "learning_rate": 0.00010232533004116822,
+      "loss": 0.9864,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5442118013497145,
+      "grad_norm": 0.4296785593032837,
+      "learning_rate": 0.00010202338342272756,
+      "loss": 1.0131,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5450770029416854,
+      "grad_norm": 0.47905465960502625,
+      "learning_rate": 0.00010172141834801823,
+      "loss": 0.9798,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5459422045336564,
+      "grad_norm": 0.43154534697532654,
+      "learning_rate": 0.00010141943757141119,
+      "loss": 0.9428,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5468074061256273,
+      "grad_norm": 0.5342848300933838,
+      "learning_rate": 0.00010111744384742064,
+      "loss": 0.947,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5476726077175982,
+      "grad_norm": 0.4587153494358063,
+      "learning_rate": 0.00010081543993067883,
+      "loss": 1.0066,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5485378093095691,
+      "grad_norm": 0.4687547981739044,
+      "learning_rate": 0.00010051342857591104,
+      "loss": 0.9945,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5494030109015401,
+      "grad_norm": 0.4243508577346802,
+      "learning_rate": 0.0001002114125379104,
+      "loss": 0.9973,
+      "step": 3175
+    },
+    {
+      "epoch": 0.550268212493511,
+      "grad_norm": 0.49467694759368896,
+      "learning_rate": 9.990939457151272e-05,
+      "loss": 0.9942,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5511334140854819,
+      "grad_norm": 0.44364285469055176,
+      "learning_rate": 9.96073774315714e-05,
+      "loss": 0.9667,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5519986156774529,
+      "grad_norm": 0.4399394094944,
+      "learning_rate": 9.930536387293235e-05,
+      "loss": 0.9902,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5528638172694238,
+      "grad_norm": 0.43766874074935913,
+      "learning_rate": 9.900335665040876e-05,
+      "loss": 0.9626,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5537290188613947,
+      "grad_norm": 0.5083682537078857,
+      "learning_rate": 9.870135851875605e-05,
+      "loss": 0.971,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5545942204533656,
+      "grad_norm": 0.41614192724227905,
+      "learning_rate": 9.839937223264674e-05,
+      "loss": 1.0278,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5554594220453366,
+      "grad_norm": 0.46571752429008484,
+      "learning_rate": 9.80974005466452e-05,
+      "loss": 0.9571,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5563246236373075,
+      "grad_norm": 0.46136540174484253,
+      "learning_rate": 9.779544621518281e-05,
+      "loss": 0.9669,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5571898252292784,
+      "grad_norm": 0.5057949423789978,
+      "learning_rate": 9.749351199253243e-05,
+      "loss": 0.9876,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5580550268212493,
+      "grad_norm": 0.44634687900543213,
+      "learning_rate": 9.719160063278371e-05,
+      "loss": 0.9805,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5589202284132203,
+      "grad_norm": 0.4209567904472351,
+      "learning_rate": 9.688971488981759e-05,
+      "loss": 0.9621,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5597854300051912,
+      "grad_norm": 0.4183271825313568,
+      "learning_rate": 9.658785751728144e-05,
+      "loss": 0.9189,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5606506315971621,
+      "grad_norm": 0.4438595175743103,
+      "learning_rate": 9.628603126856378e-05,
+      "loss": 1.0109,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5615158331891331,
+      "grad_norm": 0.4465368688106537,
+      "learning_rate": 9.598423889676933e-05,
+      "loss": 0.9747,
+      "step": 3245
+    },
+    {
+      "epoch": 0.562381034781104,
+      "grad_norm": 0.444805383682251,
+      "learning_rate": 9.568248315469373e-05,
+      "loss": 0.9712,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5632462363730749,
+      "grad_norm": 0.43554961681365967,
+      "learning_rate": 9.53807667947985e-05,
+      "loss": 1.0063,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5641114379650458,
+      "grad_norm": 0.4528721868991852,
+      "learning_rate": 9.507909256918602e-05,
+      "loss": 0.9849,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5649766395570168,
+      "grad_norm": 0.45312389731407166,
+      "learning_rate": 9.477746322957416e-05,
+      "loss": 1.0017,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5658418411489877,
+      "grad_norm": 0.45233070850372314,
+      "learning_rate": 9.447588152727161e-05,
+      "loss": 1.0283,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5667070427409586,
+      "grad_norm": 0.44334331154823303,
+      "learning_rate": 9.417435021315231e-05,
+      "loss": 0.9917,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5675722443329295,
+      "grad_norm": 0.5259175896644592,
+      "learning_rate": 9.387287203763072e-05,
+      "loss": 0.962,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5684374459249005,
+      "grad_norm": 0.4928758144378662,
+      "learning_rate": 9.357144975063661e-05,
+      "loss": 0.9807,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5693026475168714,
+      "grad_norm": 0.44045624136924744,
+      "learning_rate": 9.327008610158981e-05,
+      "loss": 0.9728,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5701678491088423,
+      "grad_norm": 0.46304404735565186,
+      "learning_rate": 9.296878383937552e-05,
+      "loss": 0.9755,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5710330507008133,
+      "grad_norm": 0.45428958535194397,
+      "learning_rate": 9.266754571231874e-05,
+      "loss": 0.9927,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5718982522927842,
+      "grad_norm": 0.4710966944694519,
+      "learning_rate": 9.236637446815969e-05,
+      "loss": 0.9391,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5727634538847551,
+      "grad_norm": 0.46206918358802795,
+      "learning_rate": 9.206527285402835e-05,
+      "loss": 0.9709,
+      "step": 3310
+    },
+    {
+      "epoch": 0.573628655476726,
+      "grad_norm": 0.4327426254749298,
+      "learning_rate": 9.176424361641965e-05,
+      "loss": 0.9467,
+      "step": 3315
+    },
+    {
+      "epoch": 0.574493857068697,
+      "grad_norm": 0.4926101565361023,
+      "learning_rate": 9.146328950116839e-05,
+      "loss": 0.9767,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5753590586606679,
+      "grad_norm": 0.4621298313140869,
+      "learning_rate": 9.116241325342399e-05,
+      "loss": 0.99,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5762242602526388,
+      "grad_norm": 0.46005651354789734,
+      "learning_rate": 9.086161761762578e-05,
+      "loss": 0.9796,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5770894618446097,
+      "grad_norm": 0.4343353807926178,
+      "learning_rate": 9.056090533747762e-05,
+      "loss": 0.9349,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5779546634365808,
+      "grad_norm": 0.4175094962120056,
+      "learning_rate": 9.026027915592321e-05,
+      "loss": 0.9857,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5788198650285516,
+      "grad_norm": 0.44060176610946655,
+      "learning_rate": 8.995974181512072e-05,
+      "loss": 0.9523,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5796850666205225,
+      "grad_norm": 0.4819500148296356,
+      "learning_rate": 8.965929605641811e-05,
+      "loss": 0.9838,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5805502682124936,
+      "grad_norm": 0.4309912919998169,
+      "learning_rate": 8.935894462032798e-05,
+      "loss": 0.9787,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5814154698044645,
+      "grad_norm": 0.440983384847641,
+      "learning_rate": 8.905869024650242e-05,
+      "loss": 0.9485,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5822806713964354,
+      "grad_norm": 0.4531494975090027,
+      "learning_rate": 8.875853567370829e-05,
+      "loss": 0.9819,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5831458729884063,
+      "grad_norm": 0.43064072728157043,
+      "learning_rate": 8.845848363980207e-05,
+      "loss": 0.9748,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5840110745803773,
+      "grad_norm": 0.46520137786865234,
+      "learning_rate": 8.815853688170498e-05,
+      "loss": 0.9634,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5848762761723482,
+      "grad_norm": 0.5070540904998779,
+      "learning_rate": 8.785869813537789e-05,
+      "loss": 1.011,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5857414777643191,
+      "grad_norm": 0.4707706868648529,
+      "learning_rate": 8.75589701357965e-05,
+      "loss": 0.9667,
+      "step": 3385
+    },
+    {
+      "epoch": 0.58660667935629,
+      "grad_norm": 0.4585541784763336,
+      "learning_rate": 8.725935561692637e-05,
+      "loss": 0.9676,
+      "step": 3390
+    },
+    {
+      "epoch": 0.587471880948261,
+      "grad_norm": 0.4634505808353424,
+      "learning_rate": 8.695985731169781e-05,
+      "loss": 0.9875,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5883370825402319,
+      "grad_norm": 0.4741441011428833,
+      "learning_rate": 8.666047795198124e-05,
+      "loss": 0.9787,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5892022841322028,
+      "grad_norm": 0.44273069500923157,
+      "learning_rate": 8.636122026856201e-05,
+      "loss": 0.9481,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5900674857241738,
+      "grad_norm": 0.45366448163986206,
+      "learning_rate": 8.606208699111573e-05,
+      "loss": 0.9577,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5909326873161447,
+      "grad_norm": 0.4686761498451233,
+      "learning_rate": 8.576308084818305e-05,
+      "loss": 0.9743,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5917978889081156,
+      "grad_norm": 0.4535995125770569,
+      "learning_rate": 8.546420456714517e-05,
+      "loss": 0.9928,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5926630905000865,
+      "grad_norm": 0.426045298576355,
+      "learning_rate": 8.516546087419856e-05,
+      "loss": 0.9829,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5935282920920575,
+      "grad_norm": 0.4387094974517822,
+      "learning_rate": 8.486685249433048e-05,
+      "loss": 0.9899,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5943934936840284,
+      "grad_norm": 0.4502699375152588,
+      "learning_rate": 8.456838215129384e-05,
+      "loss": 0.998,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5952586952759993,
+      "grad_norm": 0.4558119773864746,
+      "learning_rate": 8.427005256758236e-05,
+      "loss": 0.9493,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5961238968679702,
+      "grad_norm": 0.4319314658641815,
+      "learning_rate": 8.397186646440603e-05,
+      "loss": 0.9276,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5969890984599412,
+      "grad_norm": 0.4241703748703003,
+      "learning_rate": 8.367382656166584e-05,
+      "loss": 0.9368,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5978543000519121,
+      "grad_norm": 0.4475589394569397,
+      "learning_rate": 8.33759355779295e-05,
+      "loss": 0.9648,
+      "step": 3455
+    },
+    {
+      "epoch": 0.598719501643883,
+      "grad_norm": 0.4357362389564514,
+      "learning_rate": 8.307819623040606e-05,
+      "loss": 0.9449,
+      "step": 3460
+    },
+    {
+      "epoch": 0.599584703235854,
+      "grad_norm": 0.5170083045959473,
+      "learning_rate": 8.278061123492162e-05,
+      "loss": 0.9446,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6004499048278249,
+      "grad_norm": 0.45184993743896484,
+      "learning_rate": 8.248318330589437e-05,
+      "loss": 0.9494,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6013151064197958,
+      "grad_norm": 0.44678378105163574,
+      "learning_rate": 8.218591515630962e-05,
+      "loss": 0.951,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6021803080117667,
+      "grad_norm": 0.4784872531890869,
+      "learning_rate": 8.188880949769552e-05,
+      "loss": 1.0141,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6030455096037377,
+      "grad_norm": 0.4891086220741272,
+      "learning_rate": 8.159186904009783e-05,
+      "loss": 0.9438,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6039107111957086,
+      "grad_norm": 0.4449233412742615,
+      "learning_rate": 8.129509649205558e-05,
+      "loss": 0.9473,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6047759127876795,
+      "grad_norm": 0.4476933479309082,
+      "learning_rate": 8.099849456057613e-05,
+      "loss": 0.9215,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6056411143796505,
+      "grad_norm": 0.47453024983406067,
+      "learning_rate": 8.070206595111061e-05,
+      "loss": 0.9419,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6065063159716214,
+      "grad_norm": 0.4311204254627228,
+      "learning_rate": 8.040581336752921e-05,
+      "loss": 0.9707,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6073715175635923,
+      "grad_norm": 0.4637129008769989,
+      "learning_rate": 8.010973951209642e-05,
+      "loss": 0.994,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6082367191555632,
+      "grad_norm": 0.42446842789649963,
+      "learning_rate": 7.981384708544658e-05,
+      "loss": 0.9716,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6091019207475342,
+      "grad_norm": 0.4456198811531067,
+      "learning_rate": 7.951813878655898e-05,
+      "loss": 0.9435,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6099671223395051,
+      "grad_norm": 0.4548759162425995,
+      "learning_rate": 7.922261731273359e-05,
+      "loss": 0.9703,
+      "step": 3525
+    },
+    {
+      "epoch": 0.610832323931476,
+      "grad_norm": 0.4350235164165497,
+      "learning_rate": 7.892728535956601e-05,
+      "loss": 0.957,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6116975255234469,
+      "grad_norm": 0.4410596191883087,
+      "learning_rate": 7.863214562092336e-05,
+      "loss": 0.9588,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6125627271154179,
+      "grad_norm": 0.46627217531204224,
+      "learning_rate": 7.833720078891938e-05,
+      "loss": 0.9848,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6134279287073888,
+      "grad_norm": 0.44155973196029663,
+      "learning_rate": 7.804245355388994e-05,
+      "loss": 0.9629,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6142931302993597,
+      "grad_norm": 0.4281848669052124,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 0.9505,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6151583318913307,
+      "grad_norm": 0.4397382140159607,
+      "learning_rate": 7.745356262706189e-05,
+      "loss": 0.9638,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6160235334833016,
+      "grad_norm": 0.44392621517181396,
+      "learning_rate": 7.715942430682515e-05,
+      "loss": 1.0033,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6168887350752725,
+      "grad_norm": 0.4472537338733673,
+      "learning_rate": 7.686549432663757e-05,
+      "loss": 0.9348,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6177539366672434,
+      "grad_norm": 0.4813332259654999,
+      "learning_rate": 7.657177536757816e-05,
+      "loss": 0.9431,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6186191382592144,
+      "grad_norm": 0.4196160137653351,
+      "learning_rate": 7.627827010880108e-05,
+      "loss": 0.9661,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6194843398511853,
+      "grad_norm": 0.43365371227264404,
+      "learning_rate": 7.598498122751108e-05,
+      "loss": 0.9446,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6203495414431562,
+      "grad_norm": 0.44143837690353394,
+      "learning_rate": 7.569191139893944e-05,
+      "loss": 0.9582,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6212147430351271,
+      "grad_norm": 0.455301970243454,
+      "learning_rate": 7.539906329631911e-05,
+      "loss": 0.9487,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6220799446270981,
+      "grad_norm": 0.4444476366043091,
+      "learning_rate": 7.510643959086083e-05,
+      "loss": 0.9818,
+      "step": 3595
+    },
+    {
+      "epoch": 0.622945146219069,
+      "grad_norm": 0.4136192500591278,
+      "learning_rate": 7.481404295172827e-05,
+      "loss": 1.0015,
+      "step": 3600
+    },
+    {
+      "epoch": 0.62381034781104,
+      "grad_norm": 0.42628371715545654,
+      "learning_rate": 7.452187604601407e-05,
+      "loss": 0.98,
+      "step": 3605
+    },
+    {
+      "epoch": 0.624675549403011,
+      "grad_norm": 0.4436239004135132,
+      "learning_rate": 7.422994153871534e-05,
+      "loss": 0.9909,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6255407509949819,
+      "grad_norm": 0.4337378144264221,
+      "learning_rate": 7.393824209270925e-05,
+      "loss": 0.9714,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6264059525869528,
+      "grad_norm": 0.46607285737991333,
+      "learning_rate": 7.364678036872903e-05,
+      "loss": 0.9933,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6272711541789237,
+      "grad_norm": 0.4370154142379761,
+      "learning_rate": 7.335555902533941e-05,
+      "loss": 0.987,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6281363557708947,
+      "grad_norm": 0.44606542587280273,
+      "learning_rate": 7.306458071891258e-05,
+      "loss": 0.9482,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6290015573628656,
+      "grad_norm": 0.5535009503364563,
+      "learning_rate": 7.277384810360375e-05,
+      "loss": 0.9597,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6298667589548365,
+      "grad_norm": 0.5309751033782959,
+      "learning_rate": 7.248336383132718e-05,
+      "loss": 0.9586,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6307319605468074,
+      "grad_norm": 0.4356056749820709,
+      "learning_rate": 7.219313055173186e-05,
+      "loss": 0.9881,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6315971621387784,
+      "grad_norm": 0.4677325487136841,
+      "learning_rate": 7.190315091217728e-05,
+      "loss": 0.963,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6324623637307493,
+      "grad_norm": 0.4469841420650482,
+      "learning_rate": 7.161342755770942e-05,
+      "loss": 0.9331,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6333275653227202,
+      "grad_norm": 0.4805571138858795,
+      "learning_rate": 7.132396313103651e-05,
+      "loss": 0.9428,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6341927669146912,
+      "grad_norm": 0.43493425846099854,
+      "learning_rate": 7.1034760272505e-05,
+      "loss": 0.9357,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6350579685066621,
+      "grad_norm": 0.44341421127319336,
+      "learning_rate": 7.074582162007546e-05,
+      "loss": 0.9954,
+      "step": 3670
+    },
+    {
+      "epoch": 0.635923170098633,
+      "grad_norm": 0.47012633085250854,
+      "learning_rate": 7.045714980929845e-05,
+      "loss": 0.9575,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6367883716906039,
+      "grad_norm": 0.4581446647644043,
+      "learning_rate": 7.016874747329068e-05,
+      "loss": 0.9845,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6376535732825749,
+      "grad_norm": 0.40438956022262573,
+      "learning_rate": 6.988061724271065e-05,
+      "loss": 0.9699,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6385187748745458,
+      "grad_norm": 0.4338352680206299,
+      "learning_rate": 6.959276174573503e-05,
+      "loss": 0.9328,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6393839764665167,
+      "grad_norm": 0.4316514730453491,
+      "learning_rate": 6.930518360803438e-05,
+      "loss": 0.9436,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6402491780584876,
+      "grad_norm": 0.4489031434059143,
+      "learning_rate": 6.90178854527494e-05,
+      "loss": 0.9399,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6411143796504586,
+      "grad_norm": 0.42906907200813293,
+      "learning_rate": 6.873086990046691e-05,
+      "loss": 0.9655,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6419795812424295,
+      "grad_norm": 0.42408278584480286,
+      "learning_rate": 6.844413956919603e-05,
+      "loss": 0.9566,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6428447828344004,
+      "grad_norm": 0.4122326076030731,
+      "learning_rate": 6.815769707434411e-05,
+      "loss": 0.9992,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6437099844263714,
+      "grad_norm": 0.4537835121154785,
+      "learning_rate": 6.787154502869313e-05,
+      "loss": 0.9384,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6445751860183423,
+      "grad_norm": 0.4500395953655243,
+      "learning_rate": 6.758568604237568e-05,
+      "loss": 0.9851,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6454403876103132,
+      "grad_norm": 0.43696606159210205,
+      "learning_rate": 6.730012272285123e-05,
+      "loss": 0.942,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6463055892022841,
+      "grad_norm": 0.42087531089782715,
+      "learning_rate": 6.701485767488235e-05,
+      "loss": 0.9386,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6471707907942551,
+      "grad_norm": 0.4692328870296478,
+      "learning_rate": 6.672989350051082e-05,
+      "loss": 0.9937,
+      "step": 3740
+    },
+    {
+      "epoch": 0.648035992386226,
+      "grad_norm": 0.4377414584159851,
+      "learning_rate": 6.644523279903419e-05,
+      "loss": 0.953,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6489011939781969,
+      "grad_norm": 0.5129314661026001,
+      "learning_rate": 6.61608781669817e-05,
+      "loss": 0.9555,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6497663955701678,
+      "grad_norm": 0.452594131231308,
+      "learning_rate": 6.587683219809089e-05,
+      "loss": 0.9346,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6506315971621388,
+      "grad_norm": 0.4192207157611847,
+      "learning_rate": 6.559309748328388e-05,
+      "loss": 0.9461,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6514967987541097,
+      "grad_norm": 0.4243721067905426,
+      "learning_rate": 6.530967661064348e-05,
+      "loss": 0.9644,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6523620003460806,
+      "grad_norm": 0.4021894931793213,
+      "learning_rate": 6.502657216539007e-05,
+      "loss": 0.9608,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6532272019380516,
+      "grad_norm": 0.43916577100753784,
+      "learning_rate": 6.474378672985749e-05,
+      "loss": 0.9321,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6540924035300225,
+      "grad_norm": 0.4375644326210022,
+      "learning_rate": 6.446132288346994e-05,
+      "loss": 0.9507,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6549576051219934,
+      "grad_norm": 0.45128095149993896,
+      "learning_rate": 6.417918320271807e-05,
+      "loss": 0.9804,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6558228067139643,
+      "grad_norm": 0.45634618401527405,
+      "learning_rate": 6.389737026113577e-05,
+      "loss": 0.9579,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6566880083059353,
+      "grad_norm": 0.45252084732055664,
+      "learning_rate": 6.36158866292766e-05,
+      "loss": 0.9811,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6575532098979062,
+      "grad_norm": 0.4186381697654724,
+      "learning_rate": 6.333473487469021e-05,
+      "loss": 0.983,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6584184114898771,
+      "grad_norm": 0.43495872616767883,
+      "learning_rate": 6.305391756189916e-05,
+      "loss": 1.0123,
+      "step": 3805
+    },
+    {
+      "epoch": 0.659283613081848,
+      "grad_norm": 0.42767274379730225,
+      "learning_rate": 6.27734372523753e-05,
+      "loss": 0.9882,
+      "step": 3810
+    },
+    {
+      "epoch": 0.660148814673819,
+      "grad_norm": 0.44282782077789307,
+      "learning_rate": 6.249329650451663e-05,
+      "loss": 0.9771,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6610140162657899,
+      "grad_norm": 0.4403495490550995,
+      "learning_rate": 6.221349787362366e-05,
+      "loss": 1.0058,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6618792178577608,
+      "grad_norm": 0.4575168490409851,
+      "learning_rate": 6.193404391187647e-05,
+      "loss": 0.9948,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6627444194497318,
+      "grad_norm": 0.42425480484962463,
+      "learning_rate": 6.165493716831119e-05,
+      "loss": 0.9357,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6636096210417027,
+      "grad_norm": 0.4653586447238922,
+      "learning_rate": 6.137618018879671e-05,
+      "loss": 0.973,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6644748226336736,
+      "grad_norm": 0.4468865692615509,
+      "learning_rate": 6.109777551601165e-05,
+      "loss": 0.9675,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6653400242256445,
+      "grad_norm": 0.4487997889518738,
+      "learning_rate": 6.0819725689421046e-05,
+      "loss": 0.9482,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6662052258176155,
+      "grad_norm": 0.4328047037124634,
+      "learning_rate": 6.054203324525321e-05,
+      "loss": 0.9853,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6670704274095864,
+      "grad_norm": 0.4487217664718628,
+      "learning_rate": 6.026470071647656e-05,
+      "loss": 0.9879,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6679356290015573,
+      "grad_norm": 0.5115017294883728,
+      "learning_rate": 5.9987730632776584e-05,
+      "loss": 0.9804,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6688008305935282,
+      "grad_norm": 0.494248628616333,
+      "learning_rate": 5.9711125520532754e-05,
+      "loss": 0.9657,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6696660321854992,
+      "grad_norm": 0.4214867949485779,
+      "learning_rate": 5.943488790279539e-05,
+      "loss": 0.9687,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6705312337774701,
+      "grad_norm": 0.4864426851272583,
+      "learning_rate": 5.9159020299262793e-05,
+      "loss": 0.9391,
+      "step": 3875
+    },
+    {
+      "epoch": 0.671396435369441,
+      "grad_norm": 0.4803209602832794,
+      "learning_rate": 5.888352522625809e-05,
+      "loss": 0.9195,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6722616369614121,
+      "grad_norm": 0.42843303084373474,
+      "learning_rate": 5.860840519670648e-05,
+      "loss": 0.9757,
+      "step": 3885
+    },
+    {
+      "epoch": 0.673126838553383,
+      "grad_norm": 0.4331001341342926,
+      "learning_rate": 5.833366272011212e-05,
+      "loss": 0.9961,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6739920401453539,
+      "grad_norm": 0.43525782227516174,
+      "learning_rate": 5.805930030253539e-05,
+      "loss": 0.9764,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6748572417373248,
+      "grad_norm": 0.4345535933971405,
+      "learning_rate": 5.7785320446569924e-05,
+      "loss": 0.9753,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6757224433292958,
+      "grad_norm": 0.44036418199539185,
+      "learning_rate": 5.7511725651319834e-05,
+      "loss": 0.9439,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6765876449212667,
+      "grad_norm": 0.42969611287117004,
+      "learning_rate": 5.7238518412376906e-05,
+      "loss": 0.9827,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6774528465132376,
+      "grad_norm": 0.42602795362472534,
+      "learning_rate": 5.696570122179783e-05,
+      "loss": 0.9341,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6783180481052085,
+      "grad_norm": 0.4188039302825928,
+      "learning_rate": 5.66932765680815e-05,
+      "loss": 0.955,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6791832496971795,
+      "grad_norm": 0.49803611636161804,
+      "learning_rate": 5.6421246936146255e-05,
+      "loss": 0.9364,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6800484512891504,
+      "grad_norm": 0.46372202038764954,
+      "learning_rate": 5.614961480730727e-05,
+      "loss": 0.9426,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6809136528811213,
+      "grad_norm": 0.4493297040462494,
+      "learning_rate": 5.587838265925389e-05,
+      "loss": 0.9297,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6817788544730923,
+      "grad_norm": 0.43697822093963623,
+      "learning_rate": 5.5607552966027044e-05,
+      "loss": 0.9581,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6826440560650632,
+      "grad_norm": 0.44038650393486023,
+      "learning_rate": 5.533712819799667e-05,
+      "loss": 1.0135,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6835092576570341,
+      "grad_norm": 0.44993671774864197,
+      "learning_rate": 5.5067110821839186e-05,
+      "loss": 0.9798,
+      "step": 3950
+    },
+    {
+      "epoch": 0.684374459249005,
+      "grad_norm": 0.4506613314151764,
+      "learning_rate": 5.4797503300515e-05,
+      "loss": 0.9571,
+      "step": 3955
+    },
+    {
+      "epoch": 0.685239660840976,
+      "grad_norm": 0.4558075964450836,
+      "learning_rate": 5.452830809324601e-05,
+      "loss": 0.9611,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6861048624329469,
+      "grad_norm": 0.4847477376461029,
+      "learning_rate": 5.425952765549319e-05,
+      "loss": 1.0172,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6869700640249178,
+      "grad_norm": 0.4571908414363861,
+      "learning_rate": 5.399116443893433e-05,
+      "loss": 0.9883,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6878352656168887,
+      "grad_norm": 0.41807690262794495,
+      "learning_rate": 5.3723220891441274e-05,
+      "loss": 1.0335,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6887004672088597,
+      "grad_norm": 0.4681232273578644,
+      "learning_rate": 5.345569945705817e-05,
+      "loss": 0.9219,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6895656688008306,
+      "grad_norm": 0.4585653841495514,
+      "learning_rate": 5.3188602575978616e-05,
+      "loss": 1.0058,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6904308703928015,
+      "grad_norm": 0.41353005170822144,
+      "learning_rate": 5.292193268452377e-05,
+      "loss": 0.9544,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6912960719847725,
+      "grad_norm": 0.4379875361919403,
+      "learning_rate": 5.265569221511999e-05,
+      "loss": 0.9619,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6921612735767434,
+      "grad_norm": 0.45931750535964966,
+      "learning_rate": 5.238988359627662e-05,
+      "loss": 0.9852,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6930264751687143,
+      "grad_norm": 0.46481239795684814,
+      "learning_rate": 5.212450925256406e-05,
+      "loss": 0.9666,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6938916767606852,
+      "grad_norm": 0.4746643304824829,
+      "learning_rate": 5.185957160459119e-05,
+      "loss": 0.9852,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6947568783526562,
+      "grad_norm": 0.4331592321395874,
+      "learning_rate": 5.159507306898386e-05,
+      "loss": 0.9581,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6956220799446271,
+      "grad_norm": 0.44027334451675415,
+      "learning_rate": 5.133101605836228e-05,
+      "loss": 0.9949,
+      "step": 4020
+    },
+    {
+      "epoch": 0.696487281536598,
+      "grad_norm": 0.42148056626319885,
+      "learning_rate": 5.106740298131961e-05,
+      "loss": 0.961,
+      "step": 4025
+    },
+    {
+      "epoch": 0.697352483128569,
+      "grad_norm": 0.41860854625701904,
+      "learning_rate": 5.080423624239936e-05,
+      "loss": 0.9602,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6982176847205399,
+      "grad_norm": 0.4259825348854065,
+      "learning_rate": 5.054151824207405e-05,
+      "loss": 0.9516,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6990828863125108,
+      "grad_norm": 0.44930359721183777,
+      "learning_rate": 5.027925137672278e-05,
+      "loss": 0.9646,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6999480879044817,
+      "grad_norm": 0.45277413725852966,
+      "learning_rate": 5.0017438038609846e-05,
+      "loss": 0.9787,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7008132894964527,
+      "grad_norm": 0.45245739817619324,
+      "learning_rate": 4.975608061586258e-05,
+      "loss": 0.9823,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7016784910884236,
+      "grad_norm": 0.44991815090179443,
+      "learning_rate": 4.9495181492449724e-05,
+      "loss": 0.9416,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7025436926803945,
+      "grad_norm": 0.42899975180625916,
+      "learning_rate": 4.923474304815967e-05,
+      "loss": 0.987,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7034088942723654,
+      "grad_norm": 0.4850844442844391,
+      "learning_rate": 4.897476765857858e-05,
+      "loss": 0.9856,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7042740958643364,
+      "grad_norm": 0.42250022292137146,
+      "learning_rate": 4.871525769506916e-05,
+      "loss": 0.9327,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7051392974563073,
+      "grad_norm": 0.4618523120880127,
+      "learning_rate": 4.8456215524748396e-05,
+      "loss": 1.017,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7060044990482782,
+      "grad_norm": 0.42984339594841003,
+      "learning_rate": 4.819764351046659e-05,
+      "loss": 0.9534,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7068697006402492,
+      "grad_norm": 0.43350252509117126,
+      "learning_rate": 4.7939544010785377e-05,
+      "loss": 0.9516,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7077349022322201,
+      "grad_norm": 0.4169905185699463,
+      "learning_rate": 4.768191937995636e-05,
+      "loss": 0.9477,
+      "step": 4090
+    },
+    {
+      "epoch": 0.708600103824191,
+      "grad_norm": 0.40108850598335266,
+      "learning_rate": 4.74247719678997e-05,
+      "loss": 0.9751,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7094653054161619,
+      "grad_norm": 0.4502684772014618,
+      "learning_rate": 4.716810412018251e-05,
+      "loss": 0.9778,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7103305070081329,
+      "grad_norm": 0.4119910001754761,
+      "learning_rate": 4.691191817799765e-05,
+      "loss": 0.9507,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7111957086001038,
+      "grad_norm": 0.43756359815597534,
+      "learning_rate": 4.665621647814226e-05,
+      "loss": 0.9897,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7120609101920747,
+      "grad_norm": 0.44212594628334045,
+      "learning_rate": 4.640100135299644e-05,
+      "loss": 0.9555,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7129261117840456,
+      "grad_norm": 0.464538037776947,
+      "learning_rate": 4.614627513050205e-05,
+      "loss": 0.9533,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7137913133760166,
+      "grad_norm": 0.42958030104637146,
+      "learning_rate": 4.589204013414139e-05,
+      "loss": 1.0032,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7146565149679875,
+      "grad_norm": 0.45080655813217163,
+      "learning_rate": 4.563829868291607e-05,
+      "loss": 0.9501,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7155217165599584,
+      "grad_norm": 0.4486967623233795,
+      "learning_rate": 4.538505309132584e-05,
+      "loss": 0.9779,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7163869181519295,
+      "grad_norm": 0.4742283821105957,
+      "learning_rate": 4.513230566934743e-05,
+      "loss": 0.9393,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7172521197439004,
+      "grad_norm": 0.4522545635700226,
+      "learning_rate": 4.4880058722413566e-05,
+      "loss": 0.9623,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7181173213358713,
+      "grad_norm": 0.4180852472782135,
+      "learning_rate": 4.462831455139185e-05,
+      "loss": 0.9544,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7189825229278421,
+      "grad_norm": 0.4502962827682495,
+      "learning_rate": 4.4377075452563955e-05,
+      "loss": 0.9749,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7198477245198132,
+      "grad_norm": 0.42445746064186096,
+      "learning_rate": 4.412634371760429e-05,
+      "loss": 0.9975,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7207129261117841,
+      "grad_norm": 0.4130094051361084,
+      "learning_rate": 4.387612163355966e-05,
+      "loss": 1.0015,
+      "step": 4165
+    },
+    {
+      "epoch": 0.721578127703755,
+      "grad_norm": 0.5205526947975159,
+      "learning_rate": 4.362641148282781e-05,
+      "loss": 0.9721,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7224433292957259,
+      "grad_norm": 0.4239148199558258,
+      "learning_rate": 4.337721554313705e-05,
+      "loss": 0.9012,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7233085308876969,
+      "grad_norm": 0.4548054337501526,
+      "learning_rate": 4.312853608752531e-05,
+      "loss": 0.9368,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7241737324796678,
+      "grad_norm": 0.4600510597229004,
+      "learning_rate": 4.288037538431933e-05,
+      "loss": 0.9614,
+      "step": 4185
+    },
+    {
+      "epoch": 0.7250389340716387,
+      "grad_norm": 0.4393291175365448,
+      "learning_rate": 4.2632735697114226e-05,
+      "loss": 0.9403,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7259041356636097,
+      "grad_norm": 0.4278406798839569,
+      "learning_rate": 4.238561928475244e-05,
+      "loss": 0.9459,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7267693372555806,
+      "grad_norm": 0.43765777349472046,
+      "learning_rate": 4.2139028401303625e-05,
+      "loss": 0.9891,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7276345388475515,
+      "grad_norm": 0.4184790551662445,
+      "learning_rate": 4.189296529604355e-05,
+      "loss": 0.9551,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7284997404395224,
+      "grad_norm": 0.4558292031288147,
+      "learning_rate": 4.164743221343417e-05,
+      "loss": 0.9709,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7293649420314934,
+      "grad_norm": 0.4263213276863098,
+      "learning_rate": 4.1402431393102516e-05,
+      "loss": 0.967,
+      "step": 4215
+    },
+    {
+      "epoch": 0.7302301436234643,
+      "grad_norm": 0.44499507546424866,
+      "learning_rate": 4.115796506982087e-05,
+      "loss": 0.9841,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7310953452154352,
+      "grad_norm": 0.46089598536491394,
+      "learning_rate": 4.091403547348597e-05,
+      "loss": 1.0149,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7319605468074061,
+      "grad_norm": 0.44323694705963135,
+      "learning_rate": 4.0670644829098824e-05,
+      "loss": 0.945,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7328257483993771,
+      "grad_norm": 0.4892789125442505,
+      "learning_rate": 4.0427795356744445e-05,
+      "loss": 0.9208,
+      "step": 4235
+    },
+    {
+      "epoch": 0.733690949991348,
+      "grad_norm": 0.43195831775665283,
+      "learning_rate": 4.018548927157141e-05,
+      "loss": 0.948,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7345561515833189,
+      "grad_norm": 0.4387545585632324,
+      "learning_rate": 3.994372878377202e-05,
+      "loss": 1.029,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7354213531752899,
+      "grad_norm": 0.4202425479888916,
+      "learning_rate": 3.970251609856166e-05,
+      "loss": 0.9322,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7362865547672608,
+      "grad_norm": 0.4669027030467987,
+      "learning_rate": 3.946185341615918e-05,
+      "loss": 0.9433,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7371517563592317,
+      "grad_norm": 0.426996648311615,
+      "learning_rate": 3.9221742931766436e-05,
+      "loss": 0.9624,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7380169579512026,
+      "grad_norm": 0.4329177439212799,
+      "learning_rate": 3.898218683554845e-05,
+      "loss": 0.9852,
+      "step": 4265
+    },
+    {
+      "epoch": 0.7388821595431736,
+      "grad_norm": 0.40790703892707825,
+      "learning_rate": 3.874318731261343e-05,
+      "loss": 0.9441,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7397473611351445,
+      "grad_norm": 0.43579617142677307,
+      "learning_rate": 3.8504746542992765e-05,
+      "loss": 0.9446,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7406125627271154,
+      "grad_norm": 0.4374306797981262,
+      "learning_rate": 3.826686670162119e-05,
+      "loss": 0.932,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7414777643190863,
+      "grad_norm": 0.43435388803482056,
+      "learning_rate": 3.802954995831699e-05,
+      "loss": 0.9898,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7423429659110573,
+      "grad_norm": 0.4677979350090027,
+      "learning_rate": 3.779279847776208e-05,
+      "loss": 0.9797,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7432081675030282,
+      "grad_norm": 0.43926307559013367,
+      "learning_rate": 3.7556614419482385e-05,
+      "loss": 0.9666,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7440733690949991,
+      "grad_norm": 0.4474581778049469,
+      "learning_rate": 3.7320999937828084e-05,
+      "loss": 0.9616,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7449385706869701,
+      "grad_norm": 0.4644235074520111,
+      "learning_rate": 3.7085957181953966e-05,
+      "loss": 1.0063,
+      "step": 4305
+    },
+    {
+      "epoch": 0.745803772278941,
+      "grad_norm": 0.4108608067035675,
+      "learning_rate": 3.685148829579984e-05,
+      "loss": 0.9367,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7466689738709119,
+      "grad_norm": 0.44119784235954285,
+      "learning_rate": 3.661759541807096e-05,
+      "loss": 0.9638,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7475341754628828,
+      "grad_norm": 0.4548141360282898,
+      "learning_rate": 3.638428068221853e-05,
+      "loss": 0.9699,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7483993770548538,
+      "grad_norm": 0.4526442289352417,
+      "learning_rate": 3.615154621642023e-05,
+      "loss": 0.9521,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7492645786468247,
+      "grad_norm": 0.5478776693344116,
+      "learning_rate": 3.591939414356081e-05,
+      "loss": 0.9538,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7501297802387956,
+      "grad_norm": 0.423746258020401,
+      "learning_rate": 3.568782658121275e-05,
+      "loss": 0.9548,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7509949818307665,
+      "grad_norm": 0.44348010420799255,
+      "learning_rate": 3.545684564161697e-05,
+      "loss": 0.9325,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7518601834227375,
+      "grad_norm": 0.4436131417751312,
+      "learning_rate": 3.522645343166336e-05,
+      "loss": 0.9738,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7527253850147084,
+      "grad_norm": 0.4240371882915497,
+      "learning_rate": 3.499665205287189e-05,
+      "loss": 0.9721,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7535905866066793,
+      "grad_norm": 0.4415445327758789,
+      "learning_rate": 3.476744360137317e-05,
+      "loss": 0.9401,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7544557881986503,
+      "grad_norm": 0.43800264596939087,
+      "learning_rate": 3.453883016788948e-05,
+      "loss": 0.9331,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7553209897906212,
+      "grad_norm": 0.4613566994667053,
+      "learning_rate": 3.4310813837715625e-05,
+      "loss": 0.9906,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7561861913825921,
+      "grad_norm": 0.43836039304733276,
+      "learning_rate": 3.408339669069991e-05,
+      "loss": 0.9468,
+      "step": 4370
+    },
+    {
+      "epoch": 0.757051392974563,
+      "grad_norm": 0.43051472306251526,
+      "learning_rate": 3.3856580801225327e-05,
+      "loss": 0.9338,
+      "step": 4375
+    },
+    {
+      "epoch": 0.757916594566534,
+      "grad_norm": 0.445026159286499,
+      "learning_rate": 3.363036823819029e-05,
+      "loss": 0.9629,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7587817961585049,
+      "grad_norm": 0.4377189874649048,
+      "learning_rate": 3.340476106499023e-05,
+      "loss": 0.9837,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7596469977504758,
+      "grad_norm": 0.43655675649642944,
+      "learning_rate": 3.317976133949825e-05,
+      "loss": 0.9978,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7605121993424467,
+      "grad_norm": 0.44946399331092834,
+      "learning_rate": 3.2955371114046895e-05,
+      "loss": 0.9705,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7613774009344177,
+      "grad_norm": 0.3919507563114166,
+      "learning_rate": 3.2731592435408906e-05,
+      "loss": 0.9462,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7622426025263886,
+      "grad_norm": 0.4803646206855774,
+      "learning_rate": 3.250842734477904e-05,
+      "loss": 0.953,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7631078041183595,
+      "grad_norm": 0.4366481304168701,
+      "learning_rate": 3.22858778777551e-05,
+      "loss": 0.9663,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7639730057103306,
+      "grad_norm": 0.4419713616371155,
+      "learning_rate": 3.206394606431941e-05,
+      "loss": 0.9984,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7648382073023015,
+      "grad_norm": 0.44880411028862,
+      "learning_rate": 3.1842633928820574e-05,
+      "loss": 0.9361,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7657034088942724,
+      "grad_norm": 0.4589973986148834,
+      "learning_rate": 3.162194348995459e-05,
+      "loss": 0.9589,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7665686104862433,
+      "grad_norm": 0.46531566977500916,
+      "learning_rate": 3.1401876760746876e-05,
+      "loss": 0.9758,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7674338120782143,
+      "grad_norm": 0.4017270505428314,
+      "learning_rate": 3.1182435748533455e-05,
+      "loss": 0.9565,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7682990136701852,
+      "grad_norm": 0.4428539574146271,
+      "learning_rate": 3.096362245494311e-05,
+      "loss": 0.9959,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7691642152621561,
+      "grad_norm": 0.44016364216804504,
+      "learning_rate": 3.074543887587874e-05,
+      "loss": 0.9669,
+      "step": 4445
+    },
+    {
+      "epoch": 0.770029416854127,
+      "grad_norm": 0.454472154378891,
+      "learning_rate": 3.0527887001499364e-05,
+      "loss": 0.947,
+      "step": 4450
+    },
+    {
+      "epoch": 0.770894618446098,
+      "grad_norm": 0.4053809344768524,
+      "learning_rate": 3.0310968816201925e-05,
+      "loss": 0.9217,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7717598200380689,
+      "grad_norm": 0.43864405155181885,
+      "learning_rate": 3.0094686298603136e-05,
+      "loss": 0.9562,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7726250216300398,
+      "grad_norm": 0.4394964277744293,
+      "learning_rate": 2.9879041421521527e-05,
+      "loss": 0.9766,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7734902232220108,
+      "grad_norm": 0.44996631145477295,
+      "learning_rate": 2.966403615195934e-05,
+      "loss": 0.9347,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7743554248139817,
+      "grad_norm": 0.48114338517189026,
+      "learning_rate": 2.944967245108471e-05,
+      "loss": 0.9749,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7752206264059526,
+      "grad_norm": 0.45745283365249634,
+      "learning_rate": 2.923595227421364e-05,
+      "loss": 0.9799,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7760858279979235,
+      "grad_norm": 0.4538094103336334,
+      "learning_rate": 2.9022877570792305e-05,
+      "loss": 0.9462,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7769510295898945,
+      "grad_norm": 0.4058148264884949,
+      "learning_rate": 2.8810450284379153e-05,
+      "loss": 0.9927,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7778162311818654,
+      "grad_norm": 0.435544490814209,
+      "learning_rate": 2.8598672352627244e-05,
+      "loss": 0.9089,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7786814327738363,
+      "grad_norm": 0.4508376717567444,
+      "learning_rate": 2.838754570726655e-05,
+      "loss": 0.9786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7795466343658072,
+      "grad_norm": 0.44496816396713257,
+      "learning_rate": 2.8177072274086348e-05,
+      "loss": 0.9692,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7804118359577782,
+      "grad_norm": 0.5042269229888916,
+      "learning_rate": 2.7967253972917605e-05,
+      "loss": 0.9742,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7812770375497491,
+      "grad_norm": 0.44786500930786133,
+      "learning_rate": 2.7758092717615657e-05,
+      "loss": 0.9582,
+      "step": 4515
+    },
+    {
+      "epoch": 0.78214223914172,
+      "grad_norm": 0.4112664461135864,
+      "learning_rate": 2.754959041604239e-05,
+      "loss": 1.0003,
+      "step": 4520
+    },
+    {
+      "epoch": 0.783007440733691,
+      "grad_norm": 0.4549863636493683,
+      "learning_rate": 2.7341748970049196e-05,
+      "loss": 0.9559,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7838726423256619,
+      "grad_norm": 0.42742258310317993,
+      "learning_rate": 2.7134570275459435e-05,
+      "loss": 0.9488,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7847378439176328,
+      "grad_norm": 0.43447646498680115,
+      "learning_rate": 2.6928056222051202e-05,
+      "loss": 0.9612,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7856030455096037,
+      "grad_norm": 0.44213923811912537,
+      "learning_rate": 2.6722208693540052e-05,
+      "loss": 0.9417,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7864682471015747,
+      "grad_norm": 0.4459418058395386,
+      "learning_rate": 2.6517029567561836e-05,
+      "loss": 1.0021,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7873334486935456,
+      "grad_norm": 0.4390890598297119,
+      "learning_rate": 2.6312520715655685e-05,
+      "loss": 0.9761,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7881986502855165,
+      "grad_norm": 0.47070950269699097,
+      "learning_rate": 2.610868400324663e-05,
+      "loss": 0.956,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7890638518774875,
+      "grad_norm": 0.4373131990432739,
+      "learning_rate": 2.590552128962902e-05,
+      "loss": 0.9721,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7899290534694584,
+      "grad_norm": 0.42470088601112366,
+      "learning_rate": 2.5703034427949103e-05,
+      "loss": 0.9514,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7907942550614293,
+      "grad_norm": 0.45303651690483093,
+      "learning_rate": 2.5501225265188578e-05,
+      "loss": 1.0286,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7916594566534002,
+      "grad_norm": 0.4624740779399872,
+      "learning_rate": 2.5300095642147293e-05,
+      "loss": 0.9757,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7925246582453712,
+      "grad_norm": 0.4534415900707245,
+      "learning_rate": 2.5099647393426883e-05,
+      "loss": 1.0163,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7933898598373421,
+      "grad_norm": 0.43228623270988464,
+      "learning_rate": 2.489988234741376e-05,
+      "loss": 0.928,
+      "step": 4585
+    },
+    {
+      "epoch": 0.794255061429313,
+      "grad_norm": 0.4066708981990814,
+      "learning_rate": 2.4700802326262405e-05,
+      "loss": 0.9341,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7951202630212839,
+      "grad_norm": 0.45343056321144104,
+      "learning_rate": 2.450240914587907e-05,
+      "loss": 0.9669,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7959854646132549,
+      "grad_norm": 0.43527576327323914,
+      "learning_rate": 2.430470461590475e-05,
+      "loss": 0.9616,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7968506662052258,
+      "grad_norm": 0.44016727805137634,
+      "learning_rate": 2.4107690539699157e-05,
+      "loss": 0.9827,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7977158677971967,
+      "grad_norm": 0.4747732877731323,
+      "learning_rate": 2.3911368714323835e-05,
+      "loss": 0.955,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7985810693891677,
+      "grad_norm": 0.41703560948371887,
+      "learning_rate": 2.371574093052612e-05,
+      "loss": 0.9736,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7994462709811386,
+      "grad_norm": 0.40136978030204773,
+      "learning_rate": 2.3520808972722596e-05,
+      "loss": 0.9487,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8003114725731095,
+      "grad_norm": 0.46093153953552246,
+      "learning_rate": 2.3326574618982865e-05,
+      "loss": 0.9431,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8011766741650804,
+      "grad_norm": 0.459598183631897,
+      "learning_rate": 2.3133039641013366e-05,
+      "loss": 0.9502,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8020418757570514,
+      "grad_norm": 0.4129520058631897,
+      "learning_rate": 2.294020580414118e-05,
+      "loss": 0.9672,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8029070773490223,
+      "grad_norm": 0.460458904504776,
+      "learning_rate": 2.2748074867297918e-05,
+      "loss": 1.0069,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8037722789409932,
+      "grad_norm": 0.45732662081718445,
+      "learning_rate": 2.2556648583003746e-05,
+      "loss": 1.0105,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8046374805329641,
+      "grad_norm": 0.42087164521217346,
+      "learning_rate": 2.2365928697351314e-05,
+      "loss": 0.9292,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8055026821249351,
+      "grad_norm": 0.4139239490032196,
+      "learning_rate": 2.217591694998985e-05,
+      "loss": 0.9789,
+      "step": 4655
+    },
+    {
+      "epoch": 0.806367883716906,
+      "grad_norm": 0.4530125856399536,
+      "learning_rate": 2.198661507410935e-05,
+      "loss": 0.9588,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8072330853088769,
+      "grad_norm": 0.4521411955356598,
+      "learning_rate": 2.17980247964247e-05,
+      "loss": 0.9694,
+      "step": 4665
+    },
+    {
+      "epoch": 0.808098286900848,
+      "grad_norm": 0.41040509939193726,
+      "learning_rate": 2.1610147837159955e-05,
+      "loss": 0.9449,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8089634884928188,
+      "grad_norm": 0.45691537857055664,
+      "learning_rate": 2.1422985910032632e-05,
+      "loss": 0.9519,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8098286900847897,
+      "grad_norm": 0.4356061518192291,
+      "learning_rate": 2.1236540722238118e-05,
+      "loss": 0.9641,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8106938916767606,
+      "grad_norm": 0.44307219982147217,
+      "learning_rate": 2.105081397443406e-05,
+      "loss": 0.9275,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8115590932687317,
+      "grad_norm": 0.44193556904792786,
+      "learning_rate": 2.0865807360724832e-05,
+      "loss": 0.9502,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8124242948607026,
+      "grad_norm": 0.4376203119754791,
+      "learning_rate": 2.0681522568646162e-05,
+      "loss": 0.9811,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8132894964526735,
+      "grad_norm": 0.4987525939941406,
+      "learning_rate": 2.0497961279149658e-05,
+      "loss": 0.9999,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8141546980446444,
+      "grad_norm": 0.5082269310951233,
+      "learning_rate": 2.031512516658751e-05,
+      "loss": 0.9405,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8150198996366154,
+      "grad_norm": 0.46282055974006653,
+      "learning_rate": 2.0133015898697237e-05,
+      "loss": 0.9807,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8158851012285863,
+      "grad_norm": 0.4175664484500885,
+      "learning_rate": 1.9951635136586434e-05,
+      "loss": 0.9495,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8167503028205572,
+      "grad_norm": 0.4660712778568268,
+      "learning_rate": 1.977098453471764e-05,
+      "loss": 0.9512,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8176155044125282,
+      "grad_norm": 0.43538960814476013,
+      "learning_rate": 1.9591065740893245e-05,
+      "loss": 0.9572,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8184807060044991,
+      "grad_norm": 0.46293970942497253,
+      "learning_rate": 1.9411880396240435e-05,
+      "loss": 0.9913,
+      "step": 4730
+    },
+    {
+      "epoch": 0.81934590759647,
+      "grad_norm": 0.44890257716178894,
+      "learning_rate": 1.9233430135196362e-05,
+      "loss": 0.9729,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8202111091884409,
+      "grad_norm": 0.49698060750961304,
+      "learning_rate": 1.9055716585492922e-05,
+      "loss": 1.0063,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8210763107804119,
+      "grad_norm": 0.47737550735473633,
+      "learning_rate": 1.887874136814234e-05,
+      "loss": 0.988,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8219415123723828,
+      "grad_norm": 0.4616323709487915,
+      "learning_rate": 1.8702506097421914e-05,
+      "loss": 0.972,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8228067139643537,
+      "grad_norm": 0.41429662704467773,
+      "learning_rate": 1.8527012380859756e-05,
+      "loss": 0.9651,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8236719155563246,
+      "grad_norm": 0.47355395555496216,
+      "learning_rate": 1.8352261819219672e-05,
+      "loss": 0.957,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8245371171482956,
+      "grad_norm": 0.4533310830593109,
+      "learning_rate": 1.8178256006486993e-05,
+      "loss": 1.0007,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8254023187402665,
+      "grad_norm": 0.4279444217681885,
+      "learning_rate": 1.8004996529853756e-05,
+      "loss": 0.9577,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8262675203322374,
+      "grad_norm": 0.45256754755973816,
+      "learning_rate": 1.7832484969704188e-05,
+      "loss": 1.0217,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8271327219242084,
+      "grad_norm": 0.46098792552948,
+      "learning_rate": 1.7660722899600624e-05,
+      "loss": 0.9831,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8279979235161793,
+      "grad_norm": 0.4408135712146759,
+      "learning_rate": 1.7489711886268712e-05,
+      "loss": 0.9692,
+      "step": 4785
+    },
+    {
+      "epoch": 0.8288631251081502,
+      "grad_norm": 0.4415886700153351,
+      "learning_rate": 1.731945348958356e-05,
+      "loss": 0.9818,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8297283267001211,
+      "grad_norm": 0.4410308301448822,
+      "learning_rate": 1.7149949262555054e-05,
+      "loss": 1.0049,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8305935282920921,
+      "grad_norm": 0.4444870352745056,
+      "learning_rate": 1.6981200751314165e-05,
+      "loss": 0.9561,
+      "step": 4800
+    },
+    {
+      "epoch": 0.831458729884063,
+      "grad_norm": 0.4517926871776581,
+      "learning_rate": 1.6813209495098447e-05,
+      "loss": 0.9454,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8323239314760339,
+      "grad_norm": 0.4491353929042816,
+      "learning_rate": 1.6645977026238235e-05,
+      "loss": 0.9803,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8331891330680048,
+      "grad_norm": 0.42471835017204285,
+      "learning_rate": 1.647950487014257e-05,
+      "loss": 0.938,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8340543346599758,
+      "grad_norm": 0.4573848247528076,
+      "learning_rate": 1.631379454528531e-05,
+      "loss": 0.906,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8349195362519467,
+      "grad_norm": 0.48127248883247375,
+      "learning_rate": 1.6148847563191294e-05,
+      "loss": 0.9694,
+      "step": 4825
+    },
+    {
+      "epoch": 0.8357847378439176,
+      "grad_norm": 0.42135560512542725,
+      "learning_rate": 1.598466542842253e-05,
+      "loss": 0.9399,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8366499394358886,
+      "grad_norm": 0.4253736734390259,
+      "learning_rate": 1.5821249638564462e-05,
+      "loss": 0.9661,
+      "step": 4835
+    },
+    {
+      "epoch": 0.8375151410278595,
+      "grad_norm": 0.4118526577949524,
+      "learning_rate": 1.565860168421235e-05,
+      "loss": 0.9096,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8383803426198304,
+      "grad_norm": 0.41610080003738403,
+      "learning_rate": 1.549672304895764e-05,
+      "loss": 0.9544,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8392455442118013,
+      "grad_norm": 0.44422540068626404,
+      "learning_rate": 1.5335615209374453e-05,
+      "loss": 0.9124,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8401107458037723,
+      "grad_norm": 0.5014816522598267,
+      "learning_rate": 1.51752796350061e-05,
+      "loss": 0.977,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8409759473957432,
+      "grad_norm": 0.44182294607162476,
+      "learning_rate": 1.501571778835168e-05,
+      "loss": 0.9411,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8418411489877141,
+      "grad_norm": 0.4564247131347656,
+      "learning_rate": 1.4856931124852746e-05,
+      "loss": 0.9687,
+      "step": 4865
+    },
+    {
+      "epoch": 0.842706350579685,
+      "grad_norm": 0.4558230936527252,
+      "learning_rate": 1.4698921092880036e-05,
+      "loss": 0.9988,
+      "step": 4870
+    },
+    {
+      "epoch": 0.843571552171656,
+      "grad_norm": 0.4334022104740143,
+      "learning_rate": 1.4541689133720238e-05,
+      "loss": 0.9319,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8444367537636269,
+      "grad_norm": 0.45239973068237305,
+      "learning_rate": 1.4385236681562863e-05,
+      "loss": 0.9476,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8453019553555978,
+      "grad_norm": 0.4498237073421478,
+      "learning_rate": 1.4229565163487157e-05,
+      "loss": 0.958,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8461671569475688,
+      "grad_norm": 0.47215980291366577,
+      "learning_rate": 1.4074675999449093e-05,
+      "loss": 0.9464,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8470323585395397,
+      "grad_norm": 0.48363256454467773,
+      "learning_rate": 1.3920570602268413e-05,
+      "loss": 0.9376,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8478975601315106,
+      "grad_norm": 0.44979047775268555,
+      "learning_rate": 1.3767250377615703e-05,
+      "loss": 0.9764,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8487627617234815,
+      "grad_norm": 0.4314928352832794,
+      "learning_rate": 1.3614716723999644e-05,
+      "loss": 0.9463,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8496279633154525,
+      "grad_norm": 0.44183149933815,
+      "learning_rate": 1.3462971032754246e-05,
+      "loss": 0.9331,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8504931649074234,
+      "grad_norm": 0.42287370562553406,
+      "learning_rate": 1.331201468802602e-05,
+      "loss": 0.9585,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8513583664993943,
+      "grad_norm": 0.4583037197589874,
+      "learning_rate": 1.3161849066761611e-05,
+      "loss": 0.93,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8522235680913652,
+      "grad_norm": 0.3929138779640198,
+      "learning_rate": 1.3012475538694934e-05,
+      "loss": 0.9642,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8530887696833362,
+      "grad_norm": 0.44397783279418945,
+      "learning_rate": 1.2863895466334996e-05,
+      "loss": 0.996,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8539539712753071,
+      "grad_norm": 0.446635365486145,
+      "learning_rate": 1.2716110204953124e-05,
+      "loss": 0.9543,
+      "step": 4935
+    },
+    {
+      "epoch": 0.854819172867278,
+      "grad_norm": 0.4561355710029602,
+      "learning_rate": 1.2569121102570969e-05,
+      "loss": 0.9529,
+      "step": 4940
+    },
+    {
+      "epoch": 0.855684374459249,
+      "grad_norm": 0.4427773654460907,
+      "learning_rate": 1.2422929499947933e-05,
+      "loss": 0.9941,
+      "step": 4945
+    },
+    {
+      "epoch": 0.85654957605122,
+      "grad_norm": 0.4415104389190674,
+      "learning_rate": 1.2277536730568972e-05,
+      "loss": 0.9566,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8574147776431909,
+      "grad_norm": 0.45385250449180603,
+      "learning_rate": 1.2132944120632661e-05,
+      "loss": 0.9628,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8582799792351617,
+      "grad_norm": 0.42043399810791016,
+      "learning_rate": 1.1989152989038743e-05,
+      "loss": 0.997,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8591451808271328,
+      "grad_norm": 0.4300490617752075,
+      "learning_rate": 1.1846164647376457e-05,
+      "loss": 0.9709,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8600103824191037,
+      "grad_norm": 0.41449466347694397,
+      "learning_rate": 1.170398039991224e-05,
+      "loss": 0.972,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8608755840110746,
+      "grad_norm": 0.39637109637260437,
+      "learning_rate": 1.1562601543578167e-05,
+      "loss": 0.9513,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8617407856030455,
+      "grad_norm": 0.43413782119750977,
+      "learning_rate": 1.1422029367959753e-05,
+      "loss": 0.9851,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8626059871950165,
+      "grad_norm": 0.4083414375782013,
+      "learning_rate": 1.1282265155284566e-05,
+      "loss": 0.9846,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8634711887869874,
+      "grad_norm": 0.4629712700843811,
+      "learning_rate": 1.1143310180410238e-05,
+      "loss": 0.9431,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8643363903789583,
+      "grad_norm": 0.43061643838882446,
+      "learning_rate": 1.1005165710812982e-05,
+      "loss": 0.9519,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8652015919709293,
+      "grad_norm": 0.4178582429885864,
+      "learning_rate": 1.086783300657599e-05,
+      "loss": 0.9638,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8660667935629002,
+      "grad_norm": 0.41832438111305237,
+      "learning_rate": 1.0731313320377945e-05,
+      "loss": 0.9772,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8669319951548711,
+      "grad_norm": 0.42187732458114624,
+      "learning_rate": 1.059560789748163e-05,
+      "loss": 0.9524,
+      "step": 5010
+    },
+    {
+      "epoch": 0.867797196746842,
+      "grad_norm": 0.44064244627952576,
+      "learning_rate": 1.0460717975722446e-05,
+      "loss": 0.9601,
+      "step": 5015
+    },
+    {
+      "epoch": 0.868662398338813,
+      "grad_norm": 0.4325951635837555,
+      "learning_rate": 1.0326644785497309e-05,
+      "loss": 0.976,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8695275999307839,
+      "grad_norm": 0.42271390557289124,
+      "learning_rate": 1.0193389549753275e-05,
+      "loss": 0.9519,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8703928015227548,
+      "grad_norm": 0.4525720179080963,
+      "learning_rate": 1.0060953483976454e-05,
+      "loss": 0.9833,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8712580031147257,
+      "grad_norm": 0.44258391857147217,
+      "learning_rate": 9.929337796180904e-06,
+      "loss": 0.984,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8721232047066967,
+      "grad_norm": 0.4149342477321625,
+      "learning_rate": 9.798543686897588e-06,
+      "loss": 0.9293,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8729884062986676,
+      "grad_norm": 0.41939181089401245,
+      "learning_rate": 9.668572349163507e-06,
+      "loss": 0.979,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8738536078906385,
+      "grad_norm": 0.44124025106430054,
+      "learning_rate": 9.539424968510702e-06,
+      "loss": 0.9822,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8747188094826095,
+      "grad_norm": 0.4244748651981354,
+      "learning_rate": 9.411102722955534e-06,
+      "loss": 1.0002,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8755840110745804,
+      "grad_norm": 0.43866798281669617,
+      "learning_rate": 9.283606782987887e-06,
+      "loss": 0.9615,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8764492126665513,
+      "grad_norm": 0.454874724149704,
+      "learning_rate": 9.156938311560525e-06,
+      "loss": 0.946,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8773144142585222,
+      "grad_norm": 0.44537651538848877,
+      "learning_rate": 9.031098464078457e-06,
+      "loss": 0.9726,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8781796158504932,
+      "grad_norm": 0.44006219506263733,
+      "learning_rate": 8.906088388388411e-06,
+      "loss": 0.9902,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8790448174424641,
+      "grad_norm": 0.4104478061199188,
+      "learning_rate": 8.781909224768337e-06,
+      "loss": 0.9935,
+      "step": 5080
+    },
+    {
+      "epoch": 0.879910019034435,
+      "grad_norm": 0.415463387966156,
+      "learning_rate": 8.658562105917079e-06,
+      "loss": 0.9851,
+      "step": 5085
+    },
+    {
+      "epoch": 0.880775220626406,
+      "grad_norm": 0.4624484181404114,
+      "learning_rate": 8.536048156943932e-06,
+      "loss": 0.9356,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8816404222183769,
+      "grad_norm": 0.4278486669063568,
+      "learning_rate": 8.414368495358538e-06,
+      "loss": 0.9641,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8825056238103478,
+      "grad_norm": 0.4188804030418396,
+      "learning_rate": 8.293524231060468e-06,
+      "loss": 0.9955,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8833708254023187,
+      "grad_norm": 0.4604167342185974,
+      "learning_rate": 8.173516466329345e-06,
+      "loss": 0.984,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8842360269942897,
+      "grad_norm": 0.40931734442710876,
+      "learning_rate": 8.054346295814564e-06,
+      "loss": 0.9508,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8851012285862606,
+      "grad_norm": 0.4736385643482208,
+      "learning_rate": 7.936014806525538e-06,
+      "loss": 0.9467,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8859664301782315,
+      "grad_norm": 0.42703744769096375,
+      "learning_rate": 7.818523077821527e-06,
+      "loss": 0.9343,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8868316317702024,
+      "grad_norm": 0.4844820499420166,
+      "learning_rate": 7.701872181401992e-06,
+      "loss": 1.0121,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8876968333621734,
+      "grad_norm": 0.42825546860694885,
+      "learning_rate": 7.586063181296799e-06,
+      "loss": 0.9751,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8885620349541443,
+      "grad_norm": 0.41822490096092224,
+      "learning_rate": 7.471097133856353e-06,
+      "loss": 0.9261,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8894272365461152,
+      "grad_norm": 0.47537797689437866,
+      "learning_rate": 7.356975087742202e-06,
+      "loss": 0.9766,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8902924381380862,
+      "grad_norm": 0.45567959547042847,
+      "learning_rate": 7.243698083917194e-06,
+      "loss": 0.9857,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8911576397300571,
+      "grad_norm": 0.43980252742767334,
+      "learning_rate": 7.131267155636323e-06,
+      "loss": 0.9359,
+      "step": 5150
+    },
+    {
+      "epoch": 0.892022841322028,
+      "grad_norm": 0.4581001400947571,
+      "learning_rate": 7.01968332843691e-06,
+      "loss": 0.9722,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8928880429139989,
+      "grad_norm": 0.4442357122898102,
+      "learning_rate": 6.908947620129624e-06,
+      "loss": 1.0104,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8937532445059699,
+      "grad_norm": 0.43554165959358215,
+      "learning_rate": 6.799061040788934e-06,
+      "loss": 0.9082,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8946184460979408,
+      "grad_norm": 0.6001029014587402,
+      "learning_rate": 6.690024592744027e-06,
+      "loss": 0.9798,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8954836476899117,
+      "grad_norm": 0.44273775815963745,
+      "learning_rate": 6.581839270569601e-06,
+      "loss": 0.9635,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8963488492818826,
+      "grad_norm": 0.4063413143157959,
+      "learning_rate": 6.474506061076824e-06,
+      "loss": 0.9386,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8972140508738536,
+      "grad_norm": 0.43752190470695496,
+      "learning_rate": 6.368025943304367e-06,
+      "loss": 0.9594,
+      "step": 5185
+    },
+    {
+      "epoch": 0.8980792524658245,
+      "grad_norm": 0.42202919721603394,
+      "learning_rate": 6.262399888509351e-06,
+      "loss": 0.9876,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8989444540577954,
+      "grad_norm": 0.43374916911125183,
+      "learning_rate": 6.157628860158671e-06,
+      "loss": 0.9623,
+      "step": 5195
+    },
+    {
+      "epoch": 0.8998096556497664,
+      "grad_norm": 0.4456869661808014,
+      "learning_rate": 6.053713813920036e-06,
+      "loss": 0.9817,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9006748572417373,
+      "grad_norm": 0.4352042078971863,
+      "learning_rate": 5.950655697653362e-06,
+      "loss": 0.9868,
+      "step": 5205
+    },
+    {
+      "epoch": 0.9015400588337082,
+      "grad_norm": 0.4586144685745239,
+      "learning_rate": 5.848455451402057e-06,
+      "loss": 0.9918,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9024052604256791,
+      "grad_norm": 0.4337520897388458,
+      "learning_rate": 5.747114007384513e-06,
+      "loss": 0.9956,
+      "step": 5215
+    },
+    {
+      "epoch": 0.9032704620176502,
+      "grad_norm": 0.45494288206100464,
+      "learning_rate": 5.646632289985543e-06,
+      "loss": 0.9722,
+      "step": 5220
+    },
+    {
+      "epoch": 0.904135663609621,
+      "grad_norm": 0.4270175099372864,
+      "learning_rate": 5.547011215747977e-06,
+      "loss": 0.93,
+      "step": 5225
+    },
+    {
+      "epoch": 0.905000865201592,
+      "grad_norm": 0.4172538220882416,
+      "learning_rate": 5.448251693364304e-06,
+      "loss": 0.959,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9058660667935629,
+      "grad_norm": 0.4290398359298706,
+      "learning_rate": 5.350354623668374e-06,
+      "loss": 0.9706,
+      "step": 5235
+    },
+    {
+      "epoch": 0.9067312683855339,
+      "grad_norm": 0.4319567084312439,
+      "learning_rate": 5.253320899627179e-06,
+      "loss": 0.9898,
+      "step": 5240
+    },
+    {
+      "epoch": 0.9075964699775048,
+      "grad_norm": 0.42448529601097107,
+      "learning_rate": 5.157151406332739e-06,
+      "loss": 0.9293,
+      "step": 5245
+    },
+    {
+      "epoch": 0.9084616715694757,
+      "grad_norm": 0.4146623909473419,
+      "learning_rate": 5.0618470209939596e-06,
+      "loss": 0.9676,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9093268731614467,
+      "grad_norm": 0.45237162709236145,
+      "learning_rate": 4.9674086129287144e-06,
+      "loss": 0.94,
+      "step": 5255
+    },
+    {
+      "epoch": 0.9101920747534176,
+      "grad_norm": 0.44053730368614197,
+      "learning_rate": 4.8738370435558514e-06,
+      "loss": 0.9613,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9110572763453885,
+      "grad_norm": 0.424537718296051,
+      "learning_rate": 4.7811331663873745e-06,
+      "loss": 0.9724,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9119224779373594,
+      "grad_norm": 0.4328090250492096,
+      "learning_rate": 4.689297827020634e-06,
+      "loss": 0.9832,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9127876795293304,
+      "grad_norm": 0.4516524374485016,
+      "learning_rate": 4.598331863130611e-06,
+      "loss": 0.9805,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9136528811213013,
+      "grad_norm": 0.45256564021110535,
+      "learning_rate": 4.508236104462338e-06,
+      "loss": 0.9512,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9145180827132722,
+      "grad_norm": 0.44559070467948914,
+      "learning_rate": 4.4190113728231895e-06,
+      "loss": 0.9284,
+      "step": 5285
+    },
+    {
+      "epoch": 0.9153832843052431,
+      "grad_norm": 0.3941524624824524,
+      "learning_rate": 4.330658482075589e-06,
+      "loss": 0.9409,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9162484858972141,
+      "grad_norm": 0.4107651710510254,
+      "learning_rate": 4.243178238129386e-06,
+      "loss": 0.9803,
+      "step": 5295
+    },
+    {
+      "epoch": 0.917113687489185,
+      "grad_norm": 0.4321724772453308,
+      "learning_rate": 4.15657143893462e-06,
+      "loss": 0.9904,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9179788890811559,
+      "grad_norm": 0.49873462319374084,
+      "learning_rate": 4.070838874474225e-06,
+      "loss": 1.0426,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9188440906731269,
+      "grad_norm": 0.4886455833911896,
+      "learning_rate": 3.985981326756793e-06,
+      "loss": 1.0217,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9197092922650978,
+      "grad_norm": 0.42019376158714294,
+      "learning_rate": 3.901999569809501e-06,
+      "loss": 0.982,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9205744938570687,
+      "grad_norm": 0.4222065210342407,
+      "learning_rate": 3.818894369670933e-06,
+      "loss": 0.9921,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9214396954490396,
+      "grad_norm": 0.4316330850124359,
+      "learning_rate": 3.736666484384288e-06,
+      "loss": 1.0336,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9223048970410106,
+      "grad_norm": 0.41588303446769714,
+      "learning_rate": 3.6553166639902048e-06,
+      "loss": 0.9715,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9231700986329815,
+      "grad_norm": 0.4579836130142212,
+      "learning_rate": 3.5748456505201975e-06,
+      "loss": 0.9602,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9240353002249524,
+      "grad_norm": 0.39906853437423706,
+      "learning_rate": 3.4952541779896376e-06,
+      "loss": 0.9555,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9249005018169233,
+      "grad_norm": 0.43084821105003357,
+      "learning_rate": 3.4165429723912677e-06,
+      "loss": 0.9961,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9257657034088943,
+      "grad_norm": 0.4219900369644165,
+      "learning_rate": 3.3387127516884443e-06,
+      "loss": 0.9769,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9266309050008652,
+      "grad_norm": 0.4021409749984741,
+      "learning_rate": 3.261764225808628e-06,
+      "loss": 0.9609,
+      "step": 5355
+    },
+    {
+      "epoch": 0.9274961065928361,
+      "grad_norm": 0.4550817608833313,
+      "learning_rate": 3.185698096636902e-06,
+      "loss": 0.9429,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9283613081848071,
+      "grad_norm": 0.432493656873703,
+      "learning_rate": 3.1105150580096108e-06,
+      "loss": 0.9509,
+      "step": 5365
+    },
+    {
+      "epoch": 0.929226509776778,
+      "grad_norm": 0.43657249212265015,
+      "learning_rate": 3.036215795707975e-06,
+      "loss": 0.9454,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9300917113687489,
+      "grad_norm": 0.43158912658691406,
+      "learning_rate": 2.9628009874518304e-06,
+      "loss": 0.9408,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9309569129607198,
+      "grad_norm": 0.4231918156147003,
+      "learning_rate": 2.8902713028935545e-06,
+      "loss": 0.9776,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9318221145526908,
+      "grad_norm": 0.4418452978134155,
+      "learning_rate": 2.818627403611818e-06,
+      "loss": 0.9733,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9326873161446617,
+      "grad_norm": 0.45113956928253174,
+      "learning_rate": 2.7478699431056186e-06,
+      "loss": 0.961,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9335525177366326,
+      "grad_norm": 0.42509666085243225,
+      "learning_rate": 2.677999566788325e-06,
+      "loss": 0.9885,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9344177193286035,
+      "grad_norm": 0.42578408122062683,
+      "learning_rate": 2.6090169119817764e-06,
+      "loss": 0.9629,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9352829209205745,
+      "grad_norm": 0.5043439865112305,
+      "learning_rate": 2.5409226079104563e-06,
+      "loss": 0.9571,
+      "step": 5405
+    },
+    {
+      "epoch": 0.9361481225125454,
+      "grad_norm": 0.4643230438232422,
+      "learning_rate": 2.473717275695797e-06,
+      "loss": 0.9987,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9370133241045163,
+      "grad_norm": 0.4713646471500397,
+      "learning_rate": 2.40740152835045e-06,
+      "loss": 0.9663,
+      "step": 5415
+    },
+    {
+      "epoch": 0.9378785256964873,
+      "grad_norm": 0.5420528054237366,
+      "learning_rate": 2.3419759707727584e-06,
+      "loss": 0.957,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9387437272884582,
+      "grad_norm": 0.43974393606185913,
+      "learning_rate": 2.2774411997411814e-06,
+      "loss": 0.9602,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9396089288804291,
+      "grad_norm": 0.4206051528453827,
+      "learning_rate": 2.21379780390889e-06,
+      "loss": 0.9466,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9404741304724,
+      "grad_norm": 0.42323896288871765,
+      "learning_rate": 2.1510463637984035e-06,
+      "loss": 0.9683,
+      "step": 5435
+    },
+    {
+      "epoch": 0.941339332064371,
+      "grad_norm": 0.43212366104125977,
+      "learning_rate": 2.0891874517962594e-06,
+      "loss": 0.9524,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9422045336563419,
+      "grad_norm": 0.46724721789360046,
+      "learning_rate": 2.0282216321478085e-06,
+      "loss": 0.9736,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9430697352483128,
+      "grad_norm": 0.44604164361953735,
+      "learning_rate": 1.9681494609520736e-06,
+      "loss": 0.9384,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9439349368402837,
+      "grad_norm": 0.4035724997520447,
+      "learning_rate": 1.9089714861566964e-06,
+      "loss": 1.0091,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9448001384322547,
+      "grad_norm": 0.5425649285316467,
+      "learning_rate": 1.8506882475528565e-06,
+      "loss": 0.9617,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9456653400242256,
+      "grad_norm": 0.46280762553215027,
+      "learning_rate": 1.7933002767704931e-06,
+      "loss": 0.9462,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9465305416161965,
+      "grad_norm": 0.45342132449150085,
+      "learning_rate": 1.7368080972732792e-06,
+      "loss": 0.9315,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9473957432081676,
+      "grad_norm": 0.43039941787719727,
+      "learning_rate": 1.6812122243540119e-06,
+      "loss": 1.0075,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9482609448001385,
+      "grad_norm": 0.4240161180496216,
+      "learning_rate": 1.626513165129795e-06,
+      "loss": 0.9827,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9491261463921093,
+      "grad_norm": 0.4125286340713501,
+      "learning_rate": 1.5727114185374758e-06,
+      "loss": 0.9314,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9499913479840802,
+      "grad_norm": 0.4606507122516632,
+      "learning_rate": 1.5198074753290714e-06,
+      "loss": 0.9825,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9508565495760513,
+      "grad_norm": 0.408927857875824,
+      "learning_rate": 1.4678018180672715e-06,
+      "loss": 0.9769,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9517217511680222,
+      "grad_norm": 0.4514823257923126,
+      "learning_rate": 1.41669492112112e-06,
+      "loss": 0.9654,
+      "step": 5500
+    },
+    {
+      "epoch": 0.952586952759993,
+      "grad_norm": 0.46445760130882263,
+      "learning_rate": 1.3664872506615523e-06,
+      "loss": 0.9393,
+      "step": 5505
+    },
+    {
+      "epoch": 0.953452154351964,
+      "grad_norm": 0.4294278621673584,
+      "learning_rate": 1.3171792646572978e-06,
+      "loss": 0.9528,
+      "step": 5510
+    },
+    {
+      "epoch": 0.954317355943935,
+      "grad_norm": 0.3879070580005646,
+      "learning_rate": 1.2687714128705397e-06,
+      "loss": 0.8864,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9551825575359059,
+      "grad_norm": 0.4252847135066986,
+      "learning_rate": 1.2212641368529842e-06,
+      "loss": 0.9317,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9560477591278768,
+      "grad_norm": 0.4258163273334503,
+      "learning_rate": 1.1746578699416754e-06,
+      "loss": 0.9282,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9569129607198478,
+      "grad_norm": 0.44812753796577454,
+      "learning_rate": 1.128953037255165e-06,
+      "loss": 0.9677,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9577781623118187,
+      "grad_norm": 0.40613648295402527,
+      "learning_rate": 1.0841500556895478e-06,
+      "loss": 0.9674,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9586433639037896,
+      "grad_norm": 0.4718643128871918,
+      "learning_rate": 1.0402493339147112e-06,
+      "loss": 0.9724,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9595085654957605,
+      "grad_norm": 0.4308999180793762,
+      "learning_rate": 9.972512723705807e-07,
+      "loss": 0.9625,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9603737670877315,
+      "grad_norm": 0.4665636420249939,
+      "learning_rate": 9.551562632634792e-07,
+      "loss": 0.9491,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9612389686797024,
+      "grad_norm": 0.4143814742565155,
+      "learning_rate": 9.13964690562552e-07,
+      "loss": 0.9403,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9621041702716733,
+      "grad_norm": 0.4198594391345978,
+      "learning_rate": 8.736769299962588e-07,
+      "loss": 0.9819,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9629693718636442,
+      "grad_norm": 0.42377784848213196,
+      "learning_rate": 8.342933490489535e-07,
+      "loss": 0.931,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9638345734556152,
+      "grad_norm": 0.4581563472747803,
+      "learning_rate": 7.958143069575097e-07,
+      "loss": 0.9698,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9646997750475861,
+      "grad_norm": 0.4543211758136749,
+      "learning_rate": 7.582401547080564e-07,
+      "loss": 0.954,
+      "step": 5575
+    },
+    {
+      "epoch": 0.965564976639557,
+      "grad_norm": 0.4556075930595398,
+      "learning_rate": 7.215712350328141e-07,
+      "loss": 0.9849,
+      "step": 5580
+    },
+    {
+      "epoch": 0.966430178231528,
+      "grad_norm": 0.43737539649009705,
+      "learning_rate": 6.85807882406908e-07,
+      "loss": 0.9737,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9672953798234989,
+      "grad_norm": 0.46313777565956116,
+      "learning_rate": 6.509504230453378e-07,
+      "loss": 0.9459,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9681605814154698,
+      "grad_norm": 0.45123517513275146,
+      "learning_rate": 6.169991749000237e-07,
+      "loss": 0.9601,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9690257830074407,
+      "grad_norm": 0.4670114517211914,
+      "learning_rate": 5.839544476568981e-07,
+      "loss": 0.9651,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9698909845994117,
+      "grad_norm": 0.44797414541244507,
+      "learning_rate": 5.518165427330413e-07,
+      "loss": 0.9654,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9707561861913826,
+      "grad_norm": 0.4370104968547821,
+      "learning_rate": 5.205857532740054e-07,
+      "loss": 0.9191,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9716213877833535,
+      "grad_norm": 0.4665054678916931,
+      "learning_rate": 4.902623641510617e-07,
+      "loss": 0.9943,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9724865893753245,
+      "grad_norm": 0.41058677434921265,
+      "learning_rate": 4.608466519586574e-07,
+      "loss": 0.9637,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9733517909672954,
+      "grad_norm": 0.4567413926124573,
+      "learning_rate": 4.323388850118848e-07,
+      "loss": 0.9586,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9742169925592663,
+      "grad_norm": 0.3996705710887909,
+      "learning_rate": 4.0473932334401665e-07,
+      "loss": 0.9494,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9750821941512372,
+      "grad_norm": 0.4506674110889435,
+      "learning_rate": 3.780482187041079e-07,
+      "loss": 0.9389,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9759473957432082,
+      "grad_norm": 0.4265097677707672,
+      "learning_rate": 3.5226581455480857e-07,
+      "loss": 1.0264,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9768125973351791,
+      "grad_norm": 0.44127118587493896,
+      "learning_rate": 3.273923460699879e-07,
+      "loss": 0.9439,
+      "step": 5645
+    },
+    {
+      "epoch": 0.97767779892715,
+      "grad_norm": 0.4313807785511017,
+      "learning_rate": 3.0342804013271386e-07,
+      "loss": 0.924,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9785430005191209,
+      "grad_norm": 0.4515504539012909,
+      "learning_rate": 2.8037311533312126e-07,
+      "loss": 0.944,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9794082021110919,
+      "grad_norm": 0.4783179461956024,
+      "learning_rate": 2.5822778196645803e-07,
+      "loss": 0.9624,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9802734037030628,
+      "grad_norm": 0.43046215176582336,
+      "learning_rate": 2.3699224203111992e-07,
+      "loss": 0.9754,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9811386052950337,
+      "grad_norm": 0.4199523627758026,
+      "learning_rate": 2.1666668922682987e-07,
+      "loss": 0.9489,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9820038068870047,
+      "grad_norm": 0.4476526379585266,
+      "learning_rate": 1.97251308952906e-07,
+      "loss": 0.951,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9828690084789756,
+      "grad_norm": 0.43268635869026184,
+      "learning_rate": 1.7874627830650748e-07,
+      "loss": 0.9758,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9837342100709465,
+      "grad_norm": 0.4126809537410736,
+      "learning_rate": 1.6115176608104688e-07,
+      "loss": 0.94,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9845994116629174,
+      "grad_norm": 0.4443066120147705,
+      "learning_rate": 1.4446793276468029e-07,
+      "loss": 0.9786,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9854646132548884,
+      "grad_norm": 0.43723779916763306,
+      "learning_rate": 1.2869493053880855e-07,
+      "loss": 0.9376,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9863298148468593,
+      "grad_norm": 0.4405401051044464,
+      "learning_rate": 1.1383290327666718e-07,
+      "loss": 0.9677,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9871950164388302,
+      "grad_norm": 0.4371713399887085,
+      "learning_rate": 9.988198654209413e-08,
+      "loss": 0.9591,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9880602180308011,
+      "grad_norm": 0.4550669491291046,
+      "learning_rate": 8.684230758818634e-08,
+      "loss": 1.0056,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9889254196227721,
+      "grad_norm": 0.42430931329727173,
+      "learning_rate": 7.471398535625618e-08,
+      "loss": 0.9978,
+      "step": 5715
+    },
+    {
+      "epoch": 0.989790621214743,
+      "grad_norm": 0.4184456467628479,
+      "learning_rate": 6.349713047464345e-08,
+      "loss": 0.9465,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9906558228067139,
+      "grad_norm": 0.43265965580940247,
+      "learning_rate": 5.319184525774956e-08,
+      "loss": 1.0002,
+      "step": 5725
+    },
+    {
+      "epoch": 0.991521024398685,
+      "grad_norm": 0.47475698590278625,
+      "learning_rate": 4.37982237051271e-08,
+      "loss": 0.9465,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9923862259906558,
+      "grad_norm": 0.4656958281993866,
+      "learning_rate": 3.531635150059165e-08,
+      "loss": 0.9793,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9932514275826267,
+      "grad_norm": 0.4549233913421631,
+      "learning_rate": 2.7746306011433576e-08,
+      "loss": 0.9444,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9941166291745976,
+      "grad_norm": 0.4547722041606903,
+      "learning_rate": 2.1088156287729644e-08,
+      "loss": 0.9762,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9949818307665687,
+      "grad_norm": 0.4455709457397461,
+      "learning_rate": 1.5341963061743515e-08,
+      "loss": 0.9767,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9958470323585396,
+      "grad_norm": 0.43941980600357056,
+      "learning_rate": 1.0507778747281815e-08,
+      "loss": 0.9707,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9967122339505105,
+      "grad_norm": 0.46377772092819214,
+      "learning_rate": 6.5856474393166666e-09,
+      "loss": 0.9928,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9975774355424813,
+      "grad_norm": 0.4075023829936981,
+      "learning_rate": 3.5756049135304835e-09,
+      "loss": 0.9541,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9984426371344524,
+      "grad_norm": 0.4415639638900757,
+      "learning_rate": 1.4776786259718123e-09,
+      "loss": 0.9693,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9993078387264233,
+      "grad_norm": 0.43806275725364685,
+      "learning_rate": 2.9188771285548754e-10,
+      "loss": 0.8886,
+      "step": 5775
+    },
+    {
+      "epoch": 1.0,
+      "step": 5779,
+      "total_flos": 1.1487027058043781e+19,
+      "train_loss": 0.0,
+      "train_runtime": 0.0143,
+      "train_samples_per_second": 9715326.166,
+      "train_steps_per_second": 404831.526
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 5779,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1487027058043781e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}