diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.1549566891241578,
+  "epoch": 1.539942252165544,
   "eval_steps": 500,
-  "global_step": 30000,
+  "global_step": 40000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -42007,6 +42007,14006 @@
       "learning_rate": 7.598194135786166e-05,
       "loss": 0.8086,
       "step": 30000
+    },
+    {
+      "epoch": 1.1551491819056785,
+      "grad_norm": 1.5883291959762573,
+      "learning_rate": 7.59525880175407e-05,
+      "loss": 0.7994,
+      "step": 30005
+    },
+    {
+      "epoch": 1.1553416746871992,
+      "grad_norm": 0.8642548322677612,
+      "learning_rate": 7.59232368761963e-05,
+      "loss": 0.8237,
+      "step": 30010
+    },
+    {
+      "epoch": 1.1555341674687198,
+      "grad_norm": 1.0051319599151611,
+      "learning_rate": 7.589388793651239e-05,
+      "loss": 0.8595,
+      "step": 30015
+    },
+    {
+      "epoch": 1.1557266602502407,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.586454120117271e-05,
+      "loss": 0.8389,
+      "step": 30020
+    },
+    {
+      "epoch": 1.1559191530317614,
+      "grad_norm": 1.0464012622833252,
+      "learning_rate": 7.583519667286088e-05,
+      "loss": 0.9032,
+      "step": 30025
+    },
+    {
+      "epoch": 1.156111645813282,
+      "grad_norm": 1.8586502075195312,
+      "learning_rate": 7.580585435426024e-05,
+      "loss": 0.8104,
+      "step": 30030
+    },
+    {
+      "epoch": 1.1563041385948027,
+      "grad_norm": 1.273821234703064,
+      "learning_rate": 7.577651424805392e-05,
+      "loss": 0.7211,
+      "step": 30035
+    },
+    {
+      "epoch": 1.1564966313763234,
+      "grad_norm": 1.5293164253234863,
+      "learning_rate": 7.574717635692492e-05,
+      "loss": 0.8558,
+      "step": 30040
+    },
+    {
+      "epoch": 1.156689124157844,
+      "grad_norm": 1.440649390220642,
+      "learning_rate": 7.5717840683556e-05,
+      "loss": 0.816,
+      "step": 30045
+    },
+    {
+      "epoch": 1.1568816169393648,
+      "grad_norm": 2.2141449451446533,
+      "learning_rate": 7.568850723062967e-05,
+      "loss": 0.8884,
+      "step": 30050
+    },
+    {
+      "epoch": 1.1570741097208854,
+      "grad_norm": 1.564818024635315,
+      "learning_rate": 7.565917600082833e-05,
+      "loss": 1.0443,
+      "step": 30055
+    },
+    {
+      "epoch": 1.157266602502406,
+      "grad_norm": 0.9042505621910095,
+      "learning_rate": 7.562984699683408e-05,
+      "loss": 0.6834,
+      "step": 30060
+    },
+    {
+      "epoch": 1.1574590952839268,
+      "grad_norm": 1.246706485748291,
+      "learning_rate": 7.560052022132889e-05,
+      "loss": 0.9267,
+      "step": 30065
+    },
+    {
+      "epoch": 1.1576515880654474,
+      "grad_norm": 1.4773999452590942,
+      "learning_rate": 7.557119567699452e-05,
+      "loss": 0.7131,
+      "step": 30070
+    },
+    {
+      "epoch": 1.1578440808469683,
+      "grad_norm": 1.9604829549789429,
+      "learning_rate": 7.554187336651247e-05,
+      "loss": 0.8189,
+      "step": 30075
+    },
+    {
+      "epoch": 1.158036573628489,
+      "grad_norm": 1.0815962553024292,
+      "learning_rate": 7.551255329256402e-05,
+      "loss": 0.8375,
+      "step": 30080
+    },
+    {
+      "epoch": 1.1582290664100097,
+      "grad_norm": 1.2420703172683716,
+      "learning_rate": 7.548323545783042e-05,
+      "loss": 0.9146,
+      "step": 30085
+    },
+    {
+      "epoch": 1.1584215591915303,
+      "grad_norm": 1.9062912464141846,
+      "learning_rate": 7.545391986499252e-05,
+      "loss": 0.8919,
+      "step": 30090
+    },
+    {
+      "epoch": 1.158614051973051,
+      "grad_norm": 1.7222148180007935,
+      "learning_rate": 7.5424606516731e-05,
+      "loss": 0.7967,
+      "step": 30095
+    },
+    {
+      "epoch": 1.1588065447545717,
+      "grad_norm": 1.2561261653900146,
+      "learning_rate": 7.539529541572647e-05,
+      "loss": 0.837,
+      "step": 30100
+    },
+    {
+      "epoch": 1.1589990375360923,
+      "grad_norm": 1.5790022611618042,
+      "learning_rate": 7.536598656465918e-05,
+      "loss": 0.8983,
+      "step": 30105
+    },
+    {
+      "epoch": 1.159191530317613,
+      "grad_norm": 1.3885726928710938,
+      "learning_rate": 7.533667996620919e-05,
+      "loss": 0.8999,
+      "step": 30110
+    },
+    {
+      "epoch": 1.1593840230991337,
+      "grad_norm": 1.676992416381836,
+      "learning_rate": 7.530737562305649e-05,
+      "loss": 1.0169,
+      "step": 30115
+    },
+    {
+      "epoch": 1.1595765158806546,
+      "grad_norm": 1.5144734382629395,
+      "learning_rate": 7.52780735378807e-05,
+      "loss": 0.9365,
+      "step": 30120
+    },
+    {
+      "epoch": 1.1597690086621752,
+      "grad_norm": 1.8382542133331299,
+      "learning_rate": 7.524877371336129e-05,
+      "loss": 0.8854,
+      "step": 30125
+    },
+    {
+      "epoch": 1.159961501443696,
+      "grad_norm": 2.0379273891448975,
+      "learning_rate": 7.52194761521776e-05,
+      "loss": 0.8692,
+      "step": 30130
+    },
+    {
+      "epoch": 1.1601539942252166,
+      "grad_norm": 1.4084811210632324,
+      "learning_rate": 7.519018085700861e-05,
+      "loss": 0.9688,
+      "step": 30135
+    },
+    {
+      "epoch": 1.1603464870067373,
+      "grad_norm": 1.7691874504089355,
+      "learning_rate": 7.516088783053327e-05,
+      "loss": 1.031,
+      "step": 30140
+    },
+    {
+      "epoch": 1.160538979788258,
+      "grad_norm": 1.5990632772445679,
+      "learning_rate": 7.51315970754302e-05,
+      "loss": 0.9808,
+      "step": 30145
+    },
+    {
+      "epoch": 1.1607314725697786,
+      "grad_norm": 1.0263117551803589,
+      "learning_rate": 7.510230859437781e-05,
+      "loss": 0.85,
+      "step": 30150
+    },
+    {
+      "epoch": 1.1609239653512993,
+      "grad_norm": 0.9199804067611694,
+      "learning_rate": 7.50730223900544e-05,
+      "loss": 0.8131,
+      "step": 30155
+    },
+    {
+      "epoch": 1.16111645813282,
+      "grad_norm": 1.6857373714447021,
+      "learning_rate": 7.504373846513796e-05,
+      "loss": 0.9652,
+      "step": 30160
+    },
+    {
+      "epoch": 1.1613089509143406,
+      "grad_norm": 0.9786146283149719,
+      "learning_rate": 7.501445682230628e-05,
+      "loss": 0.8816,
+      "step": 30165
+    },
+    {
+      "epoch": 1.1615014436958615,
+      "grad_norm": 1.5863782167434692,
+      "learning_rate": 7.498517746423706e-05,
+      "loss": 0.8624,
+      "step": 30170
+    },
+    {
+      "epoch": 1.1616939364773822,
+      "grad_norm": 1.287121295928955,
+      "learning_rate": 7.495590039360763e-05,
+      "loss": 0.9224,
+      "step": 30175
+    },
+    {
+      "epoch": 1.1618864292589028,
+      "grad_norm": 2.6649892330169678,
+      "learning_rate": 7.492662561309518e-05,
+      "loss": 0.8616,
+      "step": 30180
+    },
+    {
+      "epoch": 1.1620789220404235,
+      "grad_norm": 1.1567127704620361,
+      "learning_rate": 7.489735312537676e-05,
+      "loss": 0.759,
+      "step": 30185
+    },
+    {
+      "epoch": 1.1622714148219442,
+      "grad_norm": 1.2174935340881348,
+      "learning_rate": 7.48680829331291e-05,
+      "loss": 0.7874,
+      "step": 30190
+    },
+    {
+      "epoch": 1.1624639076034649,
+      "grad_norm": 1.6268326044082642,
+      "learning_rate": 7.483881503902874e-05,
+      "loss": 0.8314,
+      "step": 30195
+    },
+    {
+      "epoch": 1.1626564003849855,
+      "grad_norm": 1.1430530548095703,
+      "learning_rate": 7.480954944575212e-05,
+      "loss": 0.8684,
+      "step": 30200
+    },
+    {
+      "epoch": 1.1628488931665062,
+      "grad_norm": 1.0805143117904663,
+      "learning_rate": 7.478028615597532e-05,
+      "loss": 0.9936,
+      "step": 30205
+    },
+    {
+      "epoch": 1.1630413859480269,
+      "grad_norm": 1.8118839263916016,
+      "learning_rate": 7.475102517237424e-05,
+      "loss": 0.7458,
+      "step": 30210
+    },
+    {
+      "epoch": 1.1632338787295478,
+      "grad_norm": 0.9031025171279907,
+      "learning_rate": 7.47217664976247e-05,
+      "loss": 0.7417,
+      "step": 30215
+    },
+    {
+      "epoch": 1.1634263715110684,
+      "grad_norm": 1.5473326444625854,
+      "learning_rate": 7.469251013440215e-05,
+      "loss": 0.8992,
+      "step": 30220
+    },
+    {
+      "epoch": 1.163618864292589,
+      "grad_norm": 1.164165735244751,
+      "learning_rate": 7.466325608538185e-05,
+      "loss": 0.9047,
+      "step": 30225
+    },
+    {
+      "epoch": 1.1638113570741098,
+      "grad_norm": 1.0037180185317993,
+      "learning_rate": 7.463400435323899e-05,
+      "loss": 0.7691,
+      "step": 30230
+    },
+    {
+      "epoch": 1.1640038498556304,
+      "grad_norm": 1.021618127822876,
+      "learning_rate": 7.460475494064841e-05,
+      "loss": 0.6932,
+      "step": 30235
+    },
+    {
+      "epoch": 1.164196342637151,
+      "grad_norm": 1.689186930656433,
+      "learning_rate": 7.457550785028472e-05,
+      "loss": 0.891,
+      "step": 30240
+    },
+    {
+      "epoch": 1.1643888354186718,
+      "grad_norm": 1.9816436767578125,
+      "learning_rate": 7.454626308482244e-05,
+      "loss": 0.8508,
+      "step": 30245
+    },
+    {
+      "epoch": 1.1645813282001924,
+      "grad_norm": 1.265624761581421,
+      "learning_rate": 7.45170206469358e-05,
+      "loss": 0.8929,
+      "step": 30250
+    },
+    {
+      "epoch": 1.1647738209817131,
+      "grad_norm": 1.2222243547439575,
+      "learning_rate": 7.44877805392988e-05,
+      "loss": 0.9135,
+      "step": 30255
+    },
+    {
+      "epoch": 1.1649663137632338,
+      "grad_norm": 2.1236441135406494,
+      "learning_rate": 7.445854276458527e-05,
+      "loss": 0.7737,
+      "step": 30260
+    },
+    {
+      "epoch": 1.1651588065447545,
+      "grad_norm": 0.9599865078926086,
+      "learning_rate": 7.44293073254688e-05,
+      "loss": 0.8891,
+      "step": 30265
+    },
+    {
+      "epoch": 1.1653512993262753,
+      "grad_norm": 1.1516355276107788,
+      "learning_rate": 7.440007422462276e-05,
+      "loss": 0.7167,
+      "step": 30270
+    },
+    {
+      "epoch": 1.165543792107796,
+      "grad_norm": 1.8902909755706787,
+      "learning_rate": 7.43708434647204e-05,
+      "loss": 0.8657,
+      "step": 30275
+    },
+    {
+      "epoch": 1.1657362848893167,
+      "grad_norm": 2.002458095550537,
+      "learning_rate": 7.434161504843461e-05,
+      "loss": 0.7881,
+      "step": 30280
+    },
+    {
+      "epoch": 1.1659287776708374,
+      "grad_norm": 1.3423494100570679,
+      "learning_rate": 7.431238897843811e-05,
+      "loss": 0.9164,
+      "step": 30285
+    },
+    {
+      "epoch": 1.166121270452358,
+      "grad_norm": 1.7833929061889648,
+      "learning_rate": 7.428316525740353e-05,
+      "loss": 0.9904,
+      "step": 30290
+    },
+    {
+      "epoch": 1.1663137632338787,
+      "grad_norm": 1.0303772687911987,
+      "learning_rate": 7.425394388800311e-05,
+      "loss": 0.7141,
+      "step": 30295
+    },
+    {
+      "epoch": 1.1665062560153994,
+      "grad_norm": 1.3508719205856323,
+      "learning_rate": 7.422472487290893e-05,
+      "loss": 0.792,
+      "step": 30300
+    },
+    {
+      "epoch": 1.16669874879692,
+      "grad_norm": 1.598431944847107,
+      "learning_rate": 7.419550821479298e-05,
+      "loss": 0.8109,
+      "step": 30305
+    },
+    {
+      "epoch": 1.166891241578441,
+      "grad_norm": 1.6909544467926025,
+      "learning_rate": 7.416629391632683e-05,
+      "loss": 0.9068,
+      "step": 30310
+    },
+    {
+      "epoch": 1.1670837343599616,
+      "grad_norm": 1.0492467880249023,
+      "learning_rate": 7.413708198018195e-05,
+      "loss": 0.9566,
+      "step": 30315
+    },
+    {
+      "epoch": 1.1672762271414823,
+      "grad_norm": 2.0640323162078857,
+      "learning_rate": 7.410787240902963e-05,
+      "loss": 0.818,
+      "step": 30320
+    },
+    {
+      "epoch": 1.167468719923003,
+      "grad_norm": 1.4553757905960083,
+      "learning_rate": 7.407866520554087e-05,
+      "loss": 0.8928,
+      "step": 30325
+    },
+    {
+      "epoch": 1.1676612127045236,
+      "grad_norm": 1.5272727012634277,
+      "learning_rate": 7.404946037238641e-05,
+      "loss": 0.8591,
+      "step": 30330
+    },
+    {
+      "epoch": 1.1678537054860443,
+      "grad_norm": 1.040766716003418,
+      "learning_rate": 7.402025791223694e-05,
+      "loss": 0.851,
+      "step": 30335
+    },
+    {
+      "epoch": 1.168046198267565,
+      "grad_norm": 0.9818170666694641,
+      "learning_rate": 7.399105782776276e-05,
+      "loss": 0.7963,
+      "step": 30340
+    },
+    {
+      "epoch": 1.1682386910490856,
+      "grad_norm": 1.3725941181182861,
+      "learning_rate": 7.396186012163404e-05,
+      "loss": 0.7677,
+      "step": 30345
+    },
+    {
+      "epoch": 1.1684311838306063,
+      "grad_norm": 1.8234846591949463,
+      "learning_rate": 7.393266479652075e-05,
+      "loss": 0.7203,
+      "step": 30350
+    },
+    {
+      "epoch": 1.168623676612127,
+      "grad_norm": 1.143553376197815,
+      "learning_rate": 7.390347185509258e-05,
+      "loss": 0.8207,
+      "step": 30355
+    },
+    {
+      "epoch": 1.1688161693936476,
+      "grad_norm": 1.3585904836654663,
+      "learning_rate": 7.387428130001904e-05,
+      "loss": 0.8421,
+      "step": 30360
+    },
+    {
+      "epoch": 1.1690086621751685,
+      "grad_norm": 2.451633930206299,
+      "learning_rate": 7.384509313396939e-05,
+      "loss": 0.8745,
+      "step": 30365
+    },
+    {
+      "epoch": 1.1692011549566892,
+      "grad_norm": 0.9068706035614014,
+      "learning_rate": 7.381590735961272e-05,
+      "loss": 1.0768,
+      "step": 30370
+    },
+    {
+      "epoch": 1.1693936477382099,
+      "grad_norm": 1.411170244216919,
+      "learning_rate": 7.378672397961788e-05,
+      "loss": 0.9386,
+      "step": 30375
+    },
+    {
+      "epoch": 1.1695861405197305,
+      "grad_norm": 2.274181842803955,
+      "learning_rate": 7.375754299665348e-05,
+      "loss": 0.9783,
+      "step": 30380
+    },
+    {
+      "epoch": 1.1697786333012512,
+      "grad_norm": 2.0918445587158203,
+      "learning_rate": 7.372836441338789e-05,
+      "loss": 0.8331,
+      "step": 30385
+    },
+    {
+      "epoch": 1.1699711260827719,
+      "grad_norm": 0.9770554900169373,
+      "learning_rate": 7.36991882324894e-05,
+      "loss": 0.8223,
+      "step": 30390
+    },
+    {
+      "epoch": 1.1701636188642925,
+      "grad_norm": 1.2381200790405273,
+      "learning_rate": 7.367001445662591e-05,
+      "loss": 0.8515,
+      "step": 30395
+    },
+    {
+      "epoch": 1.1703561116458132,
+      "grad_norm": 1.2155394554138184,
+      "learning_rate": 7.364084308846512e-05,
+      "loss": 0.9158,
+      "step": 30400
+    },
+    {
+      "epoch": 1.1705486044273339,
+      "grad_norm": 1.4437912702560425,
+      "learning_rate": 7.361167413067469e-05,
+      "loss": 0.8522,
+      "step": 30405
+    },
+    {
+      "epoch": 1.1707410972088548,
+      "grad_norm": 1.226899266242981,
+      "learning_rate": 7.358250758592184e-05,
+      "loss": 0.9088,
+      "step": 30410
+    },
+    {
+      "epoch": 1.1709335899903754,
+      "grad_norm": 0.8239467144012451,
+      "learning_rate": 7.355334345687361e-05,
+      "loss": 0.8154,
+      "step": 30415
+    },
+    {
+      "epoch": 1.1711260827718961,
+      "grad_norm": 1.354842185974121,
+      "learning_rate": 7.3524181746197e-05,
+      "loss": 0.7883,
+      "step": 30420
+    },
+    {
+      "epoch": 1.1713185755534168,
+      "grad_norm": 0.8536742329597473,
+      "learning_rate": 7.349502245655857e-05,
+      "loss": 0.6467,
+      "step": 30425
+    },
+    {
+      "epoch": 1.1715110683349375,
+      "grad_norm": 1.2554560899734497,
+      "learning_rate": 7.346586559062472e-05,
+      "loss": 0.8991,
+      "step": 30430
+    },
+    {
+      "epoch": 1.1717035611164581,
+      "grad_norm": 1.260736107826233,
+      "learning_rate": 7.343671115106172e-05,
+      "loss": 0.7645,
+      "step": 30435
+    },
+    {
+      "epoch": 1.1718960538979788,
+      "grad_norm": 1.7142833471298218,
+      "learning_rate": 7.340755914053552e-05,
+      "loss": 0.8953,
+      "step": 30440
+    },
+    {
+      "epoch": 1.1720885466794995,
+      "grad_norm": 1.4862356185913086,
+      "learning_rate": 7.337840956171184e-05,
+      "loss": 0.8362,
+      "step": 30445
+    },
+    {
+      "epoch": 1.1722810394610201,
+      "grad_norm": 1.4640238285064697,
+      "learning_rate": 7.33492624172563e-05,
+      "loss": 0.7173,
+      "step": 30450
+    },
+    {
+      "epoch": 1.1724735322425408,
+      "grad_norm": 1.3464069366455078,
+      "learning_rate": 7.332011770983417e-05,
+      "loss": 0.8988,
+      "step": 30455
+    },
+    {
+      "epoch": 1.1726660250240615,
+      "grad_norm": 1.3081461191177368,
+      "learning_rate": 7.32909754421105e-05,
+      "loss": 1.0021,
+      "step": 30460
+    },
+    {
+      "epoch": 1.1728585178055824,
+      "grad_norm": 0.9563549160957336,
+      "learning_rate": 7.326183561675022e-05,
+      "loss": 0.7904,
+      "step": 30465
+    },
+    {
+      "epoch": 1.173051010587103,
+      "grad_norm": 1.8011226654052734,
+      "learning_rate": 7.323269823641794e-05,
+      "loss": 0.9917,
+      "step": 30470
+    },
+    {
+      "epoch": 1.1732435033686237,
+      "grad_norm": 1.0634485483169556,
+      "learning_rate": 7.320356330377809e-05,
+      "loss": 0.8297,
+      "step": 30475
+    },
+    {
+      "epoch": 1.1734359961501444,
+      "grad_norm": 1.195955753326416,
+      "learning_rate": 7.317443082149488e-05,
+      "loss": 1.7867,
+      "step": 30480
+    },
+    {
+      "epoch": 1.173628488931665,
+      "grad_norm": 0.9291029572486877,
+      "learning_rate": 7.314530079223225e-05,
+      "loss": 0.7805,
+      "step": 30485
+    },
+    {
+      "epoch": 1.1738209817131857,
+      "grad_norm": 1.8289291858673096,
+      "learning_rate": 7.311617321865396e-05,
+      "loss": 0.8179,
+      "step": 30490
+    },
+    {
+      "epoch": 1.1740134744947064,
+      "grad_norm": 2.2788593769073486,
+      "learning_rate": 7.308704810342357e-05,
+      "loss": 1.0026,
+      "step": 30495
+    },
+    {
+      "epoch": 1.174205967276227,
+      "grad_norm": 0.9969417452812195,
+      "learning_rate": 7.305792544920433e-05,
+      "loss": 0.9294,
+      "step": 30500
+    },
+    {
+      "epoch": 1.174398460057748,
+      "grad_norm": 1.9698654413223267,
+      "learning_rate": 7.302880525865932e-05,
+      "loss": 1.031,
+      "step": 30505
+    },
+    {
+      "epoch": 1.1745909528392686,
+      "grad_norm": 2.6228976249694824,
+      "learning_rate": 7.299968753445142e-05,
+      "loss": 0.8685,
+      "step": 30510
+    },
+    {
+      "epoch": 1.1747834456207893,
+      "grad_norm": 1.7787238359451294,
+      "learning_rate": 7.297057227924324e-05,
+      "loss": 0.7516,
+      "step": 30515
+    },
+    {
+      "epoch": 1.17497593840231,
+      "grad_norm": 1.7885974645614624,
+      "learning_rate": 7.294145949569713e-05,
+      "loss": 0.9863,
+      "step": 30520
+    },
+    {
+      "epoch": 1.1751684311838306,
+      "grad_norm": 0.9006327986717224,
+      "learning_rate": 7.291234918647534e-05,
+      "loss": 0.8484,
+      "step": 30525
+    },
+    {
+      "epoch": 1.1753609239653513,
+      "grad_norm": 1.1137930154800415,
+      "learning_rate": 7.288324135423979e-05,
+      "loss": 0.8275,
+      "step": 30530
+    },
+    {
+      "epoch": 1.175553416746872,
+      "grad_norm": 0.8534238338470459,
+      "learning_rate": 7.285413600165214e-05,
+      "loss": 0.8755,
+      "step": 30535
+    },
+    {
+      "epoch": 1.1757459095283926,
+      "grad_norm": 1.3716295957565308,
+      "learning_rate": 7.282503313137397e-05,
+      "loss": 0.7926,
+      "step": 30540
+    },
+    {
+      "epoch": 1.1759384023099133,
+      "grad_norm": 2.3967831134796143,
+      "learning_rate": 7.27959327460665e-05,
+      "loss": 0.7856,
+      "step": 30545
+    },
+    {
+      "epoch": 1.176130895091434,
+      "grad_norm": 0.8649664521217346,
+      "learning_rate": 7.276683484839074e-05,
+      "loss": 0.6645,
+      "step": 30550
+    },
+    {
+      "epoch": 1.1763233878729547,
+      "grad_norm": 1.004795789718628,
+      "learning_rate": 7.273773944100755e-05,
+      "loss": 0.9662,
+      "step": 30555
+    },
+    {
+      "epoch": 1.1765158806544755,
+      "grad_norm": 1.4656823873519897,
+      "learning_rate": 7.27086465265775e-05,
+      "loss": 1.0357,
+      "step": 30560
+    },
+    {
+      "epoch": 1.1767083734359962,
+      "grad_norm": 1.2406914234161377,
+      "learning_rate": 7.267955610776089e-05,
+      "loss": 0.7643,
+      "step": 30565
+    },
+    {
+      "epoch": 1.1769008662175169,
+      "grad_norm": 0.761400043964386,
+      "learning_rate": 7.265046818721795e-05,
+      "loss": 0.7157,
+      "step": 30570
+    },
+    {
+      "epoch": 1.1770933589990376,
+      "grad_norm": 1.4222087860107422,
+      "learning_rate": 7.26213827676085e-05,
+      "loss": 0.7819,
+      "step": 30575
+    },
+    {
+      "epoch": 1.1772858517805582,
+      "grad_norm": 1.604976773262024,
+      "learning_rate": 7.259229985159223e-05,
+      "loss": 0.8072,
+      "step": 30580
+    },
+    {
+      "epoch": 1.177478344562079,
+      "grad_norm": 1.5902528762817383,
+      "learning_rate": 7.256321944182856e-05,
+      "loss": 0.8419,
+      "step": 30585
+    },
+    {
+      "epoch": 1.1776708373435996,
+      "grad_norm": 1.3055870532989502,
+      "learning_rate": 7.253414154097675e-05,
+      "loss": 0.8395,
+      "step": 30590
+    },
+    {
+      "epoch": 1.1778633301251202,
+      "grad_norm": 1.5313003063201904,
+      "learning_rate": 7.250506615169573e-05,
+      "loss": 0.853,
+      "step": 30595
+    },
+    {
+      "epoch": 1.1780558229066411,
+      "grad_norm": 2.564984083175659,
+      "learning_rate": 7.24759932766443e-05,
+      "loss": 0.8832,
+      "step": 30600
+    },
+    {
+      "epoch": 1.1782483156881618,
+      "grad_norm": 1.0367460250854492,
+      "learning_rate": 7.244692291848091e-05,
+      "loss": 0.7639,
+      "step": 30605
+    },
+    {
+      "epoch": 1.1784408084696825,
+      "grad_norm": 1.0948158502578735,
+      "learning_rate": 7.241785507986392e-05,
+      "loss": 0.8525,
+      "step": 30610
+    },
+    {
+      "epoch": 1.1786333012512031,
+      "grad_norm": 1.3061574697494507,
+      "learning_rate": 7.23887897634514e-05,
+      "loss": 0.7501,
+      "step": 30615
+    },
+    {
+      "epoch": 1.1788257940327238,
+      "grad_norm": 2.5141634941101074,
+      "learning_rate": 7.235972697190112e-05,
+      "loss": 0.8892,
+      "step": 30620
+    },
+    {
+      "epoch": 1.1790182868142445,
+      "grad_norm": 1.4427481889724731,
+      "learning_rate": 7.233066670787068e-05,
+      "loss": 0.8994,
+      "step": 30625
+    },
+    {
+      "epoch": 1.1792107795957651,
+      "grad_norm": 0.6231327652931213,
+      "learning_rate": 7.230160897401752e-05,
+      "loss": 0.711,
+      "step": 30630
+    },
+    {
+      "epoch": 1.1794032723772858,
+      "grad_norm": 1.6313964128494263,
+      "learning_rate": 7.227255377299873e-05,
+      "loss": 0.9548,
+      "step": 30635
+    },
+    {
+      "epoch": 1.1795957651588065,
+      "grad_norm": 1.321081280708313,
+      "learning_rate": 7.224350110747118e-05,
+      "loss": 0.7634,
+      "step": 30640
+    },
+    {
+      "epoch": 1.1797882579403272,
+      "grad_norm": 1.1919338703155518,
+      "learning_rate": 7.221445098009163e-05,
+      "loss": 0.8379,
+      "step": 30645
+    },
+    {
+      "epoch": 1.1799807507218478,
+      "grad_norm": 1.954206109046936,
+      "learning_rate": 7.218540339351643e-05,
+      "loss": 0.6995,
+      "step": 30650
+    },
+    {
+      "epoch": 1.1801732435033687,
+      "grad_norm": 1.4222872257232666,
+      "learning_rate": 7.215635835040187e-05,
+      "loss": 0.7785,
+      "step": 30655
+    },
+    {
+      "epoch": 1.1803657362848894,
+      "grad_norm": 1.8065252304077148,
+      "learning_rate": 7.21273158534039e-05,
+      "loss": 0.8699,
+      "step": 30660
+    },
+    {
+      "epoch": 1.18055822906641,
+      "grad_norm": 0.9798047542572021,
+      "learning_rate": 7.209827590517822e-05,
+      "loss": 0.9183,
+      "step": 30665
+    },
+    {
+      "epoch": 1.1807507218479307,
+      "grad_norm": 1.1454569101333618,
+      "learning_rate": 7.206923850838041e-05,
+      "loss": 0.8424,
+      "step": 30670
+    },
+    {
+      "epoch": 1.1809432146294514,
+      "grad_norm": 0.8481330871582031,
+      "learning_rate": 7.204020366566571e-05,
+      "loss": 0.8378,
+      "step": 30675
+    },
+    {
+      "epoch": 1.181135707410972,
+      "grad_norm": 1.2855151891708374,
+      "learning_rate": 7.201117137968915e-05,
+      "loss": 0.8317,
+      "step": 30680
+    },
+    {
+      "epoch": 1.1813282001924927,
+      "grad_norm": 2.0627200603485107,
+      "learning_rate": 7.198214165310555e-05,
+      "loss": 0.7981,
+      "step": 30685
+    },
+    {
+      "epoch": 1.1815206929740134,
+      "grad_norm": 0.8609490394592285,
+      "learning_rate": 7.195311448856952e-05,
+      "loss": 0.7964,
+      "step": 30690
+    },
+    {
+      "epoch": 1.181713185755534,
+      "grad_norm": 1.0228421688079834,
+      "learning_rate": 7.192408988873537e-05,
+      "loss": 0.914,
+      "step": 30695
+    },
+    {
+      "epoch": 1.181905678537055,
+      "grad_norm": 1.3062362670898438,
+      "learning_rate": 7.189506785625722e-05,
+      "loss": 0.8929,
+      "step": 30700
+    },
+    {
+      "epoch": 1.1820981713185756,
+      "grad_norm": 0.9503874182701111,
+      "learning_rate": 7.186604839378891e-05,
+      "loss": 0.8451,
+      "step": 30705
+    },
+    {
+      "epoch": 1.1822906641000963,
+      "grad_norm": 1.9427149295806885,
+      "learning_rate": 7.183703150398414e-05,
+      "loss": 1.0375,
+      "step": 30710
+    },
+    {
+      "epoch": 1.182483156881617,
+      "grad_norm": 1.4807151556015015,
+      "learning_rate": 7.180801718949626e-05,
+      "loss": 0.9003,
+      "step": 30715
+    },
+    {
+      "epoch": 1.1826756496631377,
+      "grad_norm": 1.048420786857605,
+      "learning_rate": 7.177900545297846e-05,
+      "loss": 0.775,
+      "step": 30720
+    },
+    {
+      "epoch": 1.1828681424446583,
+      "grad_norm": 1.6969749927520752,
+      "learning_rate": 7.174999629708363e-05,
+      "loss": 0.9417,
+      "step": 30725
+    },
+    {
+      "epoch": 1.183060635226179,
+      "grad_norm": 1.2834903001785278,
+      "learning_rate": 7.172098972446453e-05,
+      "loss": 0.9584,
+      "step": 30730
+    },
+    {
+      "epoch": 1.1832531280076997,
+      "grad_norm": 2.063692331314087,
+      "learning_rate": 7.169198573777361e-05,
+      "loss": 0.8202,
+      "step": 30735
+    },
+    {
+      "epoch": 1.1834456207892203,
+      "grad_norm": 0.9396731853485107,
+      "learning_rate": 7.166298433966301e-05,
+      "loss": 0.9421,
+      "step": 30740
+    },
+    {
+      "epoch": 1.183638113570741,
+      "grad_norm": 1.4655632972717285,
+      "learning_rate": 7.163398553278483e-05,
+      "loss": 0.8273,
+      "step": 30745
+    },
+    {
+      "epoch": 1.1838306063522617,
+      "grad_norm": 2.0412590503692627,
+      "learning_rate": 7.160498931979076e-05,
+      "loss": 0.8433,
+      "step": 30750
+    },
+    {
+      "epoch": 1.1840230991337826,
+      "grad_norm": 1.9432737827301025,
+      "learning_rate": 7.157599570333226e-05,
+      "loss": 0.9141,
+      "step": 30755
+    },
+    {
+      "epoch": 1.1842155919153032,
+      "grad_norm": 0.863294243812561,
+      "learning_rate": 7.154700468606073e-05,
+      "loss": 0.9083,
+      "step": 30760
+    },
+    {
+      "epoch": 1.184408084696824,
+      "grad_norm": 1.1578996181488037,
+      "learning_rate": 7.151801627062713e-05,
+      "loss": 0.8211,
+      "step": 30765
+    },
+    {
+      "epoch": 1.1846005774783446,
+      "grad_norm": 1.4375747442245483,
+      "learning_rate": 7.148903045968221e-05,
+      "loss": 0.7579,
+      "step": 30770
+    },
+    {
+      "epoch": 1.1847930702598652,
+      "grad_norm": 1.566175937652588,
+      "learning_rate": 7.146004725587664e-05,
+      "loss": 0.7958,
+      "step": 30775
+    },
+    {
+      "epoch": 1.184985563041386,
+      "grad_norm": 2.4917242527008057,
+      "learning_rate": 7.143106666186068e-05,
+      "loss": 0.9255,
+      "step": 30780
+    },
+    {
+      "epoch": 1.1851780558229066,
+      "grad_norm": 1.2194336652755737,
+      "learning_rate": 7.14020886802844e-05,
+      "loss": 0.715,
+      "step": 30785
+    },
+    {
+      "epoch": 1.1853705486044273,
+      "grad_norm": 0.9946492314338684,
+      "learning_rate": 7.137311331379769e-05,
+      "loss": 0.8673,
+      "step": 30790
+    },
+    {
+      "epoch": 1.1855630413859481,
+      "grad_norm": 1.6159164905548096,
+      "learning_rate": 7.134414056505015e-05,
+      "loss": 0.7092,
+      "step": 30795
+    },
+    {
+      "epoch": 1.1857555341674688,
+      "grad_norm": 1.3039904832839966,
+      "learning_rate": 7.131517043669108e-05,
+      "loss": 0.9336,
+      "step": 30800
+    },
+    {
+      "epoch": 1.1859480269489895,
+      "grad_norm": 0.7158260345458984,
+      "learning_rate": 7.12862029313697e-05,
+      "loss": 0.9515,
+      "step": 30805
+    },
+    {
+      "epoch": 1.1861405197305102,
+      "grad_norm": 1.4139379262924194,
+      "learning_rate": 7.12572380517348e-05,
+      "loss": 0.8344,
+      "step": 30810
+    },
+    {
+      "epoch": 1.1863330125120308,
+      "grad_norm": 1.963030219078064,
+      "learning_rate": 7.122827580043509e-05,
+      "loss": 0.8987,
+      "step": 30815
+    },
+    {
+      "epoch": 1.1865255052935515,
+      "grad_norm": 1.060595989227295,
+      "learning_rate": 7.1199316180119e-05,
+      "loss": 0.7739,
+      "step": 30820
+    },
+    {
+      "epoch": 1.1867179980750722,
+      "grad_norm": 1.1344969272613525,
+      "learning_rate": 7.117035919343464e-05,
+      "loss": 0.9343,
+      "step": 30825
+    },
+    {
+      "epoch": 1.1869104908565928,
+      "grad_norm": 1.3690916299819946,
+      "learning_rate": 7.114140484302992e-05,
+      "loss": 0.6558,
+      "step": 30830
+    },
+    {
+      "epoch": 1.1871029836381135,
+      "grad_norm": 1.053550124168396,
+      "learning_rate": 7.11124531315526e-05,
+      "loss": 0.7993,
+      "step": 30835
+    },
+    {
+      "epoch": 1.1872954764196342,
+      "grad_norm": 1.509711503982544,
+      "learning_rate": 7.108350406165007e-05,
+      "loss": 1.0615,
+      "step": 30840
+    },
+    {
+      "epoch": 1.1874879692011548,
+      "grad_norm": 1.440314769744873,
+      "learning_rate": 7.10545576359695e-05,
+      "loss": 0.7368,
+      "step": 30845
+    },
+    {
+      "epoch": 1.1876804619826757,
+      "grad_norm": 1.1829510927200317,
+      "learning_rate": 7.102561385715794e-05,
+      "loss": 0.7871,
+      "step": 30850
+    },
+    {
+      "epoch": 1.1878729547641964,
+      "grad_norm": 1.2442659139633179,
+      "learning_rate": 7.099667272786205e-05,
+      "loss": 0.7765,
+      "step": 30855
+    },
+    {
+      "epoch": 1.188065447545717,
+      "grad_norm": 1.2763118743896484,
+      "learning_rate": 7.096773425072827e-05,
+      "loss": 0.9741,
+      "step": 30860
+    },
+    {
+      "epoch": 1.1882579403272377,
+      "grad_norm": 1.0801767110824585,
+      "learning_rate": 7.093879842840289e-05,
+      "loss": 0.7907,
+      "step": 30865
+    },
+    {
+      "epoch": 1.1884504331087584,
+      "grad_norm": 1.863099217414856,
+      "learning_rate": 7.090986526353192e-05,
+      "loss": 0.7885,
+      "step": 30870
+    },
+    {
+      "epoch": 1.188642925890279,
+      "grad_norm": 2.064089298248291,
+      "learning_rate": 7.088093475876098e-05,
+      "loss": 0.9183,
+      "step": 30875
+    },
+    {
+      "epoch": 1.1888354186717998,
+      "grad_norm": 1.4984971284866333,
+      "learning_rate": 7.085200691673573e-05,
+      "loss": 0.9088,
+      "step": 30880
+    },
+    {
+      "epoch": 1.1890279114533204,
+      "grad_norm": 1.2373710870742798,
+      "learning_rate": 7.082308174010138e-05,
+      "loss": 0.8277,
+      "step": 30885
+    },
+    {
+      "epoch": 1.189220404234841,
+      "grad_norm": 0.864380955696106,
+      "learning_rate": 7.079415923150285e-05,
+      "loss": 0.7536,
+      "step": 30890
+    },
+    {
+      "epoch": 1.189412897016362,
+      "grad_norm": 1.1806195974349976,
+      "learning_rate": 7.076523939358504e-05,
+      "loss": 0.7892,
+      "step": 30895
+    },
+    {
+      "epoch": 1.1896053897978827,
+      "grad_norm": 1.486262559890747,
+      "learning_rate": 7.07363222289924e-05,
+      "loss": 0.7652,
+      "step": 30900
+    },
+    {
+      "epoch": 1.1897978825794033,
+      "grad_norm": 1.093450665473938,
+      "learning_rate": 7.070740774036926e-05,
+      "loss": 0.7741,
+      "step": 30905
+    },
+    {
+      "epoch": 1.189990375360924,
+      "grad_norm": 0.9283474683761597,
+      "learning_rate": 7.067849593035962e-05,
+      "loss": 0.7752,
+      "step": 30910
+    },
+    {
+      "epoch": 1.1901828681424447,
+      "grad_norm": 1.1739362478256226,
+      "learning_rate": 7.064958680160729e-05,
+      "loss": 0.762,
+      "step": 30915
+    },
+    {
+      "epoch": 1.1903753609239653,
+      "grad_norm": 1.4015986919403076,
+      "learning_rate": 7.062068035675584e-05,
+      "loss": 0.866,
+      "step": 30920
+    },
+    {
+      "epoch": 1.190567853705486,
+      "grad_norm": 1.121469259262085,
+      "learning_rate": 7.059177659844853e-05,
+      "loss": 0.7242,
+      "step": 30925
+    },
+    {
+      "epoch": 1.1907603464870067,
+      "grad_norm": 1.6382770538330078,
+      "learning_rate": 7.056287552932842e-05,
+      "loss": 0.8558,
+      "step": 30930
+    },
+    {
+      "epoch": 1.1909528392685274,
+      "grad_norm": 1.2235866785049438,
+      "learning_rate": 7.053397715203837e-05,
+      "loss": 1.0149,
+      "step": 30935
+    },
+    {
+      "epoch": 1.191145332050048,
+      "grad_norm": 1.0109368562698364,
+      "learning_rate": 7.050508146922093e-05,
+      "loss": 0.8432,
+      "step": 30940
+    },
+    {
+      "epoch": 1.1913378248315687,
+      "grad_norm": 1.2103551626205444,
+      "learning_rate": 7.047618848351835e-05,
+      "loss": 0.8176,
+      "step": 30945
+    },
+    {
+      "epoch": 1.1915303176130896,
+      "grad_norm": 1.0425639152526855,
+      "learning_rate": 7.044729819757279e-05,
+      "loss": 0.7266,
+      "step": 30950
+    },
+    {
+      "epoch": 1.1917228103946103,
+      "grad_norm": 1.1277648210525513,
+      "learning_rate": 7.041841061402606e-05,
+      "loss": 0.743,
+      "step": 30955
+    },
+    {
+      "epoch": 1.191915303176131,
+      "grad_norm": 1.4034432172775269,
+      "learning_rate": 7.038952573551967e-05,
+      "loss": 0.9021,
+      "step": 30960
+    },
+    {
+      "epoch": 1.1921077959576516,
+      "grad_norm": 0.9935101270675659,
+      "learning_rate": 7.036064356469504e-05,
+      "loss": 0.7794,
+      "step": 30965
+    },
+    {
+      "epoch": 1.1923002887391723,
+      "grad_norm": 1.6952626705169678,
+      "learning_rate": 7.033176410419322e-05,
+      "loss": 0.8085,
+      "step": 30970
+    },
+    {
+      "epoch": 1.192492781520693,
+      "grad_norm": 1.1734607219696045,
+      "learning_rate": 7.030288735665498e-05,
+      "loss": 0.742,
+      "step": 30975
+    },
+    {
+      "epoch": 1.1926852743022136,
+      "grad_norm": 1.4427086114883423,
+      "learning_rate": 7.027401332472102e-05,
+      "loss": 0.9636,
+      "step": 30980
+    },
+    {
+      "epoch": 1.1928777670837343,
+      "grad_norm": 1.4568814039230347,
+      "learning_rate": 7.024514201103163e-05,
+      "loss": 0.857,
+      "step": 30985
+    },
+    {
+      "epoch": 1.1930702598652552,
+      "grad_norm": 0.9864609241485596,
+      "learning_rate": 7.021627341822684e-05,
+      "loss": 0.7251,
+      "step": 30990
+    },
+    {
+      "epoch": 1.1932627526467758,
+      "grad_norm": 1.3467034101486206,
+      "learning_rate": 7.018740754894659e-05,
+      "loss": 0.8612,
+      "step": 30995
+    },
+    {
+      "epoch": 1.1934552454282965,
+      "grad_norm": 0.9315931797027588,
+      "learning_rate": 7.015854440583044e-05,
+      "loss": 0.7092,
+      "step": 31000
+    },
+    {
+      "epoch": 1.1936477382098172,
+      "grad_norm": 1.0659290552139282,
+      "learning_rate": 7.012968399151769e-05,
+      "loss": 0.8342,
+      "step": 31005
+    },
+    {
+      "epoch": 1.1938402309913378,
+      "grad_norm": 1.1278982162475586,
+      "learning_rate": 7.010082630864748e-05,
+      "loss": 0.8266,
+      "step": 31010
+    },
+    {
+      "epoch": 1.1940327237728585,
+      "grad_norm": 1.739660382270813,
+      "learning_rate": 7.007197135985865e-05,
+      "loss": 0.8522,
+      "step": 31015
+    },
+    {
+      "epoch": 1.1942252165543792,
+      "grad_norm": 1.020524263381958,
+      "learning_rate": 7.004311914778977e-05,
+      "loss": 0.7267,
+      "step": 31020
+    },
+    {
+      "epoch": 1.1944177093358999,
+      "grad_norm": 1.6146787405014038,
+      "learning_rate": 7.001426967507921e-05,
+      "loss": 0.8447,
+      "step": 31025
+    },
+    {
+      "epoch": 1.1946102021174205,
+      "grad_norm": 2.2342722415924072,
+      "learning_rate": 6.998542294436504e-05,
+      "loss": 0.9637,
+      "step": 31030
+    },
+    {
+      "epoch": 1.1948026948989412,
+      "grad_norm": 1.4456665515899658,
+      "learning_rate": 6.995657895828511e-05,
+      "loss": 0.8051,
+      "step": 31035
+    },
+    {
+      "epoch": 1.1949951876804619,
+      "grad_norm": 3.0097768306732178,
+      "learning_rate": 6.992773771947703e-05,
+      "loss": 0.9212,
+      "step": 31040
+    },
+    {
+      "epoch": 1.1951876804619828,
+      "grad_norm": 2.814483165740967,
+      "learning_rate": 6.989889923057813e-05,
+      "loss": 0.7442,
+      "step": 31045
+    },
+    {
+      "epoch": 1.1953801732435034,
+      "grad_norm": 1.2800745964050293,
+      "learning_rate": 6.987006349422546e-05,
+      "loss": 0.8654,
+      "step": 31050
+    },
+    {
+      "epoch": 1.195572666025024,
+      "grad_norm": 1.355048418045044,
+      "learning_rate": 6.98412305130559e-05,
+      "loss": 0.7145,
+      "step": 31055
+    },
+    {
+      "epoch": 1.1957651588065448,
+      "grad_norm": 1.7989205121994019,
+      "learning_rate": 6.981240028970607e-05,
+      "loss": 1.0386,
+      "step": 31060
+    },
+    {
+      "epoch": 1.1959576515880654,
+      "grad_norm": 2.1455464363098145,
+      "learning_rate": 6.97835728268122e-05,
+      "loss": 1.0023,
+      "step": 31065
+    },
+    {
+      "epoch": 1.196150144369586,
+      "grad_norm": 1.548008918762207,
+      "learning_rate": 6.975474812701047e-05,
+      "loss": 0.834,
+      "step": 31070
+    },
+    {
+      "epoch": 1.1963426371511068,
+      "grad_norm": 1.2741403579711914,
+      "learning_rate": 6.972592619293665e-05,
+      "loss": 0.6641,
+      "step": 31075
+    },
+    {
+      "epoch": 1.1965351299326275,
+      "grad_norm": 1.5108489990234375,
+      "learning_rate": 6.969710702722632e-05,
+      "loss": 0.8214,
+      "step": 31080
+    },
+    {
+      "epoch": 1.1967276227141483,
+      "grad_norm": 1.490020513534546,
+      "learning_rate": 6.966829063251484e-05,
+      "loss": 0.8304,
+      "step": 31085
+    },
+    {
+      "epoch": 1.196920115495669,
+      "grad_norm": 0.9961896538734436,
+      "learning_rate": 6.963947701143724e-05,
+      "loss": 0.804,
+      "step": 31090
+    },
+    {
+      "epoch": 1.1971126082771897,
+      "grad_norm": 1.3393563032150269,
+      "learning_rate": 6.961066616662834e-05,
+      "loss": 0.882,
+      "step": 31095
+    },
+    {
+      "epoch": 1.1973051010587104,
+      "grad_norm": 0.9488278031349182,
+      "learning_rate": 6.958185810072273e-05,
+      "loss": 0.8017,
+      "step": 31100
+    },
+    {
+      "epoch": 1.197497593840231,
+      "grad_norm": 1.5525720119476318,
+      "learning_rate": 6.955305281635469e-05,
+      "loss": 0.8404,
+      "step": 31105
+    },
+    {
+      "epoch": 1.1976900866217517,
+      "grad_norm": 1.5819134712219238,
+      "learning_rate": 6.952425031615823e-05,
+      "loss": 0.8679,
+      "step": 31110
+    },
+    {
+      "epoch": 1.1978825794032724,
+      "grad_norm": 1.090757966041565,
+      "learning_rate": 6.949545060276726e-05,
+      "loss": 0.6385,
+      "step": 31115
+    },
+    {
+      "epoch": 1.198075072184793,
+      "grad_norm": 1.378661870956421,
+      "learning_rate": 6.946665367881523e-05,
+      "loss": 0.8704,
+      "step": 31120
+    },
+    {
+      "epoch": 1.1982675649663137,
+      "grad_norm": 1.3583415746688843,
+      "learning_rate": 6.943785954693547e-05,
+      "loss": 0.992,
+      "step": 31125
+    },
+    {
+      "epoch": 1.1984600577478344,
+      "grad_norm": 1.195250153541565,
+      "learning_rate": 6.940906820976097e-05,
+      "loss": 0.8763,
+      "step": 31130
+    },
+    {
+      "epoch": 1.198652550529355,
+      "grad_norm": 1.0746605396270752,
+      "learning_rate": 6.938027966992458e-05,
+      "loss": 0.7867,
+      "step": 31135
+    },
+    {
+      "epoch": 1.198845043310876,
+      "grad_norm": 1.0139212608337402,
+      "learning_rate": 6.935149393005873e-05,
+      "loss": 0.8423,
+      "step": 31140
+    },
+    {
+      "epoch": 1.1990375360923966,
+      "grad_norm": 1.3249413967132568,
+      "learning_rate": 6.932271099279576e-05,
+      "loss": 0.8433,
+      "step": 31145
+    },
+    {
+      "epoch": 1.1992300288739173,
+      "grad_norm": 0.8639804124832153,
+      "learning_rate": 6.929393086076765e-05,
+      "loss": 0.9309,
+      "step": 31150
+    },
+    {
+      "epoch": 1.199422521655438,
+      "grad_norm": 1.7848960161209106,
+      "learning_rate": 6.926515353660614e-05,
+      "loss": 0.9963,
+      "step": 31155
+    },
+    {
+      "epoch": 1.1996150144369586,
+      "grad_norm": 1.1610658168792725,
+      "learning_rate": 6.923637902294275e-05,
+      "loss": 0.8316,
+      "step": 31160
+    },
+    {
+      "epoch": 1.1998075072184793,
+      "grad_norm": 1.335810899734497,
+      "learning_rate": 6.920760732240868e-05,
+      "loss": 0.9464,
+      "step": 31165
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3204543590545654,
+      "learning_rate": 6.917883843763497e-05,
+      "loss": 0.7098,
+      "step": 31170
+    },
+    {
+      "epoch": 1.2001924927815206,
+      "grad_norm": 1.3988627195358276,
+      "learning_rate": 6.915007237125233e-05,
+      "loss": 0.8002,
+      "step": 31175
+    },
+    {
+      "epoch": 1.2003849855630413,
+      "grad_norm": 1.4686270952224731,
+      "learning_rate": 6.912130912589116e-05,
+      "loss": 0.7664,
+      "step": 31180
+    },
+    {
+      "epoch": 1.2005774783445622,
+      "grad_norm": 1.4970102310180664,
+      "learning_rate": 6.909254870418176e-05,
+      "loss": 0.8932,
+      "step": 31185
+    },
+    {
+      "epoch": 1.2007699711260829,
+      "grad_norm": 1.4880247116088867,
+      "learning_rate": 6.906379110875403e-05,
+      "loss": 0.8988,
+      "step": 31190
+    },
+    {
+      "epoch": 1.2009624639076035,
+      "grad_norm": 1.4801645278930664,
+      "learning_rate": 6.903503634223764e-05,
+      "loss": 0.9222,
+      "step": 31195
+    },
+    {
+      "epoch": 1.2011549566891242,
+      "grad_norm": 1.630956768989563,
+      "learning_rate": 6.900628440726209e-05,
+      "loss": 0.8345,
+      "step": 31200
+    },
+    {
+      "epoch": 1.2013474494706449,
+      "grad_norm": 1.5143691301345825,
+      "learning_rate": 6.897753530645652e-05,
+      "loss": 0.7211,
+      "step": 31205
+    },
+    {
+      "epoch": 1.2015399422521655,
+      "grad_norm": 0.8206589818000793,
+      "learning_rate": 6.894878904244979e-05,
+      "loss": 0.8318,
+      "step": 31210
+    },
+    {
+      "epoch": 1.2017324350336862,
+      "grad_norm": 1.4505168199539185,
+      "learning_rate": 6.892004561787064e-05,
+      "loss": 0.8583,
+      "step": 31215
+    },
+    {
+      "epoch": 1.2019249278152069,
+      "grad_norm": 1.4092106819152832,
+      "learning_rate": 6.889130503534745e-05,
+      "loss": 0.9484,
+      "step": 31220
+    },
+    {
+      "epoch": 1.2021174205967275,
+      "grad_norm": 1.2476885318756104,
+      "learning_rate": 6.886256729750832e-05,
+      "loss": 0.7712,
+      "step": 31225
+    },
+    {
+      "epoch": 1.2023099133782482,
+      "grad_norm": 0.6989136338233948,
+      "learning_rate": 6.883383240698114e-05,
+      "loss": 0.7654,
+      "step": 31230
+    },
+    {
+      "epoch": 1.2025024061597689,
+      "grad_norm": 1.4014019966125488,
+      "learning_rate": 6.880510036639354e-05,
+      "loss": 0.9665,
+      "step": 31235
+    },
+    {
+      "epoch": 1.2026948989412898,
+      "grad_norm": 2.1451468467712402,
+      "learning_rate": 6.877637117837286e-05,
+      "loss": 0.9765,
+      "step": 31240
+    },
+    {
+      "epoch": 1.2028873917228105,
+      "grad_norm": 1.5151429176330566,
+      "learning_rate": 6.87476448455462e-05,
+      "loss": 0.8647,
+      "step": 31245
+    },
+    {
+      "epoch": 1.2030798845043311,
+      "grad_norm": 0.9325594902038574,
+      "learning_rate": 6.871892137054038e-05,
+      "loss": 0.8579,
+      "step": 31250
+    },
+    {
+      "epoch": 1.2032723772858518,
+      "grad_norm": 1.5600671768188477,
+      "learning_rate": 6.869020075598198e-05,
+      "loss": 0.7753,
+      "step": 31255
+    },
+    {
+      "epoch": 1.2034648700673725,
+      "grad_norm": 1.1633321046829224,
+      "learning_rate": 6.866148300449733e-05,
+      "loss": 0.8108,
+      "step": 31260
+    },
+    {
+      "epoch": 1.2036573628488931,
+      "grad_norm": 1.3864375352859497,
+      "learning_rate": 6.863276811871246e-05,
+      "loss": 0.8359,
+      "step": 31265
+    },
+    {
+      "epoch": 1.2038498556304138,
+      "grad_norm": 1.111289381980896,
+      "learning_rate": 6.860405610125313e-05,
+      "loss": 0.8911,
+      "step": 31270
+    },
+    {
+      "epoch": 1.2040423484119345,
+      "grad_norm": 1.1889983415603638,
+      "learning_rate": 6.857534695474492e-05,
+      "loss": 0.9569,
+      "step": 31275
+    },
+    {
+      "epoch": 1.2042348411934554,
+      "grad_norm": 1.9969587326049805,
+      "learning_rate": 6.854664068181307e-05,
+      "loss": 0.9898,
+      "step": 31280
+    },
+    {
+      "epoch": 1.204427333974976,
+      "grad_norm": 1.4256417751312256,
+      "learning_rate": 6.851793728508251e-05,
+      "loss": 0.8671,
+      "step": 31285
+    },
+    {
+      "epoch": 1.2046198267564967,
+      "grad_norm": 1.4675942659378052,
+      "learning_rate": 6.84892367671781e-05,
+      "loss": 0.8998,
+      "step": 31290
+    },
+    {
+      "epoch": 1.2048123195380174,
+      "grad_norm": 1.571075201034546,
+      "learning_rate": 6.846053913072423e-05,
+      "loss": 0.8697,
+      "step": 31295
+    },
+    {
+      "epoch": 1.205004812319538,
+      "grad_norm": 1.4728553295135498,
+      "learning_rate": 6.84318443783451e-05,
+      "loss": 0.874,
+      "step": 31300
+    },
+    {
+      "epoch": 1.2051973051010587,
+      "grad_norm": 1.1161712408065796,
+      "learning_rate": 6.840315251266472e-05,
+      "loss": 0.7356,
+      "step": 31305
+    },
+    {
+      "epoch": 1.2053897978825794,
+      "grad_norm": 1.4653940200805664,
+      "learning_rate": 6.837446353630672e-05,
+      "loss": 0.8024,
+      "step": 31310
+    },
+    {
+      "epoch": 1.2055822906641,
+      "grad_norm": 1.2116683721542358,
+      "learning_rate": 6.83457774518945e-05,
+      "loss": 0.8066,
+      "step": 31315
+    },
+    {
+      "epoch": 1.2057747834456207,
+      "grad_norm": 1.4532692432403564,
+      "learning_rate": 6.831709426205128e-05,
+      "loss": 0.8092,
+      "step": 31320
+    },
+    {
+      "epoch": 1.2059672762271414,
+      "grad_norm": 1.5983264446258545,
+      "learning_rate": 6.82884139693999e-05,
+      "loss": 1.2661,
+      "step": 31325
+    },
+    {
+      "epoch": 1.206159769008662,
+      "grad_norm": 1.1187528371810913,
+      "learning_rate": 6.825973657656297e-05,
+      "loss": 0.8836,
+      "step": 31330
+    },
+    {
+      "epoch": 1.206352261790183,
+      "grad_norm": 1.9153698682785034,
+      "learning_rate": 6.82310620861629e-05,
+      "loss": 0.8676,
+      "step": 31335
+    },
+    {
+      "epoch": 1.2065447545717036,
+      "grad_norm": 1.0382484197616577,
+      "learning_rate": 6.820239050082175e-05,
+      "loss": 0.7845,
+      "step": 31340
+    },
+    {
+      "epoch": 1.2067372473532243,
+      "grad_norm": 1.228818416595459,
+      "learning_rate": 6.817372182316135e-05,
+      "loss": 0.9074,
+      "step": 31345
+    },
+    {
+      "epoch": 1.206929740134745,
+      "grad_norm": 1.1787536144256592,
+      "learning_rate": 6.814505605580323e-05,
+      "loss": 0.8794,
+      "step": 31350
+    },
+    {
+      "epoch": 1.2071222329162656,
+      "grad_norm": 1.3236230611801147,
+      "learning_rate": 6.811639320136876e-05,
+      "loss": 0.7517,
+      "step": 31355
+    },
+    {
+      "epoch": 1.2073147256977863,
+      "grad_norm": 1.2837519645690918,
+      "learning_rate": 6.808773326247888e-05,
+      "loss": 1.0796,
+      "step": 31360
+    },
+    {
+      "epoch": 1.207507218479307,
+      "grad_norm": 1.7796859741210938,
+      "learning_rate": 6.805907624175443e-05,
+      "loss": 0.8648,
+      "step": 31365
+    },
+    {
+      "epoch": 1.2076997112608276,
+      "grad_norm": 1.2428163290023804,
+      "learning_rate": 6.803042214181586e-05,
+      "loss": 0.9098,
+      "step": 31370
+    },
+    {
+      "epoch": 1.2078922040423483,
+      "grad_norm": 0.7685776352882385,
+      "learning_rate": 6.800177096528337e-05,
+      "loss": 0.935,
+      "step": 31375
+    },
+    {
+      "epoch": 1.2080846968238692,
+      "grad_norm": 1.5274989604949951,
+      "learning_rate": 6.797312271477699e-05,
+      "loss": 0.8187,
+      "step": 31380
+    },
+    {
+      "epoch": 1.2082771896053899,
+      "grad_norm": 1.497298240661621,
+      "learning_rate": 6.794447739291639e-05,
+      "loss": 0.7576,
+      "step": 31385
+    },
+    {
+      "epoch": 1.2084696823869105,
+      "grad_norm": 1.8094556331634521,
+      "learning_rate": 6.791583500232092e-05,
+      "loss": 0.8676,
+      "step": 31390
+    },
+    {
+      "epoch": 1.2086621751684312,
+      "grad_norm": 1.844341516494751,
+      "learning_rate": 6.788719554560987e-05,
+      "loss": 0.8661,
+      "step": 31395
+    },
+    {
+      "epoch": 1.2088546679499519,
+      "grad_norm": 2.015148639678955,
+      "learning_rate": 6.785855902540205e-05,
+      "loss": 0.8985,
+      "step": 31400
+    },
+    {
+      "epoch": 1.2090471607314726,
+      "grad_norm": 1.5119047164916992,
+      "learning_rate": 6.782992544431603e-05,
+      "loss": 0.8247,
+      "step": 31405
+    },
+    {
+      "epoch": 1.2092396535129932,
+      "grad_norm": 1.3965736627578735,
+      "learning_rate": 6.780129480497028e-05,
+      "loss": 0.7286,
+      "step": 31410
+    },
+    {
+      "epoch": 1.209432146294514,
+      "grad_norm": 1.238343358039856,
+      "learning_rate": 6.777266710998283e-05,
+      "loss": 0.9662,
+      "step": 31415
+    },
+    {
+      "epoch": 1.2096246390760346,
+      "grad_norm": 0.8886631727218628,
+      "learning_rate": 6.774404236197144e-05,
+      "loss": 0.7578,
+      "step": 31420
+    },
+    {
+      "epoch": 1.2098171318575552,
+      "grad_norm": 1.1155580282211304,
+      "learning_rate": 6.771542056355373e-05,
+      "loss": 0.6954,
+      "step": 31425
+    },
+    {
+      "epoch": 1.2100096246390761,
+      "grad_norm": 1.5056712627410889,
+      "learning_rate": 6.768680171734692e-05,
+      "loss": 0.9816,
+      "step": 31430
+    },
+    {
+      "epoch": 1.2102021174205968,
+      "grad_norm": 1.3402817249298096,
+      "learning_rate": 6.765818582596805e-05,
+      "loss": 0.9098,
+      "step": 31435
+    },
+    {
+      "epoch": 1.2103946102021175,
+      "grad_norm": 1.0524393320083618,
+      "learning_rate": 6.762957289203386e-05,
+      "loss": 0.8848,
+      "step": 31440
+    },
+    {
+      "epoch": 1.2105871029836381,
+      "grad_norm": 0.9701569676399231,
+      "learning_rate": 6.760096291816078e-05,
+      "loss": 0.9192,
+      "step": 31445
+    },
+    {
+      "epoch": 1.2107795957651588,
+      "grad_norm": 2.2335519790649414,
+      "learning_rate": 6.757235590696503e-05,
+      "loss": 0.8811,
+      "step": 31450
+    },
+    {
+      "epoch": 1.2109720885466795,
+      "grad_norm": 1.1316584348678589,
+      "learning_rate": 6.75437518610625e-05,
+      "loss": 0.796,
+      "step": 31455
+    },
+    {
+      "epoch": 1.2111645813282002,
+      "grad_norm": 1.6016689538955688,
+      "learning_rate": 6.751515078306887e-05,
+      "loss": 0.8402,
+      "step": 31460
+    },
+    {
+      "epoch": 1.2113570741097208,
+      "grad_norm": 1.7846043109893799,
+      "learning_rate": 6.748655267559951e-05,
+      "loss": 0.8634,
+      "step": 31465
+    },
+    {
+      "epoch": 1.2115495668912415,
+      "grad_norm": 1.7141379117965698,
+      "learning_rate": 6.745795754126953e-05,
+      "loss": 1.0208,
+      "step": 31470
+    },
+    {
+      "epoch": 1.2117420596727624,
+      "grad_norm": 1.225437879562378,
+      "learning_rate": 6.742936538269373e-05,
+      "loss": 0.9122,
+      "step": 31475
+    },
+    {
+      "epoch": 1.211934552454283,
+      "grad_norm": 1.7134424448013306,
+      "learning_rate": 6.740077620248675e-05,
+      "loss": 0.8984,
+      "step": 31480
+    },
+    {
+      "epoch": 1.2121270452358037,
+      "grad_norm": 1.0409084558486938,
+      "learning_rate": 6.737219000326283e-05,
+      "loss": 0.9374,
+      "step": 31485
+    },
+    {
+      "epoch": 1.2123195380173244,
+      "grad_norm": 1.527655005455017,
+      "learning_rate": 6.734360678763593e-05,
+      "loss": 0.9277,
+      "step": 31490
+    },
+    {
+      "epoch": 1.212512030798845,
+      "grad_norm": 1.9206295013427734,
+      "learning_rate": 6.731502655821993e-05,
+      "loss": 0.7926,
+      "step": 31495
+    },
+    {
+      "epoch": 1.2127045235803657,
+      "grad_norm": 1.035962462425232,
+      "learning_rate": 6.728644931762824e-05,
+      "loss": 0.8946,
+      "step": 31500
+    },
+    {
+      "epoch": 1.2128970163618864,
+      "grad_norm": 1.146831750869751,
+      "learning_rate": 6.725787506847399e-05,
+      "loss": 0.9099,
+      "step": 31505
+    },
+    {
+      "epoch": 1.213089509143407,
+      "grad_norm": 1.0527504682540894,
+      "learning_rate": 6.723501782474154e-05,
+      "loss": 0.7984,
+      "step": 31510
+    },
+    {
+      "epoch": 1.2132820019249277,
+      "grad_norm": 1.390851378440857,
+      "learning_rate": 6.72064489667592e-05,
+      "loss": 0.9729,
+      "step": 31515
+    },
+    {
+      "epoch": 1.2134744947064484,
+      "grad_norm": 1.2551788091659546,
+      "learning_rate": 6.717788310752988e-05,
+      "loss": 0.945,
+      "step": 31520
+    },
+    {
+      "epoch": 1.213666987487969,
+      "grad_norm": 1.2362979650497437,
+      "learning_rate": 6.714932024966574e-05,
+      "loss": 1.068,
+      "step": 31525
+    },
+    {
+      "epoch": 1.21385948026949,
+      "grad_norm": 1.221686601638794,
+      "learning_rate": 6.712076039577862e-05,
+      "loss": 0.897,
+      "step": 31530
+    },
+    {
+      "epoch": 1.2140519730510106,
+      "grad_norm": 1.4022241830825806,
+      "learning_rate": 6.709220354848018e-05,
+      "loss": 0.8292,
+      "step": 31535
+    },
+    {
+      "epoch": 1.2142444658325313,
+      "grad_norm": 1.1567379236221313,
+      "learning_rate": 6.706364971038173e-05,
+      "loss": 0.838,
+      "step": 31540
+    },
+    {
+      "epoch": 1.214436958614052,
+      "grad_norm": 1.1204020977020264,
+      "learning_rate": 6.70350988840943e-05,
+      "loss": 0.9189,
+      "step": 31545
+    },
+    {
+      "epoch": 1.2146294513955727,
+      "grad_norm": 1.358654260635376,
+      "learning_rate": 6.700655107222874e-05,
+      "loss": 0.8147,
+      "step": 31550
+    },
+    {
+      "epoch": 1.2148219441770933,
+      "grad_norm": 1.1854889392852783,
+      "learning_rate": 6.697800627739552e-05,
+      "loss": 0.7513,
+      "step": 31555
+    },
+    {
+      "epoch": 1.215014436958614,
+      "grad_norm": 1.1493016481399536,
+      "learning_rate": 6.694946450220483e-05,
+      "loss": 1.0312,
+      "step": 31560
+    },
+    {
+      "epoch": 1.2152069297401347,
+      "grad_norm": 1.5674629211425781,
+      "learning_rate": 6.692092574926673e-05,
+      "loss": 0.7985,
+      "step": 31565
+    },
+    {
+      "epoch": 1.2153994225216556,
+      "grad_norm": 1.1671700477600098,
+      "learning_rate": 6.689239002119084e-05,
+      "loss": 0.9128,
+      "step": 31570
+    },
+    {
+      "epoch": 1.2155919153031762,
+      "grad_norm": 1.3727703094482422,
+      "learning_rate": 6.686385732058651e-05,
+      "loss": 0.8517,
+      "step": 31575
+    },
+    {
+      "epoch": 1.215784408084697,
+      "grad_norm": 0.9869505763053894,
+      "learning_rate": 6.683532765006296e-05,
+      "loss": 0.8142,
+      "step": 31580
+    },
+    {
+      "epoch": 1.2159769008662176,
+      "grad_norm": 1.681252121925354,
+      "learning_rate": 6.680680101222901e-05,
+      "loss": 0.7783,
+      "step": 31585
+    },
+    {
+      "epoch": 1.2161693936477382,
+      "grad_norm": 1.7720850706100464,
+      "learning_rate": 6.677827740969317e-05,
+      "loss": 0.922,
+      "step": 31590
+    },
+    {
+      "epoch": 1.216361886429259,
+      "grad_norm": 1.2633975744247437,
+      "learning_rate": 6.674975684506384e-05,
+      "loss": 0.8357,
+      "step": 31595
+    },
+    {
+      "epoch": 1.2165543792107796,
+      "grad_norm": 1.2901790142059326,
+      "learning_rate": 6.672123932094898e-05,
+      "loss": 0.856,
+      "step": 31600
+    },
+    {
+      "epoch": 1.2167468719923002,
+      "grad_norm": 0.9209010004997253,
+      "learning_rate": 6.66927248399563e-05,
+      "loss": 0.8126,
+      "step": 31605
+    },
+    {
+      "epoch": 1.216939364773821,
+      "grad_norm": 1.8657188415527344,
+      "learning_rate": 6.666421340469331e-05,
+      "loss": 0.8026,
+      "step": 31610
+    },
+    {
+      "epoch": 1.2171318575553416,
+      "grad_norm": 0.8723480701446533,
+      "learning_rate": 6.66357050177672e-05,
+      "loss": 0.6801,
+      "step": 31615
+    },
+    {
+      "epoch": 1.2173243503368623,
+      "grad_norm": 1.3046056032180786,
+      "learning_rate": 6.660719968178479e-05,
+      "loss": 0.7513,
+      "step": 31620
+    },
+    {
+      "epoch": 1.2175168431183832,
+      "grad_norm": 1.1734675168991089,
+      "learning_rate": 6.657869739935282e-05,
+      "loss": 0.7972,
+      "step": 31625
+    },
+    {
+      "epoch": 1.2177093358999038,
+      "grad_norm": 1.3330581188201904,
+      "learning_rate": 6.655019817307758e-05,
+      "loss": 0.8381,
+      "step": 31630
+    },
+    {
+      "epoch": 1.2179018286814245,
+      "grad_norm": 1.2588646411895752,
+      "learning_rate": 6.652170200556508e-05,
+      "loss": 0.8309,
+      "step": 31635
+    },
+    {
+      "epoch": 1.2180943214629452,
+      "grad_norm": 1.5502039194107056,
+      "learning_rate": 6.649320889942121e-05,
+      "loss": 0.9067,
+      "step": 31640
+    },
+    {
+      "epoch": 1.2182868142444658,
+      "grad_norm": 1.2267836332321167,
+      "learning_rate": 6.646471885725139e-05,
+      "loss": 0.9091,
+      "step": 31645
+    },
+    {
+      "epoch": 1.2184793070259865,
+      "grad_norm": 0.8720155954360962,
+      "learning_rate": 6.643623188166088e-05,
+      "loss": 0.9406,
+      "step": 31650
+    },
+    {
+      "epoch": 1.2186717998075072,
+      "grad_norm": 2.29374623298645,
+      "learning_rate": 6.640774797525464e-05,
+      "loss": 0.976,
+      "step": 31655
+    },
+    {
+      "epoch": 1.2188642925890278,
+      "grad_norm": 1.3115676641464233,
+      "learning_rate": 6.637926714063734e-05,
+      "loss": 0.7881,
+      "step": 31660
+    },
+    {
+      "epoch": 1.2190567853705485,
+      "grad_norm": 1.2413272857666016,
+      "learning_rate": 6.635078938041328e-05,
+      "loss": 0.8841,
+      "step": 31665
+    },
+    {
+      "epoch": 1.2192492781520694,
+      "grad_norm": 1.0983569622039795,
+      "learning_rate": 6.632231469718668e-05,
+      "loss": 0.8903,
+      "step": 31670
+    },
+    {
+      "epoch": 1.21944177093359,
+      "grad_norm": 1.3633029460906982,
+      "learning_rate": 6.62938430935613e-05,
+      "loss": 0.7965,
+      "step": 31675
+    },
+    {
+      "epoch": 1.2196342637151107,
+      "grad_norm": 1.1858634948730469,
+      "learning_rate": 6.626537457214064e-05,
+      "loss": 0.7937,
+      "step": 31680
+    },
+    {
+      "epoch": 1.2198267564966314,
+      "grad_norm": 1.2882673740386963,
+      "learning_rate": 6.623690913552804e-05,
+      "loss": 0.9946,
+      "step": 31685
+    },
+    {
+      "epoch": 1.220019249278152,
+      "grad_norm": 0.9842979311943054,
+      "learning_rate": 6.620844678632641e-05,
+      "loss": 0.7972,
+      "step": 31690
+    },
+    {
+      "epoch": 1.2202117420596728,
+      "grad_norm": 1.0719999074935913,
+      "learning_rate": 6.617998752713852e-05,
+      "loss": 0.823,
+      "step": 31695
+    },
+    {
+      "epoch": 1.2204042348411934,
+      "grad_norm": 1.3692537546157837,
+      "learning_rate": 6.615153136056674e-05,
+      "loss": 0.7786,
+      "step": 31700
+    },
+    {
+      "epoch": 1.220596727622714,
+      "grad_norm": 1.2221060991287231,
+      "learning_rate": 6.612307828921313e-05,
+      "loss": 0.7953,
+      "step": 31705
+    },
+    {
+      "epoch": 1.2207892204042348,
+      "grad_norm": 1.3177229166030884,
+      "learning_rate": 6.609462831567964e-05,
+      "loss": 0.7488,
+      "step": 31710
+    },
+    {
+      "epoch": 1.2209817131857554,
+      "grad_norm": 0.9634941816329956,
+      "learning_rate": 6.60661814425678e-05,
+      "loss": 0.7675,
+      "step": 31715
+    },
+    {
+      "epoch": 1.221174205967276,
+      "grad_norm": 0.9454881548881531,
+      "learning_rate": 6.603773767247885e-05,
+      "loss": 0.7102,
+      "step": 31720
+    },
+    {
+      "epoch": 1.221366698748797,
+      "grad_norm": 1.579321026802063,
+      "learning_rate": 6.60092970080138e-05,
+      "loss": 0.9614,
+      "step": 31725
+    },
+    {
+      "epoch": 1.2215591915303177,
+      "grad_norm": 1.7888600826263428,
+      "learning_rate": 6.59808594517734e-05,
+      "loss": 0.96,
+      "step": 31730
+    },
+    {
+      "epoch": 1.2217516843118383,
+      "grad_norm": 1.8311741352081299,
+      "learning_rate": 6.595242500635804e-05,
+      "loss": 0.9041,
+      "step": 31735
+    },
+    {
+      "epoch": 1.221944177093359,
+      "grad_norm": 1.2068012952804565,
+      "learning_rate": 6.592399367436787e-05,
+      "loss": 0.9126,
+      "step": 31740
+    },
+    {
+      "epoch": 1.2221366698748797,
+      "grad_norm": 0.8985161781311035,
+      "learning_rate": 6.589556545840274e-05,
+      "loss": 0.9111,
+      "step": 31745
+    },
+    {
+      "epoch": 1.2223291626564003,
+      "grad_norm": 1.460936427116394,
+      "learning_rate": 6.586714036106221e-05,
+      "loss": 0.6821,
+      "step": 31750
+    },
+    {
+      "epoch": 1.222521655437921,
+      "grad_norm": 1.1933438777923584,
+      "learning_rate": 6.583871838494562e-05,
+      "loss": 0.7843,
+      "step": 31755
+    },
+    {
+      "epoch": 1.2227141482194417,
+      "grad_norm": 1.5719155073165894,
+      "learning_rate": 6.581029953265192e-05,
+      "loss": 0.8344,
+      "step": 31760
+    },
+    {
+      "epoch": 1.2229066410009626,
+      "grad_norm": 1.3877325057983398,
+      "learning_rate": 6.578188380677981e-05,
+      "loss": 0.7456,
+      "step": 31765
+    },
+    {
+      "epoch": 1.2230991337824832,
+      "grad_norm": 1.7668566703796387,
+      "learning_rate": 6.575347120992779e-05,
+      "loss": 1.0394,
+      "step": 31770
+    },
+    {
+      "epoch": 1.223291626564004,
+      "grad_norm": 1.813501000404358,
+      "learning_rate": 6.572506174469398e-05,
+      "loss": 0.8027,
+      "step": 31775
+    },
+    {
+      "epoch": 1.2234841193455246,
+      "grad_norm": 0.9930652976036072,
+      "learning_rate": 6.569665541367616e-05,
+      "loss": 0.9634,
+      "step": 31780
+    },
+    {
+      "epoch": 1.2236766121270453,
+      "grad_norm": 1.6402193307876587,
+      "learning_rate": 6.566825221947201e-05,
+      "loss": 0.8719,
+      "step": 31785
+    },
+    {
+      "epoch": 1.223869104908566,
+      "grad_norm": 2.223097562789917,
+      "learning_rate": 6.563985216467879e-05,
+      "loss": 0.8185,
+      "step": 31790
+    },
+    {
+      "epoch": 1.2240615976900866,
+      "grad_norm": 1.260113000869751,
+      "learning_rate": 6.56114552518934e-05,
+      "loss": 0.835,
+      "step": 31795
+    },
+    {
+      "epoch": 1.2242540904716073,
+      "grad_norm": 2.260995864868164,
+      "learning_rate": 6.558306148371269e-05,
+      "loss": 0.8712,
+      "step": 31800
+    },
+    {
+      "epoch": 1.224446583253128,
+      "grad_norm": 1.3736824989318848,
+      "learning_rate": 6.5554670862733e-05,
+      "loss": 0.8764,
+      "step": 31805
+    },
+    {
+      "epoch": 1.2246390760346486,
+      "grad_norm": 1.0253816843032837,
+      "learning_rate": 6.552628339155044e-05,
+      "loss": 0.8936,
+      "step": 31810
+    },
+    {
+      "epoch": 1.2248315688161693,
+      "grad_norm": 1.0663061141967773,
+      "learning_rate": 6.549789907276094e-05,
+      "loss": 0.8789,
+      "step": 31815
+    },
+    {
+      "epoch": 1.2250240615976902,
+      "grad_norm": 1.781774640083313,
+      "learning_rate": 6.546951790896003e-05,
+      "loss": 0.8116,
+      "step": 31820
+    },
+    {
+      "epoch": 1.2252165543792108,
+      "grad_norm": 1.6696699857711792,
+      "learning_rate": 6.544113990274291e-05,
+      "loss": 0.9066,
+      "step": 31825
+    },
+    {
+      "epoch": 1.2254090471607315,
+      "grad_norm": 1.6726038455963135,
+      "learning_rate": 6.541276505670466e-05,
+      "loss": 0.7497,
+      "step": 31830
+    },
+    {
+      "epoch": 1.2256015399422522,
+      "grad_norm": 1.851981282234192,
+      "learning_rate": 6.538439337343991e-05,
+      "loss": 0.98,
+      "step": 31835
+    },
+    {
+      "epoch": 1.2257940327237729,
+      "grad_norm": 1.3425476551055908,
+      "learning_rate": 6.535602485554307e-05,
+      "loss": 0.8501,
+      "step": 31840
+    },
+    {
+      "epoch": 1.2259865255052935,
+      "grad_norm": 1.2194117307662964,
+      "learning_rate": 6.532765950560827e-05,
+      "loss": 0.8375,
+      "step": 31845
+    },
+    {
+      "epoch": 1.2261790182868142,
+      "grad_norm": 1.3500804901123047,
+      "learning_rate": 6.529929732622932e-05,
+      "loss": 0.9349,
+      "step": 31850
+    },
+    {
+      "epoch": 1.2263715110683349,
+      "grad_norm": 1.5260719060897827,
+      "learning_rate": 6.527093831999977e-05,
+      "loss": 0.8235,
+      "step": 31855
+    },
+    {
+      "epoch": 1.2265640038498558,
+      "grad_norm": 1.9417953491210938,
+      "learning_rate": 6.524258248951285e-05,
+      "loss": 0.8308,
+      "step": 31860
+    },
+    {
+      "epoch": 1.2267564966313764,
+      "grad_norm": 1.3070498704910278,
+      "learning_rate": 6.521422983736151e-05,
+      "loss": 0.8632,
+      "step": 31865
+    },
+    {
+      "epoch": 1.226948989412897,
+      "grad_norm": 0.9319446086883545,
+      "learning_rate": 6.51858803661384e-05,
+      "loss": 0.8022,
+      "step": 31870
+    },
+    {
+      "epoch": 1.2271414821944178,
+      "grad_norm": 1.5047639608383179,
+      "learning_rate": 6.515753407843595e-05,
+      "loss": 0.9273,
+      "step": 31875
+    },
+    {
+      "epoch": 1.2273339749759384,
+      "grad_norm": 1.0930204391479492,
+      "learning_rate": 6.512919097684617e-05,
+      "loss": 0.8992,
+      "step": 31880
+    },
+    {
+      "epoch": 1.227526467757459,
+      "grad_norm": 1.2178794145584106,
+      "learning_rate": 6.510085106396085e-05,
+      "loss": 0.8333,
+      "step": 31885
+    },
+    {
+      "epoch": 1.2277189605389798,
+      "grad_norm": 1.62590491771698,
+      "learning_rate": 6.507251434237155e-05,
+      "loss": 0.7445,
+      "step": 31890
+    },
+    {
+      "epoch": 1.2279114533205004,
+      "grad_norm": 3.1164746284484863,
+      "learning_rate": 6.504418081466945e-05,
+      "loss": 1.0109,
+      "step": 31895
+    },
+    {
+      "epoch": 1.2281039461020211,
+      "grad_norm": 1.2229804992675781,
+      "learning_rate": 6.50158504834454e-05,
+      "loss": 0.8863,
+      "step": 31900
+    },
+    {
+      "epoch": 1.2282964388835418,
+      "grad_norm": 1.2148635387420654,
+      "learning_rate": 6.498752335129014e-05,
+      "loss": 0.817,
+      "step": 31905
+    },
+    {
+      "epoch": 1.2284889316650625,
+      "grad_norm": 0.9945757389068604,
+      "learning_rate": 6.495919942079391e-05,
+      "loss": 0.819,
+      "step": 31910
+    },
+    {
+      "epoch": 1.2286814244465833,
+      "grad_norm": 2.1877007484436035,
+      "learning_rate": 6.493087869454673e-05,
+      "loss": 0.8977,
+      "step": 31915
+    },
+    {
+      "epoch": 1.228873917228104,
+      "grad_norm": 2.055363178253174,
+      "learning_rate": 6.490256117513845e-05,
+      "loss": 0.9108,
+      "step": 31920
+    },
+    {
+      "epoch": 1.2290664100096247,
+      "grad_norm": 1.7472233772277832,
+      "learning_rate": 6.487424686515842e-05,
+      "loss": 1.0029,
+      "step": 31925
+    },
+    {
+      "epoch": 1.2292589027911454,
+      "grad_norm": 0.9974002838134766,
+      "learning_rate": 6.48459357671958e-05,
+      "loss": 0.8006,
+      "step": 31930
+    },
+    {
+      "epoch": 1.229451395572666,
+      "grad_norm": 1.9604872465133667,
+      "learning_rate": 6.481762788383951e-05,
+      "loss": 0.9948,
+      "step": 31935
+    },
+    {
+      "epoch": 1.2296438883541867,
+      "grad_norm": 0.9676084518432617,
+      "learning_rate": 6.478932321767808e-05,
+      "loss": 0.7514,
+      "step": 31940
+    },
+    {
+      "epoch": 1.2298363811357074,
+      "grad_norm": 2.2556610107421875,
+      "learning_rate": 6.476102177129978e-05,
+      "loss": 0.8251,
+      "step": 31945
+    },
+    {
+      "epoch": 1.230028873917228,
+      "grad_norm": 0.8798750042915344,
+      "learning_rate": 6.473272354729263e-05,
+      "loss": 0.772,
+      "step": 31950
+    },
+    {
+      "epoch": 1.2302213666987487,
+      "grad_norm": 1.4595062732696533,
+      "learning_rate": 6.470442854824425e-05,
+      "loss": 0.9378,
+      "step": 31955
+    },
+    {
+      "epoch": 1.2304138594802696,
+      "grad_norm": 0.9831540584564209,
+      "learning_rate": 6.467613677674212e-05,
+      "loss": 0.752,
+      "step": 31960
+    },
+    {
+      "epoch": 1.2306063522617903,
+      "grad_norm": 2.3248252868652344,
+      "learning_rate": 6.464784823537324e-05,
+      "loss": 0.8944,
+      "step": 31965
+    },
+    {
+      "epoch": 1.230798845043311,
+      "grad_norm": 1.3205488920211792,
+      "learning_rate": 6.461956292672447e-05,
+      "loss": 0.7635,
+      "step": 31970
+    },
+    {
+      "epoch": 1.2309913378248316,
+      "grad_norm": 2.114743232727051,
+      "learning_rate": 6.459128085338229e-05,
+      "loss": 0.8587,
+      "step": 31975
+    },
+    {
+      "epoch": 1.2311838306063523,
+      "grad_norm": 1.4038870334625244,
+      "learning_rate": 6.456300201793292e-05,
+      "loss": 0.8453,
+      "step": 31980
+    },
+    {
+      "epoch": 1.231376323387873,
+      "grad_norm": 1.8678901195526123,
+      "learning_rate": 6.453472642296226e-05,
+      "loss": 0.9314,
+      "step": 31985
+    },
+    {
+      "epoch": 1.2315688161693936,
+      "grad_norm": 1.9399640560150146,
+      "learning_rate": 6.450645407105594e-05,
+      "loss": 0.9189,
+      "step": 31990
+    },
+    {
+      "epoch": 1.2317613089509143,
+      "grad_norm": 1.849877953529358,
+      "learning_rate": 6.447818496479927e-05,
+      "loss": 0.8623,
+      "step": 31995
+    },
+    {
+      "epoch": 1.231953801732435,
+      "grad_norm": 0.9687787890434265,
+      "learning_rate": 6.444991910677725e-05,
+      "loss": 0.7928,
+      "step": 32000
+    },
+    {
+      "epoch": 1.2321462945139556,
+      "grad_norm": 1.2318834066390991,
+      "learning_rate": 6.442165649957467e-05,
+      "loss": 0.9324,
+      "step": 32005
+    },
+    {
+      "epoch": 1.2323387872954763,
+      "grad_norm": 1.2354921102523804,
+      "learning_rate": 6.439339714577592e-05,
+      "loss": 0.9398,
+      "step": 32010
+    },
+    {
+      "epoch": 1.2325312800769972,
+      "grad_norm": 1.4701021909713745,
+      "learning_rate": 6.436514104796507e-05,
+      "loss": 0.8559,
+      "step": 32015
+    },
+    {
+      "epoch": 1.2327237728585179,
+      "grad_norm": 0.9806448221206665,
+      "learning_rate": 6.433688820872607e-05,
+      "loss": 0.8319,
+      "step": 32020
+    },
+    {
+      "epoch": 1.2329162656400385,
+      "grad_norm": 1.2079403400421143,
+      "learning_rate": 6.430863863064238e-05,
+      "loss": 0.8578,
+      "step": 32025
+    },
+    {
+      "epoch": 1.2331087584215592,
+      "grad_norm": 1.240198016166687,
+      "learning_rate": 6.428039231629723e-05,
+      "loss": 0.8681,
+      "step": 32030
+    },
+    {
+      "epoch": 1.2333012512030799,
+      "grad_norm": 1.0847597122192383,
+      "learning_rate": 6.425214926827361e-05,
+      "loss": 0.8595,
+      "step": 32035
+    },
+    {
+      "epoch": 1.2334937439846005,
+      "grad_norm": 1.1608355045318604,
+      "learning_rate": 6.422390948915414e-05,
+      "loss": 0.7778,
+      "step": 32040
+    },
+    {
+      "epoch": 1.2336862367661212,
+      "grad_norm": 0.8766851425170898,
+      "learning_rate": 6.419567298152111e-05,
+      "loss": 0.8284,
+      "step": 32045
+    },
+    {
+      "epoch": 1.2338787295476419,
+      "grad_norm": 1.4966709613800049,
+      "learning_rate": 6.416743974795665e-05,
+      "loss": 0.7442,
+      "step": 32050
+    },
+    {
+      "epoch": 1.2340712223291628,
+      "grad_norm": 1.342774748802185,
+      "learning_rate": 6.413920979104244e-05,
+      "loss": 0.875,
+      "step": 32055
+    },
+    {
+      "epoch": 1.2342637151106834,
+      "grad_norm": 1.4193564653396606,
+      "learning_rate": 6.411098311335993e-05,
+      "loss": 0.7753,
+      "step": 32060
+    },
+    {
+      "epoch": 1.2344562078922041,
+      "grad_norm": 1.7735865116119385,
+      "learning_rate": 6.408275971749027e-05,
+      "loss": 0.7883,
+      "step": 32065
+    },
+    {
+      "epoch": 1.2346487006737248,
+      "grad_norm": 1.5808089971542358,
+      "learning_rate": 6.405453960601432e-05,
+      "loss": 0.7633,
+      "step": 32070
+    },
+    {
+      "epoch": 1.2348411934552455,
+      "grad_norm": 1.0840727090835571,
+      "learning_rate": 6.402632278151259e-05,
+      "loss": 0.8083,
+      "step": 32075
+    },
+    {
+      "epoch": 1.2350336862367661,
+      "grad_norm": 1.475110650062561,
+      "learning_rate": 6.399810924656537e-05,
+      "loss": 0.6657,
+      "step": 32080
+    },
+    {
+      "epoch": 1.2352261790182868,
+      "grad_norm": 0.9028719663619995,
+      "learning_rate": 6.396989900375256e-05,
+      "loss": 0.9991,
+      "step": 32085
+    },
+    {
+      "epoch": 1.2354186717998075,
+      "grad_norm": 1.3939464092254639,
+      "learning_rate": 6.394169205565377e-05,
+      "loss": 0.8523,
+      "step": 32090
+    },
+    {
+      "epoch": 1.2356111645813281,
+      "grad_norm": 1.7557820081710815,
+      "learning_rate": 6.391348840484841e-05,
+      "loss": 0.7734,
+      "step": 32095
+    },
+    {
+      "epoch": 1.2358036573628488,
+      "grad_norm": 1.2897764444351196,
+      "learning_rate": 6.388528805391548e-05,
+      "loss": 0.7278,
+      "step": 32100
+    },
+    {
+      "epoch": 1.2359961501443695,
+      "grad_norm": 1.3506861925125122,
+      "learning_rate": 6.38570910054337e-05,
+      "loss": 0.795,
+      "step": 32105
+    },
+    {
+      "epoch": 1.2361886429258904,
+      "grad_norm": 2.072033405303955,
+      "learning_rate": 6.382889726198154e-05,
+      "loss": 0.8527,
+      "step": 32110
+    },
+    {
+      "epoch": 1.236381135707411,
+      "grad_norm": 1.8020820617675781,
+      "learning_rate": 6.380070682613711e-05,
+      "loss": 0.9461,
+      "step": 32115
+    },
+    {
+      "epoch": 1.2365736284889317,
+      "grad_norm": 2.3181891441345215,
+      "learning_rate": 6.377251970047822e-05,
+      "loss": 0.8269,
+      "step": 32120
+    },
+    {
+      "epoch": 1.2367661212704524,
+      "grad_norm": 1.5341979265213013,
+      "learning_rate": 6.374433588758246e-05,
+      "loss": 0.9718,
+      "step": 32125
+    },
+    {
+      "epoch": 1.236958614051973,
+      "grad_norm": 1.0951958894729614,
+      "learning_rate": 6.3716155390027e-05,
+      "loss": 0.8376,
+      "step": 32130
+    },
+    {
+      "epoch": 1.2371511068334937,
+      "grad_norm": 1.2054345607757568,
+      "learning_rate": 6.368797821038874e-05,
+      "loss": 0.7169,
+      "step": 32135
+    },
+    {
+      "epoch": 1.2373435996150144,
+      "grad_norm": 1.4558501243591309,
+      "learning_rate": 6.365980435124435e-05,
+      "loss": 0.8068,
+      "step": 32140
+    },
+    {
+      "epoch": 1.237536092396535,
+      "grad_norm": 1.592523217201233,
+      "learning_rate": 6.363163381517015e-05,
+      "loss": 0.8709,
+      "step": 32145
+    },
+    {
+      "epoch": 1.2377285851780557,
+      "grad_norm": 1.191009759902954,
+      "learning_rate": 6.360346660474206e-05,
+      "loss": 0.9404,
+      "step": 32150
+    },
+    {
+      "epoch": 1.2379210779595766,
+      "grad_norm": 1.2547787427902222,
+      "learning_rate": 6.357530272253587e-05,
+      "loss": 0.7077,
+      "step": 32155
+    },
+    {
+      "epoch": 1.2381135707410973,
+      "grad_norm": 1.442322850227356,
+      "learning_rate": 6.354714217112698e-05,
+      "loss": 0.8027,
+      "step": 32160
+    },
+    {
+      "epoch": 1.238306063522618,
+      "grad_norm": 1.3657112121582031,
+      "learning_rate": 6.351898495309041e-05,
+      "loss": 0.8016,
+      "step": 32165
+    },
+    {
+      "epoch": 1.2384985563041386,
+      "grad_norm": 1.7618088722229004,
+      "learning_rate": 6.349083107100104e-05,
+      "loss": 0.8553,
+      "step": 32170
+    },
+    {
+      "epoch": 1.2386910490856593,
+      "grad_norm": 1.7059003114700317,
+      "learning_rate": 6.346268052743331e-05,
+      "loss": 0.5883,
+      "step": 32175
+    },
+    {
+      "epoch": 1.23888354186718,
+      "grad_norm": 1.33328115940094,
+      "learning_rate": 6.343453332496141e-05,
+      "loss": 0.7425,
+      "step": 32180
+    },
+    {
+      "epoch": 1.2390760346487006,
+      "grad_norm": 1.3579884767532349,
+      "learning_rate": 6.340638946615922e-05,
+      "loss": 0.7521,
+      "step": 32185
+    },
+    {
+      "epoch": 1.2392685274302213,
+      "grad_norm": 1.7783567905426025,
+      "learning_rate": 6.337824895360028e-05,
+      "loss": 0.8292,
+      "step": 32190
+    },
+    {
+      "epoch": 1.239461020211742,
+      "grad_norm": 1.4413247108459473,
+      "learning_rate": 6.335011178985788e-05,
+      "loss": 0.781,
+      "step": 32195
+    },
+    {
+      "epoch": 1.2396535129932627,
+      "grad_norm": 1.817164421081543,
+      "learning_rate": 6.3321977977505e-05,
+      "loss": 0.8307,
+      "step": 32200
+    },
+    {
+      "epoch": 1.2398460057747833,
+      "grad_norm": 0.7775023579597473,
+      "learning_rate": 6.329384751911422e-05,
+      "loss": 0.8496,
+      "step": 32205
+    },
+    {
+      "epoch": 1.2400384985563042,
+      "grad_norm": 1.8241838216781616,
+      "learning_rate": 6.326572041725795e-05,
+      "loss": 0.8688,
+      "step": 32210
+    },
+    {
+      "epoch": 1.2402309913378249,
+      "grad_norm": 0.9852082133293152,
+      "learning_rate": 6.323759667450824e-05,
+      "loss": 0.8371,
+      "step": 32215
+    },
+    {
+      "epoch": 1.2404234841193456,
+      "grad_norm": 1.4170035123825073,
+      "learning_rate": 6.320947629343672e-05,
+      "loss": 1.0588,
+      "step": 32220
+    },
+    {
+      "epoch": 1.2406159769008662,
+      "grad_norm": 1.4265270233154297,
+      "learning_rate": 6.318135927661493e-05,
+      "loss": 0.8352,
+      "step": 32225
+    },
+    {
+      "epoch": 1.240808469682387,
+      "grad_norm": 0.9820341467857361,
+      "learning_rate": 6.315324562661393e-05,
+      "loss": 0.9075,
+      "step": 32230
+    },
+    {
+      "epoch": 1.2410009624639076,
+      "grad_norm": 1.0748177766799927,
+      "learning_rate": 6.312513534600448e-05,
+      "loss": 0.9917,
+      "step": 32235
+    },
+    {
+      "epoch": 1.2411934552454282,
+      "grad_norm": 1.1158158779144287,
+      "learning_rate": 6.309702843735719e-05,
+      "loss": 0.851,
+      "step": 32240
+    },
+    {
+      "epoch": 1.241385948026949,
+      "grad_norm": 1.3649555444717407,
+      "learning_rate": 6.306892490324217e-05,
+      "loss": 0.9673,
+      "step": 32245
+    },
+    {
+      "epoch": 1.2415784408084698,
+      "grad_norm": 1.7820628881454468,
+      "learning_rate": 6.30408247462293e-05,
+      "loss": 0.8305,
+      "step": 32250
+    },
+    {
+      "epoch": 1.2417709335899905,
+      "grad_norm": 1.373089075088501,
+      "learning_rate": 6.30127279688882e-05,
+      "loss": 0.881,
+      "step": 32255
+    },
+    {
+      "epoch": 1.2419634263715111,
+      "grad_norm": 1.3298135995864868,
+      "learning_rate": 6.298463457378812e-05,
+      "loss": 0.9557,
+      "step": 32260
+    },
+    {
+      "epoch": 1.2421559191530318,
+      "grad_norm": 1.0167731046676636,
+      "learning_rate": 6.295654456349794e-05,
+      "loss": 0.9262,
+      "step": 32265
+    },
+    {
+      "epoch": 1.2423484119345525,
+      "grad_norm": 0.9300015568733215,
+      "learning_rate": 6.292845794058644e-05,
+      "loss": 0.8913,
+      "step": 32270
+    },
+    {
+      "epoch": 1.2425409047160731,
+      "grad_norm": 2.170581102371216,
+      "learning_rate": 6.290037470762186e-05,
+      "loss": 0.8422,
+      "step": 32275
+    },
+    {
+      "epoch": 1.2427333974975938,
+      "grad_norm": 2.3894753456115723,
+      "learning_rate": 6.287229486717224e-05,
+      "loss": 0.8744,
+      "step": 32280
+    },
+    {
+      "epoch": 1.2429258902791145,
+      "grad_norm": 1.2343543767929077,
+      "learning_rate": 6.284421842180534e-05,
+      "loss": 0.757,
+      "step": 32285
+    },
+    {
+      "epoch": 1.2431183830606352,
+      "grad_norm": 1.0302842855453491,
+      "learning_rate": 6.281614537408849e-05,
+      "loss": 0.8185,
+      "step": 32290
+    },
+    {
+      "epoch": 1.2433108758421558,
+      "grad_norm": 1.0366548299789429,
+      "learning_rate": 6.278807572658883e-05,
+      "loss": 0.9143,
+      "step": 32295
+    },
+    {
+      "epoch": 1.2435033686236765,
+      "grad_norm": 1.3391799926757812,
+      "learning_rate": 6.276000948187317e-05,
+      "loss": 0.8389,
+      "step": 32300
+    },
+    {
+      "epoch": 1.2436958614051974,
+      "grad_norm": 1.0500547885894775,
+      "learning_rate": 6.273194664250795e-05,
+      "loss": 0.7855,
+      "step": 32305
+    },
+    {
+      "epoch": 1.243888354186718,
+      "grad_norm": 1.3002054691314697,
+      "learning_rate": 6.270388721105927e-05,
+      "loss": 0.8311,
+      "step": 32310
+    },
+    {
+      "epoch": 1.2440808469682387,
+      "grad_norm": 1.2565590143203735,
+      "learning_rate": 6.267583119009309e-05,
+      "loss": 0.7857,
+      "step": 32315
+    },
+    {
+      "epoch": 1.2442733397497594,
+      "grad_norm": 1.8590493202209473,
+      "learning_rate": 6.26477785821749e-05,
+      "loss": 0.9076,
+      "step": 32320
+    },
+    {
+      "epoch": 1.24446583253128,
+      "grad_norm": 2.736550807952881,
+      "learning_rate": 6.261972938986989e-05,
+      "loss": 0.8985,
+      "step": 32325
+    },
+    {
+      "epoch": 1.2446583253128007,
+      "grad_norm": 1.4446461200714111,
+      "learning_rate": 6.259168361574303e-05,
+      "loss": 0.848,
+      "step": 32330
+    },
+    {
+      "epoch": 1.2448508180943214,
+      "grad_norm": 2.1344213485717773,
+      "learning_rate": 6.256364126235892e-05,
+      "loss": 0.9543,
+      "step": 32335
+    },
+    {
+      "epoch": 1.245043310875842,
+      "grad_norm": 1.068670392036438,
+      "learning_rate": 6.253560233228176e-05,
+      "loss": 0.8341,
+      "step": 32340
+    },
+    {
+      "epoch": 1.245235803657363,
+      "grad_norm": 0.9258108735084534,
+      "learning_rate": 6.250756682807564e-05,
+      "loss": 0.7102,
+      "step": 32345
+    },
+    {
+      "epoch": 1.2454282964388836,
+      "grad_norm": 1.4921131134033203,
+      "learning_rate": 6.247953475230417e-05,
+      "loss": 0.854,
+      "step": 32350
+    },
+    {
+      "epoch": 1.2456207892204043,
+      "grad_norm": 1.102432131767273,
+      "learning_rate": 6.245150610753066e-05,
+      "loss": 0.9141,
+      "step": 32355
+    },
+    {
+      "epoch": 1.245813282001925,
+      "grad_norm": 1.5257620811462402,
+      "learning_rate": 6.24234808963182e-05,
+      "loss": 0.8949,
+      "step": 32360
+    },
+    {
+      "epoch": 1.2460057747834457,
+      "grad_norm": 2.053269624710083,
+      "learning_rate": 6.239545912122951e-05,
+      "loss": 0.9681,
+      "step": 32365
+    },
+    {
+      "epoch": 1.2461982675649663,
+      "grad_norm": 2.117687702178955,
+      "learning_rate": 6.236744078482693e-05,
+      "loss": 0.8536,
+      "step": 32370
+    },
+    {
+      "epoch": 1.246390760346487,
+      "grad_norm": 1.0575151443481445,
+      "learning_rate": 6.233942588967264e-05,
+      "loss": 0.6996,
+      "step": 32375
+    },
+    {
+      "epoch": 1.2465832531280077,
+      "grad_norm": 1.6774235963821411,
+      "learning_rate": 6.231141443832835e-05,
+      "loss": 0.7842,
+      "step": 32380
+    },
+    {
+      "epoch": 1.2467757459095283,
+      "grad_norm": 0.9772216081619263,
+      "learning_rate": 6.228340643335554e-05,
+      "loss": 0.825,
+      "step": 32385
+    },
+    {
+      "epoch": 1.246968238691049,
+      "grad_norm": 1.221349835395813,
+      "learning_rate": 6.225540187731538e-05,
+      "loss": 0.8491,
+      "step": 32390
+    },
+    {
+      "epoch": 1.2471607314725697,
+      "grad_norm": 1.2611428499221802,
+      "learning_rate": 6.222740077276869e-05,
+      "loss": 0.8939,
+      "step": 32395
+    },
+    {
+      "epoch": 1.2473532242540906,
+      "grad_norm": 1.5732476711273193,
+      "learning_rate": 6.219940312227596e-05,
+      "loss": 0.7606,
+      "step": 32400
+    },
+    {
+      "epoch": 1.2475457170356112,
+      "grad_norm": 1.725197196006775,
+      "learning_rate": 6.217140892839744e-05,
+      "loss": 0.8467,
+      "step": 32405
+    },
+    {
+      "epoch": 1.247738209817132,
+      "grad_norm": 1.427253246307373,
+      "learning_rate": 6.214341819369294e-05,
+      "loss": 0.7289,
+      "step": 32410
+    },
+    {
+      "epoch": 1.2479307025986526,
+      "grad_norm": 1.0685710906982422,
+      "learning_rate": 6.211543092072209e-05,
+      "loss": 0.9277,
+      "step": 32415
+    },
+    {
+      "epoch": 1.2481231953801732,
+      "grad_norm": 1.519148826599121,
+      "learning_rate": 6.208744711204413e-05,
+      "loss": 0.9213,
+      "step": 32420
+    },
+    {
+      "epoch": 1.248315688161694,
+      "grad_norm": 0.941789448261261,
+      "learning_rate": 6.205946677021797e-05,
+      "loss": 0.8961,
+      "step": 32425
+    },
+    {
+      "epoch": 1.2485081809432146,
+      "grad_norm": 1.4877368211746216,
+      "learning_rate": 6.203148989780223e-05,
+      "loss": 0.8506,
+      "step": 32430
+    },
+    {
+      "epoch": 1.2487006737247353,
+      "grad_norm": 1.1104483604431152,
+      "learning_rate": 6.200351649735524e-05,
+      "loss": 0.859,
+      "step": 32435
+    },
+    {
+      "epoch": 1.248893166506256,
+      "grad_norm": 1.6220526695251465,
+      "learning_rate": 6.197554657143496e-05,
+      "loss": 0.978,
+      "step": 32440
+    },
+    {
+      "epoch": 1.2490856592877768,
+      "grad_norm": 1.7555328607559204,
+      "learning_rate": 6.194758012259903e-05,
+      "loss": 1.0804,
+      "step": 32445
+    },
+    {
+      "epoch": 1.2492781520692975,
+      "grad_norm": 1.939076542854309,
+      "learning_rate": 6.191961715340487e-05,
+      "loss": 0.8503,
+      "step": 32450
+    },
+    {
+      "epoch": 1.2494706448508182,
+      "grad_norm": 1.9745993614196777,
+      "learning_rate": 6.18916576664094e-05,
+      "loss": 0.9226,
+      "step": 32455
+    },
+    {
+      "epoch": 1.2496631376323388,
+      "grad_norm": 0.7877691984176636,
+      "learning_rate": 6.186370166416943e-05,
+      "loss": 0.8174,
+      "step": 32460
+    },
+    {
+      "epoch": 1.2498556304138595,
+      "grad_norm": 1.0937036275863647,
+      "learning_rate": 6.183574914924135e-05,
+      "loss": 0.9914,
+      "step": 32465
+    },
+    {
+      "epoch": 1.2500481231953802,
+      "grad_norm": 1.2272846698760986,
+      "learning_rate": 6.180780012418112e-05,
+      "loss": 0.8287,
+      "step": 32470
+    },
+    {
+      "epoch": 1.2502406159769008,
+      "grad_norm": 1.2231870889663696,
+      "learning_rate": 6.177985459154462e-05,
+      "loss": 0.889,
+      "step": 32475
+    },
+    {
+      "epoch": 1.2504331087584215,
+      "grad_norm": 1.420955777168274,
+      "learning_rate": 6.175191255388723e-05,
+      "loss": 0.9343,
+      "step": 32480
+    },
+    {
+      "epoch": 1.2506256015399422,
+      "grad_norm": 0.7461833953857422,
+      "learning_rate": 6.172397401376404e-05,
+      "loss": 0.7098,
+      "step": 32485
+    },
+    {
+      "epoch": 1.2508180943214628,
+      "grad_norm": 0.9062157869338989,
+      "learning_rate": 6.169603897372988e-05,
+      "loss": 0.8073,
+      "step": 32490
+    },
+    {
+      "epoch": 1.2510105871029835,
+      "grad_norm": 0.5169772505760193,
+      "learning_rate": 6.166810743633924e-05,
+      "loss": 0.9059,
+      "step": 32495
+    },
+    {
+      "epoch": 1.2512030798845044,
+      "grad_norm": 1.1018903255462646,
+      "learning_rate": 6.164017940414621e-05,
+      "loss": 0.9242,
+      "step": 32500
+    },
+    {
+      "epoch": 1.251395572666025,
+      "grad_norm": 1.6499050855636597,
+      "learning_rate": 6.16122548797047e-05,
+      "loss": 0.7794,
+      "step": 32505
+    },
+    {
+      "epoch": 1.2515880654475458,
+      "grad_norm": 1.915618658065796,
+      "learning_rate": 6.158433386556817e-05,
+      "loss": 0.8513,
+      "step": 32510
+    },
+    {
+      "epoch": 1.2517805582290664,
+      "grad_norm": 1.2275880575180054,
+      "learning_rate": 6.155641636428981e-05,
+      "loss": 0.8407,
+      "step": 32515
+    },
+    {
+      "epoch": 1.251973051010587,
+      "grad_norm": 1.4273251295089722,
+      "learning_rate": 6.152850237842255e-05,
+      "loss": 0.7142,
+      "step": 32520
+    },
+    {
+      "epoch": 1.2521655437921078,
+      "grad_norm": 1.4185856580734253,
+      "learning_rate": 6.15005919105189e-05,
+      "loss": 0.7348,
+      "step": 32525
+    },
+    {
+      "epoch": 1.2523580365736284,
+      "grad_norm": 1.3497728109359741,
+      "learning_rate": 6.147268496313101e-05,
+      "loss": 0.9843,
+      "step": 32530
+    },
+    {
+      "epoch": 1.252550529355149,
+      "grad_norm": 1.5742275714874268,
+      "learning_rate": 6.144478153881093e-05,
+      "loss": 0.9799,
+      "step": 32535
+    },
+    {
+      "epoch": 1.25274302213667,
+      "grad_norm": 1.4493976831436157,
+      "learning_rate": 6.141688164011014e-05,
+      "loss": 0.8815,
+      "step": 32540
+    },
+    {
+      "epoch": 1.2529355149181907,
+      "grad_norm": 1.0895041227340698,
+      "learning_rate": 6.138898526957993e-05,
+      "loss": 0.8332,
+      "step": 32545
+    },
+    {
+      "epoch": 1.2531280076997113,
+      "grad_norm": 2.0370841026306152,
+      "learning_rate": 6.136109242977126e-05,
+      "loss": 1.0292,
+      "step": 32550
+    },
+    {
+      "epoch": 1.253320500481232,
+      "grad_norm": 1.0882296562194824,
+      "learning_rate": 6.133320312323473e-05,
+      "loss": 0.7347,
+      "step": 32555
+    },
+    {
+      "epoch": 1.2535129932627527,
+      "grad_norm": 1.2545933723449707,
+      "learning_rate": 6.13053173525206e-05,
+      "loss": 0.7906,
+      "step": 32560
+    },
+    {
+      "epoch": 1.2537054860442733,
+      "grad_norm": 1.7179350852966309,
+      "learning_rate": 6.127743512017892e-05,
+      "loss": 0.9072,
+      "step": 32565
+    },
+    {
+      "epoch": 1.253897978825794,
+      "grad_norm": 1.0224990844726562,
+      "learning_rate": 6.124955642875927e-05,
+      "loss": 0.8066,
+      "step": 32570
+    },
+    {
+      "epoch": 1.2540904716073147,
+      "grad_norm": 1.681639313697815,
+      "learning_rate": 6.122168128081096e-05,
+      "loss": 0.9829,
+      "step": 32575
+    },
+    {
+      "epoch": 1.2542829643888354,
+      "grad_norm": 2.0990216732025146,
+      "learning_rate": 6.119380967888304e-05,
+      "loss": 1.0276,
+      "step": 32580
+    },
+    {
+      "epoch": 1.254475457170356,
+      "grad_norm": 1.4491196870803833,
+      "learning_rate": 6.116594162552416e-05,
+      "loss": 0.8186,
+      "step": 32585
+    },
+    {
+      "epoch": 1.2546679499518767,
+      "grad_norm": 1.1606724262237549,
+      "learning_rate": 6.113807712328261e-05,
+      "loss": 0.857,
+      "step": 32590
+    },
+    {
+      "epoch": 1.2548604427333974,
+      "grad_norm": 1.064584493637085,
+      "learning_rate": 6.111021617470654e-05,
+      "loss": 0.7753,
+      "step": 32595
+    },
+    {
+      "epoch": 1.2550529355149183,
+      "grad_norm": 1.3536344766616821,
+      "learning_rate": 6.108235878234357e-05,
+      "loss": 0.8481,
+      "step": 32600
+    },
+    {
+      "epoch": 1.255245428296439,
+      "grad_norm": 1.1210103034973145,
+      "learning_rate": 6.105450494874105e-05,
+      "loss": 0.8809,
+      "step": 32605
+    },
+    {
+      "epoch": 1.2554379210779596,
+      "grad_norm": 1.9485646486282349,
+      "learning_rate": 6.102665467644607e-05,
+      "loss": 0.909,
+      "step": 32610
+    },
+    {
+      "epoch": 1.2556304138594803,
+      "grad_norm": 0.9706501960754395,
+      "learning_rate": 6.099880796800537e-05,
+      "loss": 0.9546,
+      "step": 32615
+    },
+    {
+      "epoch": 1.255822906641001,
+      "grad_norm": 0.9644052386283875,
+      "learning_rate": 6.097096482596529e-05,
+      "loss": 0.7045,
+      "step": 32620
+    },
+    {
+      "epoch": 1.2560153994225216,
+      "grad_norm": 1.0382925271987915,
+      "learning_rate": 6.094312525287197e-05,
+      "loss": 0.7066,
+      "step": 32625
+    },
+    {
+      "epoch": 1.2562078922040423,
+      "grad_norm": 1.0624562501907349,
+      "learning_rate": 6.0915289251271076e-05,
+      "loss": 0.7864,
+      "step": 32630
+    },
+    {
+      "epoch": 1.2564003849855632,
+      "grad_norm": 0.8789560794830322,
+      "learning_rate": 6.088745682370809e-05,
+      "loss": 0.9018,
+      "step": 32635
+    },
+    {
+      "epoch": 1.2565928777670838,
+      "grad_norm": 1.3996667861938477,
+      "learning_rate": 6.0859627972728086e-05,
+      "loss": 0.8411,
+      "step": 32640
+    },
+    {
+      "epoch": 1.2567853705486045,
+      "grad_norm": 1.1187728643417358,
+      "learning_rate": 6.083180270087583e-05,
+      "loss": 0.7984,
+      "step": 32645
+    },
+    {
+      "epoch": 1.2569778633301252,
+      "grad_norm": 1.481068730354309,
+      "learning_rate": 6.080398101069571e-05,
+      "loss": 0.7356,
+      "step": 32650
+    },
+    {
+      "epoch": 1.2571703561116458,
+      "grad_norm": 1.1698410511016846,
+      "learning_rate": 6.0776162904731915e-05,
+      "loss": 0.9059,
+      "step": 32655
+    },
+    {
+      "epoch": 1.2573628488931665,
+      "grad_norm": 1.0608761310577393,
+      "learning_rate": 6.0748348385528185e-05,
+      "loss": 0.8176,
+      "step": 32660
+    },
+    {
+      "epoch": 1.2575553416746872,
+      "grad_norm": 1.4496049880981445,
+      "learning_rate": 6.0720537455627944e-05,
+      "loss": 1.139,
+      "step": 32665
+    },
+    {
+      "epoch": 1.2577478344562079,
+      "grad_norm": 1.7745342254638672,
+      "learning_rate": 6.069273011757439e-05,
+      "loss": 0.924,
+      "step": 32670
+    },
+    {
+      "epoch": 1.2579403272377285,
+      "grad_norm": 1.3103641271591187,
+      "learning_rate": 6.066492637391028e-05,
+      "loss": 0.9955,
+      "step": 32675
+    },
+    {
+      "epoch": 1.2581328200192492,
+      "grad_norm": 1.4893795251846313,
+      "learning_rate": 6.063712622717803e-05,
+      "loss": 0.7122,
+      "step": 32680
+    },
+    {
+      "epoch": 1.2583253128007699,
+      "grad_norm": 1.8530791997909546,
+      "learning_rate": 6.060932967991988e-05,
+      "loss": 0.7563,
+      "step": 32685
+    },
+    {
+      "epoch": 1.2585178055822905,
+      "grad_norm": 0.8076843619346619,
+      "learning_rate": 6.058153673467759e-05,
+      "loss": 0.7344,
+      "step": 32690
+    },
+    {
+      "epoch": 1.2587102983638114,
+      "grad_norm": 1.2442243099212646,
+      "learning_rate": 6.055374739399261e-05,
+      "loss": 0.8517,
+      "step": 32695
+    },
+    {
+      "epoch": 1.258902791145332,
+      "grad_norm": 1.8538851737976074,
+      "learning_rate": 6.052596166040616e-05,
+      "loss": 0.799,
+      "step": 32700
+    },
+    {
+      "epoch": 1.2590952839268528,
+      "grad_norm": 1.2174726724624634,
+      "learning_rate": 6.049817953645902e-05,
+      "loss": 0.734,
+      "step": 32705
+    },
+    {
+      "epoch": 1.2592877767083734,
+      "grad_norm": 1.1436805725097656,
+      "learning_rate": 6.0470401024691646e-05,
+      "loss": 0.8512,
+      "step": 32710
+    },
+    {
+      "epoch": 1.259480269489894,
+      "grad_norm": 0.9243611097335815,
+      "learning_rate": 6.044262612764429e-05,
+      "loss": 0.7716,
+      "step": 32715
+    },
+    {
+      "epoch": 1.2596727622714148,
+      "grad_norm": 1.3061074018478394,
+      "learning_rate": 6.0414854847856694e-05,
+      "loss": 0.7524,
+      "step": 32720
+    },
+    {
+      "epoch": 1.2598652550529355,
+      "grad_norm": 1.4081029891967773,
+      "learning_rate": 6.038708718786843e-05,
+      "loss": 0.8902,
+      "step": 32725
+    },
+    {
+      "epoch": 1.2600577478344563,
+      "grad_norm": 1.0270228385925293,
+      "learning_rate": 6.0359323150218616e-05,
+      "loss": 1.0107,
+      "step": 32730
+    },
+    {
+      "epoch": 1.260250240615977,
+      "grad_norm": 1.0451138019561768,
+      "learning_rate": 6.033156273744607e-05,
+      "loss": 0.9819,
+      "step": 32735
+    },
+    {
+      "epoch": 1.2604427333974977,
+      "grad_norm": 0.9528657793998718,
+      "learning_rate": 6.030380595208938e-05,
+      "loss": 0.8202,
+      "step": 32740
+    },
+    {
+      "epoch": 1.2606352261790184,
+      "grad_norm": 2.1417109966278076,
+      "learning_rate": 6.027605279668666e-05,
+      "loss": 0.8817,
+      "step": 32745
+    },
+    {
+      "epoch": 1.260827718960539,
+      "grad_norm": 1.5614452362060547,
+      "learning_rate": 6.024830327377573e-05,
+      "loss": 0.8378,
+      "step": 32750
+    },
+    {
+      "epoch": 1.2610202117420597,
+      "grad_norm": 1.5928235054016113,
+      "learning_rate": 6.0220557385894184e-05,
+      "loss": 0.9096,
+      "step": 32755
+    },
+    {
+      "epoch": 1.2612127045235804,
+      "grad_norm": 0.8466683030128479,
+      "learning_rate": 6.019281513557913e-05,
+      "loss": 0.9457,
+      "step": 32760
+    },
+    {
+      "epoch": 1.261405197305101,
+      "grad_norm": 1.4196317195892334,
+      "learning_rate": 6.01650765253674e-05,
+      "loss": 0.7927,
+      "step": 32765
+    },
+    {
+      "epoch": 1.2615976900866217,
+      "grad_norm": 0.9929688572883606,
+      "learning_rate": 6.01373415577956e-05,
+      "loss": 0.8208,
+      "step": 32770
+    },
+    {
+      "epoch": 1.2617901828681424,
+      "grad_norm": 1.321770429611206,
+      "learning_rate": 6.0109610235399826e-05,
+      "loss": 0.9258,
+      "step": 32775
+    },
+    {
+      "epoch": 1.261982675649663,
+      "grad_norm": 1.1568975448608398,
+      "learning_rate": 6.008188256071592e-05,
+      "loss": 0.9201,
+      "step": 32780
+    },
+    {
+      "epoch": 1.2621751684311837,
+      "grad_norm": 1.2595579624176025,
+      "learning_rate": 6.0054158536279446e-05,
+      "loss": 0.8037,
+      "step": 32785
+    },
+    {
+      "epoch": 1.2623676612127046,
+      "grad_norm": 1.0337523221969604,
+      "learning_rate": 6.0026438164625565e-05,
+      "loss": 0.7372,
+      "step": 32790
+    },
+    {
+      "epoch": 1.2625601539942253,
+      "grad_norm": 2.097609758377075,
+      "learning_rate": 5.999872144828907e-05,
+      "loss": 0.9208,
+      "step": 32795
+    },
+    {
+      "epoch": 1.262752646775746,
+      "grad_norm": 0.9220959544181824,
+      "learning_rate": 5.997100838980456e-05,
+      "loss": 0.7774,
+      "step": 32800
+    },
+    {
+      "epoch": 1.2629451395572666,
+      "grad_norm": 0.9968023896217346,
+      "learning_rate": 5.994329899170617e-05,
+      "loss": 0.8029,
+      "step": 32805
+    },
+    {
+      "epoch": 1.2631376323387873,
+      "grad_norm": 2.015770196914673,
+      "learning_rate": 5.991559325652769e-05,
+      "loss": 0.8083,
+      "step": 32810
+    },
+    {
+      "epoch": 1.263330125120308,
+      "grad_norm": 1.5899488925933838,
+      "learning_rate": 5.988789118680272e-05,
+      "loss": 0.7113,
+      "step": 32815
+    },
+    {
+      "epoch": 1.2635226179018286,
+      "grad_norm": 1.1025646924972534,
+      "learning_rate": 5.986019278506436e-05,
+      "loss": 0.8005,
+      "step": 32820
+    },
+    {
+      "epoch": 1.2637151106833493,
+      "grad_norm": 1.4543554782867432,
+      "learning_rate": 5.983249805384545e-05,
+      "loss": 0.7893,
+      "step": 32825
+    },
+    {
+      "epoch": 1.2639076034648702,
+      "grad_norm": 1.1230753660202026,
+      "learning_rate": 5.980480699567854e-05,
+      "loss": 0.9324,
+      "step": 32830
+    },
+    {
+      "epoch": 1.2641000962463909,
+      "grad_norm": 1.583644986152649,
+      "learning_rate": 5.9777119613095755e-05,
+      "loss": 0.7477,
+      "step": 32835
+    },
+    {
+      "epoch": 1.2642925890279115,
+      "grad_norm": 1.9501726627349854,
+      "learning_rate": 5.9749435908628915e-05,
+      "loss": 0.8344,
+      "step": 32840
+    },
+    {
+      "epoch": 1.2644850818094322,
+      "grad_norm": 0.6011807918548584,
+      "learning_rate": 5.972175588480954e-05,
+      "loss": 0.7556,
+      "step": 32845
+    },
+    {
+      "epoch": 1.2646775745909529,
+      "grad_norm": 1.0539261102676392,
+      "learning_rate": 5.9694079544168766e-05,
+      "loss": 0.8686,
+      "step": 32850
+    },
+    {
+      "epoch": 1.2648700673724735,
+      "grad_norm": 1.184226393699646,
+      "learning_rate": 5.966640688923739e-05,
+      "loss": 0.908,
+      "step": 32855
+    },
+    {
+      "epoch": 1.2650625601539942,
+      "grad_norm": 0.9844930768013,
+      "learning_rate": 5.963873792254595e-05,
+      "loss": 0.724,
+      "step": 32860
+    },
+    {
+      "epoch": 1.2652550529355149,
+      "grad_norm": 1.4224261045455933,
+      "learning_rate": 5.9611072646624565e-05,
+      "loss": 0.8247,
+      "step": 32865
+    },
+    {
+      "epoch": 1.2654475457170355,
+      "grad_norm": 1.5510427951812744,
+      "learning_rate": 5.9583411064002995e-05,
+      "loss": 0.8445,
+      "step": 32870
+    },
+    {
+      "epoch": 1.2656400384985562,
+      "grad_norm": 1.473929524421692,
+      "learning_rate": 5.955575317721078e-05,
+      "loss": 0.7831,
+      "step": 32875
+    },
+    {
+      "epoch": 1.265832531280077,
+      "grad_norm": 1.2767577171325684,
+      "learning_rate": 5.9528098988777024e-05,
+      "loss": 0.7781,
+      "step": 32880
+    },
+    {
+      "epoch": 1.2660250240615976,
+      "grad_norm": 1.8396121263504028,
+      "learning_rate": 5.950044850123047e-05,
+      "loss": 0.856,
+      "step": 32885
+    },
+    {
+      "epoch": 1.2662175168431185,
+      "grad_norm": 0.8659001588821411,
+      "learning_rate": 5.947280171709966e-05,
+      "loss": 0.9269,
+      "step": 32890
+    },
+    {
+      "epoch": 1.2664100096246391,
+      "grad_norm": 2.1501402854919434,
+      "learning_rate": 5.9445158638912646e-05,
+      "loss": 0.8831,
+      "step": 32895
+    },
+    {
+      "epoch": 1.2666025024061598,
+      "grad_norm": 0.7777537703514099,
+      "learning_rate": 5.941751926919721e-05,
+      "loss": 0.7503,
+      "step": 32900
+    },
+    {
+      "epoch": 1.2667949951876805,
+      "grad_norm": 1.1741917133331299,
+      "learning_rate": 5.9389883610480835e-05,
+      "loss": 0.9508,
+      "step": 32905
+    },
+    {
+      "epoch": 1.2669874879692011,
+      "grad_norm": 1.2577232122421265,
+      "learning_rate": 5.936225166529057e-05,
+      "loss": 0.7789,
+      "step": 32910
+    },
+    {
+      "epoch": 1.2671799807507218,
+      "grad_norm": 1.0245388746261597,
+      "learning_rate": 5.933462343615317e-05,
+      "loss": 0.8104,
+      "step": 32915
+    },
+    {
+      "epoch": 1.2673724735322425,
+      "grad_norm": 1.5200345516204834,
+      "learning_rate": 5.9306998925595105e-05,
+      "loss": 0.8895,
+      "step": 32920
+    },
+    {
+      "epoch": 1.2675649663137634,
+      "grad_norm": 1.0815972089767456,
+      "learning_rate": 5.927937813614241e-05,
+      "loss": 0.746,
+      "step": 32925
+    },
+    {
+      "epoch": 1.267757459095284,
+      "grad_norm": 1.1192761659622192,
+      "learning_rate": 5.9251761070320845e-05,
+      "loss": 0.8543,
+      "step": 32930
+    },
+    {
+      "epoch": 1.2679499518768047,
+      "grad_norm": 1.0540046691894531,
+      "learning_rate": 5.9224147730655766e-05,
+      "loss": 0.9522,
+      "step": 32935
+    },
+    {
+      "epoch": 1.2681424446583254,
+      "grad_norm": 1.5147613286972046,
+      "learning_rate": 5.9196538119672297e-05,
+      "loss": 0.868,
+      "step": 32940
+    },
+    {
+      "epoch": 1.268334937439846,
+      "grad_norm": 2.390244960784912,
+      "learning_rate": 5.91689322398951e-05,
+      "loss": 0.7859,
+      "step": 32945
+    },
+    {
+      "epoch": 1.2685274302213667,
+      "grad_norm": 1.3347294330596924,
+      "learning_rate": 5.914133009384859e-05,
+      "loss": 0.8369,
+      "step": 32950
+    },
+    {
+      "epoch": 1.2687199230028874,
+      "grad_norm": 1.1025590896606445,
+      "learning_rate": 5.911373168405676e-05,
+      "loss": 0.7856,
+      "step": 32955
+    },
+    {
+      "epoch": 1.268912415784408,
+      "grad_norm": 1.335947871208191,
+      "learning_rate": 5.908613701304331e-05,
+      "loss": 0.8423,
+      "step": 32960
+    },
+    {
+      "epoch": 1.2691049085659287,
+      "grad_norm": 1.9916443824768066,
+      "learning_rate": 5.905854608333163e-05,
+      "loss": 0.7697,
+      "step": 32965
+    },
+    {
+      "epoch": 1.2692974013474494,
+      "grad_norm": 1.1374703645706177,
+      "learning_rate": 5.903095889744466e-05,
+      "loss": 0.6431,
+      "step": 32970
+    },
+    {
+      "epoch": 1.26948989412897,
+      "grad_norm": 1.484373688697815,
+      "learning_rate": 5.900337545790513e-05,
+      "loss": 0.8287,
+      "step": 32975
+    },
+    {
+      "epoch": 1.2696823869104907,
+      "grad_norm": 1.3706483840942383,
+      "learning_rate": 5.8975795767235354e-05,
+      "loss": 0.9413,
+      "step": 32980
+    },
+    {
+      "epoch": 1.2698748796920116,
+      "grad_norm": 1.4376710653305054,
+      "learning_rate": 5.894821982795724e-05,
+      "loss": 0.7843,
+      "step": 32985
+    },
+    {
+      "epoch": 1.2700673724735323,
+      "grad_norm": 1.0955665111541748,
+      "learning_rate": 5.892064764259253e-05,
+      "loss": 0.8953,
+      "step": 32990
+    },
+    {
+      "epoch": 1.270259865255053,
+      "grad_norm": 1.9772429466247559,
+      "learning_rate": 5.889307921366245e-05,
+      "loss": 0.7771,
+      "step": 32995
+    },
+    {
+      "epoch": 1.2704523580365736,
+      "grad_norm": 0.92624431848526,
+      "learning_rate": 5.886551454368794e-05,
+      "loss": 0.8009,
+      "step": 33000
+    },
+    {
+      "epoch": 1.2706448508180943,
+      "grad_norm": 1.8379896879196167,
+      "learning_rate": 5.883795363518968e-05,
+      "loss": 1.0357,
+      "step": 33005
+    },
+    {
+      "epoch": 1.270837343599615,
+      "grad_norm": 0.9164409041404724,
+      "learning_rate": 5.881039649068788e-05,
+      "loss": 0.8232,
+      "step": 33010
+    },
+    {
+      "epoch": 1.2710298363811356,
+      "grad_norm": 1.9721314907073975,
+      "learning_rate": 5.8782843112702436e-05,
+      "loss": 0.6249,
+      "step": 33015
+    },
+    {
+      "epoch": 1.2712223291626565,
+      "grad_norm": 1.4756965637207031,
+      "learning_rate": 5.8755293503752975e-05,
+      "loss": 0.7346,
+      "step": 33020
+    },
+    {
+      "epoch": 1.2714148219441772,
+      "grad_norm": 1.0704786777496338,
+      "learning_rate": 5.8727747666358704e-05,
+      "loss": 0.8372,
+      "step": 33025
+    },
+    {
+      "epoch": 1.2716073147256979,
+      "grad_norm": 0.9195441007614136,
+      "learning_rate": 5.8700205603038484e-05,
+      "loss": 0.8975,
+      "step": 33030
+    },
+    {
+      "epoch": 1.2717998075072185,
+      "grad_norm": 0.9442548155784607,
+      "learning_rate": 5.867266731631088e-05,
+      "loss": 0.8154,
+      "step": 33035
+    },
+    {
+      "epoch": 1.2719923002887392,
+      "grad_norm": 1.0243034362792969,
+      "learning_rate": 5.86451328086941e-05,
+      "loss": 0.9787,
+      "step": 33040
+    },
+    {
+      "epoch": 1.27218479307026,
+      "grad_norm": 1.9684053659439087,
+      "learning_rate": 5.8617602082705955e-05,
+      "loss": 0.8663,
+      "step": 33045
+    },
+    {
+      "epoch": 1.2723772858517806,
+      "grad_norm": 2.0619711875915527,
+      "learning_rate": 5.859007514086399e-05,
+      "loss": 0.8495,
+      "step": 33050
+    },
+    {
+      "epoch": 1.2725697786333012,
+      "grad_norm": 1.4054616689682007,
+      "learning_rate": 5.856255198568532e-05,
+      "loss": 0.8247,
+      "step": 33055
+    },
+    {
+      "epoch": 1.272762271414822,
+      "grad_norm": 1.5287714004516602,
+      "learning_rate": 5.853503261968676e-05,
+      "loss": 0.9497,
+      "step": 33060
+    },
+    {
+      "epoch": 1.2729547641963426,
+      "grad_norm": 2.1714093685150146,
+      "learning_rate": 5.850751704538482e-05,
+      "loss": 0.8473,
+      "step": 33065
+    },
+    {
+      "epoch": 1.2731472569778632,
+      "grad_norm": 2.1262662410736084,
+      "learning_rate": 5.848000526529558e-05,
+      "loss": 0.9971,
+      "step": 33070
+    },
+    {
+      "epoch": 1.273339749759384,
+      "grad_norm": 0.8755722641944885,
+      "learning_rate": 5.845249728193476e-05,
+      "loss": 0.8989,
+      "step": 33075
+    },
+    {
+      "epoch": 1.2735322425409046,
+      "grad_norm": 1.1650525331497192,
+      "learning_rate": 5.842499309781789e-05,
+      "loss": 0.8994,
+      "step": 33080
+    },
+    {
+      "epoch": 1.2737247353224255,
+      "grad_norm": 1.2620556354522705,
+      "learning_rate": 5.839749271545999e-05,
+      "loss": 0.8267,
+      "step": 33085
+    },
+    {
+      "epoch": 1.2739172281039461,
+      "grad_norm": 1.1734013557434082,
+      "learning_rate": 5.836999613737577e-05,
+      "loss": 0.9163,
+      "step": 33090
+    },
+    {
+      "epoch": 1.2741097208854668,
+      "grad_norm": 2.227886199951172,
+      "learning_rate": 5.8342503366079624e-05,
+      "loss": 0.878,
+      "step": 33095
+    },
+    {
+      "epoch": 1.2743022136669875,
+      "grad_norm": 0.975045382976532,
+      "learning_rate": 5.83150144040856e-05,
+      "loss": 0.873,
+      "step": 33100
+    },
+    {
+      "epoch": 1.2744947064485082,
+      "grad_norm": 0.9172561764717102,
+      "learning_rate": 5.828752925390737e-05,
+      "loss": 0.7909,
+      "step": 33105
+    },
+    {
+      "epoch": 1.2746871992300288,
+      "grad_norm": 1.7539204359054565,
+      "learning_rate": 5.82600479180583e-05,
+      "loss": 0.915,
+      "step": 33110
+    },
+    {
+      "epoch": 1.2748796920115495,
+      "grad_norm": 2.9440081119537354,
+      "learning_rate": 5.823257039905131e-05,
+      "loss": 0.8617,
+      "step": 33115
+    },
+    {
+      "epoch": 1.2750721847930704,
+      "grad_norm": 1.3572540283203125,
+      "learning_rate": 5.8205096699399074e-05,
+      "loss": 0.8082,
+      "step": 33120
+    },
+    {
+      "epoch": 1.275264677574591,
+      "grad_norm": 1.449960470199585,
+      "learning_rate": 5.8177626821613876e-05,
+      "loss": 0.8901,
+      "step": 33125
+    },
+    {
+      "epoch": 1.2754571703561117,
+      "grad_norm": 1.7100377082824707,
+      "learning_rate": 5.8150160768207704e-05,
+      "loss": 0.9157,
+      "step": 33130
+    },
+    {
+      "epoch": 1.2756496631376324,
+      "grad_norm": 0.9856832027435303,
+      "learning_rate": 5.812269854169201e-05,
+      "loss": 0.7319,
+      "step": 33135
+    },
+    {
+      "epoch": 1.275842155919153,
+      "grad_norm": 2.1571691036224365,
+      "learning_rate": 5.809524014457821e-05,
+      "loss": 0.8611,
+      "step": 33140
+    },
+    {
+      "epoch": 1.2760346487006737,
+      "grad_norm": 1.537744402885437,
+      "learning_rate": 5.806778557937707e-05,
+      "loss": 0.967,
+      "step": 33145
+    },
+    {
+      "epoch": 1.2762271414821944,
+      "grad_norm": 2.2577595710754395,
+      "learning_rate": 5.8040334848599166e-05,
+      "loss": 0.8516,
+      "step": 33150
+    },
+    {
+      "epoch": 1.276419634263715,
+      "grad_norm": 1.4101483821868896,
+      "learning_rate": 5.801288795475466e-05,
+      "loss": 0.8992,
+      "step": 33155
+    },
+    {
+      "epoch": 1.2766121270452357,
+      "grad_norm": 1.332582950592041,
+      "learning_rate": 5.798544490035347e-05,
+      "loss": 0.8865,
+      "step": 33160
+    },
+    {
+      "epoch": 1.2768046198267564,
+      "grad_norm": 1.6516834497451782,
+      "learning_rate": 5.795800568790495e-05,
+      "loss": 0.8555,
+      "step": 33165
+    },
+    {
+      "epoch": 1.276997112608277,
+      "grad_norm": 1.0843029022216797,
+      "learning_rate": 5.793057031991836e-05,
+      "loss": 0.7221,
+      "step": 33170
+    },
+    {
+      "epoch": 1.2771896053897978,
+      "grad_norm": 0.715171754360199,
+      "learning_rate": 5.790313879890241e-05,
+      "loss": 0.8228,
+      "step": 33175
+    },
+    {
+      "epoch": 1.2773820981713186,
+      "grad_norm": 1.1554157733917236,
+      "learning_rate": 5.787571112736554e-05,
+      "loss": 0.8982,
+      "step": 33180
+    },
+    {
+      "epoch": 1.2775745909528393,
+      "grad_norm": 1.4646668434143066,
+      "learning_rate": 5.784828730781585e-05,
+      "loss": 0.9118,
+      "step": 33185
+    },
+    {
+      "epoch": 1.27776708373436,
+      "grad_norm": 1.4582480192184448,
+      "learning_rate": 5.782086734276109e-05,
+      "loss": 0.6899,
+      "step": 33190
+    },
+    {
+      "epoch": 1.2779595765158807,
+      "grad_norm": 0.8803904056549072,
+      "learning_rate": 5.779345123470853e-05,
+      "loss": 0.918,
+      "step": 33195
+    },
+    {
+      "epoch": 1.2781520692974013,
+      "grad_norm": 1.423751711845398,
+      "learning_rate": 5.7766038986165325e-05,
+      "loss": 0.7573,
+      "step": 33200
+    },
+    {
+      "epoch": 1.278344562078922,
+      "grad_norm": 1.1513595581054688,
+      "learning_rate": 5.7738630599638054e-05,
+      "loss": 0.8907,
+      "step": 33205
+    },
+    {
+      "epoch": 1.2785370548604427,
+      "grad_norm": 0.7591530084609985,
+      "learning_rate": 5.7711226077633065e-05,
+      "loss": 0.7822,
+      "step": 33210
+    },
+    {
+      "epoch": 1.2787295476419636,
+      "grad_norm": 1.3467390537261963,
+      "learning_rate": 5.768382542265629e-05,
+      "loss": 0.97,
+      "step": 33215
+    },
+    {
+      "epoch": 1.2789220404234842,
+      "grad_norm": 0.9523210525512695,
+      "learning_rate": 5.765642863721341e-05,
+      "loss": 0.861,
+      "step": 33220
+    },
+    {
+      "epoch": 1.279114533205005,
+      "grad_norm": 1.2797752618789673,
+      "learning_rate": 5.76290357238096e-05,
+      "loss": 0.8446,
+      "step": 33225
+    },
+    {
+      "epoch": 1.2793070259865256,
+      "grad_norm": 2.0043842792510986,
+      "learning_rate": 5.7601646684949784e-05,
+      "loss": 0.8505,
+      "step": 33230
+    },
+    {
+      "epoch": 1.2794995187680462,
+      "grad_norm": 1.1926679611206055,
+      "learning_rate": 5.7574261523138514e-05,
+      "loss": 1.039,
+      "step": 33235
+    },
+    {
+      "epoch": 1.279692011549567,
+      "grad_norm": 1.4371801614761353,
+      "learning_rate": 5.754688024088e-05,
+      "loss": 0.975,
+      "step": 33240
+    },
+    {
+      "epoch": 1.2798845043310876,
+      "grad_norm": 0.7704469561576843,
+      "learning_rate": 5.751950284067809e-05,
+      "loss": 0.6496,
+      "step": 33245
+    },
+    {
+      "epoch": 1.2800769971126083,
+      "grad_norm": 1.707034707069397,
+      "learning_rate": 5.749212932503614e-05,
+      "loss": 0.8091,
+      "step": 33250
+    },
+    {
+      "epoch": 1.280269489894129,
+      "grad_norm": 1.1753286123275757,
+      "learning_rate": 5.746475969645747e-05,
+      "loss": 0.7916,
+      "step": 33255
+    },
+    {
+      "epoch": 1.2804619826756496,
+      "grad_norm": 1.4952518939971924,
+      "learning_rate": 5.7437393957444716e-05,
+      "loss": 0.8581,
+      "step": 33260
+    },
+    {
+      "epoch": 1.2806544754571703,
+      "grad_norm": 1.7860945463180542,
+      "learning_rate": 5.7410032110500335e-05,
+      "loss": 0.6714,
+      "step": 33265
+    },
+    {
+      "epoch": 1.280846968238691,
+      "grad_norm": 1.6719321012496948,
+      "learning_rate": 5.7382674158126385e-05,
+      "loss": 0.9253,
+      "step": 33270
+    },
+    {
+      "epoch": 1.2810394610202118,
+      "grad_norm": 1.7569420337677002,
+      "learning_rate": 5.735532010282461e-05,
+      "loss": 0.8774,
+      "step": 33275
+    },
+    {
+      "epoch": 1.2812319538017325,
+      "grad_norm": 0.898158073425293,
+      "learning_rate": 5.732796994709623e-05,
+      "loss": 0.8034,
+      "step": 33280
+    },
+    {
+      "epoch": 1.2814244465832532,
+      "grad_norm": 1.5631740093231201,
+      "learning_rate": 5.73006236934424e-05,
+      "loss": 0.8423,
+      "step": 33285
+    },
+    {
+      "epoch": 1.2816169393647738,
+      "grad_norm": 1.7265678644180298,
+      "learning_rate": 5.727328134436364e-05,
+      "loss": 0.8724,
+      "step": 33290
+    },
+    {
+      "epoch": 1.2818094321462945,
+      "grad_norm": 0.971207320690155,
+      "learning_rate": 5.7245942902360274e-05,
+      "loss": 0.7923,
+      "step": 33295
+    },
+    {
+      "epoch": 1.2820019249278152,
+      "grad_norm": 1.0335525274276733,
+      "learning_rate": 5.721860836993218e-05,
+      "loss": 0.6989,
+      "step": 33300
+    },
+    {
+      "epoch": 1.2821944177093358,
+      "grad_norm": 0.9981820583343506,
+      "learning_rate": 5.719127774957902e-05,
+      "loss": 0.9665,
+      "step": 33305
+    },
+    {
+      "epoch": 1.2823869104908565,
+      "grad_norm": 2.178272247314453,
+      "learning_rate": 5.716395104379988e-05,
+      "loss": 0.8747,
+      "step": 33310
+    },
+    {
+      "epoch": 1.2825794032723774,
+      "grad_norm": 1.31416654586792,
+      "learning_rate": 5.713662825509365e-05,
+      "loss": 0.8943,
+      "step": 33315
+    },
+    {
+      "epoch": 1.282771896053898,
+      "grad_norm": 1.1709874868392944,
+      "learning_rate": 5.710930938595882e-05,
+      "loss": 0.8149,
+      "step": 33320
+    },
+    {
+      "epoch": 1.2829643888354187,
+      "grad_norm": 1.7127902507781982,
+      "learning_rate": 5.708199443889353e-05,
+      "loss": 0.8275,
+      "step": 33325
+    },
+    {
+      "epoch": 1.2831568816169394,
+      "grad_norm": 0.7726919054985046,
+      "learning_rate": 5.705468341639557e-05,
+      "loss": 0.7617,
+      "step": 33330
+    },
+    {
+      "epoch": 1.28334937439846,
+      "grad_norm": 1.722057819366455,
+      "learning_rate": 5.702737632096229e-05,
+      "loss": 0.8446,
+      "step": 33335
+    },
+    {
+      "epoch": 1.2835418671799808,
+      "grad_norm": 1.3741867542266846,
+      "learning_rate": 5.700007315509078e-05,
+      "loss": 0.9602,
+      "step": 33340
+    },
+    {
+      "epoch": 1.2837343599615014,
+      "grad_norm": 2.013511896133423,
+      "learning_rate": 5.6972773921277734e-05,
+      "loss": 0.8998,
+      "step": 33345
+    },
+    {
+      "epoch": 1.283926852743022,
+      "grad_norm": 2.3441731929779053,
+      "learning_rate": 5.6945478622019524e-05,
+      "loss": 0.857,
+      "step": 33350
+    },
+    {
+      "epoch": 1.2841193455245428,
+      "grad_norm": 1.1220638751983643,
+      "learning_rate": 5.691818725981199e-05,
+      "loss": 0.9617,
+      "step": 33355
+    },
+    {
+      "epoch": 1.2843118383060634,
+      "grad_norm": 1.3120940923690796,
+      "learning_rate": 5.6890899837150944e-05,
+      "loss": 0.8752,
+      "step": 33360
+    },
+    {
+      "epoch": 1.284504331087584,
+      "grad_norm": 1.2576730251312256,
+      "learning_rate": 5.686361635653148e-05,
+      "loss": 0.8307,
+      "step": 33365
+    },
+    {
+      "epoch": 1.2846968238691048,
+      "grad_norm": 1.0636831521987915,
+      "learning_rate": 5.6836336820448556e-05,
+      "loss": 0.7714,
+      "step": 33370
+    },
+    {
+      "epoch": 1.2848893166506257,
+      "grad_norm": 0.9599676728248596,
+      "learning_rate": 5.680906123139669e-05,
+      "loss": 0.8351,
+      "step": 33375
+    },
+    {
+      "epoch": 1.2850818094321463,
+      "grad_norm": 1.8764914274215698,
+      "learning_rate": 5.67817895918701e-05,
+      "loss": 0.9712,
+      "step": 33380
+    },
+    {
+      "epoch": 1.285274302213667,
+      "grad_norm": 1.3599872589111328,
+      "learning_rate": 5.675452190436248e-05,
+      "loss": 0.9119,
+      "step": 33385
+    },
+    {
+      "epoch": 1.2854667949951877,
+      "grad_norm": 0.9683537483215332,
+      "learning_rate": 5.672725817136744e-05,
+      "loss": 0.8115,
+      "step": 33390
+    },
+    {
+      "epoch": 1.2856592877767083,
+      "grad_norm": 1.1662638187408447,
+      "learning_rate": 5.669999839537794e-05,
+      "loss": 0.9159,
+      "step": 33395
+    },
+    {
+      "epoch": 1.285851780558229,
+      "grad_norm": 2.2945556640625,
+      "learning_rate": 5.667274257888675e-05,
+      "loss": 0.8131,
+      "step": 33400
+    },
+    {
+      "epoch": 1.2860442733397497,
+      "grad_norm": 0.9132623076438904,
+      "learning_rate": 5.664549072438624e-05,
+      "loss": 0.799,
+      "step": 33405
+    },
+    {
+      "epoch": 1.2862367661212706,
+      "grad_norm": 0.9698240756988525,
+      "learning_rate": 5.661824283436844e-05,
+      "loss": 0.9786,
+      "step": 33410
+    },
+    {
+      "epoch": 1.2864292589027913,
+      "grad_norm": 0.7265766263008118,
+      "learning_rate": 5.659099891132488e-05,
+      "loss": 0.7017,
+      "step": 33415
+    },
+    {
+      "epoch": 1.286621751684312,
+      "grad_norm": 0.8965588212013245,
+      "learning_rate": 5.656375895774699e-05,
+      "loss": 0.8478,
+      "step": 33420
+    },
+    {
+      "epoch": 1.2868142444658326,
+      "grad_norm": 1.5300185680389404,
+      "learning_rate": 5.653652297612556e-05,
+      "loss": 0.851,
+      "step": 33425
+    },
+    {
+      "epoch": 1.2870067372473533,
+      "grad_norm": 1.6665253639221191,
+      "learning_rate": 5.650929096895119e-05,
+      "loss": 0.8494,
+      "step": 33430
+    },
+    {
+      "epoch": 1.287199230028874,
+      "grad_norm": 1.2362685203552246,
+      "learning_rate": 5.6482062938714095e-05,
+      "loss": 0.7511,
+      "step": 33435
+    },
+    {
+      "epoch": 1.2873917228103946,
+      "grad_norm": 1.0025197267532349,
+      "learning_rate": 5.645483888790404e-05,
+      "loss": 0.8445,
+      "step": 33440
+    },
+    {
+      "epoch": 1.2875842155919153,
+      "grad_norm": 3.0761525630950928,
+      "learning_rate": 5.6427618819010486e-05,
+      "loss": 0.9236,
+      "step": 33445
+    },
+    {
+      "epoch": 1.287776708373436,
+      "grad_norm": 1.8657230138778687,
+      "learning_rate": 5.640040273452256e-05,
+      "loss": 0.925,
+      "step": 33450
+    },
+    {
+      "epoch": 1.2879692011549566,
+      "grad_norm": 1.3410662412643433,
+      "learning_rate": 5.637319063692903e-05,
+      "loss": 0.7961,
+      "step": 33455
+    },
+    {
+      "epoch": 1.2881616939364773,
+      "grad_norm": 1.056356430053711,
+      "learning_rate": 5.6345982528718125e-05,
+      "loss": 0.8616,
+      "step": 33460
+    },
+    {
+      "epoch": 1.288354186717998,
+      "grad_norm": 2.1360607147216797,
+      "learning_rate": 5.6318778412378024e-05,
+      "loss": 0.8894,
+      "step": 33465
+    },
+    {
+      "epoch": 1.2885466794995188,
+      "grad_norm": 1.3970694541931152,
+      "learning_rate": 5.629157829039623e-05,
+      "loss": 1.0804,
+      "step": 33470
+    },
+    {
+      "epoch": 1.2887391722810395,
+      "grad_norm": 1.7947211265563965,
+      "learning_rate": 5.6264382165260065e-05,
+      "loss": 0.6816,
+      "step": 33475
+    },
+    {
+      "epoch": 1.2889316650625602,
+      "grad_norm": 1.3382033109664917,
+      "learning_rate": 5.6237190039456425e-05,
+      "loss": 0.9228,
+      "step": 33480
+    },
+    {
+      "epoch": 1.2891241578440809,
+      "grad_norm": 1.0116480588912964,
+      "learning_rate": 5.6210001915471896e-05,
+      "loss": 0.9483,
+      "step": 33485
+    },
+    {
+      "epoch": 1.2893166506256015,
+      "grad_norm": 1.6364978551864624,
+      "learning_rate": 5.618281779579253e-05,
+      "loss": 1.0702,
+      "step": 33490
+    },
+    {
+      "epoch": 1.2895091434071222,
+      "grad_norm": 1.1366705894470215,
+      "learning_rate": 5.6155637682904284e-05,
+      "loss": 0.8214,
+      "step": 33495
+    },
+    {
+      "epoch": 1.2897016361886429,
+      "grad_norm": 1.1931660175323486,
+      "learning_rate": 5.6128461579292456e-05,
+      "loss": 0.6575,
+      "step": 33500
+    },
+    {
+      "epoch": 1.2898941289701638,
+      "grad_norm": 1.7498856782913208,
+      "learning_rate": 5.610128948744229e-05,
+      "loss": 0.843,
+      "step": 33505
+    },
+    {
+      "epoch": 1.2900866217516844,
+      "grad_norm": 1.446738839149475,
+      "learning_rate": 5.6074121409838345e-05,
+      "loss": 0.7902,
+      "step": 33510
+    },
+    {
+      "epoch": 1.290279114533205,
+      "grad_norm": 1.2919175624847412,
+      "learning_rate": 5.6046957348965014e-05,
+      "loss": 0.8396,
+      "step": 33515
+    },
+    {
+      "epoch": 1.2904716073147258,
+      "grad_norm": 1.1369450092315674,
+      "learning_rate": 5.6019797307306264e-05,
+      "loss": 0.9185,
+      "step": 33520
+    },
+    {
+      "epoch": 1.2906641000962464,
+      "grad_norm": 1.1856272220611572,
+      "learning_rate": 5.599264128734576e-05,
+      "loss": 0.8774,
+      "step": 33525
+    },
+    {
+      "epoch": 1.290856592877767,
+      "grad_norm": 2.4234628677368164,
+      "learning_rate": 5.5965489291566645e-05,
+      "loss": 1.0206,
+      "step": 33530
+    },
+    {
+      "epoch": 1.2910490856592878,
+      "grad_norm": 1.700831651687622,
+      "learning_rate": 5.593834132245183e-05,
+      "loss": 0.8308,
+      "step": 33535
+    },
+    {
+      "epoch": 1.2912415784408084,
+      "grad_norm": 0.7165193557739258,
+      "learning_rate": 5.591119738248382e-05,
+      "loss": 0.7948,
+      "step": 33540
+    },
+    {
+      "epoch": 1.2914340712223291,
+      "grad_norm": 1.512609601020813,
+      "learning_rate": 5.5884057474144736e-05,
+      "loss": 0.8691,
+      "step": 33545
+    },
+    {
+      "epoch": 1.2916265640038498,
+      "grad_norm": 1.0967670679092407,
+      "learning_rate": 5.585692159991641e-05,
+      "loss": 0.7975,
+      "step": 33550
+    },
+    {
+      "epoch": 1.2918190567853705,
+      "grad_norm": 1.0149413347244263,
+      "learning_rate": 5.582978976228015e-05,
+      "loss": 0.6579,
+      "step": 33555
+    },
+    {
+      "epoch": 1.2920115495668911,
+      "grad_norm": 1.2751359939575195,
+      "learning_rate": 5.5802661963716994e-05,
+      "loss": 0.8688,
+      "step": 33560
+    },
+    {
+      "epoch": 1.2922040423484118,
+      "grad_norm": 0.7386071681976318,
+      "learning_rate": 5.5775538206707636e-05,
+      "loss": 0.819,
+      "step": 33565
+    },
+    {
+      "epoch": 1.2923965351299327,
+      "grad_norm": 1.0043171644210815,
+      "learning_rate": 5.574841849373238e-05,
+      "loss": 0.7907,
+      "step": 33570
+    },
+    {
+      "epoch": 1.2925890279114534,
+      "grad_norm": 0.9456561803817749,
+      "learning_rate": 5.5721302827271035e-05,
+      "loss": 1.0571,
+      "step": 33575
+    },
+    {
+      "epoch": 1.292781520692974,
+      "grad_norm": 2.149979591369629,
+      "learning_rate": 5.5694191209803313e-05,
+      "loss": 0.8367,
+      "step": 33580
+    },
+    {
+      "epoch": 1.2929740134744947,
+      "grad_norm": 1.2153743505477905,
+      "learning_rate": 5.566708364380826e-05,
+      "loss": 0.7986,
+      "step": 33585
+    },
+    {
+      "epoch": 1.2931665062560154,
+      "grad_norm": 1.090076208114624,
+      "learning_rate": 5.563998013176474e-05,
+      "loss": 0.7111,
+      "step": 33590
+    },
+    {
+      "epoch": 1.293358999037536,
+      "grad_norm": 1.1182845830917358,
+      "learning_rate": 5.5612880676151154e-05,
+      "loss": 0.6748,
+      "step": 33595
+    },
+    {
+      "epoch": 1.2935514918190567,
+      "grad_norm": 1.2427207231521606,
+      "learning_rate": 5.5585785279445654e-05,
+      "loss": 0.8749,
+      "step": 33600
+    },
+    {
+      "epoch": 1.2937439846005776,
+      "grad_norm": 1.2770304679870605,
+      "learning_rate": 5.555869394412578e-05,
+      "loss": 0.9485,
+      "step": 33605
+    },
+    {
+      "epoch": 1.2939364773820983,
+      "grad_norm": 1.2712188959121704,
+      "learning_rate": 5.5531606672669045e-05,
+      "loss": 0.7984,
+      "step": 33610
+    },
+    {
+      "epoch": 1.294128970163619,
+      "grad_norm": 1.0220839977264404,
+      "learning_rate": 5.550452346755225e-05,
+      "loss": 0.8363,
+      "step": 33615
+    },
+    {
+      "epoch": 1.2943214629451396,
+      "grad_norm": 2.0478289127349854,
+      "learning_rate": 5.547744433125204e-05,
+      "loss": 0.8417,
+      "step": 33620
+    },
+    {
+      "epoch": 1.2945139557266603,
+      "grad_norm": 0.9348976612091064,
+      "learning_rate": 5.5450369266244595e-05,
+      "loss": 0.8596,
+      "step": 33625
+    },
+    {
+      "epoch": 1.294706448508181,
+      "grad_norm": 1.4484078884124756,
+      "learning_rate": 5.542329827500581e-05,
+      "loss": 0.8122,
+      "step": 33630
+    },
+    {
+      "epoch": 1.2948989412897016,
+      "grad_norm": 1.7183796167373657,
+      "learning_rate": 5.5396231360011074e-05,
+      "loss": 0.857,
+      "step": 33635
+    },
+    {
+      "epoch": 1.2950914340712223,
+      "grad_norm": 0.8953412175178528,
+      "learning_rate": 5.5369168523735505e-05,
+      "loss": 1.0084,
+      "step": 33640
+    },
+    {
+      "epoch": 1.295283926852743,
+      "grad_norm": 0.8453150987625122,
+      "learning_rate": 5.5342109768653815e-05,
+      "loss": 0.8156,
+      "step": 33645
+    },
+    {
+      "epoch": 1.2954764196342636,
+      "grad_norm": 1.296000599861145,
+      "learning_rate": 5.531505509724036e-05,
+      "loss": 0.808,
+      "step": 33650
+    },
+    {
+      "epoch": 1.2956689124157843,
+      "grad_norm": Infinity,
+      "learning_rate": 5.529341430201327e-05,
+      "loss": 0.8415,
+      "step": 33655
+    },
+    {
+      "epoch": 1.295861405197305,
+      "grad_norm": 1.4675425291061401,
+      "learning_rate": 5.526636698743678e-05,
+      "loss": 0.8757,
+      "step": 33660
+    },
+    {
+      "epoch": 1.2960538979788259,
+      "grad_norm": 1.3060146570205688,
+      "learning_rate": 5.5239323763454696e-05,
+      "loss": 0.8328,
+      "step": 33665
+    },
+    {
+      "epoch": 1.2962463907603465,
+      "grad_norm": 1.034527063369751,
+      "learning_rate": 5.5212284632539976e-05,
+      "loss": 0.9086,
+      "step": 33670
+    },
+    {
+      "epoch": 1.2964388835418672,
+      "grad_norm": 1.3816114664077759,
+      "learning_rate": 5.518524959716507e-05,
+      "loss": 0.8794,
+      "step": 33675
+    },
+    {
+      "epoch": 1.2966313763233879,
+      "grad_norm": 1.4892889261245728,
+      "learning_rate": 5.515821865980228e-05,
+      "loss": 0.8859,
+      "step": 33680
+    },
+    {
+      "epoch": 1.2968238691049085,
+      "grad_norm": 1.8635764122009277,
+      "learning_rate": 5.513119182292332e-05,
+      "loss": 0.8079,
+      "step": 33685
+    },
+    {
+      "epoch": 1.2970163618864292,
+      "grad_norm": 1.203260898590088,
+      "learning_rate": 5.5104169088999644e-05,
+      "loss": 0.7971,
+      "step": 33690
+    },
+    {
+      "epoch": 1.2972088546679499,
+      "grad_norm": 1.3224972486495972,
+      "learning_rate": 5.507715046050228e-05,
+      "loss": 0.916,
+      "step": 33695
+    },
+    {
+      "epoch": 1.2974013474494708,
+      "grad_norm": 1.774021029472351,
+      "learning_rate": 5.505013593990197e-05,
+      "loss": 0.9424,
+      "step": 33700
+    },
+    {
+      "epoch": 1.2975938402309914,
+      "grad_norm": 0.8528000712394714,
+      "learning_rate": 5.502312552966892e-05,
+      "loss": 0.7671,
+      "step": 33705
+    },
+    {
+      "epoch": 1.2977863330125121,
+      "grad_norm": 2.2666587829589844,
+      "learning_rate": 5.49961192322731e-05,
+      "loss": 0.8172,
+      "step": 33710
+    },
+    {
+      "epoch": 1.2979788257940328,
+      "grad_norm": 1.4332960844039917,
+      "learning_rate": 5.496911705018404e-05,
+      "loss": 0.9298,
+      "step": 33715
+    },
+    {
+      "epoch": 1.2981713185755535,
+      "grad_norm": 2.359644889831543,
+      "learning_rate": 5.494211898587094e-05,
+      "loss": 0.9154,
+      "step": 33720
+    },
+    {
+      "epoch": 1.2983638113570741,
+      "grad_norm": 1.5822657346725464,
+      "learning_rate": 5.491512504180261e-05,
+      "loss": 0.8183,
+      "step": 33725
+    },
+    {
+      "epoch": 1.2985563041385948,
+      "grad_norm": 1.935848593711853,
+      "learning_rate": 5.488813522044739e-05,
+      "loss": 0.813,
+      "step": 33730
+    },
+    {
+      "epoch": 1.2987487969201155,
+      "grad_norm": 2.000776529312134,
+      "learning_rate": 5.486114952427337e-05,
+      "loss": 0.7426,
+      "step": 33735
+    },
+    {
+      "epoch": 1.2989412897016361,
+      "grad_norm": 1.2055097818374634,
+      "learning_rate": 5.4834167955748204e-05,
+      "loss": 0.82,
+      "step": 33740
+    },
+    {
+      "epoch": 1.2991337824831568,
+      "grad_norm": 1.9020949602127075,
+      "learning_rate": 5.480719051733918e-05,
+      "loss": 0.788,
+      "step": 33745
+    },
+    {
+      "epoch": 1.2993262752646775,
+      "grad_norm": 1.5985333919525146,
+      "learning_rate": 5.478021721151323e-05,
+      "loss": 0.7221,
+      "step": 33750
+    },
+    {
+      "epoch": 1.2995187680461981,
+      "grad_norm": 1.5660638809204102,
+      "learning_rate": 5.475324804073687e-05,
+      "loss": 0.8852,
+      "step": 33755
+    },
+    {
+      "epoch": 1.299711260827719,
+      "grad_norm": 1.3215242624282837,
+      "learning_rate": 5.4726283007476196e-05,
+      "loss": 0.7963,
+      "step": 33760
+    },
+    {
+      "epoch": 1.2999037536092397,
+      "grad_norm": 1.6176220178604126,
+      "learning_rate": 5.4699322114197084e-05,
+      "loss": 0.8704,
+      "step": 33765
+    },
+    {
+      "epoch": 1.3000962463907604,
+      "grad_norm": 1.5978866815567017,
+      "learning_rate": 5.4672365363364855e-05,
+      "loss": 0.8481,
+      "step": 33770
+    },
+    {
+      "epoch": 1.300288739172281,
+      "grad_norm": 0.9393962621688843,
+      "learning_rate": 5.4645412757444525e-05,
+      "loss": 0.7539,
+      "step": 33775
+    },
+    {
+      "epoch": 1.3004812319538017,
+      "grad_norm": 1.5960298776626587,
+      "learning_rate": 5.461846429890077e-05,
+      "loss": 0.8225,
+      "step": 33780
+    },
+    {
+      "epoch": 1.3006737247353224,
+      "grad_norm": 1.2632452249526978,
+      "learning_rate": 5.459151999019787e-05,
+      "loss": 0.8851,
+      "step": 33785
+    },
+    {
+      "epoch": 1.300866217516843,
+      "grad_norm": 0.8271371722221375,
+      "learning_rate": 5.456457983379957e-05,
+      "loss": 0.8399,
+      "step": 33790
+    },
+    {
+      "epoch": 1.3010587102983637,
+      "grad_norm": 1.2971397638320923,
+      "learning_rate": 5.453764383216955e-05,
+      "loss": 0.7467,
+      "step": 33795
+    },
+    {
+      "epoch": 1.3012512030798846,
+      "grad_norm": 1.0965392589569092,
+      "learning_rate": 5.45107119877708e-05,
+      "loss": 0.7249,
+      "step": 33800
+    },
+    {
+      "epoch": 1.3014436958614053,
+      "grad_norm": 1.258906364440918,
+      "learning_rate": 5.4483784303066096e-05,
+      "loss": 0.8473,
+      "step": 33805
+    },
+    {
+      "epoch": 1.301636188642926,
+      "grad_norm": 1.6706708669662476,
+      "learning_rate": 5.44568607805178e-05,
+      "loss": 0.8842,
+      "step": 33810
+    },
+    {
+      "epoch": 1.3018286814244466,
+      "grad_norm": 1.3727566003799438,
+      "learning_rate": 5.442994142258794e-05,
+      "loss": 0.8538,
+      "step": 33815
+    },
+    {
+      "epoch": 1.3020211742059673,
+      "grad_norm": 0.9679449796676636,
+      "learning_rate": 5.440302623173801e-05,
+      "loss": 0.739,
+      "step": 33820
+    },
+    {
+      "epoch": 1.302213666987488,
+      "grad_norm": 1.0687127113342285,
+      "learning_rate": 5.437611521042929e-05,
+      "loss": 0.665,
+      "step": 33825
+    },
+    {
+      "epoch": 1.3024061597690086,
+      "grad_norm": 1.3849055767059326,
+      "learning_rate": 5.4349208361122604e-05,
+      "loss": 0.9492,
+      "step": 33830
+    },
+    {
+      "epoch": 1.3025986525505293,
+      "grad_norm": 2.082472562789917,
+      "learning_rate": 5.4322305686278386e-05,
+      "loss": 0.8852,
+      "step": 33835
+    },
+    {
+      "epoch": 1.30279114533205,
+      "grad_norm": 1.0276436805725098,
+      "learning_rate": 5.4295407188356784e-05,
+      "loss": 0.7791,
+      "step": 33840
+    },
+    {
+      "epoch": 1.3029836381135707,
+      "grad_norm": 1.8117910623550415,
+      "learning_rate": 5.426851286981738e-05,
+      "loss": 1.0149,
+      "step": 33845
+    },
+    {
+      "epoch": 1.3031761308950913,
+      "grad_norm": 1.362642526626587,
+      "learning_rate": 5.4241622733119545e-05,
+      "loss": 0.8697,
+      "step": 33850
+    },
+    {
+      "epoch": 1.303368623676612,
+      "grad_norm": 1.1220524311065674,
+      "learning_rate": 5.421473678072217e-05,
+      "loss": 0.7154,
+      "step": 33855
+    },
+    {
+      "epoch": 1.3035611164581329,
+      "grad_norm": 1.0112050771713257,
+      "learning_rate": 5.4187855015083875e-05,
+      "loss": 0.7251,
+      "step": 33860
+    },
+    {
+      "epoch": 1.3037536092396536,
+      "grad_norm": 1.8056666851043701,
+      "learning_rate": 5.4160977438662665e-05,
+      "loss": 1.0671,
+      "step": 33865
+    },
+    {
+      "epoch": 1.3039461020211742,
+      "grad_norm": 1.237334966659546,
+      "learning_rate": 5.41341040539165e-05,
+      "loss": 0.7798,
+      "step": 33870
+    },
+    {
+      "epoch": 1.304138594802695,
+      "grad_norm": 1.337863802909851,
+      "learning_rate": 5.410723486330265e-05,
+      "loss": 0.8027,
+      "step": 33875
+    },
+    {
+      "epoch": 1.3043310875842156,
+      "grad_norm": 1.7887320518493652,
+      "learning_rate": 5.408036986927816e-05,
+      "loss": 0.8194,
+      "step": 33880
+    },
+    {
+      "epoch": 1.3045235803657362,
+      "grad_norm": 1.657067894935608,
+      "learning_rate": 5.405350907429965e-05,
+      "loss": 0.9871,
+      "step": 33885
+    },
+    {
+      "epoch": 1.304716073147257,
+      "grad_norm": 1.5351759195327759,
+      "learning_rate": 5.40266524808234e-05,
+      "loss": 0.9261,
+      "step": 33890
+    },
+    {
+      "epoch": 1.3049085659287778,
+      "grad_norm": 1.6100406646728516,
+      "learning_rate": 5.399980009130516e-05,
+      "loss": 0.8397,
+      "step": 33895
+    },
+    {
+      "epoch": 1.3051010587102985,
+      "grad_norm": 1.196283221244812,
+      "learning_rate": 5.397295190820058e-05,
+      "loss": 0.832,
+      "step": 33900
+    },
+    {
+      "epoch": 1.3052935514918191,
+      "grad_norm": 1.709205150604248,
+      "learning_rate": 5.3946107933964576e-05,
+      "loss": 0.9753,
+      "step": 33905
+    },
+    {
+      "epoch": 1.3054860442733398,
+      "grad_norm": 1.0625784397125244,
+      "learning_rate": 5.391926817105194e-05,
+      "loss": 0.7821,
+      "step": 33910
+    },
+    {
+      "epoch": 1.3056785370548605,
+      "grad_norm": 1.3273873329162598,
+      "learning_rate": 5.389243262191697e-05,
+      "loss": 0.9206,
+      "step": 33915
+    },
+    {
+      "epoch": 1.3058710298363811,
+      "grad_norm": 2.023766040802002,
+      "learning_rate": 5.3865601289013636e-05,
+      "loss": 0.8539,
+      "step": 33920
+    },
+    {
+      "epoch": 1.3060635226179018,
+      "grad_norm": 1.1720694303512573,
+      "learning_rate": 5.383877417479541e-05,
+      "loss": 0.8685,
+      "step": 33925
+    },
+    {
+      "epoch": 1.3062560153994225,
+      "grad_norm": 1.6969107389450073,
+      "learning_rate": 5.38119512817155e-05,
+      "loss": 0.8275,
+      "step": 33930
+    },
+    {
+      "epoch": 1.3064485081809432,
+      "grad_norm": 1.802259922027588,
+      "learning_rate": 5.378513261222669e-05,
+      "loss": 0.8782,
+      "step": 33935
+    },
+    {
+      "epoch": 1.3066410009624638,
+      "grad_norm": 0.5480902791023254,
+      "learning_rate": 5.375831816878134e-05,
+      "loss": 0.8165,
+      "step": 33940
+    },
+    {
+      "epoch": 1.3068334937439845,
+      "grad_norm": 2.5565297603607178,
+      "learning_rate": 5.3731507953831514e-05,
+      "loss": 0.8427,
+      "step": 33945
+    },
+    {
+      "epoch": 1.3070259865255052,
+      "grad_norm": 1.1750694513320923,
+      "learning_rate": 5.3704701969828754e-05,
+      "loss": 0.9518,
+      "step": 33950
+    },
+    {
+      "epoch": 1.307218479307026,
+      "grad_norm": 0.9000421762466431,
+      "learning_rate": 5.367790021922431e-05,
+      "loss": 0.7622,
+      "step": 33955
+    },
+    {
+      "epoch": 1.3074109720885467,
+      "grad_norm": 1.3734017610549927,
+      "learning_rate": 5.365110270446904e-05,
+      "loss": 0.835,
+      "step": 33960
+    },
+    {
+      "epoch": 1.3076034648700674,
+      "grad_norm": 1.1932064294815063,
+      "learning_rate": 5.362430942801343e-05,
+      "loss": 0.8661,
+      "step": 33965
+    },
+    {
+      "epoch": 1.307795957651588,
+      "grad_norm": 1.3349897861480713,
+      "learning_rate": 5.3597520392307434e-05,
+      "loss": 0.7571,
+      "step": 33970
+    },
+    {
+      "epoch": 1.3079884504331087,
+      "grad_norm": 1.5428650379180908,
+      "learning_rate": 5.357073559980089e-05,
+      "loss": 0.7853,
+      "step": 33975
+    },
+    {
+      "epoch": 1.3081809432146294,
+      "grad_norm": 1.3346575498580933,
+      "learning_rate": 5.354395505294297e-05,
+      "loss": 0.7985,
+      "step": 33980
+    },
+    {
+      "epoch": 1.30837343599615,
+      "grad_norm": 1.840759038925171,
+      "learning_rate": 5.351717875418263e-05,
+      "loss": 0.9832,
+      "step": 33985
+    },
+    {
+      "epoch": 1.308565928777671,
+      "grad_norm": 1.7884091138839722,
+      "learning_rate": 5.349040670596835e-05,
+      "loss": 0.8081,
+      "step": 33990
+    },
+    {
+      "epoch": 1.3087584215591916,
+      "grad_norm": 1.082604169845581,
+      "learning_rate": 5.346363891074833e-05,
+      "loss": 0.7615,
+      "step": 33995
+    },
+    {
+      "epoch": 1.3089509143407123,
+      "grad_norm": 1.071516990661621,
+      "learning_rate": 5.3436875370970176e-05,
+      "loss": 0.9025,
+      "step": 34000
+    },
+    {
+      "epoch": 1.309143407122233,
+      "grad_norm": 0.9769271612167358,
+      "learning_rate": 5.3410116089081394e-05,
+      "loss": 0.7979,
+      "step": 34005
+    },
+    {
+      "epoch": 1.3093358999037537,
+      "grad_norm": 1.50300133228302,
+      "learning_rate": 5.3383361067528795e-05,
+      "loss": 0.9068,
+      "step": 34010
+    },
+    {
+      "epoch": 1.3095283926852743,
+      "grad_norm": 1.6212974786758423,
+      "learning_rate": 5.335661030875909e-05,
+      "loss": 0.8189,
+      "step": 34015
+    },
+    {
+      "epoch": 1.309720885466795,
+      "grad_norm": 1.2834120988845825,
+      "learning_rate": 5.3329863815218354e-05,
+      "loss": 0.9268,
+      "step": 34020
+    },
+    {
+      "epoch": 1.3099133782483157,
+      "grad_norm": 1.2864872217178345,
+      "learning_rate": 5.33031215893524e-05,
+      "loss": 0.8804,
+      "step": 34025
+    },
+    {
+      "epoch": 1.3101058710298363,
+      "grad_norm": 1.1591218709945679,
+      "learning_rate": 5.3276383633606656e-05,
+      "loss": 0.9498,
+      "step": 34030
+    },
+    {
+      "epoch": 1.310298363811357,
+      "grad_norm": 1.4199464321136475,
+      "learning_rate": 5.324964995042614e-05,
+      "loss": 0.7704,
+      "step": 34035
+    },
+    {
+      "epoch": 1.3104908565928777,
+      "grad_norm": 1.1240346431732178,
+      "learning_rate": 5.322292054225539e-05,
+      "loss": 0.846,
+      "step": 34040
+    },
+    {
+      "epoch": 1.3106833493743983,
+      "grad_norm": 1.1645770072937012,
+      "learning_rate": 5.3196195411538706e-05,
+      "loss": 0.8291,
+      "step": 34045
+    },
+    {
+      "epoch": 1.3108758421559192,
+      "grad_norm": 1.1007463932037354,
+      "learning_rate": 5.316947456071994e-05,
+      "loss": 0.7795,
+      "step": 34050
+    },
+    {
+      "epoch": 1.31106833493744,
+      "grad_norm": 1.4447979927062988,
+      "learning_rate": 5.314275799224243e-05,
+      "loss": 0.8513,
+      "step": 34055
+    },
+    {
+      "epoch": 1.3112608277189606,
+      "grad_norm": 1.0073038339614868,
+      "learning_rate": 5.311604570854938e-05,
+      "loss": 0.7007,
+      "step": 34060
+    },
+    {
+      "epoch": 1.3114533205004812,
+      "grad_norm": 1.1897684335708618,
+      "learning_rate": 5.308933771208332e-05,
+      "loss": 0.8345,
+      "step": 34065
+    },
+    {
+      "epoch": 1.311645813282002,
+      "grad_norm": 0.9960253834724426,
+      "learning_rate": 5.3062634005286586e-05,
+      "loss": 0.7782,
+      "step": 34070
+    },
+    {
+      "epoch": 1.3118383060635226,
+      "grad_norm": 1.1287857294082642,
+      "learning_rate": 5.303593459060103e-05,
+      "loss": 0.8248,
+      "step": 34075
+    },
+    {
+      "epoch": 1.3120307988450433,
+      "grad_norm": 1.133621096611023,
+      "learning_rate": 5.300923947046819e-05,
+      "loss": 0.7083,
+      "step": 34080
+    },
+    {
+      "epoch": 1.312223291626564,
+      "grad_norm": 0.9852873086929321,
+      "learning_rate": 5.2982548647329034e-05,
+      "loss": 0.8222,
+      "step": 34085
+    },
+    {
+      "epoch": 1.3124157844080848,
+      "grad_norm": 1.1880803108215332,
+      "learning_rate": 5.295586212362442e-05,
+      "loss": 1.0247,
+      "step": 34090
+    },
+    {
+      "epoch": 1.3126082771896055,
+      "grad_norm": 2.150893449783325,
+      "learning_rate": 5.292917990179453e-05,
+      "loss": 0.8991,
+      "step": 34095
+    },
+    {
+      "epoch": 1.3128007699711262,
+      "grad_norm": 1.5206340551376343,
+      "learning_rate": 5.290250198427934e-05,
+      "loss": 0.8353,
+      "step": 34100
+    },
+    {
+      "epoch": 1.3129932627526468,
+      "grad_norm": 1.733988642692566,
+      "learning_rate": 5.2875828373518344e-05,
+      "loss": 0.8553,
+      "step": 34105
+    },
+    {
+      "epoch": 1.3131857555341675,
+      "grad_norm": 2.8545048236846924,
+      "learning_rate": 5.2849159071950716e-05,
+      "loss": 0.835,
+      "step": 34110
+    },
+    {
+      "epoch": 1.3133782483156882,
+      "grad_norm": 1.2156729698181152,
+      "learning_rate": 5.282249408201505e-05,
+      "loss": 0.8605,
+      "step": 34115
+    },
+    {
+      "epoch": 1.3135707410972088,
+      "grad_norm": 1.5454976558685303,
+      "learning_rate": 5.2795833406149876e-05,
+      "loss": 0.7758,
+      "step": 34120
+    },
+    {
+      "epoch": 1.3137632338787295,
+      "grad_norm": 1.7499946355819702,
+      "learning_rate": 5.276917704679299e-05,
+      "loss": 0.8126,
+      "step": 34125
+    },
+    {
+      "epoch": 1.3139557266602502,
+      "grad_norm": 1.7527188062667847,
+      "learning_rate": 5.2742525006381994e-05,
+      "loss": 0.8095,
+      "step": 34130
+    },
+    {
+      "epoch": 1.3141482194417708,
+      "grad_norm": 1.1449981927871704,
+      "learning_rate": 5.271587728735402e-05,
+      "loss": 0.9137,
+      "step": 34135
+    },
+    {
+      "epoch": 1.3143407122232915,
+      "grad_norm": 1.3999851942062378,
+      "learning_rate": 5.268923389214588e-05,
+      "loss": 0.8175,
+      "step": 34140
+    },
+    {
+      "epoch": 1.3145332050048122,
+      "grad_norm": 1.119730830192566,
+      "learning_rate": 5.2662594823193865e-05,
+      "loss": 0.9228,
+      "step": 34145
+    },
+    {
+      "epoch": 1.314725697786333,
+      "grad_norm": 0.8965536952018738,
+      "learning_rate": 5.263596008293398e-05,
+      "loss": 0.8623,
+      "step": 34150
+    },
+    {
+      "epoch": 1.3149181905678538,
+      "grad_norm": 1.6485599279403687,
+      "learning_rate": 5.260932967380178e-05,
+      "loss": 0.7832,
+      "step": 34155
+    },
+    {
+      "epoch": 1.3151106833493744,
+      "grad_norm": 1.0954639911651611,
+      "learning_rate": 5.2582703598232444e-05,
+      "loss": 0.7878,
+      "step": 34160
+    },
+    {
+      "epoch": 1.315303176130895,
+      "grad_norm": 1.3316006660461426,
+      "learning_rate": 5.255608185866079e-05,
+      "loss": 0.8128,
+      "step": 34165
+    },
+    {
+      "epoch": 1.3154956689124158,
+      "grad_norm": 0.8699563145637512,
+      "learning_rate": 5.252946445752113e-05,
+      "loss": 0.8295,
+      "step": 34170
+    },
+    {
+      "epoch": 1.3156881616939364,
+      "grad_norm": 1.7148654460906982,
+      "learning_rate": 5.2502851397247476e-05,
+      "loss": 0.9354,
+      "step": 34175
+    },
+    {
+      "epoch": 1.315880654475457,
+      "grad_norm": 1.602002501487732,
+      "learning_rate": 5.247624268027342e-05,
+      "loss": 0.6457,
+      "step": 34180
+    },
+    {
+      "epoch": 1.316073147256978,
+      "grad_norm": 0.8863468170166016,
+      "learning_rate": 5.24496383090322e-05,
+      "loss": 0.8983,
+      "step": 34185
+    },
+    {
+      "epoch": 1.3162656400384987,
+      "grad_norm": 1.1766276359558105,
+      "learning_rate": 5.242303828595649e-05,
+      "loss": 0.7426,
+      "step": 34190
+    },
+    {
+      "epoch": 1.3164581328200193,
+      "grad_norm": 1.0617400407791138,
+      "learning_rate": 5.2396442613478825e-05,
+      "loss": 0.8074,
+      "step": 34195
+    },
+    {
+      "epoch": 1.31665062560154,
+      "grad_norm": 1.5441138744354248,
+      "learning_rate": 5.236985129403112e-05,
+      "loss": 0.7365,
+      "step": 34200
+    },
+    {
+      "epoch": 1.3168431183830607,
+      "grad_norm": 1.1525336503982544,
+      "learning_rate": 5.234326433004497e-05,
+      "loss": 0.8009,
+      "step": 34205
+    },
+    {
+      "epoch": 1.3170356111645813,
+      "grad_norm": 1.6178624629974365,
+      "learning_rate": 5.231668172395161e-05,
+      "loss": 0.8423,
+      "step": 34210
+    },
+    {
+      "epoch": 1.317228103946102,
+      "grad_norm": 0.7218144536018372,
+      "learning_rate": 5.229010347818187e-05,
+      "loss": 0.5935,
+      "step": 34215
+    },
+    {
+      "epoch": 1.3174205967276227,
+      "grad_norm": 1.203621506690979,
+      "learning_rate": 5.226352959516605e-05,
+      "loss": 0.812,
+      "step": 34220
+    },
+    {
+      "epoch": 1.3176130895091434,
+      "grad_norm": 1.475656509399414,
+      "learning_rate": 5.2236960077334296e-05,
+      "loss": 1.0383,
+      "step": 34225
+    },
+    {
+      "epoch": 1.317805582290664,
+      "grad_norm": 1.2111153602600098,
+      "learning_rate": 5.2210394927116105e-05,
+      "loss": 0.8406,
+      "step": 34230
+    },
+    {
+      "epoch": 1.3179980750721847,
+      "grad_norm": 2.3121533393859863,
+      "learning_rate": 5.21838341469407e-05,
+      "loss": 0.8353,
+      "step": 34235
+    },
+    {
+      "epoch": 1.3181905678537054,
+      "grad_norm": 1.2453131675720215,
+      "learning_rate": 5.215727773923693e-05,
+      "loss": 0.8723,
+      "step": 34240
+    },
+    {
+      "epoch": 1.3183830606352263,
+      "grad_norm": 1.351366639137268,
+      "learning_rate": 5.2130725706433224e-05,
+      "loss": 0.905,
+      "step": 34245
+    },
+    {
+      "epoch": 1.318575553416747,
+      "grad_norm": 1.8964182138442993,
+      "learning_rate": 5.21041780509575e-05,
+      "loss": 0.9963,
+      "step": 34250
+    },
+    {
+      "epoch": 1.3187680461982676,
+      "grad_norm": 1.1896933317184448,
+      "learning_rate": 5.2077634775237415e-05,
+      "loss": 0.8957,
+      "step": 34255
+    },
+    {
+      "epoch": 1.3189605389797883,
+      "grad_norm": 2.013298511505127,
+      "learning_rate": 5.205109588170016e-05,
+      "loss": 0.8777,
+      "step": 34260
+    },
+    {
+      "epoch": 1.319153031761309,
+      "grad_norm": 1.3011314868927002,
+      "learning_rate": 5.202456137277256e-05,
+      "loss": 1.0396,
+      "step": 34265
+    },
+    {
+      "epoch": 1.3193455245428296,
+      "grad_norm": 1.5230119228363037,
+      "learning_rate": 5.199803125088107e-05,
+      "loss": 1.0502,
+      "step": 34270
+    },
+    {
+      "epoch": 1.3195380173243503,
+      "grad_norm": 1.011749505996704,
+      "learning_rate": 5.197150551845155e-05,
+      "loss": 0.7362,
+      "step": 34275
+    },
+    {
+      "epoch": 1.3197305101058712,
+      "grad_norm": 1.0205798149108887,
+      "learning_rate": 5.1944984177909765e-05,
+      "loss": 0.8891,
+      "step": 34280
+    },
+    {
+      "epoch": 1.3199230028873918,
+      "grad_norm": 1.5326029062271118,
+      "learning_rate": 5.1918467231680815e-05,
+      "loss": 0.7624,
+      "step": 34285
+    },
+    {
+      "epoch": 1.3201154956689125,
+      "grad_norm": 1.5170354843139648,
+      "learning_rate": 5.1891954682189505e-05,
+      "loss": 0.8149,
+      "step": 34290
+    },
+    {
+      "epoch": 1.3203079884504332,
+      "grad_norm": 0.9941301345825195,
+      "learning_rate": 5.186544653186026e-05,
+      "loss": 0.9011,
+      "step": 34295
+    },
+    {
+      "epoch": 1.3205004812319538,
+      "grad_norm": 3.1168370246887207,
+      "learning_rate": 5.183894278311712e-05,
+      "loss": 0.9253,
+      "step": 34300
+    },
+    {
+      "epoch": 1.3206929740134745,
+      "grad_norm": 0.9092430472373962,
+      "learning_rate": 5.181244343838353e-05,
+      "loss": 0.8621,
+      "step": 34305
+    },
+    {
+      "epoch": 1.3208854667949952,
+      "grad_norm": 0.8857255578041077,
+      "learning_rate": 5.178594850008286e-05,
+      "loss": 0.7869,
+      "step": 34310
+    },
+    {
+      "epoch": 1.3210779595765159,
+      "grad_norm": 0.8203316330909729,
+      "learning_rate": 5.175945797063777e-05,
+      "loss": 0.9274,
+      "step": 34315
+    },
+    {
+      "epoch": 1.3212704523580365,
+      "grad_norm": 1.180347204208374,
+      "learning_rate": 5.173297185247068e-05,
+      "loss": 0.7436,
+      "step": 34320
+    },
+    {
+      "epoch": 1.3214629451395572,
+      "grad_norm": 2.1619889736175537,
+      "learning_rate": 5.170649014800358e-05,
+      "loss": 0.7501,
+      "step": 34325
+    },
+    {
+      "epoch": 1.3216554379210779,
+      "grad_norm": 1.5437921285629272,
+      "learning_rate": 5.168001285965808e-05,
+      "loss": 0.8124,
+      "step": 34330
+    },
+    {
+      "epoch": 1.3218479307025985,
+      "grad_norm": 1.1066104173660278,
+      "learning_rate": 5.165353998985523e-05,
+      "loss": 0.8066,
+      "step": 34335
+    },
+    {
+      "epoch": 1.3220404234841192,
+      "grad_norm": 0.8616192936897278,
+      "learning_rate": 5.162707154101597e-05,
+      "loss": 0.7576,
+      "step": 34340
+    },
+    {
+      "epoch": 1.32223291626564,
+      "grad_norm": 0.889975368976593,
+      "learning_rate": 5.160060751556053e-05,
+      "loss": 0.7436,
+      "step": 34345
+    },
+    {
+      "epoch": 1.3224254090471608,
+      "grad_norm": 2.0296168327331543,
+      "learning_rate": 5.157414791590891e-05,
+      "loss": 0.6742,
+      "step": 34350
+    },
+    {
+      "epoch": 1.3226179018286814,
+      "grad_norm": 1.0081722736358643,
+      "learning_rate": 5.154769274448068e-05,
+      "loss": 0.7889,
+      "step": 34355
+    },
+    {
+      "epoch": 1.3228103946102021,
+      "grad_norm": 1.0939741134643555,
+      "learning_rate": 5.152124200369503e-05,
+      "loss": 0.8367,
+      "step": 34360
+    },
+    {
+      "epoch": 1.3230028873917228,
+      "grad_norm": 1.5666990280151367,
+      "learning_rate": 5.14947956959706e-05,
+      "loss": 0.8282,
+      "step": 34365
+    },
+    {
+      "epoch": 1.3231953801732435,
+      "grad_norm": 1.164833903312683,
+      "learning_rate": 5.146835382372579e-05,
+      "loss": 0.8148,
+      "step": 34370
+    },
+    {
+      "epoch": 1.3233878729547641,
+      "grad_norm": 1.489606499671936,
+      "learning_rate": 5.144191638937854e-05,
+      "loss": 0.8357,
+      "step": 34375
+    },
+    {
+      "epoch": 1.323580365736285,
+      "grad_norm": 1.4213351011276245,
+      "learning_rate": 5.1415483395346356e-05,
+      "loss": 0.9297,
+      "step": 34380
+    },
+    {
+      "epoch": 1.3237728585178057,
+      "grad_norm": 1.414014220237732,
+      "learning_rate": 5.138905484404641e-05,
+      "loss": 0.9808,
+      "step": 34385
+    },
+    {
+      "epoch": 1.3239653512993264,
+      "grad_norm": 1.255434274673462,
+      "learning_rate": 5.136263073789536e-05,
+      "loss": 0.8107,
+      "step": 34390
+    },
+    {
+      "epoch": 1.324157844080847,
+      "grad_norm": 1.3263583183288574,
+      "learning_rate": 5.133621107930951e-05,
+      "loss": 0.837,
+      "step": 34395
+    },
+    {
+      "epoch": 1.3243503368623677,
+      "grad_norm": 1.525387167930603,
+      "learning_rate": 5.1309795870704815e-05,
+      "loss": 0.8036,
+      "step": 34400
+    },
+    {
+      "epoch": 1.3245428296438884,
+      "grad_norm": 0.9029211401939392,
+      "learning_rate": 5.128338511449676e-05,
+      "loss": 0.9465,
+      "step": 34405
+    },
+    {
+      "epoch": 1.324735322425409,
+      "grad_norm": 1.5629209280014038,
+      "learning_rate": 5.1256978813100354e-05,
+      "loss": 0.8984,
+      "step": 34410
+    },
+    {
+      "epoch": 1.3249278152069297,
+      "grad_norm": 1.4381479024887085,
+      "learning_rate": 5.123057696893042e-05,
+      "loss": 0.7762,
+      "step": 34415
+    },
+    {
+      "epoch": 1.3251203079884504,
+      "grad_norm": 1.8477442264556885,
+      "learning_rate": 5.1204179584401115e-05,
+      "loss": 0.8943,
+      "step": 34420
+    },
+    {
+      "epoch": 1.325312800769971,
+      "grad_norm": 1.091605305671692,
+      "learning_rate": 5.117778666192634e-05,
+      "loss": 0.7989,
+      "step": 34425
+    },
+    {
+      "epoch": 1.3255052935514917,
+      "grad_norm": 1.6211384534835815,
+      "learning_rate": 5.1151398203919564e-05,
+      "loss": 0.7188,
+      "step": 34430
+    },
+    {
+      "epoch": 1.3256977863330124,
+      "grad_norm": 1.434945821762085,
+      "learning_rate": 5.1125014212793854e-05,
+      "loss": 0.8346,
+      "step": 34435
+    },
+    {
+      "epoch": 1.3258902791145333,
+      "grad_norm": 1.272206425666809,
+      "learning_rate": 5.1098634690961765e-05,
+      "loss": 0.8576,
+      "step": 34440
+    },
+    {
+      "epoch": 1.326082771896054,
+      "grad_norm": 0.9219827055931091,
+      "learning_rate": 5.107225964083566e-05,
+      "loss": 0.7725,
+      "step": 34445
+    },
+    {
+      "epoch": 1.3262752646775746,
+      "grad_norm": 1.1530919075012207,
+      "learning_rate": 5.1045889064827255e-05,
+      "loss": 0.9481,
+      "step": 34450
+    },
+    {
+      "epoch": 1.3264677574590953,
+      "grad_norm": 0.9978432655334473,
+      "learning_rate": 5.101952296534802e-05,
+      "loss": 0.8516,
+      "step": 34455
+    },
+    {
+      "epoch": 1.326660250240616,
+      "grad_norm": 1.018656611442566,
+      "learning_rate": 5.0993161344808924e-05,
+      "loss": 0.9756,
+      "step": 34460
+    },
+    {
+      "epoch": 1.3268527430221366,
+      "grad_norm": 1.7046258449554443,
+      "learning_rate": 5.0966804205620635e-05,
+      "loss": 1.018,
+      "step": 34465
+    },
+    {
+      "epoch": 1.3270452358036573,
+      "grad_norm": 1.1494272947311401,
+      "learning_rate": 5.094045155019325e-05,
+      "loss": 0.9536,
+      "step": 34470
+    },
+    {
+      "epoch": 1.3272377285851782,
+      "grad_norm": 2.197441339492798,
+      "learning_rate": 5.0914103380936564e-05,
+      "loss": 0.7673,
+      "step": 34475
+    },
+    {
+      "epoch": 1.3274302213666989,
+      "grad_norm": 1.758847713470459,
+      "learning_rate": 5.0887759700259965e-05,
+      "loss": 0.8132,
+      "step": 34480
+    },
+    {
+      "epoch": 1.3276227141482195,
+      "grad_norm": 1.2169286012649536,
+      "learning_rate": 5.086142051057241e-05,
+      "loss": 0.7716,
+      "step": 34485
+    },
+    {
+      "epoch": 1.3278152069297402,
+      "grad_norm": 1.7917555570602417,
+      "learning_rate": 5.083508581428247e-05,
+      "loss": 0.9653,
+      "step": 34490
+    },
+    {
+      "epoch": 1.3280076997112609,
+      "grad_norm": 1.4146066904067993,
+      "learning_rate": 5.080875561379821e-05,
+      "loss": 0.8737,
+      "step": 34495
+    },
+    {
+      "epoch": 1.3282001924927815,
+      "grad_norm": 1.2591365575790405,
+      "learning_rate": 5.0782429911527374e-05,
+      "loss": 0.7709,
+      "step": 34500
+    },
+    {
+      "epoch": 1.3283926852743022,
+      "grad_norm": 1.0664514303207397,
+      "learning_rate": 5.07561087098773e-05,
+      "loss": 0.8534,
+      "step": 34505
+    },
+    {
+      "epoch": 1.3285851780558229,
+      "grad_norm": 2.1911308765411377,
+      "learning_rate": 5.072979201125491e-05,
+      "loss": 0.944,
+      "step": 34510
+    },
+    {
+      "epoch": 1.3287776708373435,
+      "grad_norm": 0.8972740769386292,
+      "learning_rate": 5.070347981806657e-05,
+      "loss": 0.7574,
+      "step": 34515
+    },
+    {
+      "epoch": 1.3289701636188642,
+      "grad_norm": 1.0241605043411255,
+      "learning_rate": 5.067717213271852e-05,
+      "loss": 0.9743,
+      "step": 34520
+    },
+    {
+      "epoch": 1.329162656400385,
+      "grad_norm": 1.0732407569885254,
+      "learning_rate": 5.065086895761628e-05,
+      "loss": 0.7564,
+      "step": 34525
+    },
+    {
+      "epoch": 1.3293551491819056,
+      "grad_norm": 1.0235189199447632,
+      "learning_rate": 5.062457029516523e-05,
+      "loss": 0.9152,
+      "step": 34530
+    },
+    {
+      "epoch": 1.3295476419634265,
+      "grad_norm": 2.0122148990631104,
+      "learning_rate": 5.059827614777011e-05,
+      "loss": 1.0821,
+      "step": 34535
+    },
+    {
+      "epoch": 1.3297401347449471,
+      "grad_norm": 0.8569071292877197,
+      "learning_rate": 5.057198651783538e-05,
+      "loss": 0.7275,
+      "step": 34540
+    },
+    {
+      "epoch": 1.3299326275264678,
+      "grad_norm": 1.707134485244751,
+      "learning_rate": 5.0545701407765045e-05,
+      "loss": 0.9075,
+      "step": 34545
+    },
+    {
+      "epoch": 1.3301251203079885,
+      "grad_norm": 1.5125768184661865,
+      "learning_rate": 5.051942081996276e-05,
+      "loss": 0.6996,
+      "step": 34550
+    },
+    {
+      "epoch": 1.3303176130895091,
+      "grad_norm": 0.8091332316398621,
+      "learning_rate": 5.049314475683158e-05,
+      "loss": 0.6558,
+      "step": 34555
+    },
+    {
+      "epoch": 1.3305101058710298,
+      "grad_norm": 1.32486093044281,
+      "learning_rate": 5.046687322077444e-05,
+      "loss": 0.8852,
+      "step": 34560
+    },
+    {
+      "epoch": 1.3307025986525505,
+      "grad_norm": 1.4415032863616943,
+      "learning_rate": 5.0440606214193574e-05,
+      "loss": 0.8404,
+      "step": 34565
+    },
+    {
+      "epoch": 1.3308950914340711,
+      "grad_norm": 1.403547763824463,
+      "learning_rate": 5.0414343739490975e-05,
+      "loss": 0.7527,
+      "step": 34570
+    },
+    {
+      "epoch": 1.331087584215592,
+      "grad_norm": 2.1569442749023438,
+      "learning_rate": 5.038808579906816e-05,
+      "loss": 0.7985,
+      "step": 34575
+    },
+    {
+      "epoch": 1.3312800769971127,
+      "grad_norm": 1.6453166007995605,
+      "learning_rate": 5.036183239532629e-05,
+      "loss": 1.0465,
+      "step": 34580
+    },
+    {
+      "epoch": 1.3314725697786334,
+      "grad_norm": 1.2388137578964233,
+      "learning_rate": 5.0335583530665985e-05,
+      "loss": 0.8442,
+      "step": 34585
+    },
+    {
+      "epoch": 1.331665062560154,
+      "grad_norm": 1.0725327730178833,
+      "learning_rate": 5.0309339207487574e-05,
+      "loss": 0.7288,
+      "step": 34590
+    },
+    {
+      "epoch": 1.3318575553416747,
+      "grad_norm": 2.725146532058716,
+      "learning_rate": 5.028309942819091e-05,
+      "loss": 0.89,
+      "step": 34595
+    },
+    {
+      "epoch": 1.3320500481231954,
+      "grad_norm": 0.9890113472938538,
+      "learning_rate": 5.025686419517548e-05,
+      "loss": 0.7072,
+      "step": 34600
+    },
+    {
+      "epoch": 1.332242540904716,
+      "grad_norm": 1.8415369987487793,
+      "learning_rate": 5.023063351084033e-05,
+      "loss": 0.8123,
+      "step": 34605
+    },
+    {
+      "epoch": 1.3324350336862367,
+      "grad_norm": 0.9944572448730469,
+      "learning_rate": 5.020440737758401e-05,
+      "loss": 0.9136,
+      "step": 34610
+    },
+    {
+      "epoch": 1.3326275264677574,
+      "grad_norm": 1.131773591041565,
+      "learning_rate": 5.017818579780478e-05,
+      "loss": 0.8766,
+      "step": 34615
+    },
+    {
+      "epoch": 1.332820019249278,
+      "grad_norm": 1.1717694997787476,
+      "learning_rate": 5.01519687739004e-05,
+      "loss": 0.8695,
+      "step": 34620
+    },
+    {
+      "epoch": 1.3330125120307987,
+      "grad_norm": 2.2559797763824463,
+      "learning_rate": 5.0125756308268324e-05,
+      "loss": 0.7504,
+      "step": 34625
+    },
+    {
+      "epoch": 1.3332050048123194,
+      "grad_norm": 1.2070955038070679,
+      "learning_rate": 5.0099548403305354e-05,
+      "loss": 0.7427,
+      "step": 34630
+    },
+    {
+      "epoch": 1.3333974975938403,
+      "grad_norm": 1.3688490390777588,
+      "learning_rate": 5.0073345061408205e-05,
+      "loss": 0.8539,
+      "step": 34635
+    },
+    {
+      "epoch": 1.333589990375361,
+      "grad_norm": 0.8127152919769287,
+      "learning_rate": 5.004714628497288e-05,
+      "loss": 0.8662,
+      "step": 34640
+    },
+    {
+      "epoch": 1.3337824831568816,
+      "grad_norm": 1.9359740018844604,
+      "learning_rate": 5.0020952076395124e-05,
+      "loss": 0.9428,
+      "step": 34645
+    },
+    {
+      "epoch": 1.3339749759384023,
+      "grad_norm": 2.281846523284912,
+      "learning_rate": 4.999476243807021e-05,
+      "loss": 0.9462,
+      "step": 34650
+    },
+    {
+      "epoch": 1.334167468719923,
+      "grad_norm": 0.9672248959541321,
+      "learning_rate": 4.996857737239305e-05,
+      "loss": 0.6946,
+      "step": 34655
+    },
+    {
+      "epoch": 1.3343599615014436,
+      "grad_norm": 1.2799361944198608,
+      "learning_rate": 4.994239688175799e-05,
+      "loss": 0.8645,
+      "step": 34660
+    },
+    {
+      "epoch": 1.3345524542829643,
+      "grad_norm": 1.4629347324371338,
+      "learning_rate": 4.991622096855923e-05,
+      "loss": 0.7476,
+      "step": 34665
+    },
+    {
+      "epoch": 1.3347449470644852,
+      "grad_norm": 1.2526726722717285,
+      "learning_rate": 4.9890049635190216e-05,
+      "loss": 0.6983,
+      "step": 34670
+    },
+    {
+      "epoch": 1.3349374398460059,
+      "grad_norm": 1.9672549962997437,
+      "learning_rate": 4.9863882884044234e-05,
+      "loss": 0.9105,
+      "step": 34675
+    },
+    {
+      "epoch": 1.3351299326275265,
+      "grad_norm": 1.3764963150024414,
+      "learning_rate": 4.983772071751405e-05,
+      "loss": 0.8268,
+      "step": 34680
+    },
+    {
+      "epoch": 1.3353224254090472,
+      "grad_norm": 1.1114428043365479,
+      "learning_rate": 4.9811563137992036e-05,
+      "loss": 0.6683,
+      "step": 34685
+    },
+    {
+      "epoch": 1.335514918190568,
+      "grad_norm": 1.1706123352050781,
+      "learning_rate": 4.978541014787006e-05,
+      "loss": 0.8398,
+      "step": 34690
+    },
+    {
+      "epoch": 1.3357074109720886,
+      "grad_norm": 1.0030032396316528,
+      "learning_rate": 4.9759261749539695e-05,
+      "loss": 0.7736,
+      "step": 34695
+    },
+    {
+      "epoch": 1.3358999037536092,
+      "grad_norm": 2.052016258239746,
+      "learning_rate": 4.9733117945392026e-05,
+      "loss": 0.9667,
+      "step": 34700
+    },
+    {
+      "epoch": 1.33609239653513,
+      "grad_norm": 2.7888071537017822,
+      "learning_rate": 4.970697873781774e-05,
+      "loss": 1.0169,
+      "step": 34705
+    },
+    {
+      "epoch": 1.3362848893166506,
+      "grad_norm": 1.691266655921936,
+      "learning_rate": 4.968084412920712e-05,
+      "loss": 0.6926,
+      "step": 34710
+    },
+    {
+      "epoch": 1.3364773820981712,
+      "grad_norm": 1.3733407258987427,
+      "learning_rate": 4.965471412194993e-05,
+      "loss": 0.7997,
+      "step": 34715
+    },
+    {
+      "epoch": 1.336669874879692,
+      "grad_norm": 2.068976879119873,
+      "learning_rate": 4.9628588718435634e-05,
+      "loss": 0.9425,
+      "step": 34720
+    },
+    {
+      "epoch": 1.3368623676612126,
+      "grad_norm": 1.9062495231628418,
+      "learning_rate": 4.960246792105322e-05,
+      "loss": 0.8659,
+      "step": 34725
+    },
+    {
+      "epoch": 1.3370548604427335,
+      "grad_norm": 1.3141814470291138,
+      "learning_rate": 4.957635173219129e-05,
+      "loss": 0.7289,
+      "step": 34730
+    },
+    {
+      "epoch": 1.3372473532242541,
+      "grad_norm": 0.9774174094200134,
+      "learning_rate": 4.955024015423789e-05,
+      "loss": 0.7244,
+      "step": 34735
+    },
+    {
+      "epoch": 1.3374398460057748,
+      "grad_norm": 1.086158037185669,
+      "learning_rate": 4.952413318958092e-05,
+      "loss": 0.9397,
+      "step": 34740
+    },
+    {
+      "epoch": 1.3376323387872955,
+      "grad_norm": 1.092448353767395,
+      "learning_rate": 4.9498030840607547e-05,
+      "loss": 0.7571,
+      "step": 34745
+    },
+    {
+      "epoch": 1.3378248315688162,
+      "grad_norm": 1.8962229490280151,
+      "learning_rate": 4.947193310970471e-05,
+      "loss": 0.8264,
+      "step": 34750
+    },
+    {
+      "epoch": 1.3380173243503368,
+      "grad_norm": 1.5247009992599487,
+      "learning_rate": 4.944583999925888e-05,
+      "loss": 0.7856,
+      "step": 34755
+    },
+    {
+      "epoch": 1.3382098171318575,
+      "grad_norm": 1.6550134420394897,
+      "learning_rate": 4.941975151165613e-05,
+      "loss": 1.024,
+      "step": 34760
+    },
+    {
+      "epoch": 1.3384023099133784,
+      "grad_norm": 1.2675317525863647,
+      "learning_rate": 4.939366764928196e-05,
+      "loss": 0.8908,
+      "step": 34765
+    },
+    {
+      "epoch": 1.338594802694899,
+      "grad_norm": 1.3318907022476196,
+      "learning_rate": 4.9367588414521714e-05,
+      "loss": 0.7813,
+      "step": 34770
+    },
+    {
+      "epoch": 1.3387872954764197,
+      "grad_norm": 1.605064034461975,
+      "learning_rate": 4.934151380976007e-05,
+      "loss": 0.9145,
+      "step": 34775
+    },
+    {
+      "epoch": 1.3389797882579404,
+      "grad_norm": 1.6478382349014282,
+      "learning_rate": 4.9315443837381417e-05,
+      "loss": 1.0189,
+      "step": 34780
+    },
+    {
+      "epoch": 1.339172281039461,
+      "grad_norm": 1.617727279663086,
+      "learning_rate": 4.9289378499769655e-05,
+      "loss": 0.7663,
+      "step": 34785
+    },
+    {
+      "epoch": 1.3393647738209817,
+      "grad_norm": 2.157989025115967,
+      "learning_rate": 4.9263317799308305e-05,
+      "loss": 0.9308,
+      "step": 34790
+    },
+    {
+      "epoch": 1.3395572666025024,
+      "grad_norm": 1.5868369340896606,
+      "learning_rate": 4.923726173838048e-05,
+      "loss": 1.0234,
+      "step": 34795
+    },
+    {
+      "epoch": 1.339749759384023,
+      "grad_norm": 1.0689306259155273,
+      "learning_rate": 4.921121031936876e-05,
+      "loss": 0.9376,
+      "step": 34800
+    },
+    {
+      "epoch": 1.3399422521655437,
+      "grad_norm": 1.227166771888733,
+      "learning_rate": 4.918516354465541e-05,
+      "loss": 0.9385,
+      "step": 34805
+    },
+    {
+      "epoch": 1.3401347449470644,
+      "grad_norm": 0.9989410638809204,
+      "learning_rate": 4.915912141662225e-05,
+      "loss": 0.865,
+      "step": 34810
+    },
+    {
+      "epoch": 1.340327237728585,
+      "grad_norm": 1.1852025985717773,
+      "learning_rate": 4.913308393765066e-05,
+      "loss": 0.9824,
+      "step": 34815
+    },
+    {
+      "epoch": 1.3405197305101058,
+      "grad_norm": 1.3979310989379883,
+      "learning_rate": 4.910705111012153e-05,
+      "loss": 0.9297,
+      "step": 34820
+    },
+    {
+      "epoch": 1.3407122232916264,
+      "grad_norm": 1.6673941612243652,
+      "learning_rate": 4.90810229364155e-05,
+      "loss": 0.777,
+      "step": 34825
+    },
+    {
+      "epoch": 1.3409047160731473,
+      "grad_norm": 1.139458417892456,
+      "learning_rate": 4.9054999418912586e-05,
+      "loss": 0.7229,
+      "step": 34830
+    },
+    {
+      "epoch": 1.341097208854668,
+      "grad_norm": 0.8161736726760864,
+      "learning_rate": 4.902898055999249e-05,
+      "loss": 0.7305,
+      "step": 34835
+    },
+    {
+      "epoch": 1.3412897016361887,
+      "grad_norm": 1.6782662868499756,
+      "learning_rate": 4.9002966362034464e-05,
+      "loss": 0.8107,
+      "step": 34840
+    },
+    {
+      "epoch": 1.3414821944177093,
+      "grad_norm": 1.5833643674850464,
+      "learning_rate": 4.897695682741739e-05,
+      "loss": 0.7968,
+      "step": 34845
+    },
+    {
+      "epoch": 1.34167468719923,
+      "grad_norm": 1.409138798713684,
+      "learning_rate": 4.895095195851953e-05,
+      "loss": 0.9225,
+      "step": 34850
+    },
+    {
+      "epoch": 1.3418671799807507,
+      "grad_norm": 1.4280591011047363,
+      "learning_rate": 4.892495175771903e-05,
+      "loss": 0.7304,
+      "step": 34855
+    },
+    {
+      "epoch": 1.3420596727622713,
+      "grad_norm": 1.0369322299957275,
+      "learning_rate": 4.889895622739331e-05,
+      "loss": 0.884,
+      "step": 34860
+    },
+    {
+      "epoch": 1.3422521655437922,
+      "grad_norm": 0.9493328928947449,
+      "learning_rate": 4.887296536991953e-05,
+      "loss": 0.8159,
+      "step": 34865
+    },
+    {
+      "epoch": 1.342444658325313,
+      "grad_norm": 1.258876085281372,
+      "learning_rate": 4.884697918767438e-05,
+      "loss": 0.8475,
+      "step": 34870
+    },
+    {
+      "epoch": 1.3426371511068336,
+      "grad_norm": 1.2037972211837769,
+      "learning_rate": 4.8820997683034166e-05,
+      "loss": 0.8062,
+      "step": 34875
+    },
+    {
+      "epoch": 1.3428296438883542,
+      "grad_norm": 1.7686017751693726,
+      "learning_rate": 4.879502085837461e-05,
+      "loss": 0.8544,
+      "step": 34880
+    },
+    {
+      "epoch": 1.343022136669875,
+      "grad_norm": 1.291748285293579,
+      "learning_rate": 4.8769048716071264e-05,
+      "loss": 0.8541,
+      "step": 34885
+    },
+    {
+      "epoch": 1.3432146294513956,
+      "grad_norm": 1.2345646619796753,
+      "learning_rate": 4.8743081258499005e-05,
+      "loss": 0.9746,
+      "step": 34890
+    },
+    {
+      "epoch": 1.3434071222329163,
+      "grad_norm": 2.3298258781433105,
+      "learning_rate": 4.871711848803241e-05,
+      "loss": 0.9702,
+      "step": 34895
+    },
+    {
+      "epoch": 1.343599615014437,
+      "grad_norm": 1.7724379301071167,
+      "learning_rate": 4.869116040704562e-05,
+      "loss": 0.8757,
+      "step": 34900
+    },
+    {
+      "epoch": 1.3437921077959576,
+      "grad_norm": 1.737610101699829,
+      "learning_rate": 4.866520701791235e-05,
+      "loss": 0.7819,
+      "step": 34905
+    },
+    {
+      "epoch": 1.3439846005774783,
+      "grad_norm": 1.1426078081130981,
+      "learning_rate": 4.863925832300581e-05,
+      "loss": 0.8039,
+      "step": 34910
+    },
+    {
+      "epoch": 1.344177093358999,
+      "grad_norm": 1.7456549406051636,
+      "learning_rate": 4.8613314324698855e-05,
+      "loss": 0.899,
+      "step": 34915
+    },
+    {
+      "epoch": 1.3443695861405196,
+      "grad_norm": 1.9409685134887695,
+      "learning_rate": 4.8587375025363914e-05,
+      "loss": 0.9121,
+      "step": 34920
+    },
+    {
+      "epoch": 1.3445620789220405,
+      "grad_norm": 0.8889964818954468,
+      "learning_rate": 4.856144042737293e-05,
+      "loss": 0.8982,
+      "step": 34925
+    },
+    {
+      "epoch": 1.3447545717035612,
+      "grad_norm": 1.207217812538147,
+      "learning_rate": 4.8535510533097516e-05,
+      "loss": 0.9708,
+      "step": 34930
+    },
+    {
+      "epoch": 1.3449470644850818,
+      "grad_norm": 1.7888151407241821,
+      "learning_rate": 4.8509585344908705e-05,
+      "loss": 0.7838,
+      "step": 34935
+    },
+    {
+      "epoch": 1.3451395572666025,
+      "grad_norm": 1.8165149688720703,
+      "learning_rate": 4.8483664865177226e-05,
+      "loss": 0.8495,
+      "step": 34940
+    },
+    {
+      "epoch": 1.3453320500481232,
+      "grad_norm": 1.4321086406707764,
+      "learning_rate": 4.845774909627332e-05,
+      "loss": 0.8093,
+      "step": 34945
+    },
+    {
+      "epoch": 1.3455245428296438,
+      "grad_norm": 1.2221522331237793,
+      "learning_rate": 4.843183804056687e-05,
+      "loss": 0.906,
+      "step": 34950
+    },
+    {
+      "epoch": 1.3457170356111645,
+      "grad_norm": 1.2078337669372559,
+      "learning_rate": 4.8405931700427145e-05,
+      "loss": 0.7406,
+      "step": 34955
+    },
+    {
+      "epoch": 1.3459095283926854,
+      "grad_norm": 1.0744653940200806,
+      "learning_rate": 4.838003007822326e-05,
+      "loss": 0.8133,
+      "step": 34960
+    },
+    {
+      "epoch": 1.346102021174206,
+      "grad_norm": 1.2182228565216064,
+      "learning_rate": 4.835413317632363e-05,
+      "loss": 0.8024,
+      "step": 34965
+    },
+    {
+      "epoch": 1.3462945139557267,
+      "grad_norm": 1.1697922945022583,
+      "learning_rate": 4.8328240997096406e-05,
+      "loss": 0.9592,
+      "step": 34970
+    },
+    {
+      "epoch": 1.3464870067372474,
+      "grad_norm": 1.4923276901245117,
+      "learning_rate": 4.830235354290925e-05,
+      "loss": 0.7919,
+      "step": 34975
+    },
+    {
+      "epoch": 1.346679499518768,
+      "grad_norm": 1.7563990354537964,
+      "learning_rate": 4.827647081612944e-05,
+      "loss": 0.8882,
+      "step": 34980
+    },
+    {
+      "epoch": 1.3468719923002888,
+      "grad_norm": 2.2947778701782227,
+      "learning_rate": 4.825059281912365e-05,
+      "loss": 0.9188,
+      "step": 34985
+    },
+    {
+      "epoch": 1.3470644850818094,
+      "grad_norm": 1.5248280763626099,
+      "learning_rate": 4.822471955425841e-05,
+      "loss": 0.965,
+      "step": 34990
+    },
+    {
+      "epoch": 1.34725697786333,
+      "grad_norm": 0.8243402242660522,
+      "learning_rate": 4.819885102389956e-05,
+      "loss": 0.8224,
+      "step": 34995
+    },
+    {
+      "epoch": 1.3474494706448508,
+      "grad_norm": 0.7794885039329529,
+      "learning_rate": 4.817298723041264e-05,
+      "loss": 0.7754,
+      "step": 35000
+    },
+    {
+      "epoch": 1.3476419634263714,
+      "grad_norm": 1.2712163925170898,
+      "learning_rate": 4.8147128176162695e-05,
+      "loss": 0.8612,
+      "step": 35005
+    },
+    {
+      "epoch": 1.347834456207892,
+      "grad_norm": 0.9318971633911133,
+      "learning_rate": 4.8121273863514435e-05,
+      "loss": 0.7837,
+      "step": 35010
+    },
+    {
+      "epoch": 1.3480269489894128,
+      "grad_norm": 1.127020001411438,
+      "learning_rate": 4.809542429483197e-05,
+      "loss": 0.7845,
+      "step": 35015
+    },
+    {
+      "epoch": 1.3482194417709337,
+      "grad_norm": 1.16558039188385,
+      "learning_rate": 4.806957947247912e-05,
+      "loss": 0.8318,
+      "step": 35020
+    },
+    {
+      "epoch": 1.3484119345524543,
+      "grad_norm": 1.0364631414413452,
+      "learning_rate": 4.804373939881922e-05,
+      "loss": 0.9216,
+      "step": 35025
+    },
+    {
+      "epoch": 1.348604427333975,
+      "grad_norm": 1.3050929307937622,
+      "learning_rate": 4.801790407621518e-05,
+      "loss": 0.9627,
+      "step": 35030
+    },
+    {
+      "epoch": 1.3487969201154957,
+      "grad_norm": 1.1998785734176636,
+      "learning_rate": 4.799207350702949e-05,
+      "loss": 0.884,
+      "step": 35035
+    },
+    {
+      "epoch": 1.3489894128970163,
+      "grad_norm": 0.8209556341171265,
+      "learning_rate": 4.796624769362409e-05,
+      "loss": 0.6925,
+      "step": 35040
+    },
+    {
+      "epoch": 1.349181905678537,
+      "grad_norm": 2.4278218746185303,
+      "learning_rate": 4.794042663836071e-05,
+      "loss": 0.7594,
+      "step": 35045
+    },
+    {
+      "epoch": 1.3493743984600577,
+      "grad_norm": 1.8275748491287231,
+      "learning_rate": 4.791461034360043e-05,
+      "loss": 0.9516,
+      "step": 35050
+    },
+    {
+      "epoch": 1.3495668912415784,
+      "grad_norm": 1.21636962890625,
+      "learning_rate": 4.7888798811703985e-05,
+      "loss": 0.9819,
+      "step": 35055
+    },
+    {
+      "epoch": 1.3497593840230993,
+      "grad_norm": 1.5131886005401611,
+      "learning_rate": 4.7862992045031684e-05,
+      "loss": 0.955,
+      "step": 35060
+    },
+    {
+      "epoch": 1.34995187680462,
+      "grad_norm": 1.186279296875,
+      "learning_rate": 4.7837190045943436e-05,
+      "loss": 0.8632,
+      "step": 35065
+    },
+    {
+      "epoch": 1.3501443695861406,
+      "grad_norm": 1.2006460428237915,
+      "learning_rate": 4.7811392816798525e-05,
+      "loss": 0.8132,
+      "step": 35070
+    },
+    {
+      "epoch": 1.3503368623676613,
+      "grad_norm": 1.6792789697647095,
+      "learning_rate": 4.7785600359956096e-05,
+      "loss": 0.8756,
+      "step": 35075
+    },
+    {
+      "epoch": 1.350529355149182,
+      "grad_norm": 1.5003536939620972,
+      "learning_rate": 4.77598126777746e-05,
+      "loss": 0.7711,
+      "step": 35080
+    },
+    {
+      "epoch": 1.3507218479307026,
+      "grad_norm": 1.4191226959228516,
+      "learning_rate": 4.7734029772612165e-05,
+      "loss": 0.7895,
+      "step": 35085
+    },
+    {
+      "epoch": 1.3509143407122233,
+      "grad_norm": 1.391128659248352,
+      "learning_rate": 4.7708251646826476e-05,
+      "loss": 0.8829,
+      "step": 35090
+    },
+    {
+      "epoch": 1.351106833493744,
+      "grad_norm": 1.1470931768417358,
+      "learning_rate": 4.7682478302774816e-05,
+      "loss": 0.8812,
+      "step": 35095
+    },
+    {
+      "epoch": 1.3512993262752646,
+      "grad_norm": 1.3453543186187744,
+      "learning_rate": 4.765670974281386e-05,
+      "loss": 0.9689,
+      "step": 35100
+    },
+    {
+      "epoch": 1.3514918190567853,
+      "grad_norm": 1.4172368049621582,
+      "learning_rate": 4.763094596930014e-05,
+      "loss": 0.9129,
+      "step": 35105
+    },
+    {
+      "epoch": 1.351684311838306,
+      "grad_norm": 1.895308494567871,
+      "learning_rate": 4.7605186984589456e-05,
+      "loss": 0.8221,
+      "step": 35110
+    },
+    {
+      "epoch": 1.3518768046198266,
+      "grad_norm": 1.4038105010986328,
+      "learning_rate": 4.7579432791037335e-05,
+      "loss": 0.9491,
+      "step": 35115
+    },
+    {
+      "epoch": 1.3520692974013475,
+      "grad_norm": 1.111763834953308,
+      "learning_rate": 4.755368339099884e-05,
+      "loss": 0.7322,
+      "step": 35120
+    },
+    {
+      "epoch": 1.3522617901828682,
+      "grad_norm": 1.7646887302398682,
+      "learning_rate": 4.752793878682861e-05,
+      "loss": 0.8792,
+      "step": 35125
+    },
+    {
+      "epoch": 1.3524542829643889,
+      "grad_norm": 1.1743024587631226,
+      "learning_rate": 4.750219898088073e-05,
+      "loss": 0.9847,
+      "step": 35130
+    },
+    {
+      "epoch": 1.3526467757459095,
+      "grad_norm": 1.2739955186843872,
+      "learning_rate": 4.7476463975509e-05,
+      "loss": 0.7783,
+      "step": 35135
+    },
+    {
+      "epoch": 1.3528392685274302,
+      "grad_norm": 1.317272424697876,
+      "learning_rate": 4.74507337730667e-05,
+      "loss": 0.9875,
+      "step": 35140
+    },
+    {
+      "epoch": 1.3530317613089509,
+      "grad_norm": 0.9840208888053894,
+      "learning_rate": 4.74250083759067e-05,
+      "loss": 0.8396,
+      "step": 35145
+    },
+    {
+      "epoch": 1.3532242540904715,
+      "grad_norm": 1.4527745246887207,
+      "learning_rate": 4.739928778638143e-05,
+      "loss": 0.8583,
+      "step": 35150
+    },
+    {
+      "epoch": 1.3534167468719924,
+      "grad_norm": 1.6825000047683716,
+      "learning_rate": 4.7373572006842806e-05,
+      "loss": 0.9443,
+      "step": 35155
+    },
+    {
+      "epoch": 1.353609239653513,
+      "grad_norm": 1.2413737773895264,
+      "learning_rate": 4.734786103964242e-05,
+      "loss": 0.9722,
+      "step": 35160
+    },
+    {
+      "epoch": 1.3538017324350338,
+      "grad_norm": 1.2684810161590576,
+      "learning_rate": 4.732215488713133e-05,
+      "loss": 0.7281,
+      "step": 35165
+    },
+    {
+      "epoch": 1.3539942252165544,
+      "grad_norm": 1.8992153406143188,
+      "learning_rate": 4.729645355166027e-05,
+      "loss": 0.8221,
+      "step": 35170
+    },
+    {
+      "epoch": 1.354186717998075,
+      "grad_norm": 1.6467632055282593,
+      "learning_rate": 4.7270757035579325e-05,
+      "loss": 0.8439,
+      "step": 35175
+    },
+    {
+      "epoch": 1.3543792107795958,
+      "grad_norm": 1.0282069444656372,
+      "learning_rate": 4.724506534123843e-05,
+      "loss": 0.8158,
+      "step": 35180
+    },
+    {
+      "epoch": 1.3545717035611164,
+      "grad_norm": 1.7379052639007568,
+      "learning_rate": 4.72193784709868e-05,
+      "loss": 0.8245,
+      "step": 35185
+    },
+    {
+      "epoch": 1.3547641963426371,
+      "grad_norm": 2.1506638526916504,
+      "learning_rate": 4.719369642717336e-05,
+      "loss": 0.8818,
+      "step": 35190
+    },
+    {
+      "epoch": 1.3549566891241578,
+      "grad_norm": 1.284745454788208,
+      "learning_rate": 4.7168019212146576e-05,
+      "loss": 0.8457,
+      "step": 35195
+    },
+    {
+      "epoch": 1.3551491819056785,
+      "grad_norm": 0.897658109664917,
+      "learning_rate": 4.71423468282545e-05,
+      "loss": 0.8173,
+      "step": 35200
+    },
+    {
+      "epoch": 1.3553416746871991,
+      "grad_norm": 2.404270887374878,
+      "learning_rate": 4.711667927784458e-05,
+      "loss": 0.8396,
+      "step": 35205
+    },
+    {
+      "epoch": 1.3555341674687198,
+      "grad_norm": 1.8264572620391846,
+      "learning_rate": 4.7091016563264087e-05,
+      "loss": 0.9482,
+      "step": 35210
+    },
+    {
+      "epoch": 1.3557266602502407,
+      "grad_norm": 2.137230396270752,
+      "learning_rate": 4.70653586868596e-05,
+      "loss": 0.8982,
+      "step": 35215
+    },
+    {
+      "epoch": 1.3559191530317614,
+      "grad_norm": 1.1221061944961548,
+      "learning_rate": 4.703970565097742e-05,
+      "loss": 0.8983,
+      "step": 35220
+    },
+    {
+      "epoch": 1.356111645813282,
+      "grad_norm": 1.6483477354049683,
+      "learning_rate": 4.7014057457963315e-05,
+      "loss": 0.801,
+      "step": 35225
+    },
+    {
+      "epoch": 1.3563041385948027,
+      "grad_norm": 1.8643723726272583,
+      "learning_rate": 4.698841411016269e-05,
+      "loss": 1.0251,
+      "step": 35230
+    },
+    {
+      "epoch": 1.3564966313763234,
+      "grad_norm": 1.128015398979187,
+      "learning_rate": 4.6962775609920394e-05,
+      "loss": 0.844,
+      "step": 35235
+    },
+    {
+      "epoch": 1.356689124157844,
+      "grad_norm": 0.9045073390007019,
+      "learning_rate": 4.693714195958092e-05,
+      "loss": 0.869,
+      "step": 35240
+    },
+    {
+      "epoch": 1.3568816169393647,
+      "grad_norm": 1.8441210985183716,
+      "learning_rate": 4.691151316148832e-05,
+      "loss": 0.9395,
+      "step": 35245
+    },
+    {
+      "epoch": 1.3570741097208856,
+      "grad_norm": 1.437767505645752,
+      "learning_rate": 4.688588921798616e-05,
+      "loss": 0.6946,
+      "step": 35250
+    },
+    {
+      "epoch": 1.3572666025024063,
+      "grad_norm": 1.447969675064087,
+      "learning_rate": 4.68602701314176e-05,
+      "loss": 0.8183,
+      "step": 35255
+    },
+    {
+      "epoch": 1.357459095283927,
+      "grad_norm": 2.16900634765625,
+      "learning_rate": 4.68346559041253e-05,
+      "loss": 0.7718,
+      "step": 35260
+    },
+    {
+      "epoch": 1.3576515880654476,
+      "grad_norm": 1.2576295137405396,
+      "learning_rate": 4.680904653845152e-05,
+      "loss": 0.8408,
+      "step": 35265
+    },
+    {
+      "epoch": 1.3578440808469683,
+      "grad_norm": 1.3232860565185547,
+      "learning_rate": 4.678344203673808e-05,
+      "loss": 0.8454,
+      "step": 35270
+    },
+    {
+      "epoch": 1.358036573628489,
+      "grad_norm": 1.5161672830581665,
+      "learning_rate": 4.675784240132638e-05,
+      "loss": 0.8271,
+      "step": 35275
+    },
+    {
+      "epoch": 1.3582290664100096,
+      "grad_norm": 1.339381217956543,
+      "learning_rate": 4.6732247634557214e-05,
+      "loss": 0.8369,
+      "step": 35280
+    },
+    {
+      "epoch": 1.3584215591915303,
+      "grad_norm": 1.2155077457427979,
+      "learning_rate": 4.670665773877121e-05,
+      "loss": 0.8823,
+      "step": 35285
+    },
+    {
+      "epoch": 1.358614051973051,
+      "grad_norm": 1.6625285148620605,
+      "learning_rate": 4.6681072716308285e-05,
+      "loss": 0.9456,
+      "step": 35290
+    },
+    {
+      "epoch": 1.3588065447545716,
+      "grad_norm": 0.9515489339828491,
+      "learning_rate": 4.6655492569508056e-05,
+      "loss": 0.889,
+      "step": 35295
+    },
+    {
+      "epoch": 1.3589990375360923,
+      "grad_norm": 1.9020076990127563,
+      "learning_rate": 4.662991730070966e-05,
+      "loss": 0.8547,
+      "step": 35300
+    },
+    {
+      "epoch": 1.359191530317613,
+      "grad_norm": 1.41860032081604,
+      "learning_rate": 4.660434691225177e-05,
+      "loss": 0.8674,
+      "step": 35305
+    },
+    {
+      "epoch": 1.3593840230991339,
+      "grad_norm": 1.2650240659713745,
+      "learning_rate": 4.657878140647265e-05,
+      "loss": 0.834,
+      "step": 35310
+    },
+    {
+      "epoch": 1.3595765158806545,
+      "grad_norm": 1.1325031518936157,
+      "learning_rate": 4.655322078571013e-05,
+      "loss": 0.8641,
+      "step": 35315
+    },
+    {
+      "epoch": 1.3597690086621752,
+      "grad_norm": 1.4202800989151,
+      "learning_rate": 4.652766505230143e-05,
+      "loss": 0.8613,
+      "step": 35320
+    },
+    {
+      "epoch": 1.3599615014436959,
+      "grad_norm": 1.1674113273620605,
+      "learning_rate": 4.650211420858361e-05,
+      "loss": 0.9188,
+      "step": 35325
+    },
+    {
+      "epoch": 1.3601539942252165,
+      "grad_norm": 1.0588910579681396,
+      "learning_rate": 4.6476568256893025e-05,
+      "loss": 0.9119,
+      "step": 35330
+    },
+    {
+      "epoch": 1.3603464870067372,
+      "grad_norm": 2.010552167892456,
+      "learning_rate": 4.645102719956572e-05,
+      "loss": 0.7566,
+      "step": 35335
+    },
+    {
+      "epoch": 1.3605389797882579,
+      "grad_norm": 1.1530786752700806,
+      "learning_rate": 4.6425491038937244e-05,
+      "loss": 0.8898,
+      "step": 35340
+    },
+    {
+      "epoch": 1.3607314725697786,
+      "grad_norm": 1.016413927078247,
+      "learning_rate": 4.6399959777342746e-05,
+      "loss": 0.9265,
+      "step": 35345
+    },
+    {
+      "epoch": 1.3609239653512994,
+      "grad_norm": 0.9319851398468018,
+      "learning_rate": 4.6374433417116826e-05,
+      "loss": 0.8908,
+      "step": 35350
+    },
+    {
+      "epoch": 1.3611164581328201,
+      "grad_norm": 1.6301521062850952,
+      "learning_rate": 4.6348911960593736e-05,
+      "loss": 0.8063,
+      "step": 35355
+    },
+    {
+      "epoch": 1.3613089509143408,
+      "grad_norm": 1.0370879173278809,
+      "learning_rate": 4.632339541010726e-05,
+      "loss": 0.8456,
+      "step": 35360
+    },
+    {
+      "epoch": 1.3615014436958615,
+      "grad_norm": 1.1396170854568481,
+      "learning_rate": 4.629788376799065e-05,
+      "loss": 0.7626,
+      "step": 35365
+    },
+    {
+      "epoch": 1.3616939364773821,
+      "grad_norm": 1.1902436017990112,
+      "learning_rate": 4.6272377036576886e-05,
+      "loss": 0.8911,
+      "step": 35370
+    },
+    {
+      "epoch": 1.3618864292589028,
+      "grad_norm": 1.496978521347046,
+      "learning_rate": 4.6246875218198294e-05,
+      "loss": 0.7299,
+      "step": 35375
+    },
+    {
+      "epoch": 1.3620789220404235,
+      "grad_norm": 1.5510118007659912,
+      "learning_rate": 4.622137831518688e-05,
+      "loss": 0.7804,
+      "step": 35380
+    },
+    {
+      "epoch": 1.3622714148219441,
+      "grad_norm": 1.1670730113983154,
+      "learning_rate": 4.619588632987416e-05,
+      "loss": 0.8348,
+      "step": 35385
+    },
+    {
+      "epoch": 1.3624639076034648,
+      "grad_norm": 1.505469799041748,
+      "learning_rate": 4.617039926459127e-05,
+      "loss": 0.7221,
+      "step": 35390
+    },
+    {
+      "epoch": 1.3626564003849855,
+      "grad_norm": 1.3715969324111938,
+      "learning_rate": 4.61449171216687e-05,
+      "loss": 0.9265,
+      "step": 35395
+    },
+    {
+      "epoch": 1.3628488931665061,
+      "grad_norm": 0.8129710555076599,
+      "learning_rate": 4.611943990343677e-05,
+      "loss": 0.8726,
+      "step": 35400
+    },
+    {
+      "epoch": 1.3630413859480268,
+      "grad_norm": 0.745360791683197,
+      "learning_rate": 4.60939676122251e-05,
+      "loss": 0.8237,
+      "step": 35405
+    },
+    {
+      "epoch": 1.3632338787295477,
+      "grad_norm": 1.287010908126831,
+      "learning_rate": 4.606850025036299e-05,
+      "loss": 0.8159,
+      "step": 35410
+    },
+    {
+      "epoch": 1.3634263715110684,
+      "grad_norm": 1.7442882061004639,
+      "learning_rate": 4.604303782017928e-05,
+      "loss": 0.6109,
+      "step": 35415
+    },
+    {
+      "epoch": 1.363618864292589,
+      "grad_norm": 2.1345295906066895,
+      "learning_rate": 4.6017580324002364e-05,
+      "loss": 0.9068,
+      "step": 35420
+    },
+    {
+      "epoch": 1.3638113570741097,
+      "grad_norm": 1.987751841545105,
+      "learning_rate": 4.5992127764160054e-05,
+      "loss": 0.932,
+      "step": 35425
+    },
+    {
+      "epoch": 1.3640038498556304,
+      "grad_norm": 1.3692355155944824,
+      "learning_rate": 4.5966680142979954e-05,
+      "loss": 0.7202,
+      "step": 35430
+    },
+    {
+      "epoch": 1.364196342637151,
+      "grad_norm": 1.3024131059646606,
+      "learning_rate": 4.594123746278899e-05,
+      "loss": 0.9063,
+      "step": 35435
+    },
+    {
+      "epoch": 1.3643888354186717,
+      "grad_norm": 1.8773390054702759,
+      "learning_rate": 4.591579972591376e-05,
+      "loss": 0.9294,
+      "step": 35440
+    },
+    {
+      "epoch": 1.3645813282001926,
+      "grad_norm": 1.8753331899642944,
+      "learning_rate": 4.589036693468035e-05,
+      "loss": 0.9329,
+      "step": 35445
+    },
+    {
+      "epoch": 1.3647738209817133,
+      "grad_norm": 0.8658643364906311,
+      "learning_rate": 4.5864939091414495e-05,
+      "loss": 0.7045,
+      "step": 35450
+    },
+    {
+      "epoch": 1.364966313763234,
+      "grad_norm": 1.2768117189407349,
+      "learning_rate": 4.5839516198441304e-05,
+      "loss": 0.9337,
+      "step": 35455
+    },
+    {
+      "epoch": 1.3651588065447546,
+      "grad_norm": 1.145644187927246,
+      "learning_rate": 4.581409825808557e-05,
+      "loss": 0.8107,
+      "step": 35460
+    },
+    {
+      "epoch": 1.3653512993262753,
+      "grad_norm": 1.1863698959350586,
+      "learning_rate": 4.5788685272671605e-05,
+      "loss": 0.6387,
+      "step": 35465
+    },
+    {
+      "epoch": 1.365543792107796,
+      "grad_norm": 1.2812061309814453,
+      "learning_rate": 4.576327724452326e-05,
+      "loss": 0.756,
+      "step": 35470
+    },
+    {
+      "epoch": 1.3657362848893166,
+      "grad_norm": 1.6907789707183838,
+      "learning_rate": 4.5737874175963956e-05,
+      "loss": 0.9023,
+      "step": 35475
+    },
+    {
+      "epoch": 1.3659287776708373,
+      "grad_norm": 1.9304261207580566,
+      "learning_rate": 4.5712476069316576e-05,
+      "loss": 0.8313,
+      "step": 35480
+    },
+    {
+      "epoch": 1.366121270452358,
+      "grad_norm": 1.2450244426727295,
+      "learning_rate": 4.568708292690364e-05,
+      "loss": 0.8494,
+      "step": 35485
+    },
+    {
+      "epoch": 1.3663137632338787,
+      "grad_norm": 1.577697515487671,
+      "learning_rate": 4.566169475104717e-05,
+      "loss": 0.8129,
+      "step": 35490
+    },
+    {
+      "epoch": 1.3665062560153993,
+      "grad_norm": 2.367910146713257,
+      "learning_rate": 4.56363115440688e-05,
+      "loss": 0.8501,
+      "step": 35495
+    },
+    {
+      "epoch": 1.36669874879692,
+      "grad_norm": 1.588900089263916,
+      "learning_rate": 4.561093330828954e-05,
+      "loss": 0.9732,
+      "step": 35500
+    },
+    {
+      "epoch": 1.3668912415784409,
+      "grad_norm": 1.3308700323104858,
+      "learning_rate": 4.558556004603019e-05,
+      "loss": 0.8681,
+      "step": 35505
+    },
+    {
+      "epoch": 1.3670837343599616,
+      "grad_norm": 1.5678372383117676,
+      "learning_rate": 4.556019175961091e-05,
+      "loss": 0.8269,
+      "step": 35510
+    },
+    {
+      "epoch": 1.3672762271414822,
+      "grad_norm": 1.0886204242706299,
+      "learning_rate": 4.553482845135143e-05,
+      "loss": 0.9315,
+      "step": 35515
+    },
+    {
+      "epoch": 1.367468719923003,
+      "grad_norm": 1.2117904424667358,
+      "learning_rate": 4.5509470123571095e-05,
+      "loss": 0.8139,
+      "step": 35520
+    },
+    {
+      "epoch": 1.3676612127045236,
+      "grad_norm": 1.5415525436401367,
+      "learning_rate": 4.5484116778588807e-05,
+      "loss": 0.8667,
+      "step": 35525
+    },
+    {
+      "epoch": 1.3678537054860442,
+      "grad_norm": 1.6939245462417603,
+      "learning_rate": 4.545876841872281e-05,
+      "loss": 0.8011,
+      "step": 35530
+    },
+    {
+      "epoch": 1.368046198267565,
+      "grad_norm": 1.9691619873046875,
+      "learning_rate": 4.5433425046291224e-05,
+      "loss": 0.9255,
+      "step": 35535
+    },
+    {
+      "epoch": 1.3682386910490858,
+      "grad_norm": 2.5566859245300293,
+      "learning_rate": 4.54080866636114e-05,
+      "loss": 0.998,
+      "step": 35540
+    },
+    {
+      "epoch": 1.3684311838306065,
+      "grad_norm": 1.200829029083252,
+      "learning_rate": 4.538275327300042e-05,
+      "loss": 0.7522,
+      "step": 35545
+    },
+    {
+      "epoch": 1.3686236766121271,
+      "grad_norm": 1.0914881229400635,
+      "learning_rate": 4.535742487677485e-05,
+      "loss": 0.8378,
+      "step": 35550
+    },
+    {
+      "epoch": 1.3688161693936478,
+      "grad_norm": 1.1408578157424927,
+      "learning_rate": 4.5332101477250796e-05,
+      "loss": 0.8735,
+      "step": 35555
+    },
+    {
+      "epoch": 1.3690086621751685,
+      "grad_norm": 0.8551640510559082,
+      "learning_rate": 4.5306783076743955e-05,
+      "loss": 0.8452,
+      "step": 35560
+    },
+    {
+      "epoch": 1.3692011549566891,
+      "grad_norm": 2.295720100402832,
+      "learning_rate": 4.5281469677569456e-05,
+      "loss": 1.0684,
+      "step": 35565
+    },
+    {
+      "epoch": 1.3693936477382098,
+      "grad_norm": 1.2390626668930054,
+      "learning_rate": 4.5256161282042085e-05,
+      "loss": 0.7534,
+      "step": 35570
+    },
+    {
+      "epoch": 1.3695861405197305,
+      "grad_norm": 1.0765421390533447,
+      "learning_rate": 4.5230857892476106e-05,
+      "loss": 0.8554,
+      "step": 35575
+    },
+    {
+      "epoch": 1.3697786333012512,
+      "grad_norm": 1.3477513790130615,
+      "learning_rate": 4.5205559511185415e-05,
+      "loss": 0.8043,
+      "step": 35580
+    },
+    {
+      "epoch": 1.3699711260827718,
+      "grad_norm": 1.3972609043121338,
+      "learning_rate": 4.518026614048324e-05,
+      "loss": 0.7081,
+      "step": 35585
+    },
+    {
+      "epoch": 1.3701636188642925,
+      "grad_norm": 0.9186348915100098,
+      "learning_rate": 4.515497778268266e-05,
+      "loss": 0.6965,
+      "step": 35590
+    },
+    {
+      "epoch": 1.3703561116458132,
+      "grad_norm": 1.2708579301834106,
+      "learning_rate": 4.5129694440096005e-05,
+      "loss": 0.801,
+      "step": 35595
+    },
+    {
+      "epoch": 1.3705486044273338,
+      "grad_norm": 1.0227978229522705,
+      "learning_rate": 4.5104416115035306e-05,
+      "loss": 0.7764,
+      "step": 35600
+    },
+    {
+      "epoch": 1.3707410972088547,
+      "grad_norm": 1.2434272766113281,
+      "learning_rate": 4.507914280981211e-05,
+      "loss": 0.7463,
+      "step": 35605
+    },
+    {
+      "epoch": 1.3709335899903754,
+      "grad_norm": 1.1386263370513916,
+      "learning_rate": 4.505387452673753e-05,
+      "loss": 0.8806,
+      "step": 35610
+    },
+    {
+      "epoch": 1.371126082771896,
+      "grad_norm": 0.9424572587013245,
+      "learning_rate": 4.502861126812205e-05,
+      "loss": 0.9189,
+      "step": 35615
+    },
+    {
+      "epoch": 1.3713185755534167,
+      "grad_norm": 0.7779520153999329,
+      "learning_rate": 4.500335303627601e-05,
+      "loss": 0.8486,
+      "step": 35620
+    },
+    {
+      "epoch": 1.3715110683349374,
+      "grad_norm": 1.1644350290298462,
+      "learning_rate": 4.497809983350897e-05,
+      "loss": 0.7089,
+      "step": 35625
+    },
+    {
+      "epoch": 1.371703561116458,
+      "grad_norm": 1.2244153022766113,
+      "learning_rate": 4.4952851662130216e-05,
+      "loss": 0.9906,
+      "step": 35630
+    },
+    {
+      "epoch": 1.3718960538979788,
+      "grad_norm": 2.0696568489074707,
+      "learning_rate": 4.4927608524448515e-05,
+      "loss": 0.7632,
+      "step": 35635
+    },
+    {
+      "epoch": 1.3720885466794996,
+      "grad_norm": 1.5171512365341187,
+      "learning_rate": 4.4902370422772233e-05,
+      "loss": 0.9423,
+      "step": 35640
+    },
+    {
+      "epoch": 1.3722810394610203,
+      "grad_norm": 1.1163785457611084,
+      "learning_rate": 4.4877137359409116e-05,
+      "loss": 0.7411,
+      "step": 35645
+    },
+    {
+      "epoch": 1.372473532242541,
+      "grad_norm": 1.2916820049285889,
+      "learning_rate": 4.485190933666671e-05,
+      "loss": 0.8892,
+      "step": 35650
+    },
+    {
+      "epoch": 1.3726660250240617,
+      "grad_norm": 1.1933842897415161,
+      "learning_rate": 4.4826686356851834e-05,
+      "loss": 0.8643,
+      "step": 35655
+    },
+    {
+      "epoch": 1.3728585178055823,
+      "grad_norm": 1.0002435445785522,
+      "learning_rate": 4.4801468422271e-05,
+      "loss": 0.7632,
+      "step": 35660
+    },
+    {
+      "epoch": 1.373051010587103,
+      "grad_norm": 1.1150939464569092,
+      "learning_rate": 4.4776255535230216e-05,
+      "loss": 0.6594,
+      "step": 35665
+    },
+    {
+      "epoch": 1.3732435033686237,
+      "grad_norm": 1.2958896160125732,
+      "learning_rate": 4.4751047698035075e-05,
+      "loss": 0.8391,
+      "step": 35670
+    },
+    {
+      "epoch": 1.3734359961501443,
+      "grad_norm": 1.687971591949463,
+      "learning_rate": 4.47258449129906e-05,
+      "loss": 0.6359,
+      "step": 35675
+    },
+    {
+      "epoch": 1.373628488931665,
+      "grad_norm": 1.6192028522491455,
+      "learning_rate": 4.4700647182401456e-05,
+      "loss": 0.8449,
+      "step": 35680
+    },
+    {
+      "epoch": 1.3738209817131857,
+      "grad_norm": 1.6572599411010742,
+      "learning_rate": 4.467545450857179e-05,
+      "loss": 0.8733,
+      "step": 35685
+    },
+    {
+      "epoch": 1.3740134744947063,
+      "grad_norm": 1.175234317779541,
+      "learning_rate": 4.465026689380532e-05,
+      "loss": 0.8851,
+      "step": 35690
+    },
+    {
+      "epoch": 1.374205967276227,
+      "grad_norm": 0.9541772603988647,
+      "learning_rate": 4.4625084340405333e-05,
+      "loss": 0.852,
+      "step": 35695
+    },
+    {
+      "epoch": 1.374398460057748,
+      "grad_norm": 1.063652515411377,
+      "learning_rate": 4.4599906850674514e-05,
+      "loss": 0.8568,
+      "step": 35700
+    },
+    {
+      "epoch": 1.3745909528392686,
+      "grad_norm": 1.2114187479019165,
+      "learning_rate": 4.457473442691522e-05,
+      "loss": 0.731,
+      "step": 35705
+    },
+    {
+      "epoch": 1.3747834456207892,
+      "grad_norm": 1.2288764715194702,
+      "learning_rate": 4.454956707142931e-05,
+      "loss": 0.7752,
+      "step": 35710
+    },
+    {
+      "epoch": 1.37497593840231,
+      "grad_norm": 1.9316095113754272,
+      "learning_rate": 4.452440478651819e-05,
+      "loss": 0.7302,
+      "step": 35715
+    },
+    {
+      "epoch": 1.3751684311838306,
+      "grad_norm": 1.061178207397461,
+      "learning_rate": 4.449924757448269e-05,
+      "loss": 0.8386,
+      "step": 35720
+    },
+    {
+      "epoch": 1.3753609239653513,
+      "grad_norm": 1.0967941284179688,
+      "learning_rate": 4.447409543762342e-05,
+      "loss": 0.7719,
+      "step": 35725
+    },
+    {
+      "epoch": 1.375553416746872,
+      "grad_norm": 2.174457311630249,
+      "learning_rate": 4.4448948378240264e-05,
+      "loss": 0.9424,
+      "step": 35730
+    },
+    {
+      "epoch": 1.3757459095283928,
+      "grad_norm": 1.3424992561340332,
+      "learning_rate": 4.442380639863277e-05,
+      "loss": 0.7166,
+      "step": 35735
+    },
+    {
+      "epoch": 1.3759384023099135,
+      "grad_norm": 1.0383292436599731,
+      "learning_rate": 4.4398669501100044e-05,
+      "loss": 0.8558,
+      "step": 35740
+    },
+    {
+      "epoch": 1.3761308950914342,
+      "grad_norm": 0.9943356513977051,
+      "learning_rate": 4.437353768794069e-05,
+      "loss": 0.6549,
+      "step": 35745
+    },
+    {
+      "epoch": 1.3763233878729548,
+      "grad_norm": 2.162888526916504,
+      "learning_rate": 4.4348410961452744e-05,
+      "loss": 0.8268,
+      "step": 35750
+    },
+    {
+      "epoch": 1.3765158806544755,
+      "grad_norm": 0.9995039701461792,
+      "learning_rate": 4.432328932393405e-05,
+      "loss": 0.8245,
+      "step": 35755
+    },
+    {
+      "epoch": 1.3767083734359962,
+      "grad_norm": 1.1602789163589478,
+      "learning_rate": 4.429817277768167e-05,
+      "loss": 0.8199,
+      "step": 35760
+    },
+    {
+      "epoch": 1.3769008662175168,
+      "grad_norm": 1.8660534620285034,
+      "learning_rate": 4.42730613249924e-05,
+      "loss": 0.9195,
+      "step": 35765
+    },
+    {
+      "epoch": 1.3770933589990375,
+      "grad_norm": 1.5461866855621338,
+      "learning_rate": 4.424795496816252e-05,
+      "loss": 0.801,
+      "step": 35770
+    },
+    {
+      "epoch": 1.3772858517805582,
+      "grad_norm": 1.3390212059020996,
+      "learning_rate": 4.4222853709487866e-05,
+      "loss": 0.8652,
+      "step": 35775
+    },
+    {
+      "epoch": 1.3774783445620788,
+      "grad_norm": 1.2544982433319092,
+      "learning_rate": 4.419775755126372e-05,
+      "loss": 0.7289,
+      "step": 35780
+    },
+    {
+      "epoch": 1.3776708373435995,
+      "grad_norm": 1.8691519498825073,
+      "learning_rate": 4.4172666495784984e-05,
+      "loss": 0.8905,
+      "step": 35785
+    },
+    {
+      "epoch": 1.3778633301251202,
+      "grad_norm": 2.0388388633728027,
+      "learning_rate": 4.414758054534608e-05,
+      "loss": 1.0152,
+      "step": 35790
+    },
+    {
+      "epoch": 1.378055822906641,
+      "grad_norm": 1.040433645248413,
+      "learning_rate": 4.4122499702240946e-05,
+      "loss": 0.8018,
+      "step": 35795
+    },
+    {
+      "epoch": 1.3782483156881618,
+      "grad_norm": 1.2068662643432617,
+      "learning_rate": 4.409742396876309e-05,
+      "loss": 0.6547,
+      "step": 35800
+    },
+    {
+      "epoch": 1.3784408084696824,
+      "grad_norm": 1.445438265800476,
+      "learning_rate": 4.4072353347205466e-05,
+      "loss": 0.8732,
+      "step": 35805
+    },
+    {
+      "epoch": 1.378633301251203,
+      "grad_norm": 0.9814010858535767,
+      "learning_rate": 4.404728783986063e-05,
+      "loss": 0.8397,
+      "step": 35810
+    },
+    {
+      "epoch": 1.3788257940327238,
+      "grad_norm": 2.0966150760650635,
+      "learning_rate": 4.4022227449020684e-05,
+      "loss": 0.936,
+      "step": 35815
+    },
+    {
+      "epoch": 1.3790182868142444,
+      "grad_norm": 1.3904740810394287,
+      "learning_rate": 4.39971721769772e-05,
+      "loss": 0.9165,
+      "step": 35820
+    },
+    {
+      "epoch": 1.379210779595765,
+      "grad_norm": 1.3555712699890137,
+      "learning_rate": 4.3972122026021346e-05,
+      "loss": 1.0844,
+      "step": 35825
+    },
+    {
+      "epoch": 1.3794032723772858,
+      "grad_norm": 1.2570695877075195,
+      "learning_rate": 4.394707699844381e-05,
+      "loss": 0.9331,
+      "step": 35830
+    },
+    {
+      "epoch": 1.3795957651588067,
+      "grad_norm": 0.8406626582145691,
+      "learning_rate": 4.392203709653471e-05,
+      "loss": 0.8221,
+      "step": 35835
+    },
+    {
+      "epoch": 1.3797882579403273,
+      "grad_norm": 1.3139702081680298,
+      "learning_rate": 4.3897002322583894e-05,
+      "loss": 0.9063,
+      "step": 35840
+    },
+    {
+      "epoch": 1.379980750721848,
+      "grad_norm": 0.7752715945243835,
+      "learning_rate": 4.3871972678880535e-05,
+      "loss": 0.7454,
+      "step": 35845
+    },
+    {
+      "epoch": 1.3801732435033687,
+      "grad_norm": 1.247069239616394,
+      "learning_rate": 4.384694816771345e-05,
+      "loss": 0.7338,
+      "step": 35850
+    },
+    {
+      "epoch": 1.3803657362848893,
+      "grad_norm": 1.1150643825531006,
+      "learning_rate": 4.3821928791370995e-05,
+      "loss": 0.9423,
+      "step": 35855
+    },
+    {
+      "epoch": 1.38055822906641,
+      "grad_norm": 1.191602349281311,
+      "learning_rate": 4.3796914552141035e-05,
+      "loss": 0.7278,
+      "step": 35860
+    },
+    {
+      "epoch": 1.3807507218479307,
+      "grad_norm": 1.868516206741333,
+      "learning_rate": 4.3771905452310844e-05,
+      "loss": 0.8649,
+      "step": 35865
+    },
+    {
+      "epoch": 1.3809432146294514,
+      "grad_norm": 1.443765640258789,
+      "learning_rate": 4.37469014941675e-05,
+      "loss": 0.8188,
+      "step": 35870
+    },
+    {
+      "epoch": 1.381135707410972,
+      "grad_norm": 1.2555128335952759,
+      "learning_rate": 4.372190267999734e-05,
+      "loss": 0.8372,
+      "step": 35875
+    },
+    {
+      "epoch": 1.3813282001924927,
+      "grad_norm": 1.0106040239334106,
+      "learning_rate": 4.369690901208637e-05,
+      "loss": 0.769,
+      "step": 35880
+    },
+    {
+      "epoch": 1.3815206929740134,
+      "grad_norm": 0.8901906609535217,
+      "learning_rate": 4.3671920492720095e-05,
+      "loss": 0.8127,
+      "step": 35885
+    },
+    {
+      "epoch": 1.381713185755534,
+      "grad_norm": 1.0237152576446533,
+      "learning_rate": 4.3646937124183594e-05,
+      "loss": 0.8046,
+      "step": 35890
+    },
+    {
+      "epoch": 1.381905678537055,
+      "grad_norm": 2.2792160511016846,
+      "learning_rate": 4.362195890876135e-05,
+      "loss": 0.8081,
+      "step": 35895
+    },
+    {
+      "epoch": 1.3820981713185756,
+      "grad_norm": 1.6603785753250122,
+      "learning_rate": 4.359698584873749e-05,
+      "loss": 0.8406,
+      "step": 35900
+    },
+    {
+      "epoch": 1.3822906641000963,
+      "grad_norm": 0.9667415618896484,
+      "learning_rate": 4.357201794639568e-05,
+      "loss": 0.7888,
+      "step": 35905
+    },
+    {
+      "epoch": 1.382483156881617,
+      "grad_norm": 1.3200799226760864,
+      "learning_rate": 4.354705520401895e-05,
+      "loss": 0.7056,
+      "step": 35910
+    },
+    {
+      "epoch": 1.3826756496631376,
+      "grad_norm": 1.4623115062713623,
+      "learning_rate": 4.352209762389013e-05,
+      "loss": 0.8487,
+      "step": 35915
+    },
+    {
+      "epoch": 1.3828681424446583,
+      "grad_norm": 1.7032670974731445,
+      "learning_rate": 4.3497145208291314e-05,
+      "loss": 0.9328,
+      "step": 35920
+    },
+    {
+      "epoch": 1.383060635226179,
+      "grad_norm": 0.9201510548591614,
+      "learning_rate": 4.347219795950427e-05,
+      "loss": 0.8634,
+      "step": 35925
+    },
+    {
+      "epoch": 1.3832531280076998,
+      "grad_norm": 1.2735239267349243,
+      "learning_rate": 4.3447255879810266e-05,
+      "loss": 0.7324,
+      "step": 35930
+    },
+    {
+      "epoch": 1.3834456207892205,
+      "grad_norm": 1.2430258989334106,
+      "learning_rate": 4.3422318971490116e-05,
+      "loss": 0.8445,
+      "step": 35935
+    },
+    {
+      "epoch": 1.3836381135707412,
+      "grad_norm": 0.880877673625946,
+      "learning_rate": 4.3397387236824025e-05,
+      "loss": 0.7624,
+      "step": 35940
+    },
+    {
+      "epoch": 1.3838306063522618,
+      "grad_norm": 0.9846094250679016,
+      "learning_rate": 4.3372460678091984e-05,
+      "loss": 0.7539,
+      "step": 35945
+    },
+    {
+      "epoch": 1.3840230991337825,
+      "grad_norm": 2.2208034992218018,
+      "learning_rate": 4.334753929757327e-05,
+      "loss": 0.7707,
+      "step": 35950
+    },
+    {
+      "epoch": 1.3842155919153032,
+      "grad_norm": 1.152341365814209,
+      "learning_rate": 4.332262309754679e-05,
+      "loss": 0.86,
+      "step": 35955
+    },
+    {
+      "epoch": 1.3844080846968239,
+      "grad_norm": 1.2790590524673462,
+      "learning_rate": 4.3297712080290975e-05,
+      "loss": 0.7621,
+      "step": 35960
+    },
+    {
+      "epoch": 1.3846005774783445,
+      "grad_norm": 1.3545982837677002,
+      "learning_rate": 4.327280624808381e-05,
+      "loss": 0.7659,
+      "step": 35965
+    },
+    {
+      "epoch": 1.3847930702598652,
+      "grad_norm": 1.7217894792556763,
+      "learning_rate": 4.324790560320265e-05,
+      "loss": 0.8812,
+      "step": 35970
+    },
+    {
+      "epoch": 1.3849855630413859,
+      "grad_norm": 1.3170634508132935,
+      "learning_rate": 4.322798882370276e-05,
+      "loss": 0.9111,
+      "step": 35975
+    },
+    {
+      "epoch": 1.3851780558229065,
+      "grad_norm": 1.05010986328125,
+      "learning_rate": 4.320309752174627e-05,
+      "loss": 0.8722,
+      "step": 35980
+    },
+    {
+      "epoch": 1.3853705486044272,
+      "grad_norm": 1.8216160535812378,
+      "learning_rate": 4.317821141349036e-05,
+      "loss": 0.9751,
+      "step": 35985
+    },
+    {
+      "epoch": 1.385563041385948,
+      "grad_norm": 1.5039763450622559,
+      "learning_rate": 4.315333050121055e-05,
+      "loss": 0.9311,
+      "step": 35990
+    },
+    {
+      "epoch": 1.3857555341674688,
+      "grad_norm": 1.1297041177749634,
+      "learning_rate": 4.312845478718211e-05,
+      "loss": 0.8074,
+      "step": 35995
+    },
+    {
+      "epoch": 1.3859480269489894,
+      "grad_norm": 1.2953757047653198,
+      "learning_rate": 4.310358427367972e-05,
+      "loss": 0.7828,
+      "step": 36000
+    },
+    {
+      "epoch": 1.3861405197305101,
+      "grad_norm": 1.7586846351623535,
+      "learning_rate": 4.3078718962977684e-05,
+      "loss": 0.9055,
+      "step": 36005
+    },
+    {
+      "epoch": 1.3863330125120308,
+      "grad_norm": 1.59012770652771,
+      "learning_rate": 4.305385885734966e-05,
+      "loss": 0.7722,
+      "step": 36010
+    },
+    {
+      "epoch": 1.3865255052935515,
+      "grad_norm": 0.7984145879745483,
+      "learning_rate": 4.302900395906909e-05,
+      "loss": 0.803,
+      "step": 36015
+    },
+    {
+      "epoch": 1.3867179980750721,
+      "grad_norm": 2.04715895652771,
+      "learning_rate": 4.3004154270408666e-05,
+      "loss": 0.9151,
+      "step": 36020
+    },
+    {
+      "epoch": 1.386910490856593,
+      "grad_norm": 1.13097083568573,
+      "learning_rate": 4.2979309793640776e-05,
+      "loss": 0.7337,
+      "step": 36025
+    },
+    {
+      "epoch": 1.3871029836381137,
+      "grad_norm": 1.4123458862304688,
+      "learning_rate": 4.295447053103727e-05,
+      "loss": 0.7956,
+      "step": 36030
+    },
+    {
+      "epoch": 1.3872954764196344,
+      "grad_norm": 1.3871421813964844,
+      "learning_rate": 4.2929636484869586e-05,
+      "loss": 0.8623,
+      "step": 36035
+    },
+    {
+      "epoch": 1.387487969201155,
+      "grad_norm": 0.9964479804039001,
+      "learning_rate": 4.290480765740851e-05,
+      "loss": 0.7375,
+      "step": 36040
+    },
+    {
+      "epoch": 1.3876804619826757,
+      "grad_norm": 2.126817464828491,
+      "learning_rate": 4.287998405092463e-05,
+      "loss": 0.9251,
+      "step": 36045
+    },
+    {
+      "epoch": 1.3878729547641964,
+      "grad_norm": 1.7415242195129395,
+      "learning_rate": 4.285516566768779e-05,
+      "loss": 0.88,
+      "step": 36050
+    },
+    {
+      "epoch": 1.388065447545717,
+      "grad_norm": 1.341607689857483,
+      "learning_rate": 4.2830352509967486e-05,
+      "loss": 0.8688,
+      "step": 36055
+    },
+    {
+      "epoch": 1.3882579403272377,
+      "grad_norm": 2.5146100521087646,
+      "learning_rate": 4.280554458003274e-05,
+      "loss": 0.7789,
+      "step": 36060
+    },
+    {
+      "epoch": 1.3884504331087584,
+      "grad_norm": 1.7495235204696655,
+      "learning_rate": 4.2780741880152106e-05,
+      "loss": 0.749,
+      "step": 36065
+    },
+    {
+      "epoch": 1.388642925890279,
+      "grad_norm": 1.4821887016296387,
+      "learning_rate": 4.275594441259354e-05,
+      "loss": 0.8738,
+      "step": 36070
+    },
+    {
+      "epoch": 1.3888354186717997,
+      "grad_norm": 1.3802404403686523,
+      "learning_rate": 4.273115217962466e-05,
+      "loss": 0.8245,
+      "step": 36075
+    },
+    {
+      "epoch": 1.3890279114533204,
+      "grad_norm": 1.0245096683502197,
+      "learning_rate": 4.270636518351252e-05,
+      "loss": 0.8295,
+      "step": 36080
+    },
+    {
+      "epoch": 1.389220404234841,
+      "grad_norm": 1.633852243423462,
+      "learning_rate": 4.268158342652376e-05,
+      "loss": 0.8341,
+      "step": 36085
+    },
+    {
+      "epoch": 1.389412897016362,
+      "grad_norm": 1.1715857982635498,
+      "learning_rate": 4.265680691092454e-05,
+      "loss": 0.8579,
+      "step": 36090
+    },
+    {
+      "epoch": 1.3896053897978826,
+      "grad_norm": 1.3279740810394287,
+      "learning_rate": 4.263203563898038e-05,
+      "loss": 0.8271,
+      "step": 36095
+    },
+    {
+      "epoch": 1.3897978825794033,
+      "grad_norm": 0.87353515625,
+      "learning_rate": 4.2607269612956615e-05,
+      "loss": 0.7767,
+      "step": 36100
+    },
+    {
+      "epoch": 1.389990375360924,
+      "grad_norm": 0.9114276170730591,
+      "learning_rate": 4.258250883511782e-05,
+      "loss": 0.6756,
+      "step": 36105
+    },
+    {
+      "epoch": 1.3901828681424446,
+      "grad_norm": 1.5339330434799194,
+      "learning_rate": 4.255775330772822e-05,
+      "loss": 0.6985,
+      "step": 36110
+    },
+    {
+      "epoch": 1.3903753609239653,
+      "grad_norm": 1.8716709613800049,
+      "learning_rate": 4.253300303305157e-05,
+      "loss": 0.8967,
+      "step": 36115
+    },
+    {
+      "epoch": 1.390567853705486,
+      "grad_norm": 1.4840834140777588,
+      "learning_rate": 4.250825801335114e-05,
+      "loss": 0.7657,
+      "step": 36120
+    },
+    {
+      "epoch": 1.3907603464870069,
+      "grad_norm": 1.0822495222091675,
+      "learning_rate": 4.24835182508896e-05,
+      "loss": 0.7959,
+      "step": 36125
+    },
+    {
+      "epoch": 1.3909528392685275,
+      "grad_norm": 1.4410738945007324,
+      "learning_rate": 4.2458783747929375e-05,
+      "loss": 0.8108,
+      "step": 36130
+    },
+    {
+      "epoch": 1.3911453320500482,
+      "grad_norm": 1.0785428285598755,
+      "learning_rate": 4.2434054506732157e-05,
+      "loss": 0.8385,
+      "step": 36135
+    },
+    {
+      "epoch": 1.3913378248315689,
+      "grad_norm": 1.1617215871810913,
+      "learning_rate": 4.240933052955932e-05,
+      "loss": 0.8655,
+      "step": 36140
+    },
+    {
+      "epoch": 1.3915303176130895,
+      "grad_norm": 1.7867512702941895,
+      "learning_rate": 4.238461181867171e-05,
+      "loss": 0.9933,
+      "step": 36145
+    },
+    {
+      "epoch": 1.3917228103946102,
+      "grad_norm": 1.6082379817962646,
+      "learning_rate": 4.235989837632971e-05,
+      "loss": 0.7886,
+      "step": 36150
+    },
+    {
+      "epoch": 1.3919153031761309,
+      "grad_norm": 1.7708418369293213,
+      "learning_rate": 4.23351902047931e-05,
+      "loss": 0.9419,
+      "step": 36155
+    },
+    {
+      "epoch": 1.3921077959576516,
+      "grad_norm": 2.8401410579681396,
+      "learning_rate": 4.231048730632142e-05,
+      "loss": 0.9881,
+      "step": 36160
+    },
+    {
+      "epoch": 1.3923002887391722,
+      "grad_norm": 2.1983916759490967,
+      "learning_rate": 4.228578968317349e-05,
+      "loss": 0.8503,
+      "step": 36165
+    },
+    {
+      "epoch": 1.392492781520693,
+      "grad_norm": 1.9838260412216187,
+      "learning_rate": 4.226109733760777e-05,
+      "loss": 0.9066,
+      "step": 36170
+    },
+    {
+      "epoch": 1.3926852743022136,
+      "grad_norm": 1.032500147819519,
+      "learning_rate": 4.223641027188226e-05,
+      "loss": 0.7624,
+      "step": 36175
+    },
+    {
+      "epoch": 1.3928777670837342,
+      "grad_norm": 2.5539069175720215,
+      "learning_rate": 4.221172848825432e-05,
+      "loss": 0.971,
+      "step": 36180
+    },
+    {
+      "epoch": 1.3930702598652551,
+      "grad_norm": 1.1746537685394287,
+      "learning_rate": 4.218705198898102e-05,
+      "loss": 0.8605,
+      "step": 36185
+    },
+    {
+      "epoch": 1.3932627526467758,
+      "grad_norm": 1.7174090147018433,
+      "learning_rate": 4.216238077631882e-05,
+      "loss": 0.8074,
+      "step": 36190
+    },
+    {
+      "epoch": 1.3934552454282965,
+      "grad_norm": 1.3414026498794556,
+      "learning_rate": 4.2137714852523814e-05,
+      "loss": 0.8017,
+      "step": 36195
+    },
+    {
+      "epoch": 1.3936477382098171,
+      "grad_norm": 1.387629747390747,
+      "learning_rate": 4.21130542198514e-05,
+      "loss": 0.8292,
+      "step": 36200
+    },
+    {
+      "epoch": 1.3938402309913378,
+      "grad_norm": 1.5130243301391602,
+      "learning_rate": 4.2088398880556786e-05,
+      "loss": 0.8602,
+      "step": 36205
+    },
+    {
+      "epoch": 1.3940327237728585,
+      "grad_norm": 1.6352779865264893,
+      "learning_rate": 4.206374883689444e-05,
+      "loss": 0.7357,
+      "step": 36210
+    },
+    {
+      "epoch": 1.3942252165543791,
+      "grad_norm": 1.0258371829986572,
+      "learning_rate": 4.203910409111845e-05,
+      "loss": 0.9261,
+      "step": 36215
+    },
+    {
+      "epoch": 1.3944177093359,
+      "grad_norm": 1.0647536516189575,
+      "learning_rate": 4.2014464645482444e-05,
+      "loss": 0.7186,
+      "step": 36220
+    },
+    {
+      "epoch": 1.3946102021174207,
+      "grad_norm": 1.0326642990112305,
+      "learning_rate": 4.198983050223957e-05,
+      "loss": 0.7769,
+      "step": 36225
+    },
+    {
+      "epoch": 1.3948026948989414,
+      "grad_norm": 2.0613622665405273,
+      "learning_rate": 4.196520166364233e-05,
+      "loss": 0.8248,
+      "step": 36230
+    },
+    {
+      "epoch": 1.394995187680462,
+      "grad_norm": 2.12548565864563,
+      "learning_rate": 4.194057813194303e-05,
+      "loss": 0.872,
+      "step": 36235
+    },
+    {
+      "epoch": 1.3951876804619827,
+      "grad_norm": 2.1166744232177734,
+      "learning_rate": 4.191595990939322e-05,
+      "loss": 0.8744,
+      "step": 36240
+    },
+    {
+      "epoch": 1.3953801732435034,
+      "grad_norm": 1.4747822284698486,
+      "learning_rate": 4.18913469982441e-05,
+      "loss": 0.8514,
+      "step": 36245
+    },
+    {
+      "epoch": 1.395572666025024,
+      "grad_norm": 1.3667523860931396,
+      "learning_rate": 4.186673940074637e-05,
+      "loss": 0.9452,
+      "step": 36250
+    },
+    {
+      "epoch": 1.3957651588065447,
+      "grad_norm": 1.1476492881774902,
+      "learning_rate": 4.1842137119150246e-05,
+      "loss": 0.7274,
+      "step": 36255
+    },
+    {
+      "epoch": 1.3959576515880654,
+      "grad_norm": 2.258908987045288,
+      "learning_rate": 4.181754015570536e-05,
+      "loss": 1.1584,
+      "step": 36260
+    },
+    {
+      "epoch": 1.396150144369586,
+      "grad_norm": 1.5475572347640991,
+      "learning_rate": 4.179294851266107e-05,
+      "loss": 0.8509,
+      "step": 36265
+    },
+    {
+      "epoch": 1.3963426371511067,
+      "grad_norm": 0.9235596656799316,
+      "learning_rate": 4.176836219226602e-05,
+      "loss": 0.6882,
+      "step": 36270
+    },
+    {
+      "epoch": 1.3965351299326274,
+      "grad_norm": 2.50726580619812,
+      "learning_rate": 4.17437811967685e-05,
+      "loss": 0.9346,
+      "step": 36275
+    },
+    {
+      "epoch": 1.3967276227141483,
+      "grad_norm": 2.3889918327331543,
+      "learning_rate": 4.171920552841627e-05,
+      "loss": 0.7456,
+      "step": 36280
+    },
+    {
+      "epoch": 1.396920115495669,
+      "grad_norm": 1.233176350593567,
+      "learning_rate": 4.169463518945667e-05,
+      "loss": 0.8134,
+      "step": 36285
+    },
+    {
+      "epoch": 1.3971126082771896,
+      "grad_norm": 1.5915166139602661,
+      "learning_rate": 4.167007018213639e-05,
+      "loss": 0.9035,
+      "step": 36290
+    },
+    {
+      "epoch": 1.3973051010587103,
+      "grad_norm": 1.1252813339233398,
+      "learning_rate": 4.1645510508701805e-05,
+      "loss": 0.7778,
+      "step": 36295
+    },
+    {
+      "epoch": 1.397497593840231,
+      "grad_norm": 1.1730810403823853,
+      "learning_rate": 4.162095617139871e-05,
+      "loss": 0.8332,
+      "step": 36300
+    },
+    {
+      "epoch": 1.3976900866217516,
+      "grad_norm": 1.0456668138504028,
+      "learning_rate": 4.159640717247245e-05,
+      "loss": 0.8277,
+      "step": 36305
+    },
+    {
+      "epoch": 1.3978825794032723,
+      "grad_norm": 1.0216035842895508,
+      "learning_rate": 4.157186351416791e-05,
+      "loss": 0.8409,
+      "step": 36310
+    },
+    {
+      "epoch": 1.398075072184793,
+      "grad_norm": 2.1484153270721436,
+      "learning_rate": 4.154732519872936e-05,
+      "loss": 0.9005,
+      "step": 36315
+    },
+    {
+      "epoch": 1.3982675649663139,
+      "grad_norm": 0.9150620698928833,
+      "learning_rate": 4.15227922284007e-05,
+      "loss": 0.8182,
+      "step": 36320
+    },
+    {
+      "epoch": 1.3984600577478346,
+      "grad_norm": 2.366243362426758,
+      "learning_rate": 4.149826460542532e-05,
+      "loss": 0.8423,
+      "step": 36325
+    },
+    {
+      "epoch": 1.3986525505293552,
+      "grad_norm": 1.0907059907913208,
+      "learning_rate": 4.147374233204611e-05,
+      "loss": 0.8498,
+      "step": 36330
+    },
+    {
+      "epoch": 1.398845043310876,
+      "grad_norm": 1.077556848526001,
+      "learning_rate": 4.1449225410505456e-05,
+      "loss": 0.8998,
+      "step": 36335
+    },
+    {
+      "epoch": 1.3990375360923966,
+      "grad_norm": 1.197625756263733,
+      "learning_rate": 4.1424713843045305e-05,
+      "loss": 0.7545,
+      "step": 36340
+    },
+    {
+      "epoch": 1.3992300288739172,
+      "grad_norm": 1.4147288799285889,
+      "learning_rate": 4.1400207631906985e-05,
+      "loss": 0.8045,
+      "step": 36345
+    },
+    {
+      "epoch": 1.399422521655438,
+      "grad_norm": 2.3983185291290283,
+      "learning_rate": 4.1375706779331556e-05,
+      "loss": 1.0072,
+      "step": 36350
+    },
+    {
+      "epoch": 1.3996150144369586,
+      "grad_norm": 1.1587196588516235,
+      "learning_rate": 4.1351211287559366e-05,
+      "loss": 0.7277,
+      "step": 36355
+    },
+    {
+      "epoch": 1.3998075072184792,
+      "grad_norm": 1.1984206438064575,
+      "learning_rate": 4.132672115883037e-05,
+      "loss": 0.8874,
+      "step": 36360
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4345682859420776,
+      "learning_rate": 4.130223639538406e-05,
+      "loss": 0.9031,
+      "step": 36365
+    },
+    {
+      "epoch": 1.4001924927815206,
+      "grad_norm": 1.10869300365448,
+      "learning_rate": 4.127775699945944e-05,
+      "loss": 0.7777,
+      "step": 36370
+    },
+    {
+      "epoch": 1.4003849855630413,
+      "grad_norm": 0.9464773535728455,
+      "learning_rate": 4.12532829732949e-05,
+      "loss": 0.7748,
+      "step": 36375
+    },
+    {
+      "epoch": 1.4005774783445621,
+      "grad_norm": 1.2995601892471313,
+      "learning_rate": 4.122881431912847e-05,
+      "loss": 0.9252,
+      "step": 36380
+    },
+    {
+      "epoch": 1.4007699711260828,
+      "grad_norm": 1.4381591081619263,
+      "learning_rate": 4.120435103919765e-05,
+      "loss": 0.7513,
+      "step": 36385
+    },
+    {
+      "epoch": 1.4009624639076035,
+      "grad_norm": 1.7021411657333374,
+      "learning_rate": 4.117989313573943e-05,
+      "loss": 0.9105,
+      "step": 36390
+    },
+    {
+      "epoch": 1.4011549566891242,
+      "grad_norm": 1.6943551301956177,
+      "learning_rate": 4.115544061099039e-05,
+      "loss": 0.7316,
+      "step": 36395
+    },
+    {
+      "epoch": 1.4013474494706448,
+      "grad_norm": 1.5277372598648071,
+      "learning_rate": 4.1130993467186454e-05,
+      "loss": 0.8301,
+      "step": 36400
+    },
+    {
+      "epoch": 1.4015399422521655,
+      "grad_norm": 1.3319824934005737,
+      "learning_rate": 4.110655170656319e-05,
+      "loss": 0.78,
+      "step": 36405
+    },
+    {
+      "epoch": 1.4017324350336862,
+      "grad_norm": 1.3265659809112549,
+      "learning_rate": 4.108211533135563e-05,
+      "loss": 0.8302,
+      "step": 36410
+    },
+    {
+      "epoch": 1.401924927815207,
+      "grad_norm": 1.4805601835250854,
+      "learning_rate": 4.1057684343798376e-05,
+      "loss": 0.8165,
+      "step": 36415
+    },
+    {
+      "epoch": 1.4021174205967277,
+      "grad_norm": 1.1178621053695679,
+      "learning_rate": 4.103325874612536e-05,
+      "loss": 0.7693,
+      "step": 36420
+    },
+    {
+      "epoch": 1.4023099133782484,
+      "grad_norm": 2.39166522026062,
+      "learning_rate": 4.1008838540570284e-05,
+      "loss": 0.8679,
+      "step": 36425
+    },
+    {
+      "epoch": 1.402502406159769,
+      "grad_norm": 1.941292405128479,
+      "learning_rate": 4.098442372936611e-05,
+      "loss": 0.8451,
+      "step": 36430
+    },
+    {
+      "epoch": 1.4026948989412897,
+      "grad_norm": 0.9209627509117126,
+      "learning_rate": 4.0960014314745435e-05,
+      "loss": 0.8751,
+      "step": 36435
+    },
+    {
+      "epoch": 1.4028873917228104,
+      "grad_norm": 1.5222151279449463,
+      "learning_rate": 4.093561029894034e-05,
+      "loss": 0.8304,
+      "step": 36440
+    },
+    {
+      "epoch": 1.403079884504331,
+      "grad_norm": 1.1667454242706299,
+      "learning_rate": 4.091121168418247e-05,
+      "loss": 0.8462,
+      "step": 36445
+    },
+    {
+      "epoch": 1.4032723772858517,
+      "grad_norm": 1.9360030889511108,
+      "learning_rate": 4.088681847270278e-05,
+      "loss": 0.9054,
+      "step": 36450
+    },
+    {
+      "epoch": 1.4034648700673724,
+      "grad_norm": 0.9800208806991577,
+      "learning_rate": 4.086243066673203e-05,
+      "loss": 0.8068,
+      "step": 36455
+    },
+    {
+      "epoch": 1.403657362848893,
+      "grad_norm": 1.7991399765014648,
+      "learning_rate": 4.083804826850022e-05,
+      "loss": 0.6918,
+      "step": 36460
+    },
+    {
+      "epoch": 1.4038498556304138,
+      "grad_norm": 1.230858564376831,
+      "learning_rate": 4.081367128023697e-05,
+      "loss": 0.7422,
+      "step": 36465
+    },
+    {
+      "epoch": 1.4040423484119344,
+      "grad_norm": 1.1397539377212524,
+      "learning_rate": 4.0789299704171405e-05,
+      "loss": 0.8296,
+      "step": 36470
+    },
+    {
+      "epoch": 1.4042348411934553,
+      "grad_norm": 0.9111887812614441,
+      "learning_rate": 4.076493354253219e-05,
+      "loss": 0.7259,
+      "step": 36475
+    },
+    {
+      "epoch": 1.404427333974976,
+      "grad_norm": 1.0511916875839233,
+      "learning_rate": 4.074057279754733e-05,
+      "loss": 0.7495,
+      "step": 36480
+    },
+    {
+      "epoch": 1.4046198267564967,
+      "grad_norm": 1.596725344657898,
+      "learning_rate": 4.071621747144461e-05,
+      "loss": 0.8516,
+      "step": 36485
+    },
+    {
+      "epoch": 1.4048123195380173,
+      "grad_norm": 1.4566305875778198,
+      "learning_rate": 4.069186756645104e-05,
+      "loss": 0.9033,
+      "step": 36490
+    },
+    {
+      "epoch": 1.405004812319538,
+      "grad_norm": 1.7779717445373535,
+      "learning_rate": 4.06675230847933e-05,
+      "loss": 0.9614,
+      "step": 36495
+    },
+    {
+      "epoch": 1.4051973051010587,
+      "grad_norm": 1.1408967971801758,
+      "learning_rate": 4.064318402869753e-05,
+      "loss": 0.8547,
+      "step": 36500
+    },
+    {
+      "epoch": 1.4053897978825793,
+      "grad_norm": 0.9641178846359253,
+      "learning_rate": 4.061885040038942e-05,
+      "loss": 0.7535,
+      "step": 36505
+    },
+    {
+      "epoch": 1.4055822906641002,
+      "grad_norm": 1.1024551391601562,
+      "learning_rate": 4.059452220209403e-05,
+      "loss": 0.7627,
+      "step": 36510
+    },
+    {
+      "epoch": 1.405774783445621,
+      "grad_norm": 1.3117225170135498,
+      "learning_rate": 4.057019943603607e-05,
+      "loss": 0.96,
+      "step": 36515
+    },
+    {
+      "epoch": 1.4059672762271416,
+      "grad_norm": 1.1285604238510132,
+      "learning_rate": 4.054588210443969e-05,
+      "loss": 0.8481,
+      "step": 36520
+    },
+    {
+      "epoch": 1.4061597690086622,
+      "grad_norm": 1.3916871547698975,
+      "learning_rate": 4.0521570209528534e-05,
+      "loss": 0.8622,
+      "step": 36525
+    },
+    {
+      "epoch": 1.406352261790183,
+      "grad_norm": 1.8372492790222168,
+      "learning_rate": 4.049726375352582e-05,
+      "loss": 0.8732,
+      "step": 36530
+    },
+    {
+      "epoch": 1.4065447545717036,
+      "grad_norm": 1.3047436475753784,
+      "learning_rate": 4.047296273865412e-05,
+      "loss": 0.8297,
+      "step": 36535
+    },
+    {
+      "epoch": 1.4067372473532243,
+      "grad_norm": 1.984266757965088,
+      "learning_rate": 4.044866716713565e-05,
+      "loss": 0.7249,
+      "step": 36540
+    },
+    {
+      "epoch": 1.406929740134745,
+      "grad_norm": 1.0713486671447754,
+      "learning_rate": 4.0424377041192075e-05,
+      "loss": 0.7389,
+      "step": 36545
+    },
+    {
+      "epoch": 1.4071222329162656,
+      "grad_norm": 1.0382349491119385,
+      "learning_rate": 4.0400092363044594e-05,
+      "loss": 0.7965,
+      "step": 36550
+    },
+    {
+      "epoch": 1.4073147256977863,
+      "grad_norm": 2.4277896881103516,
+      "learning_rate": 4.03758131349138e-05,
+      "loss": 0.7452,
+      "step": 36555
+    },
+    {
+      "epoch": 1.407507218479307,
+      "grad_norm": 1.7688560485839844,
+      "learning_rate": 4.0351539359019985e-05,
+      "loss": 0.8284,
+      "step": 36560
+    },
+    {
+      "epoch": 1.4076997112608276,
+      "grad_norm": 1.1814532279968262,
+      "learning_rate": 4.0327271037582726e-05,
+      "loss": 0.872,
+      "step": 36565
+    },
+    {
+      "epoch": 1.4078922040423485,
+      "grad_norm": 2.220003128051758,
+      "learning_rate": 4.0303008172821235e-05,
+      "loss": 0.8533,
+      "step": 36570
+    },
+    {
+      "epoch": 1.4080846968238692,
+      "grad_norm": 1.1941807270050049,
+      "learning_rate": 4.0278750766954176e-05,
+      "loss": 0.7296,
+      "step": 36575
+    },
+    {
+      "epoch": 1.4082771896053898,
+      "grad_norm": 1.4177144765853882,
+      "learning_rate": 4.025449882219979e-05,
+      "loss": 0.8877,
+      "step": 36580
+    },
+    {
+      "epoch": 1.4084696823869105,
+      "grad_norm": 1.0108013153076172,
+      "learning_rate": 4.023025234077564e-05,
+      "loss": 0.9344,
+      "step": 36585
+    },
+    {
+      "epoch": 1.4086621751684312,
+      "grad_norm": 1.2472012042999268,
+      "learning_rate": 4.0206011324899054e-05,
+      "loss": 0.9062,
+      "step": 36590
+    },
+    {
+      "epoch": 1.4088546679499518,
+      "grad_norm": 1.353990077972412,
+      "learning_rate": 4.018177577678658e-05,
+      "loss": 0.7711,
+      "step": 36595
+    },
+    {
+      "epoch": 1.4090471607314725,
+      "grad_norm": 1.1395666599273682,
+      "learning_rate": 4.0157545698654465e-05,
+      "loss": 0.6574,
+      "step": 36600
+    },
+    {
+      "epoch": 1.4092396535129932,
+      "grad_norm": 1.177996039390564,
+      "learning_rate": 4.013332109271837e-05,
+      "loss": 0.8385,
+      "step": 36605
+    },
+    {
+      "epoch": 1.409432146294514,
+      "grad_norm": 1.8291305303573608,
+      "learning_rate": 4.0109101961193474e-05,
+      "loss": 0.8298,
+      "step": 36610
+    },
+    {
+      "epoch": 1.4096246390760347,
+      "grad_norm": 1.83578622341156,
+      "learning_rate": 4.008488830629451e-05,
+      "loss": 1.0385,
+      "step": 36615
+    },
+    {
+      "epoch": 1.4098171318575554,
+      "grad_norm": 1.2734320163726807,
+      "learning_rate": 4.0060680130235564e-05,
+      "loss": 0.8062,
+      "step": 36620
+    },
+    {
+      "epoch": 1.410009624639076,
+      "grad_norm": 0.9485237002372742,
+      "learning_rate": 4.003647743523037e-05,
+      "loss": 0.958,
+      "step": 36625
+    },
+    {
+      "epoch": 1.4102021174205968,
+      "grad_norm": 1.1491374969482422,
+      "learning_rate": 4.0012280223492084e-05,
+      "loss": 0.7825,
+      "step": 36630
+    },
+    {
+      "epoch": 1.4103946102021174,
+      "grad_norm": 1.9539427757263184,
+      "learning_rate": 3.9988088497233424e-05,
+      "loss": 0.7944,
+      "step": 36635
+    },
+    {
+      "epoch": 1.410587102983638,
+      "grad_norm": 1.9131194353103638,
+      "learning_rate": 3.9963902258666465e-05,
+      "loss": 0.9911,
+      "step": 36640
+    },
+    {
+      "epoch": 1.4107795957651588,
+      "grad_norm": 0.8639621734619141,
+      "learning_rate": 3.9939721510003e-05,
+      "loss": 0.7923,
+      "step": 36645
+    },
+    {
+      "epoch": 1.4109720885466794,
+      "grad_norm": 1.7485369443893433,
+      "learning_rate": 3.991554625345412e-05,
+      "loss": 0.7266,
+      "step": 36650
+    },
+    {
+      "epoch": 1.4111645813282,
+      "grad_norm": 0.921247124671936,
+      "learning_rate": 3.989137649123051e-05,
+      "loss": 0.7662,
+      "step": 36655
+    },
+    {
+      "epoch": 1.4113570741097208,
+      "grad_norm": 0.8794036507606506,
+      "learning_rate": 3.986721222554232e-05,
+      "loss": 0.6367,
+      "step": 36660
+    },
+    {
+      "epoch": 1.4115495668912414,
+      "grad_norm": 1.2151905298233032,
+      "learning_rate": 3.984305345859928e-05,
+      "loss": 0.7589,
+      "step": 36665
+    },
+    {
+      "epoch": 1.4117420596727623,
+      "grad_norm": 1.3154112100601196,
+      "learning_rate": 3.981890019261042e-05,
+      "loss": 1.0049,
+      "step": 36670
+    },
+    {
+      "epoch": 1.411934552454283,
+      "grad_norm": 1.4974151849746704,
+      "learning_rate": 3.979475242978454e-05,
+      "loss": 0.7273,
+      "step": 36675
+    },
+    {
+      "epoch": 1.4121270452358037,
+      "grad_norm": 1.543226718902588,
+      "learning_rate": 3.97706101723297e-05,
+      "loss": 0.8911,
+      "step": 36680
+    },
+    {
+      "epoch": 1.4123195380173243,
+      "grad_norm": 1.6745073795318604,
+      "learning_rate": 3.974647342245358e-05,
+      "loss": 0.7785,
+      "step": 36685
+    },
+    {
+      "epoch": 1.412512030798845,
+      "grad_norm": 0.8186720609664917,
+      "learning_rate": 3.972234218236331e-05,
+      "loss": 0.8083,
+      "step": 36690
+    },
+    {
+      "epoch": 1.4127045235803657,
+      "grad_norm": 1.7771086692810059,
+      "learning_rate": 3.969821645426559e-05,
+      "loss": 0.6911,
+      "step": 36695
+    },
+    {
+      "epoch": 1.4128970163618864,
+      "grad_norm": 1.0022178888320923,
+      "learning_rate": 3.9674096240366445e-05,
+      "loss": 0.9787,
+      "step": 36700
+    },
+    {
+      "epoch": 1.4130895091434073,
+      "grad_norm": 0.9852539896965027,
+      "learning_rate": 3.964998154287164e-05,
+      "loss": 0.7145,
+      "step": 36705
+    },
+    {
+      "epoch": 1.413282001924928,
+      "grad_norm": 1.1110121011734009,
+      "learning_rate": 3.9625872363986205e-05,
+      "loss": 0.7933,
+      "step": 36710
+    },
+    {
+      "epoch": 1.4134744947064486,
+      "grad_norm": 1.6256468296051025,
+      "learning_rate": 3.96017687059148e-05,
+      "loss": 1.0778,
+      "step": 36715
+    },
+    {
+      "epoch": 1.4136669874879693,
+      "grad_norm": 0.9198029041290283,
+      "learning_rate": 3.957767057086158e-05,
+      "loss": 0.6291,
+      "step": 36720
+    },
+    {
+      "epoch": 1.41385948026949,
+      "grad_norm": 0.8681755065917969,
+      "learning_rate": 3.9553577961030094e-05,
+      "loss": 0.8225,
+      "step": 36725
+    },
+    {
+      "epoch": 1.4140519730510106,
+      "grad_norm": 0.9288125038146973,
+      "learning_rate": 3.952949087862349e-05,
+      "loss": 0.7758,
+      "step": 36730
+    },
+    {
+      "epoch": 1.4142444658325313,
+      "grad_norm": 1.3477758169174194,
+      "learning_rate": 3.9505409325844346e-05,
+      "loss": 0.8854,
+      "step": 36735
+    },
+    {
+      "epoch": 1.414436958614052,
+      "grad_norm": 1.035380482673645,
+      "learning_rate": 3.948133330489483e-05,
+      "loss": 0.9035,
+      "step": 36740
+    },
+    {
+      "epoch": 1.4146294513955726,
+      "grad_norm": 1.4395008087158203,
+      "learning_rate": 3.9457262817976405e-05,
+      "loss": 0.8584,
+      "step": 36745
+    },
+    {
+      "epoch": 1.4148219441770933,
+      "grad_norm": 1.6281189918518066,
+      "learning_rate": 3.943319786729031e-05,
+      "loss": 0.8451,
+      "step": 36750
+    },
+    {
+      "epoch": 1.415014436958614,
+      "grad_norm": 1.3493483066558838,
+      "learning_rate": 3.940913845503701e-05,
+      "loss": 0.9527,
+      "step": 36755
+    },
+    {
+      "epoch": 1.4152069297401346,
+      "grad_norm": 1.200760841369629,
+      "learning_rate": 3.938508458341663e-05,
+      "loss": 0.812,
+      "step": 36760
+    },
+    {
+      "epoch": 1.4153994225216555,
+      "grad_norm": 1.3952566385269165,
+      "learning_rate": 3.9361036254628726e-05,
+      "loss": 0.9364,
+      "step": 36765
+    },
+    {
+      "epoch": 1.4155919153031762,
+      "grad_norm": 1.607851266860962,
+      "learning_rate": 3.9336993470872385e-05,
+      "loss": 0.8368,
+      "step": 36770
+    },
+    {
+      "epoch": 1.4157844080846969,
+      "grad_norm": 1.7063467502593994,
+      "learning_rate": 3.931295623434608e-05,
+      "loss": 0.8981,
+      "step": 36775
+    },
+    {
+      "epoch": 1.4159769008662175,
+      "grad_norm": 2.502572536468506,
+      "learning_rate": 3.928892454724797e-05,
+      "loss": 0.8994,
+      "step": 36780
+    },
+    {
+      "epoch": 1.4161693936477382,
+      "grad_norm": 2.07956862449646,
+      "learning_rate": 3.926489841177551e-05,
+      "loss": 1.1487,
+      "step": 36785
+    },
+    {
+      "epoch": 1.4163618864292589,
+      "grad_norm": 1.2496376037597656,
+      "learning_rate": 3.924087783012575e-05,
+      "loss": 0.7492,
+      "step": 36790
+    },
+    {
+      "epoch": 1.4165543792107795,
+      "grad_norm": 1.3253594636917114,
+      "learning_rate": 3.921686280449523e-05,
+      "loss": 0.8016,
+      "step": 36795
+    },
+    {
+      "epoch": 1.4167468719923004,
+      "grad_norm": 0.9799445271492004,
+      "learning_rate": 3.919285333707998e-05,
+      "loss": 0.7401,
+      "step": 36800
+    },
+    {
+      "epoch": 1.416939364773821,
+      "grad_norm": 0.9721969366073608,
+      "learning_rate": 3.916884943007541e-05,
+      "loss": 0.6816,
+      "step": 36805
+    },
+    {
+      "epoch": 1.4171318575553418,
+      "grad_norm": 0.9141532778739929,
+      "learning_rate": 3.914485108567667e-05,
+      "loss": 0.8165,
+      "step": 36810
+    },
+    {
+      "epoch": 1.4173243503368624,
+      "grad_norm": 1.3266597986221313,
+      "learning_rate": 3.912085830607813e-05,
+      "loss": 0.8702,
+      "step": 36815
+    },
+    {
+      "epoch": 1.417516843118383,
+      "grad_norm": 1.6577297449111938,
+      "learning_rate": 3.909687109347382e-05,
+      "loss": 0.7582,
+      "step": 36820
+    },
+    {
+      "epoch": 1.4177093358999038,
+      "grad_norm": 1.3083659410476685,
+      "learning_rate": 3.9072889450057194e-05,
+      "loss": 0.7892,
+      "step": 36825
+    },
+    {
+      "epoch": 1.4179018286814244,
+      "grad_norm": 1.1180342435836792,
+      "learning_rate": 3.904891337802126e-05,
+      "loss": 0.9944,
+      "step": 36830
+    },
+    {
+      "epoch": 1.4180943214629451,
+      "grad_norm": 1.2931442260742188,
+      "learning_rate": 3.902494287955839e-05,
+      "loss": 0.769,
+      "step": 36835
+    },
+    {
+      "epoch": 1.4182868142444658,
+      "grad_norm": 1.012529730796814,
+      "learning_rate": 3.900097795686057e-05,
+      "loss": 0.6058,
+      "step": 36840
+    },
+    {
+      "epoch": 1.4184793070259865,
+      "grad_norm": 1.3702960014343262,
+      "learning_rate": 3.897701861211924e-05,
+      "loss": 0.8077,
+      "step": 36845
+    },
+    {
+      "epoch": 1.4186717998075071,
+      "grad_norm": 1.4221330881118774,
+      "learning_rate": 3.8953064847525324e-05,
+      "loss": 0.7043,
+      "step": 36850
+    },
+    {
+      "epoch": 1.4188642925890278,
+      "grad_norm": 1.2348641157150269,
+      "learning_rate": 3.892911666526926e-05,
+      "loss": 0.9082,
+      "step": 36855
+    },
+    {
+      "epoch": 1.4190567853705485,
+      "grad_norm": 1.2621009349822998,
+      "learning_rate": 3.890517406754085e-05,
+      "loss": 0.8831,
+      "step": 36860
+    },
+    {
+      "epoch": 1.4192492781520694,
+      "grad_norm": 1.2773290872573853,
+      "learning_rate": 3.888123705652962e-05,
+      "loss": 0.7864,
+      "step": 36865
+    },
+    {
+      "epoch": 1.41944177093359,
+      "grad_norm": 1.2234888076782227,
+      "learning_rate": 3.8857305634424356e-05,
+      "loss": 0.7859,
+      "step": 36870
+    },
+    {
+      "epoch": 1.4196342637151107,
+      "grad_norm": 1.1090830564498901,
+      "learning_rate": 3.8833379803413463e-05,
+      "loss": 0.8045,
+      "step": 36875
+    },
+    {
+      "epoch": 1.4198267564966314,
+      "grad_norm": 1.4224649667739868,
+      "learning_rate": 3.8809459565684784e-05,
+      "loss": 0.8793,
+      "step": 36880
+    },
+    {
+      "epoch": 1.420019249278152,
+      "grad_norm": 1.293678879737854,
+      "learning_rate": 3.878554492342572e-05,
+      "loss": 0.7928,
+      "step": 36885
+    },
+    {
+      "epoch": 1.4202117420596727,
+      "grad_norm": 1.27838134765625,
+      "learning_rate": 3.876163587882299e-05,
+      "loss": 0.8078,
+      "step": 36890
+    },
+    {
+      "epoch": 1.4204042348411934,
+      "grad_norm": 0.9022307991981506,
+      "learning_rate": 3.873773243406306e-05,
+      "loss": 0.8416,
+      "step": 36895
+    },
+    {
+      "epoch": 1.4205967276227143,
+      "grad_norm": 1.2606749534606934,
+      "learning_rate": 3.8713834591331646e-05,
+      "loss": 0.816,
+      "step": 36900
+    },
+    {
+      "epoch": 1.420789220404235,
+      "grad_norm": 1.7780064344406128,
+      "learning_rate": 3.868994235281407e-05,
+      "loss": 0.9733,
+      "step": 36905
+    },
+    {
+      "epoch": 1.4209817131857556,
+      "grad_norm": 1.242295742034912,
+      "learning_rate": 3.8666055720695126e-05,
+      "loss": 0.6507,
+      "step": 36910
+    },
+    {
+      "epoch": 1.4211742059672763,
+      "grad_norm": 1.5196983814239502,
+      "learning_rate": 3.864217469715912e-05,
+      "loss": 0.8056,
+      "step": 36915
+    },
+    {
+      "epoch": 1.421366698748797,
+      "grad_norm": 0.9677949547767639,
+      "learning_rate": 3.8618299284389747e-05,
+      "loss": 0.8748,
+      "step": 36920
+    },
+    {
+      "epoch": 1.4215591915303176,
+      "grad_norm": 0.9580836892127991,
+      "learning_rate": 3.859442948457028e-05,
+      "loss": 0.8544,
+      "step": 36925
+    },
+    {
+      "epoch": 1.4217516843118383,
+      "grad_norm": 1.5484107732772827,
+      "learning_rate": 3.857056529988347e-05,
+      "loss": 0.8237,
+      "step": 36930
+    },
+    {
+      "epoch": 1.421944177093359,
+      "grad_norm": 0.9140622019767761,
+      "learning_rate": 3.854670673251153e-05,
+      "loss": 0.6624,
+      "step": 36935
+    },
+    {
+      "epoch": 1.4221366698748796,
+      "grad_norm": 1.0796998739242554,
+      "learning_rate": 3.852285378463619e-05,
+      "loss": 0.7723,
+      "step": 36940
+    },
+    {
+      "epoch": 1.4223291626564003,
+      "grad_norm": 0.9047008752822876,
+      "learning_rate": 3.84990064584386e-05,
+      "loss": 0.787,
+      "step": 36945
+    },
+    {
+      "epoch": 1.422521655437921,
+      "grad_norm": 2.0904667377471924,
+      "learning_rate": 3.847516475609947e-05,
+      "loss": 0.8196,
+      "step": 36950
+    },
+    {
+      "epoch": 1.4227141482194416,
+      "grad_norm": 1.2479698657989502,
+      "learning_rate": 3.845132867979895e-05,
+      "loss": 0.8103,
+      "step": 36955
+    },
+    {
+      "epoch": 1.4229066410009625,
+      "grad_norm": 1.6270309686660767,
+      "learning_rate": 3.842749823171673e-05,
+      "loss": 0.8852,
+      "step": 36960
+    },
+    {
+      "epoch": 1.4230991337824832,
+      "grad_norm": 1.9030733108520508,
+      "learning_rate": 3.840367341403185e-05,
+      "loss": 0.9027,
+      "step": 36965
+    },
+    {
+      "epoch": 1.4232916265640039,
+      "grad_norm": 1.098138689994812,
+      "learning_rate": 3.837985422892307e-05,
+      "loss": 0.8737,
+      "step": 36970
+    },
+    {
+      "epoch": 1.4234841193455245,
+      "grad_norm": 1.9470241069793701,
+      "learning_rate": 3.8356040678568393e-05,
+      "loss": 0.7401,
+      "step": 36975
+    },
+    {
+      "epoch": 1.4236766121270452,
+      "grad_norm": 1.122239589691162,
+      "learning_rate": 3.833223276514544e-05,
+      "loss": 0.873,
+      "step": 36980
+    },
+    {
+      "epoch": 1.4238691049085659,
+      "grad_norm": 1.3224730491638184,
+      "learning_rate": 3.830843049083128e-05,
+      "loss": 0.7331,
+      "step": 36985
+    },
+    {
+      "epoch": 1.4240615976900866,
+      "grad_norm": 1.4504386186599731,
+      "learning_rate": 3.8284633857802524e-05,
+      "loss": 0.7491,
+      "step": 36990
+    },
+    {
+      "epoch": 1.4242540904716074,
+      "grad_norm": 1.9830818176269531,
+      "learning_rate": 3.82608428682351e-05,
+      "loss": 0.8795,
+      "step": 36995
+    },
+    {
+      "epoch": 1.4244465832531281,
+      "grad_norm": 0.9479761719703674,
+      "learning_rate": 3.823705752430469e-05,
+      "loss": 0.972,
+      "step": 37000
+    },
+    {
+      "epoch": 1.4246390760346488,
+      "grad_norm": 2.649916172027588,
+      "learning_rate": 3.821327782818619e-05,
+      "loss": 0.8017,
+      "step": 37005
+    },
+    {
+      "epoch": 1.4248315688161695,
+      "grad_norm": 1.3527096509933472,
+      "learning_rate": 3.8189503782054135e-05,
+      "loss": 0.796,
+      "step": 37010
+    },
+    {
+      "epoch": 1.4250240615976901,
+      "grad_norm": 1.6791696548461914,
+      "learning_rate": 3.816573538808249e-05,
+      "loss": 0.7578,
+      "step": 37015
+    },
+    {
+      "epoch": 1.4252165543792108,
+      "grad_norm": 1.8477129936218262,
+      "learning_rate": 3.814197264844478e-05,
+      "loss": 0.7141,
+      "step": 37020
+    },
+    {
+      "epoch": 1.4254090471607315,
+      "grad_norm": 1.5426512956619263,
+      "learning_rate": 3.811821556531382e-05,
+      "loss": 0.8485,
+      "step": 37025
+    },
+    {
+      "epoch": 1.4256015399422521,
+      "grad_norm": 1.2095710039138794,
+      "learning_rate": 3.809446414086218e-05,
+      "loss": 0.8202,
+      "step": 37030
+    },
+    {
+      "epoch": 1.4257940327237728,
+      "grad_norm": 1.1557666063308716,
+      "learning_rate": 3.8070718377261696e-05,
+      "loss": 0.7613,
+      "step": 37035
+    },
+    {
+      "epoch": 1.4259865255052935,
+      "grad_norm": 1.4314980506896973,
+      "learning_rate": 3.8046978276683756e-05,
+      "loss": 0.8334,
+      "step": 37040
+    },
+    {
+      "epoch": 1.4261790182868141,
+      "grad_norm": 0.7171056866645813,
+      "learning_rate": 3.802324384129925e-05,
+      "loss": 0.886,
+      "step": 37045
+    },
+    {
+      "epoch": 1.4263715110683348,
+      "grad_norm": 2.3960771560668945,
+      "learning_rate": 3.799951507327858e-05,
+      "loss": 0.857,
+      "step": 37050
+    },
+    {
+      "epoch": 1.4265640038498557,
+      "grad_norm": 1.7112245559692383,
+      "learning_rate": 3.7975791974791494e-05,
+      "loss": 0.7698,
+      "step": 37055
+    },
+    {
+      "epoch": 1.4267564966313764,
+      "grad_norm": 1.118683934211731,
+      "learning_rate": 3.795207454800737e-05,
+      "loss": 0.6801,
+      "step": 37060
+    },
+    {
+      "epoch": 1.426948989412897,
+      "grad_norm": 1.9905716180801392,
+      "learning_rate": 3.7928362795095e-05,
+      "loss": 0.793,
+      "step": 37065
+    },
+    {
+      "epoch": 1.4271414821944177,
+      "grad_norm": 1.2646980285644531,
+      "learning_rate": 3.790465671822265e-05,
+      "loss": 0.8383,
+      "step": 37070
+    },
+    {
+      "epoch": 1.4273339749759384,
+      "grad_norm": 1.1731975078582764,
+      "learning_rate": 3.7880956319558146e-05,
+      "loss": 0.7778,
+      "step": 37075
+    },
+    {
+      "epoch": 1.427526467757459,
+      "grad_norm": 1.5719605684280396,
+      "learning_rate": 3.785726160126864e-05,
+      "loss": 0.8979,
+      "step": 37080
+    },
+    {
+      "epoch": 1.4277189605389797,
+      "grad_norm": 1.1388436555862427,
+      "learning_rate": 3.7833572565520924e-05,
+      "loss": 0.8414,
+      "step": 37085
+    },
+    {
+      "epoch": 1.4279114533205004,
+      "grad_norm": 1.2077659368515015,
+      "learning_rate": 3.7809889214481165e-05,
+      "loss": 0.8771,
+      "step": 37090
+    },
+    {
+      "epoch": 1.4281039461020213,
+      "grad_norm": 1.342726469039917,
+      "learning_rate": 3.778621155031511e-05,
+      "loss": 0.724,
+      "step": 37095
+    },
+    {
+      "epoch": 1.428296438883542,
+      "grad_norm": 0.8151179552078247,
+      "learning_rate": 3.77625395751878e-05,
+      "loss": 0.7926,
+      "step": 37100
+    },
+    {
+      "epoch": 1.4284889316650626,
+      "grad_norm": 1.3730394840240479,
+      "learning_rate": 3.773887329126404e-05,
+      "loss": 0.9403,
+      "step": 37105
+    },
+    {
+      "epoch": 1.4286814244465833,
+      "grad_norm": 1.6528422832489014,
+      "learning_rate": 3.77152127007078e-05,
+      "loss": 0.796,
+      "step": 37110
+    },
+    {
+      "epoch": 1.428873917228104,
+      "grad_norm": 1.3892961740493774,
+      "learning_rate": 3.769155780568283e-05,
+      "loss": 0.8341,
+      "step": 37115
+    },
+    {
+      "epoch": 1.4290664100096246,
+      "grad_norm": 0.6289898753166199,
+      "learning_rate": 3.766790860835211e-05,
+      "loss": 0.7489,
+      "step": 37120
+    },
+    {
+      "epoch": 1.4292589027911453,
+      "grad_norm": 1.1175183057785034,
+      "learning_rate": 3.764426511087823e-05,
+      "loss": 0.9842,
+      "step": 37125
+    },
+    {
+      "epoch": 1.429451395572666,
+      "grad_norm": 1.6422615051269531,
+      "learning_rate": 3.762062731542324e-05,
+      "loss": 0.8007,
+      "step": 37130
+    },
+    {
+      "epoch": 1.4296438883541867,
+      "grad_norm": 2.0091776847839355,
+      "learning_rate": 3.759699522414869e-05,
+      "loss": 0.7446,
+      "step": 37135
+    },
+    {
+      "epoch": 1.4298363811357073,
+      "grad_norm": 1.1759101152420044,
+      "learning_rate": 3.757336883921552e-05,
+      "loss": 0.7502,
+      "step": 37140
+    },
+    {
+      "epoch": 1.430028873917228,
+      "grad_norm": 1.3325209617614746,
+      "learning_rate": 3.754974816278422e-05,
+      "loss": 0.7311,
+      "step": 37145
+    },
+    {
+      "epoch": 1.4302213666987487,
+      "grad_norm": 0.9399605989456177,
+      "learning_rate": 3.752613319701476e-05,
+      "loss": 0.7261,
+      "step": 37150
+    },
+    {
+      "epoch": 1.4304138594802696,
+      "grad_norm": 0.8700345754623413,
+      "learning_rate": 3.750252394406656e-05,
+      "loss": 0.7377,
+      "step": 37155
+    },
+    {
+      "epoch": 1.4306063522617902,
+      "grad_norm": 1.1583389043807983,
+      "learning_rate": 3.747892040609858e-05,
+      "loss": 0.6968,
+      "step": 37160
+    },
+    {
+      "epoch": 1.430798845043311,
+      "grad_norm": 0.9834827184677124,
+      "learning_rate": 3.7455322585269125e-05,
+      "loss": 0.8028,
+      "step": 37165
+    },
+    {
+      "epoch": 1.4309913378248316,
+      "grad_norm": 1.4735242128372192,
+      "learning_rate": 3.74317304837361e-05,
+      "loss": 0.8179,
+      "step": 37170
+    },
+    {
+      "epoch": 1.4311838306063522,
+      "grad_norm": 1.5621691942214966,
+      "learning_rate": 3.740814410365685e-05,
+      "loss": 0.8056,
+      "step": 37175
+    },
+    {
+      "epoch": 1.431376323387873,
+      "grad_norm": 3.325570583343506,
+      "learning_rate": 3.7384563447188226e-05,
+      "loss": 0.9204,
+      "step": 37180
+    },
+    {
+      "epoch": 1.4315688161693936,
+      "grad_norm": 0.9473647475242615,
+      "learning_rate": 3.736098851648641e-05,
+      "loss": 0.9339,
+      "step": 37185
+    },
+    {
+      "epoch": 1.4317613089509145,
+      "grad_norm": 1.8556329011917114,
+      "learning_rate": 3.733741931370731e-05,
+      "loss": 0.8082,
+      "step": 37190
+    },
+    {
+      "epoch": 1.4319538017324351,
+      "grad_norm": 1.158944845199585,
+      "learning_rate": 3.73138558410061e-05,
+      "loss": 0.8515,
+      "step": 37195
+    },
+    {
+      "epoch": 1.4321462945139558,
+      "grad_norm": 1.7924774885177612,
+      "learning_rate": 3.729029810053749e-05,
+      "loss": 0.904,
+      "step": 37200
+    },
+    {
+      "epoch": 1.4323387872954765,
+      "grad_norm": 1.753733515739441,
+      "learning_rate": 3.726674609445572e-05,
+      "loss": 0.9027,
+      "step": 37205
+    },
+    {
+      "epoch": 1.4325312800769971,
+      "grad_norm": 1.4341967105865479,
+      "learning_rate": 3.7243199824914484e-05,
+      "loss": 1.0106,
+      "step": 37210
+    },
+    {
+      "epoch": 1.4327237728585178,
+      "grad_norm": 1.0536222457885742,
+      "learning_rate": 3.721965929406682e-05,
+      "loss": 0.8197,
+      "step": 37215
+    },
+    {
+      "epoch": 1.4329162656400385,
+      "grad_norm": 1.952141284942627,
+      "learning_rate": 3.719612450406551e-05,
+      "loss": 0.8157,
+      "step": 37220
+    },
+    {
+      "epoch": 1.4331087584215592,
+      "grad_norm": 2.054227113723755,
+      "learning_rate": 3.717259545706254e-05,
+      "loss": 0.859,
+      "step": 37225
+    },
+    {
+      "epoch": 1.4333012512030798,
+      "grad_norm": 0.9729071259498596,
+      "learning_rate": 3.714907215520952e-05,
+      "loss": 0.8556,
+      "step": 37230
+    },
+    {
+      "epoch": 1.4334937439846005,
+      "grad_norm": 1.8992177248001099,
+      "learning_rate": 3.712555460065751e-05,
+      "loss": 0.97,
+      "step": 37235
+    },
+    {
+      "epoch": 1.4336862367661212,
+      "grad_norm": 1.2505042552947998,
+      "learning_rate": 3.710204279555705e-05,
+      "loss": 0.8497,
+      "step": 37240
+    },
+    {
+      "epoch": 1.4338787295476418,
+      "grad_norm": 1.4273964166641235,
+      "learning_rate": 3.707853674205806e-05,
+      "loss": 0.8451,
+      "step": 37245
+    },
+    {
+      "epoch": 1.4340712223291627,
+      "grad_norm": 1.104282259941101,
+      "learning_rate": 3.705503644231012e-05,
+      "loss": 0.8491,
+      "step": 37250
+    },
+    {
+      "epoch": 1.4342637151106834,
+      "grad_norm": 1.683293104171753,
+      "learning_rate": 3.7031541898462096e-05,
+      "loss": 0.8464,
+      "step": 37255
+    },
+    {
+      "epoch": 1.434456207892204,
+      "grad_norm": 1.0647557973861694,
+      "learning_rate": 3.700805311266244e-05,
+      "loss": 0.7783,
+      "step": 37260
+    },
+    {
+      "epoch": 1.4346487006737247,
+      "grad_norm": 1.2208470106124878,
+      "learning_rate": 3.6984570087059045e-05,
+      "loss": 0.7758,
+      "step": 37265
+    },
+    {
+      "epoch": 1.4348411934552454,
+      "grad_norm": 1.9771546125411987,
+      "learning_rate": 3.6961092823799314e-05,
+      "loss": 0.9314,
+      "step": 37270
+    },
+    {
+      "epoch": 1.435033686236766,
+      "grad_norm": 0.9171108603477478,
+      "learning_rate": 3.6937621325030016e-05,
+      "loss": 0.7365,
+      "step": 37275
+    },
+    {
+      "epoch": 1.4352261790182868,
+      "grad_norm": 1.75246000289917,
+      "learning_rate": 3.691415559289749e-05,
+      "loss": 0.8452,
+      "step": 37280
+    },
+    {
+      "epoch": 1.4354186717998076,
+      "grad_norm": 1.3954969644546509,
+      "learning_rate": 3.6890695629547564e-05,
+      "loss": 0.8532,
+      "step": 37285
+    },
+    {
+      "epoch": 1.4356111645813283,
+      "grad_norm": 1.296316146850586,
+      "learning_rate": 3.68672414371254e-05,
+      "loss": 0.7224,
+      "step": 37290
+    },
+    {
+      "epoch": 1.435803657362849,
+      "grad_norm": 1.0747795104980469,
+      "learning_rate": 3.684379301777585e-05,
+      "loss": 0.9024,
+      "step": 37295
+    },
+    {
+      "epoch": 1.4359961501443697,
+      "grad_norm": 1.5699114799499512,
+      "learning_rate": 3.682035037364303e-05,
+      "loss": 0.9244,
+      "step": 37300
+    },
+    {
+      "epoch": 1.4361886429258903,
+      "grad_norm": 2.6069118976593018,
+      "learning_rate": 3.679691350687064e-05,
+      "loss": 0.645,
+      "step": 37305
+    },
+    {
+      "epoch": 1.436381135707411,
+      "grad_norm": 1.1996586322784424,
+      "learning_rate": 3.6773482419601826e-05,
+      "loss": 0.7952,
+      "step": 37310
+    },
+    {
+      "epoch": 1.4365736284889317,
+      "grad_norm": 1.3153495788574219,
+      "learning_rate": 3.675005711397924e-05,
+      "loss": 0.9398,
+      "step": 37315
+    },
+    {
+      "epoch": 1.4367661212704523,
+      "grad_norm": 1.404503583908081,
+      "learning_rate": 3.672663759214487e-05,
+      "loss": 0.845,
+      "step": 37320
+    },
+    {
+      "epoch": 1.436958614051973,
+      "grad_norm": 0.9732365012168884,
+      "learning_rate": 3.670322385624042e-05,
+      "loss": 0.8258,
+      "step": 37325
+    },
+    {
+      "epoch": 1.4371511068334937,
+      "grad_norm": 1.2123562097549438,
+      "learning_rate": 3.667981590840681e-05,
+      "loss": 0.8841,
+      "step": 37330
+    },
+    {
+      "epoch": 1.4373435996150143,
+      "grad_norm": 2.156731367111206,
+      "learning_rate": 3.6656413750784565e-05,
+      "loss": 0.9473,
+      "step": 37335
+    },
+    {
+      "epoch": 1.437536092396535,
+      "grad_norm": 1.4987213611602783,
+      "learning_rate": 3.6633017385513676e-05,
+      "loss": 0.7732,
+      "step": 37340
+    },
+    {
+      "epoch": 1.4377285851780557,
+      "grad_norm": 1.517491102218628,
+      "learning_rate": 3.660962681473362e-05,
+      "loss": 0.8491,
+      "step": 37345
+    },
+    {
+      "epoch": 1.4379210779595766,
+      "grad_norm": 1.0631245374679565,
+      "learning_rate": 3.65862420405832e-05,
+      "loss": 0.6431,
+      "step": 37350
+    },
+    {
+      "epoch": 1.4381135707410972,
+      "grad_norm": 1.114632248878479,
+      "learning_rate": 3.656286306520094e-05,
+      "loss": 0.8981,
+      "step": 37355
+    },
+    {
+      "epoch": 1.438306063522618,
+      "grad_norm": 2.538287401199341,
+      "learning_rate": 3.6539489890724575e-05,
+      "loss": 0.8646,
+      "step": 37360
+    },
+    {
+      "epoch": 1.4384985563041386,
+      "grad_norm": 1.4105244874954224,
+      "learning_rate": 3.651612251929147e-05,
+      "loss": 0.9686,
+      "step": 37365
+    },
+    {
+      "epoch": 1.4386910490856593,
+      "grad_norm": 1.3305569887161255,
+      "learning_rate": 3.649276095303843e-05,
+      "loss": 0.9057,
+      "step": 37370
+    },
+    {
+      "epoch": 1.43888354186718,
+      "grad_norm": 1.292600154876709,
+      "learning_rate": 3.646940519410169e-05,
+      "loss": 0.8515,
+      "step": 37375
+    },
+    {
+      "epoch": 1.4390760346487006,
+      "grad_norm": 1.3273907899856567,
+      "learning_rate": 3.644605524461704e-05,
+      "loss": 0.9926,
+      "step": 37380
+    },
+    {
+      "epoch": 1.4392685274302215,
+      "grad_norm": 1.1350206136703491,
+      "learning_rate": 3.6422711106719595e-05,
+      "loss": 0.9411,
+      "step": 37385
+    },
+    {
+      "epoch": 1.4394610202117422,
+      "grad_norm": 1.1308443546295166,
+      "learning_rate": 3.639937278254406e-05,
+      "loss": 0.7784,
+      "step": 37390
+    },
+    {
+      "epoch": 1.4396535129932628,
+      "grad_norm": 1.2704259157180786,
+      "learning_rate": 3.637604027422457e-05,
+      "loss": 0.7807,
+      "step": 37395
+    },
+    {
+      "epoch": 1.4398460057747835,
+      "grad_norm": 1.5792492628097534,
+      "learning_rate": 3.6352713583894746e-05,
+      "loss": 0.7525,
+      "step": 37400
+    },
+    {
+      "epoch": 1.4400384985563042,
+      "grad_norm": 1.706715703010559,
+      "learning_rate": 3.632939271368758e-05,
+      "loss": 0.856,
+      "step": 37405
+    },
+    {
+      "epoch": 1.4402309913378248,
+      "grad_norm": 1.0315080881118774,
+      "learning_rate": 3.630607766573574e-05,
+      "loss": 0.907,
+      "step": 37410
+    },
+    {
+      "epoch": 1.4404234841193455,
+      "grad_norm": 1.4221409559249878,
+      "learning_rate": 3.628276844217113e-05,
+      "loss": 0.7504,
+      "step": 37415
+    },
+    {
+      "epoch": 1.4406159769008662,
+      "grad_norm": 0.8785042762756348,
+      "learning_rate": 3.6259465045125265e-05,
+      "loss": 0.8181,
+      "step": 37420
+    },
+    {
+      "epoch": 1.4408084696823868,
+      "grad_norm": 1.8334077596664429,
+      "learning_rate": 3.623616747672907e-05,
+      "loss": 0.8067,
+      "step": 37425
+    },
+    {
+      "epoch": 1.4410009624639075,
+      "grad_norm": 0.9262105226516724,
+      "learning_rate": 3.621287573911299e-05,
+      "loss": 0.8618,
+      "step": 37430
+    },
+    {
+      "epoch": 1.4411934552454282,
+      "grad_norm": 1.228710651397705,
+      "learning_rate": 3.618958983440682e-05,
+      "loss": 0.8697,
+      "step": 37435
+    },
+    {
+      "epoch": 1.4413859480269489,
+      "grad_norm": 1.6522730588912964,
+      "learning_rate": 3.616630976474003e-05,
+      "loss": 0.9191,
+      "step": 37440
+    },
+    {
+      "epoch": 1.4415784408084698,
+      "grad_norm": 1.1182198524475098,
+      "learning_rate": 3.6143035532241296e-05,
+      "loss": 0.818,
+      "step": 37445
+    },
+    {
+      "epoch": 1.4417709335899904,
+      "grad_norm": 0.9378697872161865,
+      "learning_rate": 3.611976713903895e-05,
+      "loss": 0.8066,
+      "step": 37450
+    },
+    {
+      "epoch": 1.441963426371511,
+      "grad_norm": 1.22636079788208,
+      "learning_rate": 3.609650458726075e-05,
+      "loss": 0.6601,
+      "step": 37455
+    },
+    {
+      "epoch": 1.4421559191530318,
+      "grad_norm": 1.1185452938079834,
+      "learning_rate": 3.60732478790339e-05,
+      "loss": 0.7541,
+      "step": 37460
+    },
+    {
+      "epoch": 1.4423484119345524,
+      "grad_norm": 0.9403701424598694,
+      "learning_rate": 3.6049997016485024e-05,
+      "loss": 0.7677,
+      "step": 37465
+    },
+    {
+      "epoch": 1.442540904716073,
+      "grad_norm": 0.9934571981430054,
+      "learning_rate": 3.602675200174031e-05,
+      "loss": 0.8691,
+      "step": 37470
+    },
+    {
+      "epoch": 1.4427333974975938,
+      "grad_norm": 1.63606858253479,
+      "learning_rate": 3.600351283692531e-05,
+      "loss": 0.7052,
+      "step": 37475
+    },
+    {
+      "epoch": 1.4429258902791147,
+      "grad_norm": 1.6457669734954834,
+      "learning_rate": 3.598027952416515e-05,
+      "loss": 0.8816,
+      "step": 37480
+    },
+    {
+      "epoch": 1.4431183830606353,
+      "grad_norm": 1.1504909992218018,
+      "learning_rate": 3.595705206558435e-05,
+      "loss": 0.8399,
+      "step": 37485
+    },
+    {
+      "epoch": 1.443310875842156,
+      "grad_norm": 1.7006773948669434,
+      "learning_rate": 3.593383046330687e-05,
+      "loss": 0.8129,
+      "step": 37490
+    },
+    {
+      "epoch": 1.4435033686236767,
+      "grad_norm": 0.8277364373207092,
+      "learning_rate": 3.591061471945619e-05,
+      "loss": 0.7758,
+      "step": 37495
+    },
+    {
+      "epoch": 1.4436958614051973,
+      "grad_norm": 1.3335161209106445,
+      "learning_rate": 3.5887404836155244e-05,
+      "loss": 0.8057,
+      "step": 37500
+    },
+    {
+      "epoch": 1.443888354186718,
+      "grad_norm": 1.1840698719024658,
+      "learning_rate": 3.586420081552646e-05,
+      "loss": 0.681,
+      "step": 37505
+    },
+    {
+      "epoch": 1.4440808469682387,
+      "grad_norm": 1.2329528331756592,
+      "learning_rate": 3.584100265969157e-05,
+      "loss": 0.899,
+      "step": 37510
+    },
+    {
+      "epoch": 1.4442733397497594,
+      "grad_norm": 1.9120075702667236,
+      "learning_rate": 3.5817810370772046e-05,
+      "loss": 0.8751,
+      "step": 37515
+    },
+    {
+      "epoch": 1.44446583253128,
+      "grad_norm": 1.4686700105667114,
+      "learning_rate": 3.579462395088856e-05,
+      "loss": 0.9473,
+      "step": 37520
+    },
+    {
+      "epoch": 1.4446583253128007,
+      "grad_norm": 0.9519988298416138,
+      "learning_rate": 3.5771443402161396e-05,
+      "loss": 0.782,
+      "step": 37525
+    },
+    {
+      "epoch": 1.4448508180943214,
+      "grad_norm": 1.1774522066116333,
+      "learning_rate": 3.5748268726710254e-05,
+      "loss": 0.914,
+      "step": 37530
+    },
+    {
+      "epoch": 1.445043310875842,
+      "grad_norm": 1.1277164220809937,
+      "learning_rate": 3.5725099926654334e-05,
+      "loss": 0.8537,
+      "step": 37535
+    },
+    {
+      "epoch": 1.445235803657363,
+      "grad_norm": 1.815774917602539,
+      "learning_rate": 3.570193700411219e-05,
+      "loss": 0.7308,
+      "step": 37540
+    },
+    {
+      "epoch": 1.4454282964388836,
+      "grad_norm": 1.1836425065994263,
+      "learning_rate": 3.5678779961202035e-05,
+      "loss": 0.7268,
+      "step": 37545
+    },
+    {
+      "epoch": 1.4456207892204043,
+      "grad_norm": 0.964604914188385,
+      "learning_rate": 3.565562880004133e-05,
+      "loss": 0.689,
+      "step": 37550
+    },
+    {
+      "epoch": 1.445813282001925,
+      "grad_norm": 1.3372588157653809,
+      "learning_rate": 3.563248352274712e-05,
+      "loss": 0.8704,
+      "step": 37555
+    },
+    {
+      "epoch": 1.4460057747834456,
+      "grad_norm": 1.824810266494751,
+      "learning_rate": 3.56093441314359e-05,
+      "loss": 0.8523,
+      "step": 37560
+    },
+    {
+      "epoch": 1.4461982675649663,
+      "grad_norm": 1.0827716588974,
+      "learning_rate": 3.558621062822365e-05,
+      "loss": 0.9851,
+      "step": 37565
+    },
+    {
+      "epoch": 1.446390760346487,
+      "grad_norm": 1.1135975122451782,
+      "learning_rate": 3.556308301522566e-05,
+      "loss": 0.8532,
+      "step": 37570
+    },
+    {
+      "epoch": 1.4465832531280076,
+      "grad_norm": 1.0616776943206787,
+      "learning_rate": 3.553996129455694e-05,
+      "loss": 0.6833,
+      "step": 37575
+    },
+    {
+      "epoch": 1.4467757459095285,
+      "grad_norm": 0.9578964114189148,
+      "learning_rate": 3.551684546833173e-05,
+      "loss": 0.7893,
+      "step": 37580
+    },
+    {
+      "epoch": 1.4469682386910492,
+      "grad_norm": 1.2795662879943848,
+      "learning_rate": 3.549373553866383e-05,
+      "loss": 0.8313,
+      "step": 37585
+    },
+    {
+      "epoch": 1.4471607314725699,
+      "grad_norm": 0.9473350048065186,
+      "learning_rate": 3.547063150766651e-05,
+      "loss": 0.7281,
+      "step": 37590
+    },
+    {
+      "epoch": 1.4473532242540905,
+      "grad_norm": 2.0121681690216064,
+      "learning_rate": 3.544753337745249e-05,
+      "loss": 0.9083,
+      "step": 37595
+    },
+    {
+      "epoch": 1.4475457170356112,
+      "grad_norm": 1.5119409561157227,
+      "learning_rate": 3.542444115013389e-05,
+      "loss": 0.9774,
+      "step": 37600
+    },
+    {
+      "epoch": 1.4477382098171319,
+      "grad_norm": 1.390446424484253,
+      "learning_rate": 3.5401354827822386e-05,
+      "loss": 0.8337,
+      "step": 37605
+    },
+    {
+      "epoch": 1.4479307025986525,
+      "grad_norm": 2.891700267791748,
+      "learning_rate": 3.537827441262904e-05,
+      "loss": 0.7639,
+      "step": 37610
+    },
+    {
+      "epoch": 1.4481231953801732,
+      "grad_norm": 0.9751268625259399,
+      "learning_rate": 3.5355199906664426e-05,
+      "loss": 0.7479,
+      "step": 37615
+    },
+    {
+      "epoch": 1.4483156881616939,
+      "grad_norm": 1.9345892667770386,
+      "learning_rate": 3.533213131203859e-05,
+      "loss": 0.8192,
+      "step": 37620
+    },
+    {
+      "epoch": 1.4485081809432145,
+      "grad_norm": 0.8891172409057617,
+      "learning_rate": 3.5309068630860886e-05,
+      "loss": 0.7572,
+      "step": 37625
+    },
+    {
+      "epoch": 1.4487006737247352,
+      "grad_norm": 1.0947364568710327,
+      "learning_rate": 3.528601186524038e-05,
+      "loss": 0.8355,
+      "step": 37630
+    },
+    {
+      "epoch": 1.4488931665062559,
+      "grad_norm": 1.883512020111084,
+      "learning_rate": 3.5262961017285365e-05,
+      "loss": 0.705,
+      "step": 37635
+    },
+    {
+      "epoch": 1.4490856592877768,
+      "grad_norm": 1.5759938955307007,
+      "learning_rate": 3.523991608910373e-05,
+      "loss": 0.7667,
+      "step": 37640
+    },
+    {
+      "epoch": 1.4492781520692974,
+      "grad_norm": 1.6118040084838867,
+      "learning_rate": 3.521687708280277e-05,
+      "loss": 0.799,
+      "step": 37645
+    },
+    {
+      "epoch": 1.4494706448508181,
+      "grad_norm": 1.8567378520965576,
+      "learning_rate": 3.5193844000489275e-05,
+      "loss": 0.8733,
+      "step": 37650
+    },
+    {
+      "epoch": 1.4496631376323388,
+      "grad_norm": 1.5645298957824707,
+      "learning_rate": 3.5170816844269386e-05,
+      "loss": 0.833,
+      "step": 37655
+    },
+    {
+      "epoch": 1.4498556304138595,
+      "grad_norm": 1.3709477186203003,
+      "learning_rate": 3.5147795616248916e-05,
+      "loss": 0.8262,
+      "step": 37660
+    },
+    {
+      "epoch": 1.4500481231953801,
+      "grad_norm": 1.078057885169983,
+      "learning_rate": 3.512478031853288e-05,
+      "loss": 0.7562,
+      "step": 37665
+    },
+    {
+      "epoch": 1.4502406159769008,
+      "grad_norm": 2.0203640460968018,
+      "learning_rate": 3.510177095322593e-05,
+      "loss": 0.7648,
+      "step": 37670
+    },
+    {
+      "epoch": 1.4504331087584217,
+      "grad_norm": 2.2243845462799072,
+      "learning_rate": 3.5078767522432124e-05,
+      "loss": 0.6971,
+      "step": 37675
+    },
+    {
+      "epoch": 1.4506256015399424,
+      "grad_norm": 1.4907315969467163,
+      "learning_rate": 3.5055770028254995e-05,
+      "loss": 0.7486,
+      "step": 37680
+    },
+    {
+      "epoch": 1.450818094321463,
+      "grad_norm": 1.167829155921936,
+      "learning_rate": 3.503277847279745e-05,
+      "loss": 1.0481,
+      "step": 37685
+    },
+    {
+      "epoch": 1.4510105871029837,
+      "grad_norm": 2.3201534748077393,
+      "learning_rate": 3.500979285816194e-05,
+      "loss": 0.8633,
+      "step": 37690
+    },
+    {
+      "epoch": 1.4512030798845044,
+      "grad_norm": 1.0884878635406494,
+      "learning_rate": 3.498681318645035e-05,
+      "loss": 0.7201,
+      "step": 37695
+    },
+    {
+      "epoch": 1.451395572666025,
+      "grad_norm": 1.033949375152588,
+      "learning_rate": 3.496383945976403e-05,
+      "loss": 0.828,
+      "step": 37700
+    },
+    {
+      "epoch": 1.4515880654475457,
+      "grad_norm": 1.7160718441009521,
+      "learning_rate": 3.494087168020378e-05,
+      "loss": 1.0329,
+      "step": 37705
+    },
+    {
+      "epoch": 1.4517805582290664,
+      "grad_norm": 1.1474641561508179,
+      "learning_rate": 3.491790984986982e-05,
+      "loss": 0.7973,
+      "step": 37710
+    },
+    {
+      "epoch": 1.451973051010587,
+      "grad_norm": 1.491860270500183,
+      "learning_rate": 3.4894953970861875e-05,
+      "loss": 0.8382,
+      "step": 37715
+    },
+    {
+      "epoch": 1.4521655437921077,
+      "grad_norm": 1.2758644819259644,
+      "learning_rate": 3.487200404527911e-05,
+      "loss": 0.7288,
+      "step": 37720
+    },
+    {
+      "epoch": 1.4523580365736284,
+      "grad_norm": 1.046302318572998,
+      "learning_rate": 3.484906007522017e-05,
+      "loss": 0.787,
+      "step": 37725
+    },
+    {
+      "epoch": 1.452550529355149,
+      "grad_norm": 1.3662068843841553,
+      "learning_rate": 3.4826122062783026e-05,
+      "loss": 0.859,
+      "step": 37730
+    },
+    {
+      "epoch": 1.45274302213667,
+      "grad_norm": 1.0735856294631958,
+      "learning_rate": 3.480319001006536e-05,
+      "loss": 0.882,
+      "step": 37735
+    },
+    {
+      "epoch": 1.4529355149181906,
+      "grad_norm": 3.124530553817749,
+      "learning_rate": 3.4780263919164046e-05,
+      "loss": 1.0391,
+      "step": 37740
+    },
+    {
+      "epoch": 1.4531280076997113,
+      "grad_norm": 1.3612589836120605,
+      "learning_rate": 3.475734379217555e-05,
+      "loss": 0.731,
+      "step": 37745
+    },
+    {
+      "epoch": 1.453320500481232,
+      "grad_norm": 1.0555542707443237,
+      "learning_rate": 3.473442963119576e-05,
+      "loss": 0.9877,
+      "step": 37750
+    },
+    {
+      "epoch": 1.4535129932627526,
+      "grad_norm": 1.4260815382003784,
+      "learning_rate": 3.4711521438320094e-05,
+      "loss": 0.8009,
+      "step": 37755
+    },
+    {
+      "epoch": 1.4537054860442733,
+      "grad_norm": 1.2460854053497314,
+      "learning_rate": 3.4688619215643215e-05,
+      "loss": 0.8803,
+      "step": 37760
+    },
+    {
+      "epoch": 1.453897978825794,
+      "grad_norm": 0.8238735198974609,
+      "learning_rate": 3.4665722965259525e-05,
+      "loss": 0.7875,
+      "step": 37765
+    },
+    {
+      "epoch": 1.4540904716073149,
+      "grad_norm": 1.1429873704910278,
+      "learning_rate": 3.464283268926264e-05,
+      "loss": 0.7947,
+      "step": 37770
+    },
+    {
+      "epoch": 1.4542829643888355,
+      "grad_norm": 1.3068310022354126,
+      "learning_rate": 3.461994838974576e-05,
+      "loss": 0.826,
+      "step": 37775
+    },
+    {
+      "epoch": 1.4544754571703562,
+      "grad_norm": 1.26819908618927,
+      "learning_rate": 3.4597070068801484e-05,
+      "loss": 0.9455,
+      "step": 37780
+    },
+    {
+      "epoch": 1.4546679499518769,
+      "grad_norm": 1.3983910083770752,
+      "learning_rate": 3.457419772852194e-05,
+      "loss": 0.8876,
+      "step": 37785
+    },
+    {
+      "epoch": 1.4548604427333975,
+      "grad_norm": 1.6775517463684082,
+      "learning_rate": 3.455133137099853e-05,
+      "loss": 0.7722,
+      "step": 37790
+    },
+    {
+      "epoch": 1.4550529355149182,
+      "grad_norm": 1.2544530630111694,
+      "learning_rate": 3.452847099832238e-05,
+      "loss": 0.8891,
+      "step": 37795
+    },
+    {
+      "epoch": 1.4552454282964389,
+      "grad_norm": 0.9618650674819946,
+      "learning_rate": 3.4505616612583824e-05,
+      "loss": 0.7674,
+      "step": 37800
+    },
+    {
+      "epoch": 1.4554379210779596,
+      "grad_norm": 1.694305658340454,
+      "learning_rate": 3.448276821587275e-05,
+      "loss": 0.8635,
+      "step": 37805
+    },
+    {
+      "epoch": 1.4556304138594802,
+      "grad_norm": 1.1212447881698608,
+      "learning_rate": 3.445992581027853e-05,
+      "loss": 0.7715,
+      "step": 37810
+    },
+    {
+      "epoch": 1.455822906641001,
+      "grad_norm": 1.3266422748565674,
+      "learning_rate": 3.443708939788995e-05,
+      "loss": 0.8759,
+      "step": 37815
+    },
+    {
+      "epoch": 1.4560153994225216,
+      "grad_norm": 1.0106375217437744,
+      "learning_rate": 3.44142589807952e-05,
+      "loss": 0.639,
+      "step": 37820
+    },
+    {
+      "epoch": 1.4562078922040422,
+      "grad_norm": 1.1478214263916016,
+      "learning_rate": 3.4391434561082e-05,
+      "loss": 0.9026,
+      "step": 37825
+    },
+    {
+      "epoch": 1.4564003849855631,
+      "grad_norm": 1.7715827226638794,
+      "learning_rate": 3.436861614083753e-05,
+      "loss": 0.7835,
+      "step": 37830
+    },
+    {
+      "epoch": 1.4565928777670838,
+      "grad_norm": 1.5840331315994263,
+      "learning_rate": 3.434580372214829e-05,
+      "loss": 0.9205,
+      "step": 37835
+    },
+    {
+      "epoch": 1.4567853705486045,
+      "grad_norm": 1.0657918453216553,
+      "learning_rate": 3.432299730710042e-05,
+      "loss": 0.8256,
+      "step": 37840
+    },
+    {
+      "epoch": 1.4569778633301251,
+      "grad_norm": 1.3026148080825806,
+      "learning_rate": 3.430019689777936e-05,
+      "loss": 0.8559,
+      "step": 37845
+    },
+    {
+      "epoch": 1.4571703561116458,
+      "grad_norm": 1.7707239389419556,
+      "learning_rate": 3.4277402496270075e-05,
+      "loss": 0.8683,
+      "step": 37850
+    },
+    {
+      "epoch": 1.4573628488931665,
+      "grad_norm": 1.134018063545227,
+      "learning_rate": 3.4254614104656945e-05,
+      "loss": 0.7637,
+      "step": 37855
+    },
+    {
+      "epoch": 1.4575553416746871,
+      "grad_norm": 1.713789463043213,
+      "learning_rate": 3.423183172502388e-05,
+      "loss": 1.0235,
+      "step": 37860
+    },
+    {
+      "epoch": 1.4577478344562078,
+      "grad_norm": 1.0064624547958374,
+      "learning_rate": 3.420905535945406e-05,
+      "loss": 0.8631,
+      "step": 37865
+    },
+    {
+      "epoch": 1.4579403272377287,
+      "grad_norm": 1.2187049388885498,
+      "learning_rate": 3.418628501003036e-05,
+      "loss": 0.7412,
+      "step": 37870
+    },
+    {
+      "epoch": 1.4581328200192494,
+      "grad_norm": 0.9734949469566345,
+      "learning_rate": 3.416352067883489e-05,
+      "loss": 0.7388,
+      "step": 37875
+    },
+    {
+      "epoch": 1.45832531280077,
+      "grad_norm": 2.005460262298584,
+      "learning_rate": 3.414076236794933e-05,
+      "loss": 0.9135,
+      "step": 37880
+    },
+    {
+      "epoch": 1.4585178055822907,
+      "grad_norm": 1.063448429107666,
+      "learning_rate": 3.4118010079454775e-05,
+      "loss": 0.8584,
+      "step": 37885
+    },
+    {
+      "epoch": 1.4587102983638114,
+      "grad_norm": 1.4160773754119873,
+      "learning_rate": 3.4095263815431765e-05,
+      "loss": 0.871,
+      "step": 37890
+    },
+    {
+      "epoch": 1.458902791145332,
+      "grad_norm": 2.008450508117676,
+      "learning_rate": 3.40725235779603e-05,
+      "loss": 0.886,
+      "step": 37895
+    },
+    {
+      "epoch": 1.4590952839268527,
+      "grad_norm": 3.2115590572357178,
+      "learning_rate": 3.4049789369119866e-05,
+      "loss": 0.7679,
+      "step": 37900
+    },
+    {
+      "epoch": 1.4592877767083734,
+      "grad_norm": 1.1200793981552124,
+      "learning_rate": 3.402706119098927e-05,
+      "loss": 0.7649,
+      "step": 37905
+    },
+    {
+      "epoch": 1.459480269489894,
+      "grad_norm": 1.700742483139038,
+      "learning_rate": 3.4004339045646906e-05,
+      "loss": 0.7951,
+      "step": 37910
+    },
+    {
+      "epoch": 1.4596727622714147,
+      "grad_norm": 1.5388984680175781,
+      "learning_rate": 3.3981622935170554e-05,
+      "loss": 0.6965,
+      "step": 37915
+    },
+    {
+      "epoch": 1.4598652550529354,
+      "grad_norm": 1.1668038368225098,
+      "learning_rate": 3.3958912861637446e-05,
+      "loss": 0.807,
+      "step": 37920
+    },
+    {
+      "epoch": 1.460057747834456,
+      "grad_norm": 1.614430546760559,
+      "learning_rate": 3.3936208827124316e-05,
+      "loss": 0.7649,
+      "step": 37925
+    },
+    {
+      "epoch": 1.460250240615977,
+      "grad_norm": 1.4020650386810303,
+      "learning_rate": 3.3913510833707215e-05,
+      "loss": 0.8886,
+      "step": 37930
+    },
+    {
+      "epoch": 1.4604427333974976,
+      "grad_norm": 1.0239553451538086,
+      "learning_rate": 3.3890818883461774e-05,
+      "loss": 0.8201,
+      "step": 37935
+    },
+    {
+      "epoch": 1.4606352261790183,
+      "grad_norm": 1.2521649599075317,
+      "learning_rate": 3.386813297846301e-05,
+      "loss": 0.8144,
+      "step": 37940
+    },
+    {
+      "epoch": 1.460827718960539,
+      "grad_norm": 1.6655640602111816,
+      "learning_rate": 3.384545312078543e-05,
+      "loss": 0.7052,
+      "step": 37945
+    },
+    {
+      "epoch": 1.4610202117420596,
+      "grad_norm": 1.3284400701522827,
+      "learning_rate": 3.382277931250287e-05,
+      "loss": 0.7968,
+      "step": 37950
+    },
+    {
+      "epoch": 1.4612127045235803,
+      "grad_norm": 1.4124058485031128,
+      "learning_rate": 3.380011155568882e-05,
+      "loss": 0.7928,
+      "step": 37955
+    },
+    {
+      "epoch": 1.461405197305101,
+      "grad_norm": 1.6810020208358765,
+      "learning_rate": 3.3777449852416e-05,
+      "loss": 0.8213,
+      "step": 37960
+    },
+    {
+      "epoch": 1.4615976900866219,
+      "grad_norm": 0.6957064867019653,
+      "learning_rate": 3.375479420475671e-05,
+      "loss": 0.6974,
+      "step": 37965
+    },
+    {
+      "epoch": 1.4617901828681426,
+      "grad_norm": 1.094000220298767,
+      "learning_rate": 3.3732144614782655e-05,
+      "loss": 0.7123,
+      "step": 37970
+    },
+    {
+      "epoch": 1.4619826756496632,
+      "grad_norm": 1.638156771659851,
+      "learning_rate": 3.370950108456502e-05,
+      "loss": 0.9115,
+      "step": 37975
+    },
+    {
+      "epoch": 1.462175168431184,
+      "grad_norm": 1.41256582736969,
+      "learning_rate": 3.368686361617431e-05,
+      "loss": 0.7888,
+      "step": 37980
+    },
+    {
+      "epoch": 1.4623676612127046,
+      "grad_norm": 1.2503211498260498,
+      "learning_rate": 3.366423221168071e-05,
+      "loss": 0.7508,
+      "step": 37985
+    },
+    {
+      "epoch": 1.4625601539942252,
+      "grad_norm": 1.3764451742172241,
+      "learning_rate": 3.3641606873153596e-05,
+      "loss": 0.8037,
+      "step": 37990
+    },
+    {
+      "epoch": 1.462752646775746,
+      "grad_norm": 0.9157146215438843,
+      "learning_rate": 3.3618987602661966e-05,
+      "loss": 0.7469,
+      "step": 37995
+    },
+    {
+      "epoch": 1.4629451395572666,
+      "grad_norm": 1.5734952688217163,
+      "learning_rate": 3.359637440227418e-05,
+      "loss": 0.8036,
+      "step": 38000
+    },
+    {
+      "epoch": 1.4631376323387872,
+      "grad_norm": 1.2893953323364258,
+      "learning_rate": 3.357376727405809e-05,
+      "loss": 0.7856,
+      "step": 38005
+    },
+    {
+      "epoch": 1.463330125120308,
+      "grad_norm": 1.133774995803833,
+      "learning_rate": 3.3551166220080896e-05,
+      "loss": 0.8766,
+      "step": 38010
+    },
+    {
+      "epoch": 1.4635226179018286,
+      "grad_norm": 1.4209762811660767,
+      "learning_rate": 3.3528571242409435e-05,
+      "loss": 0.8307,
+      "step": 38015
+    },
+    {
+      "epoch": 1.4637151106833493,
+      "grad_norm": 1.2726255655288696,
+      "learning_rate": 3.350598234310977e-05,
+      "loss": 0.8027,
+      "step": 38020
+    },
+    {
+      "epoch": 1.4639076034648701,
+      "grad_norm": 1.623399257659912,
+      "learning_rate": 3.3483399524247525e-05,
+      "loss": 0.8838,
+      "step": 38025
+    },
+    {
+      "epoch": 1.4641000962463908,
+      "grad_norm": 1.2219460010528564,
+      "learning_rate": 3.3460822787887805e-05,
+      "loss": 1.0114,
+      "step": 38030
+    },
+    {
+      "epoch": 1.4642925890279115,
+      "grad_norm": 0.8296359777450562,
+      "learning_rate": 3.3438252136095016e-05,
+      "loss": 0.7348,
+      "step": 38035
+    },
+    {
+      "epoch": 1.4644850818094322,
+      "grad_norm": 1.6310055255889893,
+      "learning_rate": 3.341568757093314e-05,
+      "loss": 0.892,
+      "step": 38040
+    },
+    {
+      "epoch": 1.4646775745909528,
+      "grad_norm": 0.8464129567146301,
+      "learning_rate": 3.339312909446557e-05,
+      "loss": 0.8913,
+      "step": 38045
+    },
+    {
+      "epoch": 1.4648700673724735,
+      "grad_norm": 1.2375149726867676,
+      "learning_rate": 3.337057670875513e-05,
+      "loss": 0.6927,
+      "step": 38050
+    },
+    {
+      "epoch": 1.4650625601539942,
+      "grad_norm": 2.361973285675049,
+      "learning_rate": 3.334803041586402e-05,
+      "loss": 0.7511,
+      "step": 38055
+    },
+    {
+      "epoch": 1.465255052935515,
+      "grad_norm": 2.223909616470337,
+      "learning_rate": 3.332999776976661e-05,
+      "loss": 0.9986,
+      "step": 38060
+    },
+    {
+      "epoch": 1.4654475457170357,
+      "grad_norm": 0.8594219088554382,
+      "learning_rate": 3.330746244914553e-05,
+      "loss": 0.6887,
+      "step": 38065
+    },
+    {
+      "epoch": 1.4656400384985564,
+      "grad_norm": 1.6795843839645386,
+      "learning_rate": 3.3284933227115236e-05,
+      "loss": 0.7934,
+      "step": 38070
+    },
+    {
+      "epoch": 1.465832531280077,
+      "grad_norm": 0.9740133881568909,
+      "learning_rate": 3.3262410105735864e-05,
+      "loss": 0.7993,
+      "step": 38075
+    },
+    {
+      "epoch": 1.4660250240615977,
+      "grad_norm": 1.578036904335022,
+      "learning_rate": 3.323989308706693e-05,
+      "loss": 0.8793,
+      "step": 38080
+    },
+    {
+      "epoch": 1.4662175168431184,
+      "grad_norm": 1.1103001832962036,
+      "learning_rate": 3.32173821731676e-05,
+      "loss": 0.8391,
+      "step": 38085
+    },
+    {
+      "epoch": 1.466410009624639,
+      "grad_norm": 1.6756675243377686,
+      "learning_rate": 3.3194877366096246e-05,
+      "loss": 0.955,
+      "step": 38090
+    },
+    {
+      "epoch": 1.4666025024061597,
+      "grad_norm": 1.257323980331421,
+      "learning_rate": 3.3172378667910796e-05,
+      "loss": 0.7475,
+      "step": 38095
+    },
+    {
+      "epoch": 1.4667949951876804,
+      "grad_norm": 1.2056502103805542,
+      "learning_rate": 3.314988608066867e-05,
+      "loss": 0.8694,
+      "step": 38100
+    },
+    {
+      "epoch": 1.466987487969201,
+      "grad_norm": 1.552820086479187,
+      "learning_rate": 3.312739960642659e-05,
+      "loss": 0.8827,
+      "step": 38105
+    },
+    {
+      "epoch": 1.4671799807507218,
+      "grad_norm": 0.8146559596061707,
+      "learning_rate": 3.310491924724082e-05,
+      "loss": 0.592,
+      "step": 38110
+    },
+    {
+      "epoch": 1.4673724735322424,
+      "grad_norm": 0.8254255056381226,
+      "learning_rate": 3.3082445005167053e-05,
+      "loss": 0.7371,
+      "step": 38115
+    },
+    {
+      "epoch": 1.467564966313763,
+      "grad_norm": 1.7715586423873901,
+      "learning_rate": 3.3059976882260424e-05,
+      "loss": 0.7059,
+      "step": 38120
+    },
+    {
+      "epoch": 1.467757459095284,
+      "grad_norm": 1.2485114336013794,
+      "learning_rate": 3.303751488057541e-05,
+      "loss": 0.7397,
+      "step": 38125
+    },
+    {
+      "epoch": 1.4679499518768047,
+      "grad_norm": 1.529093623161316,
+      "learning_rate": 3.301505900216614e-05,
+      "loss": 0.8473,
+      "step": 38130
+    },
+    {
+      "epoch": 1.4681424446583253,
+      "grad_norm": 1.5459060668945312,
+      "learning_rate": 3.299260924908596e-05,
+      "loss": 0.906,
+      "step": 38135
+    },
+    {
+      "epoch": 1.468334937439846,
+      "grad_norm": 1.5474683046340942,
+      "learning_rate": 3.2970165623387785e-05,
+      "loss": 0.8314,
+      "step": 38140
+    },
+    {
+      "epoch": 1.4685274302213667,
+      "grad_norm": 2.399840831756592,
+      "learning_rate": 3.2947728127123924e-05,
+      "loss": 0.8761,
+      "step": 38145
+    },
+    {
+      "epoch": 1.4687199230028873,
+      "grad_norm": 0.8112540245056152,
+      "learning_rate": 3.292529676234615e-05,
+      "loss": 0.7489,
+      "step": 38150
+    },
+    {
+      "epoch": 1.468912415784408,
+      "grad_norm": 1.178676962852478,
+      "learning_rate": 3.290287153110565e-05,
+      "loss": 0.8148,
+      "step": 38155
+    },
+    {
+      "epoch": 1.469104908565929,
+      "grad_norm": 1.379562258720398,
+      "learning_rate": 3.28804524354531e-05,
+      "loss": 0.9524,
+      "step": 38160
+    },
+    {
+      "epoch": 1.4692974013474496,
+      "grad_norm": 2.1788909435272217,
+      "learning_rate": 3.28580394774385e-05,
+      "loss": 0.7175,
+      "step": 38165
+    },
+    {
+      "epoch": 1.4694898941289702,
+      "grad_norm": 1.2615634202957153,
+      "learning_rate": 3.283563265911147e-05,
+      "loss": 1.0649,
+      "step": 38170
+    },
+    {
+      "epoch": 1.469682386910491,
+      "grad_norm": 1.2906595468521118,
+      "learning_rate": 3.281323198252087e-05,
+      "loss": 0.6275,
+      "step": 38175
+    },
+    {
+      "epoch": 1.4698748796920116,
+      "grad_norm": 0.8177803158760071,
+      "learning_rate": 3.279083744971515e-05,
+      "loss": 0.786,
+      "step": 38180
+    },
+    {
+      "epoch": 1.4700673724735323,
+      "grad_norm": 1.317015290260315,
+      "learning_rate": 3.276844906274211e-05,
+      "loss": 0.8074,
+      "step": 38185
+    },
+    {
+      "epoch": 1.470259865255053,
+      "grad_norm": 1.0974969863891602,
+      "learning_rate": 3.274606682364908e-05,
+      "loss": 0.851,
+      "step": 38190
+    },
+    {
+      "epoch": 1.4704523580365736,
+      "grad_norm": 0.8745707273483276,
+      "learning_rate": 3.272369073448269e-05,
+      "loss": 0.9159,
+      "step": 38195
+    },
+    {
+      "epoch": 1.4706448508180943,
+      "grad_norm": 1.3758927583694458,
+      "learning_rate": 3.2701320797289114e-05,
+      "loss": 0.7093,
+      "step": 38200
+    },
+    {
+      "epoch": 1.470837343599615,
+      "grad_norm": 1.6409497261047363,
+      "learning_rate": 3.267895701411393e-05,
+      "loss": 0.7493,
+      "step": 38205
+    },
+    {
+      "epoch": 1.4710298363811356,
+      "grad_norm": 1.0862737894058228,
+      "learning_rate": 3.2656599387002176e-05,
+      "loss": 0.9189,
+      "step": 38210
+    },
+    {
+      "epoch": 1.4712223291626563,
+      "grad_norm": 0.9804477095603943,
+      "learning_rate": 3.263424791799833e-05,
+      "loss": 0.8635,
+      "step": 38215
+    },
+    {
+      "epoch": 1.4714148219441772,
+      "grad_norm": 1.2791975736618042,
+      "learning_rate": 3.2611902609146215e-05,
+      "loss": 0.8745,
+      "step": 38220
+    },
+    {
+      "epoch": 1.4716073147256978,
+      "grad_norm": 1.2094603776931763,
+      "learning_rate": 3.2589563462489214e-05,
+      "loss": 0.8492,
+      "step": 38225
+    },
+    {
+      "epoch": 1.4717998075072185,
+      "grad_norm": 0.9160768389701843,
+      "learning_rate": 3.256723048007006e-05,
+      "loss": 0.8906,
+      "step": 38230
+    },
+    {
+      "epoch": 1.4719923002887392,
+      "grad_norm": 2.3353841304779053,
+      "learning_rate": 3.254490366393104e-05,
+      "loss": 0.8709,
+      "step": 38235
+    },
+    {
+      "epoch": 1.4721847930702598,
+      "grad_norm": 1.8757665157318115,
+      "learning_rate": 3.2522583016113636e-05,
+      "loss": 0.6895,
+      "step": 38240
+    },
+    {
+      "epoch": 1.4723772858517805,
+      "grad_norm": 1.9597855806350708,
+      "learning_rate": 3.250026853865911e-05,
+      "loss": 0.7915,
+      "step": 38245
+    },
+    {
+      "epoch": 1.4725697786333012,
+      "grad_norm": 1.1252086162567139,
+      "learning_rate": 3.247796023360783e-05,
+      "loss": 0.8083,
+      "step": 38250
+    },
+    {
+      "epoch": 1.472762271414822,
+      "grad_norm": 1.7507824897766113,
+      "learning_rate": 3.2455658102999796e-05,
+      "loss": 0.8343,
+      "step": 38255
+    },
+    {
+      "epoch": 1.4729547641963427,
+      "grad_norm": 0.8124133944511414,
+      "learning_rate": 3.243336214887439e-05,
+      "loss": 0.7325,
+      "step": 38260
+    },
+    {
+      "epoch": 1.4731472569778634,
+      "grad_norm": 0.9057835936546326,
+      "learning_rate": 3.241107237327047e-05,
+      "loss": 0.8326,
+      "step": 38265
+    },
+    {
+      "epoch": 1.473339749759384,
+      "grad_norm": 2.141209602355957,
+      "learning_rate": 3.238878877822616e-05,
+      "loss": 0.8098,
+      "step": 38270
+    },
+    {
+      "epoch": 1.4735322425409048,
+      "grad_norm": 2.2603015899658203,
+      "learning_rate": 3.236651136577932e-05,
+      "loss": 0.7639,
+      "step": 38275
+    },
+    {
+      "epoch": 1.4737247353224254,
+      "grad_norm": 1.785561203956604,
+      "learning_rate": 3.234424013796694e-05,
+      "loss": 0.9348,
+      "step": 38280
+    },
+    {
+      "epoch": 1.473917228103946,
+      "grad_norm": 1.9484220743179321,
+      "learning_rate": 3.232197509682562e-05,
+      "loss": 1.0465,
+      "step": 38285
+    },
+    {
+      "epoch": 1.4741097208854668,
+      "grad_norm": 1.0257092714309692,
+      "learning_rate": 3.2299716244391356e-05,
+      "loss": 0.7771,
+      "step": 38290
+    },
+    {
+      "epoch": 1.4743022136669874,
+      "grad_norm": 0.8890891075134277,
+      "learning_rate": 3.2277463582699595e-05,
+      "loss": 0.7485,
+      "step": 38295
+    },
+    {
+      "epoch": 1.474494706448508,
+      "grad_norm": 1.7376580238342285,
+      "learning_rate": 3.225521711378514e-05,
+      "loss": 0.7943,
+      "step": 38300
+    },
+    {
+      "epoch": 1.4746871992300288,
+      "grad_norm": 0.7426727414131165,
+      "learning_rate": 3.2232976839682316e-05,
+      "loss": 0.7996,
+      "step": 38305
+    },
+    {
+      "epoch": 1.4748796920115494,
+      "grad_norm": 1.1797425746917725,
+      "learning_rate": 3.221074276242484e-05,
+      "loss": 0.7686,
+      "step": 38310
+    },
+    {
+      "epoch": 1.4750721847930703,
+      "grad_norm": 1.5532947778701782,
+      "learning_rate": 3.2188514884045885e-05,
+      "loss": 0.6781,
+      "step": 38315
+    },
+    {
+      "epoch": 1.475264677574591,
+      "grad_norm": 1.4531184434890747,
+      "learning_rate": 3.216629320657806e-05,
+      "loss": 0.9722,
+      "step": 38320
+    },
+    {
+      "epoch": 1.4754571703561117,
+      "grad_norm": 1.0439348220825195,
+      "learning_rate": 3.214407773205333e-05,
+      "loss": 0.9018,
+      "step": 38325
+    },
+    {
+      "epoch": 1.4756496631376324,
+      "grad_norm": 1.6608530282974243,
+      "learning_rate": 3.212186846250318e-05,
+      "loss": 0.8168,
+      "step": 38330
+    },
+    {
+      "epoch": 1.475842155919153,
+      "grad_norm": 1.7269642353057861,
+      "learning_rate": 3.209966539995851e-05,
+      "loss": 0.9625,
+      "step": 38335
+    },
+    {
+      "epoch": 1.4760346487006737,
+      "grad_norm": 1.6941852569580078,
+      "learning_rate": 3.2077468546449684e-05,
+      "loss": 0.8007,
+      "step": 38340
+    },
+    {
+      "epoch": 1.4762271414821944,
+      "grad_norm": 1.0302637815475464,
+      "learning_rate": 3.205527790400634e-05,
+      "loss": 0.9428,
+      "step": 38345
+    },
+    {
+      "epoch": 1.476419634263715,
+      "grad_norm": 1.4871275424957275,
+      "learning_rate": 3.2033093474657806e-05,
+      "loss": 0.9217,
+      "step": 38350
+    },
+    {
+      "epoch": 1.476612127045236,
+      "grad_norm": 1.700111746788025,
+      "learning_rate": 3.201091526043261e-05,
+      "loss": 0.791,
+      "step": 38355
+    },
+    {
+      "epoch": 1.4768046198267566,
+      "grad_norm": 0.7985191345214844,
+      "learning_rate": 3.198874326335881e-05,
+      "loss": 0.6741,
+      "step": 38360
+    },
+    {
+      "epoch": 1.4769971126082773,
+      "grad_norm": 1.5285080671310425,
+      "learning_rate": 3.1966577485463913e-05,
+      "loss": 0.8607,
+      "step": 38365
+    },
+    {
+      "epoch": 1.477189605389798,
+      "grad_norm": 1.746016263961792,
+      "learning_rate": 3.1944417928774864e-05,
+      "loss": 0.71,
+      "step": 38370
+    },
+    {
+      "epoch": 1.4773820981713186,
+      "grad_norm": 2.4140141010284424,
+      "learning_rate": 3.1922264595317895e-05,
+      "loss": 0.9232,
+      "step": 38375
+    },
+    {
+      "epoch": 1.4775745909528393,
+      "grad_norm": 1.8177671432495117,
+      "learning_rate": 3.190011748711892e-05,
+      "loss": 0.7186,
+      "step": 38380
+    },
+    {
+      "epoch": 1.47776708373436,
+      "grad_norm": 1.3410675525665283,
+      "learning_rate": 3.187797660620305e-05,
+      "loss": 0.8414,
+      "step": 38385
+    },
+    {
+      "epoch": 1.4779595765158806,
+      "grad_norm": 0.9143676161766052,
+      "learning_rate": 3.185584195459496e-05,
+      "loss": 0.7884,
+      "step": 38390
+    },
+    {
+      "epoch": 1.4781520692974013,
+      "grad_norm": 1.212738037109375,
+      "learning_rate": 3.18337135343187e-05,
+      "loss": 0.7888,
+      "step": 38395
+    },
+    {
+      "epoch": 1.478344562078922,
+      "grad_norm": 1.165589690208435,
+      "learning_rate": 3.181159134739777e-05,
+      "loss": 0.8875,
+      "step": 38400
+    },
+    {
+      "epoch": 1.4785370548604426,
+      "grad_norm": 1.0738500356674194,
+      "learning_rate": 3.178947539585512e-05,
+      "loss": 0.7586,
+      "step": 38405
+    },
+    {
+      "epoch": 1.4787295476419633,
+      "grad_norm": 1.069633960723877,
+      "learning_rate": 3.1767365681713123e-05,
+      "loss": 0.9111,
+      "step": 38410
+    },
+    {
+      "epoch": 1.4789220404234842,
+      "grad_norm": 1.1516032218933105,
+      "learning_rate": 3.17452622069935e-05,
+      "loss": 0.9261,
+      "step": 38415
+    },
+    {
+      "epoch": 1.4791145332050049,
+      "grad_norm": 1.2986743450164795,
+      "learning_rate": 3.172316497371749e-05,
+      "loss": 0.8498,
+      "step": 38420
+    },
+    {
+      "epoch": 1.4793070259865255,
+      "grad_norm": 1.1140365600585938,
+      "learning_rate": 3.170107398390576e-05,
+      "loss": 0.8164,
+      "step": 38425
+    },
+    {
+      "epoch": 1.4794995187680462,
+      "grad_norm": 0.8522436022758484,
+      "learning_rate": 3.167898923957838e-05,
+      "loss": 0.7513,
+      "step": 38430
+    },
+    {
+      "epoch": 1.4796920115495669,
+      "grad_norm": 1.4208568334579468,
+      "learning_rate": 3.1656910742754876e-05,
+      "loss": 0.8429,
+      "step": 38435
+    },
+    {
+      "epoch": 1.4798845043310875,
+      "grad_norm": 1.4376440048217773,
+      "learning_rate": 3.163483849545412e-05,
+      "loss": 0.7375,
+      "step": 38440
+    },
+    {
+      "epoch": 1.4800769971126082,
+      "grad_norm": 1.1318036317825317,
+      "learning_rate": 3.161277249969451e-05,
+      "loss": 0.8024,
+      "step": 38445
+    },
+    {
+      "epoch": 1.480269489894129,
+      "grad_norm": 1.1824818849563599,
+      "learning_rate": 3.159071275749382e-05,
+      "loss": 0.7278,
+      "step": 38450
+    },
+    {
+      "epoch": 1.4804619826756498,
+      "grad_norm": 1.4010226726531982,
+      "learning_rate": 3.1568659270869315e-05,
+      "loss": 0.7994,
+      "step": 38455
+    },
+    {
+      "epoch": 1.4806544754571704,
+      "grad_norm": 1.4055135250091553,
+      "learning_rate": 3.154661204183755e-05,
+      "loss": 0.8096,
+      "step": 38460
+    },
+    {
+      "epoch": 1.480846968238691,
+      "grad_norm": 1.6588884592056274,
+      "learning_rate": 3.152457107241471e-05,
+      "loss": 0.858,
+      "step": 38465
+    },
+    {
+      "epoch": 1.4810394610202118,
+      "grad_norm": 0.7389982342720032,
+      "learning_rate": 3.15025363646162e-05,
+      "loss": 0.8122,
+      "step": 38470
+    },
+    {
+      "epoch": 1.4812319538017324,
+      "grad_norm": 1.2149773836135864,
+      "learning_rate": 3.148050792045699e-05,
+      "loss": 0.8371,
+      "step": 38475
+    },
+    {
+      "epoch": 1.4814244465832531,
+      "grad_norm": 0.945489227771759,
+      "learning_rate": 3.1458485741951425e-05,
+      "loss": 0.9891,
+      "step": 38480
+    },
+    {
+      "epoch": 1.4816169393647738,
+      "grad_norm": 1.013566255569458,
+      "learning_rate": 3.1436469831113334e-05,
+      "loss": 0.8877,
+      "step": 38485
+    },
+    {
+      "epoch": 1.4818094321462945,
+      "grad_norm": 1.1482200622558594,
+      "learning_rate": 3.1414460189955805e-05,
+      "loss": 0.9323,
+      "step": 38490
+    },
+    {
+      "epoch": 1.4820019249278151,
+      "grad_norm": 1.5675556659698486,
+      "learning_rate": 3.139245682049163e-05,
+      "loss": 0.7917,
+      "step": 38495
+    },
+    {
+      "epoch": 1.4821944177093358,
+      "grad_norm": 1.299532413482666,
+      "learning_rate": 3.137045972473277e-05,
+      "loss": 0.7983,
+      "step": 38500
+    },
+    {
+      "epoch": 1.4823869104908565,
+      "grad_norm": 1.051574945449829,
+      "learning_rate": 3.134846890469073e-05,
+      "loss": 0.7102,
+      "step": 38505
+    },
+    {
+      "epoch": 1.4825794032723774,
+      "grad_norm": 1.790773868560791,
+      "learning_rate": 3.132648436237644e-05,
+      "loss": 0.8816,
+      "step": 38510
+    },
+    {
+      "epoch": 1.482771896053898,
+      "grad_norm": 1.0724399089813232,
+      "learning_rate": 3.1304506099800256e-05,
+      "loss": 0.9583,
+      "step": 38515
+    },
+    {
+      "epoch": 1.4829643888354187,
+      "grad_norm": 2.1715545654296875,
+      "learning_rate": 3.1282534118971893e-05,
+      "loss": 0.907,
+      "step": 38520
+    },
+    {
+      "epoch": 1.4831568816169394,
+      "grad_norm": 1.1639543771743774,
+      "learning_rate": 3.126056842190058e-05,
+      "loss": 0.8888,
+      "step": 38525
+    },
+    {
+      "epoch": 1.48334937439846,
+      "grad_norm": 1.2022185325622559,
+      "learning_rate": 3.123860901059493e-05,
+      "loss": 0.7886,
+      "step": 38530
+    },
+    {
+      "epoch": 1.4835418671799807,
+      "grad_norm": 0.8449472188949585,
+      "learning_rate": 3.121665588706297e-05,
+      "loss": 0.7561,
+      "step": 38535
+    },
+    {
+      "epoch": 1.4837343599615014,
+      "grad_norm": 1.3795840740203857,
+      "learning_rate": 3.119470905331223e-05,
+      "loss": 0.9203,
+      "step": 38540
+    },
+    {
+      "epoch": 1.4839268527430223,
+      "grad_norm": 1.0770927667617798,
+      "learning_rate": 3.1172768511349514e-05,
+      "loss": 0.9815,
+      "step": 38545
+    },
+    {
+      "epoch": 1.484119345524543,
+      "grad_norm": 1.2355643510818481,
+      "learning_rate": 3.1150834263181187e-05,
+      "loss": 0.7474,
+      "step": 38550
+    },
+    {
+      "epoch": 1.4843118383060636,
+      "grad_norm": 1.1445075273513794,
+      "learning_rate": 3.112890631081297e-05,
+      "loss": 1.0919,
+      "step": 38555
+    },
+    {
+      "epoch": 1.4845043310875843,
+      "grad_norm": 1.6212882995605469,
+      "learning_rate": 3.1106984656250074e-05,
+      "loss": 0.7342,
+      "step": 38560
+    },
+    {
+      "epoch": 1.484696823869105,
+      "grad_norm": 1.2851066589355469,
+      "learning_rate": 3.1085069301496993e-05,
+      "loss": 0.7242,
+      "step": 38565
+    },
+    {
+      "epoch": 1.4848893166506256,
+      "grad_norm": 1.7971413135528564,
+      "learning_rate": 3.1063160248557876e-05,
+      "loss": 0.7274,
+      "step": 38570
+    },
+    {
+      "epoch": 1.4850818094321463,
+      "grad_norm": 0.9864533543586731,
+      "learning_rate": 3.104125749943605e-05,
+      "loss": 0.7071,
+      "step": 38575
+    },
+    {
+      "epoch": 1.485274302213667,
+      "grad_norm": 1.3613252639770508,
+      "learning_rate": 3.101936105613442e-05,
+      "loss": 0.8422,
+      "step": 38580
+    },
+    {
+      "epoch": 1.4854667949951876,
+      "grad_norm": 1.0878537893295288,
+      "learning_rate": 3.099747092065527e-05,
+      "loss": 0.8569,
+      "step": 38585
+    },
+    {
+      "epoch": 1.4856592877767083,
+      "grad_norm": 2.1633172035217285,
+      "learning_rate": 3.0975587095000335e-05,
+      "loss": 0.904,
+      "step": 38590
+    },
+    {
+      "epoch": 1.485851780558229,
+      "grad_norm": 1.3097665309906006,
+      "learning_rate": 3.095370958117064e-05,
+      "loss": 0.8018,
+      "step": 38595
+    },
+    {
+      "epoch": 1.4860442733397496,
+      "grad_norm": 1.019307017326355,
+      "learning_rate": 3.093183838116689e-05,
+      "loss": 0.8721,
+      "step": 38600
+    },
+    {
+      "epoch": 1.4862367661212703,
+      "grad_norm": 1.0038857460021973,
+      "learning_rate": 3.090997349698895e-05,
+      "loss": 0.7122,
+      "step": 38605
+    },
+    {
+      "epoch": 1.4864292589027912,
+      "grad_norm": 1.9515637159347534,
+      "learning_rate": 3.0888114930636256e-05,
+      "loss": 0.7896,
+      "step": 38610
+    },
+    {
+      "epoch": 1.4866217516843119,
+      "grad_norm": 1.1651268005371094,
+      "learning_rate": 3.0866262684107626e-05,
+      "loss": 0.7926,
+      "step": 38615
+    },
+    {
+      "epoch": 1.4868142444658325,
+      "grad_norm": 1.1827764511108398,
+      "learning_rate": 3.084441675940134e-05,
+      "loss": 0.8771,
+      "step": 38620
+    },
+    {
+      "epoch": 1.4870067372473532,
+      "grad_norm": 1.2885782718658447,
+      "learning_rate": 3.0822577158514954e-05,
+      "loss": 0.8086,
+      "step": 38625
+    },
+    {
+      "epoch": 1.4871992300288739,
+      "grad_norm": 1.6408069133758545,
+      "learning_rate": 3.0800743883445703e-05,
+      "loss": 0.8421,
+      "step": 38630
+    },
+    {
+      "epoch": 1.4873917228103946,
+      "grad_norm": 1.5079654455184937,
+      "learning_rate": 3.077891693618998e-05,
+      "loss": 0.951,
+      "step": 38635
+    },
+    {
+      "epoch": 1.4875842155919152,
+      "grad_norm": 1.0820412635803223,
+      "learning_rate": 3.075709631874376e-05,
+      "loss": 0.9189,
+      "step": 38640
+    },
+    {
+      "epoch": 1.4877767083734361,
+      "grad_norm": 1.5331305265426636,
+      "learning_rate": 3.073528203310242e-05,
+      "loss": 0.8078,
+      "step": 38645
+    },
+    {
+      "epoch": 1.4879692011549568,
+      "grad_norm": 1.087971568107605,
+      "learning_rate": 3.0713474081260674e-05,
+      "loss": 0.9791,
+      "step": 38650
+    },
+    {
+      "epoch": 1.4881616939364775,
+      "grad_norm": 1.125899076461792,
+      "learning_rate": 3.069167246521273e-05,
+      "loss": 0.855,
+      "step": 38655
+    },
+    {
+      "epoch": 1.4883541867179981,
+      "grad_norm": 1.5379077196121216,
+      "learning_rate": 3.0669877186952226e-05,
+      "loss": 0.8884,
+      "step": 38660
+    },
+    {
+      "epoch": 1.4885466794995188,
+      "grad_norm": 1.4419198036193848,
+      "learning_rate": 3.064808824847217e-05,
+      "loss": 0.9177,
+      "step": 38665
+    },
+    {
+      "epoch": 1.4887391722810395,
+      "grad_norm": 1.242751955986023,
+      "learning_rate": 3.062630565176504e-05,
+      "loss": 0.7661,
+      "step": 38670
+    },
+    {
+      "epoch": 1.4889316650625601,
+      "grad_norm": 0.9697130918502808,
+      "learning_rate": 3.060452939882273e-05,
+      "loss": 0.7302,
+      "step": 38675
+    },
+    {
+      "epoch": 1.4891241578440808,
+      "grad_norm": 1.1777362823486328,
+      "learning_rate": 3.0582759491636445e-05,
+      "loss": 0.8163,
+      "step": 38680
+    },
+    {
+      "epoch": 1.4893166506256015,
+      "grad_norm": 1.943617582321167,
+      "learning_rate": 3.0560995932197015e-05,
+      "loss": 0.8965,
+      "step": 38685
+    },
+    {
+      "epoch": 1.4895091434071221,
+      "grad_norm": 1.0683708190917969,
+      "learning_rate": 3.053923872249448e-05,
+      "loss": 0.9112,
+      "step": 38690
+    },
+    {
+      "epoch": 1.4897016361886428,
+      "grad_norm": 1.1417663097381592,
+      "learning_rate": 3.0517487864518436e-05,
+      "loss": 0.8912,
+      "step": 38695
+    },
+    {
+      "epoch": 1.4898941289701635,
+      "grad_norm": 1.4740487337112427,
+      "learning_rate": 3.0495743360257845e-05,
+      "loss": 0.8111,
+      "step": 38700
+    },
+    {
+      "epoch": 1.4900866217516844,
+      "grad_norm": 2.3702943325042725,
+      "learning_rate": 3.047400521170113e-05,
+      "loss": 0.9583,
+      "step": 38705
+    },
+    {
+      "epoch": 1.490279114533205,
+      "grad_norm": 1.1470707654953003,
+      "learning_rate": 3.0452273420836007e-05,
+      "loss": 0.9308,
+      "step": 38710
+    },
+    {
+      "epoch": 1.4904716073147257,
+      "grad_norm": 1.7966952323913574,
+      "learning_rate": 3.0430547989649827e-05,
+      "loss": 0.9539,
+      "step": 38715
+    },
+    {
+      "epoch": 1.4906641000962464,
+      "grad_norm": 1.2520999908447266,
+      "learning_rate": 3.0408828920129152e-05,
+      "loss": 0.9079,
+      "step": 38720
+    },
+    {
+      "epoch": 1.490856592877767,
+      "grad_norm": 0.930090606212616,
+      "learning_rate": 3.038711621426007e-05,
+      "loss": 0.7571,
+      "step": 38725
+    },
+    {
+      "epoch": 1.4910490856592877,
+      "grad_norm": 1.0837591886520386,
+      "learning_rate": 3.0365409874028074e-05,
+      "loss": 0.756,
+      "step": 38730
+    },
+    {
+      "epoch": 1.4912415784408084,
+      "grad_norm": 1.3901774883270264,
+      "learning_rate": 3.0343709901418084e-05,
+      "loss": 0.8403,
+      "step": 38735
+    },
+    {
+      "epoch": 1.4914340712223293,
+      "grad_norm": 1.2014509439468384,
+      "learning_rate": 3.032201629841437e-05,
+      "loss": 0.9118,
+      "step": 38740
+    },
+    {
+      "epoch": 1.49162656400385,
+      "grad_norm": 1.7022885084152222,
+      "learning_rate": 3.0300329067000677e-05,
+      "loss": 0.7848,
+      "step": 38745
+    },
+    {
+      "epoch": 1.4918190567853706,
+      "grad_norm": 1.2013940811157227,
+      "learning_rate": 3.0278648209160176e-05,
+      "loss": 0.8609,
+      "step": 38750
+    },
+    {
+      "epoch": 1.4920115495668913,
+      "grad_norm": 1.0092146396636963,
+      "learning_rate": 3.0256973726875436e-05,
+      "loss": 0.8774,
+      "step": 38755
+    },
+    {
+      "epoch": 1.492204042348412,
+      "grad_norm": 1.953743577003479,
+      "learning_rate": 3.0235305622128483e-05,
+      "loss": 0.8166,
+      "step": 38760
+    },
+    {
+      "epoch": 1.4923965351299326,
+      "grad_norm": 2.1745972633361816,
+      "learning_rate": 3.0213643896900646e-05,
+      "loss": 0.7682,
+      "step": 38765
+    },
+    {
+      "epoch": 1.4925890279114533,
+      "grad_norm": 1.1667450666427612,
+      "learning_rate": 3.0191988553172778e-05,
+      "loss": 0.8615,
+      "step": 38770
+    },
+    {
+      "epoch": 1.492781520692974,
+      "grad_norm": 1.3305031061172485,
+      "learning_rate": 3.0170339592925125e-05,
+      "loss": 0.8538,
+      "step": 38775
+    },
+    {
+      "epoch": 1.4929740134744947,
+      "grad_norm": 1.1499173641204834,
+      "learning_rate": 3.0148697018137374e-05,
+      "loss": 0.8741,
+      "step": 38780
+    },
+    {
+      "epoch": 1.4931665062560153,
+      "grad_norm": 1.1782283782958984,
+      "learning_rate": 3.0127060830788477e-05,
+      "loss": 0.7251,
+      "step": 38785
+    },
+    {
+      "epoch": 1.493358999037536,
+      "grad_norm": 0.9500432014465332,
+      "learning_rate": 3.010543103285708e-05,
+      "loss": 0.8529,
+      "step": 38790
+    },
+    {
+      "epoch": 1.4935514918190567,
+      "grad_norm": 1.2924103736877441,
+      "learning_rate": 3.008380762632097e-05,
+      "loss": 0.8166,
+      "step": 38795
+    },
+    {
+      "epoch": 1.4937439846005776,
+      "grad_norm": 1.125877022743225,
+      "learning_rate": 3.0062190613157505e-05,
+      "loss": 0.8048,
+      "step": 38800
+    },
+    {
+      "epoch": 1.4939364773820982,
+      "grad_norm": 1.3560032844543457,
+      "learning_rate": 3.00405799953434e-05,
+      "loss": 0.8974,
+      "step": 38805
+    },
+    {
+      "epoch": 1.494128970163619,
+      "grad_norm": 1.8842239379882812,
+      "learning_rate": 3.0018975774854864e-05,
+      "loss": 0.8511,
+      "step": 38810
+    },
+    {
+      "epoch": 1.4943214629451396,
+      "grad_norm": 1.4684277772903442,
+      "learning_rate": 2.9997377953667337e-05,
+      "loss": 0.7795,
+      "step": 38815
+    },
+    {
+      "epoch": 1.4945139557266602,
+      "grad_norm": 1.0859155654907227,
+      "learning_rate": 2.9975786533755935e-05,
+      "loss": 0.8679,
+      "step": 38820
+    },
+    {
+      "epoch": 1.494706448508181,
+      "grad_norm": 1.0024845600128174,
+      "learning_rate": 2.995420151709496e-05,
+      "loss": 0.99,
+      "step": 38825
+    },
+    {
+      "epoch": 1.4948989412897016,
+      "grad_norm": 1.1407462358474731,
+      "learning_rate": 2.9932622905658238e-05,
+      "loss": 0.9279,
+      "step": 38830
+    },
+    {
+      "epoch": 1.4950914340712222,
+      "grad_norm": 1.0193146467208862,
+      "learning_rate": 2.9911050701419007e-05,
+      "loss": 0.7641,
+      "step": 38835
+    },
+    {
+      "epoch": 1.4952839268527431,
+      "grad_norm": 0.9654618501663208,
+      "learning_rate": 2.9889484906349918e-05,
+      "loss": 0.745,
+      "step": 38840
+    },
+    {
+      "epoch": 1.4954764196342638,
+      "grad_norm": 0.9131496548652649,
+      "learning_rate": 2.9867925522422957e-05,
+      "loss": 0.8413,
+      "step": 38845
+    },
+    {
+      "epoch": 1.4956689124157845,
+      "grad_norm": 1.6795357465744019,
+      "learning_rate": 2.9846372551609637e-05,
+      "loss": 0.8643,
+      "step": 38850
+    },
+    {
+      "epoch": 1.4958614051973051,
+      "grad_norm": 0.6970282793045044,
+      "learning_rate": 2.982482599588081e-05,
+      "loss": 0.671,
+      "step": 38855
+    },
+    {
+      "epoch": 1.4960538979788258,
+      "grad_norm": 2.4015212059020996,
+      "learning_rate": 2.9803285857206787e-05,
+      "loss": 1.0025,
+      "step": 38860
+    },
+    {
+      "epoch": 1.4962463907603465,
+      "grad_norm": 0.9597976207733154,
+      "learning_rate": 2.9781752137557296e-05,
+      "loss": 0.8127,
+      "step": 38865
+    },
+    {
+      "epoch": 1.4964388835418672,
+      "grad_norm": 1.1318256855010986,
+      "learning_rate": 2.9760224838901372e-05,
+      "loss": 0.7509,
+      "step": 38870
+    },
+    {
+      "epoch": 1.4966313763233878,
+      "grad_norm": 1.4549952745437622,
+      "learning_rate": 2.9738703963207604e-05,
+      "loss": 0.8471,
+      "step": 38875
+    },
+    {
+      "epoch": 1.4968238691049085,
+      "grad_norm": 1.4533936977386475,
+      "learning_rate": 2.9717189512443922e-05,
+      "loss": 0.9046,
+      "step": 38880
+    },
+    {
+      "epoch": 1.4970163618864292,
+      "grad_norm": 1.5385301113128662,
+      "learning_rate": 2.9695681488577697e-05,
+      "loss": 0.8941,
+      "step": 38885
+    },
+    {
+      "epoch": 1.4972088546679498,
+      "grad_norm": 1.1736408472061157,
+      "learning_rate": 2.9674179893575626e-05,
+      "loss": 0.7051,
+      "step": 38890
+    },
+    {
+      "epoch": 1.4974013474494705,
+      "grad_norm": 1.6953740119934082,
+      "learning_rate": 2.965268472940399e-05,
+      "loss": 0.8457,
+      "step": 38895
+    },
+    {
+      "epoch": 1.4975938402309914,
+      "grad_norm": 0.6899955868721008,
+      "learning_rate": 2.9631195998028305e-05,
+      "loss": 0.7057,
+      "step": 38900
+    },
+    {
+      "epoch": 1.497786333012512,
+      "grad_norm": 1.721205472946167,
+      "learning_rate": 2.960971370141359e-05,
+      "loss": 0.9267,
+      "step": 38905
+    },
+    {
+      "epoch": 1.4979788257940327,
+      "grad_norm": 1.408977746963501,
+      "learning_rate": 2.958823784152426e-05,
+      "loss": 0.6861,
+      "step": 38910
+    },
+    {
+      "epoch": 1.4981713185755534,
+      "grad_norm": 1.3313335180282593,
+      "learning_rate": 2.9566768420324188e-05,
+      "loss": 0.8315,
+      "step": 38915
+    },
+    {
+      "epoch": 1.498363811357074,
+      "grad_norm": 1.7257198095321655,
+      "learning_rate": 2.9545305439776494e-05,
+      "loss": 0.7739,
+      "step": 38920
+    },
+    {
+      "epoch": 1.4985563041385948,
+      "grad_norm": 1.2223293781280518,
+      "learning_rate": 2.9523848901843953e-05,
+      "loss": 0.8514,
+      "step": 38925
+    },
+    {
+      "epoch": 1.4987487969201154,
+      "grad_norm": 0.9924798607826233,
+      "learning_rate": 2.950239880848852e-05,
+      "loss": 0.8442,
+      "step": 38930
+    },
+    {
+      "epoch": 1.4989412897016363,
+      "grad_norm": 1.178801417350769,
+      "learning_rate": 2.9480955161671774e-05,
+      "loss": 0.9034,
+      "step": 38935
+    },
+    {
+      "epoch": 1.499133782483157,
+      "grad_norm": 0.8374056220054626,
+      "learning_rate": 2.94595179633545e-05,
+      "loss": 0.7436,
+      "step": 38940
+    },
+    {
+      "epoch": 1.4993262752646777,
+      "grad_norm": 1.187247633934021,
+      "learning_rate": 2.9438087215497013e-05,
+      "loss": 0.8352,
+      "step": 38945
+    },
+    {
+      "epoch": 1.4995187680461983,
+      "grad_norm": 0.9191497564315796,
+      "learning_rate": 2.9416662920059024e-05,
+      "loss": 0.7596,
+      "step": 38950
+    },
+    {
+      "epoch": 1.499711260827719,
+      "grad_norm": 0.9378131628036499,
+      "learning_rate": 2.9395245078999665e-05,
+      "loss": 0.7416,
+      "step": 38955
+    },
+    {
+      "epoch": 1.4999037536092397,
+      "grad_norm": 1.5277785062789917,
+      "learning_rate": 2.937383369427741e-05,
+      "loss": 0.6751,
+      "step": 38960
+    },
+    {
+      "epoch": 1.5000962463907603,
+      "grad_norm": 0.8396899104118347,
+      "learning_rate": 2.9352428767850204e-05,
+      "loss": 0.8376,
+      "step": 38965
+    },
+    {
+      "epoch": 1.500288739172281,
+      "grad_norm": 1.3302197456359863,
+      "learning_rate": 2.9331030301675388e-05,
+      "loss": 0.875,
+      "step": 38970
+    },
+    {
+      "epoch": 1.5004812319538017,
+      "grad_norm": 1.8585106134414673,
+      "learning_rate": 2.9309638297709708e-05,
+      "loss": 0.804,
+      "step": 38975
+    },
+    {
+      "epoch": 1.5006737247353223,
+      "grad_norm": 1.9096689224243164,
+      "learning_rate": 2.928825275790936e-05,
+      "loss": 0.8378,
+      "step": 38980
+    },
+    {
+      "epoch": 1.500866217516843,
+      "grad_norm": 1.947161078453064,
+      "learning_rate": 2.9266873684229846e-05,
+      "loss": 0.9819,
+      "step": 38985
+    },
+    {
+      "epoch": 1.5010587102983637,
+      "grad_norm": 1.3019959926605225,
+      "learning_rate": 2.924550107862617e-05,
+      "loss": 0.7941,
+      "step": 38990
+    },
+    {
+      "epoch": 1.5012512030798844,
+      "grad_norm": 1.5726149082183838,
+      "learning_rate": 2.9224134943052717e-05,
+      "loss": 0.8298,
+      "step": 38995
+    },
+    {
+      "epoch": 1.5014436958614052,
+      "grad_norm": 1.7521274089813232,
+      "learning_rate": 2.92027752794633e-05,
+      "loss": 0.7795,
+      "step": 39000
+    },
+    {
+      "epoch": 1.501636188642926,
+      "grad_norm": 1.5771510601043701,
+      "learning_rate": 2.918142208981104e-05,
+      "loss": 0.7677,
+      "step": 39005
+    },
+    {
+      "epoch": 1.5018286814244466,
+      "grad_norm": 1.1696386337280273,
+      "learning_rate": 2.9160075376048656e-05,
+      "loss": 0.735,
+      "step": 39010
+    },
+    {
+      "epoch": 1.5020211742059673,
+      "grad_norm": 1.4408166408538818,
+      "learning_rate": 2.913873514012807e-05,
+      "loss": 0.8714,
+      "step": 39015
+    },
+    {
+      "epoch": 1.502213666987488,
+      "grad_norm": 0.9149922728538513,
+      "learning_rate": 2.9117401384000753e-05,
+      "loss": 0.6836,
+      "step": 39020
+    },
+    {
+      "epoch": 1.5024061597690088,
+      "grad_norm": 1.5898869037628174,
+      "learning_rate": 2.9096074109617522e-05,
+      "loss": 0.7895,
+      "step": 39025
+    },
+    {
+      "epoch": 1.5025986525505295,
+      "grad_norm": 1.2088052034378052,
+      "learning_rate": 2.907475331892865e-05,
+      "loss": 0.7443,
+      "step": 39030
+    },
+    {
+      "epoch": 1.5027911453320502,
+      "grad_norm": 1.8943687677383423,
+      "learning_rate": 2.905343901388369e-05,
+      "loss": 0.9301,
+      "step": 39035
+    },
+    {
+      "epoch": 1.5029836381135708,
+      "grad_norm": 1.088670253753662,
+      "learning_rate": 2.903213119643181e-05,
+      "loss": 0.5987,
+      "step": 39040
+    },
+    {
+      "epoch": 1.5031761308950915,
+      "grad_norm": 1.1137605905532837,
+      "learning_rate": 2.9010829868521393e-05,
+      "loss": 0.8678,
+      "step": 39045
+    },
+    {
+      "epoch": 1.5033686236766122,
+      "grad_norm": 1.1481714248657227,
+      "learning_rate": 2.8989535032100312e-05,
+      "loss": 0.8033,
+      "step": 39050
+    },
+    {
+      "epoch": 1.5035611164581328,
+      "grad_norm": 1.2706315517425537,
+      "learning_rate": 2.8968246689115862e-05,
+      "loss": 0.7202,
+      "step": 39055
+    },
+    {
+      "epoch": 1.5037536092396535,
+      "grad_norm": 1.0178372859954834,
+      "learning_rate": 2.8946964841514735e-05,
+      "loss": 0.8806,
+      "step": 39060
+    },
+    {
+      "epoch": 1.5039461020211742,
+      "grad_norm": 1.0535964965820312,
+      "learning_rate": 2.8925689491242958e-05,
+      "loss": 0.8236,
+      "step": 39065
+    },
+    {
+      "epoch": 1.5041385948026949,
+      "grad_norm": 0.9003209471702576,
+      "learning_rate": 2.8904420640246054e-05,
+      "loss": 0.7783,
+      "step": 39070
+    },
+    {
+      "epoch": 1.5043310875842155,
+      "grad_norm": 1.5490949153900146,
+      "learning_rate": 2.8883158290468916e-05,
+      "loss": 0.8359,
+      "step": 39075
+    },
+    {
+      "epoch": 1.5045235803657362,
+      "grad_norm": 0.9985781908035278,
+      "learning_rate": 2.886190244385585e-05,
+      "loss": 1.091,
+      "step": 39080
+    },
+    {
+      "epoch": 1.5047160731472569,
+      "grad_norm": 1.7629387378692627,
+      "learning_rate": 2.884065310235059e-05,
+      "loss": 0.9146,
+      "step": 39085
+    },
+    {
+      "epoch": 1.5049085659287775,
+      "grad_norm": 1.2450439929962158,
+      "learning_rate": 2.8819410267896173e-05,
+      "loss": 0.7949,
+      "step": 39090
+    },
+    {
+      "epoch": 1.5051010587102982,
+      "grad_norm": 1.3784174919128418,
+      "learning_rate": 2.879817394243517e-05,
+      "loss": 0.8738,
+      "step": 39095
+    },
+    {
+      "epoch": 1.505293551491819,
+      "grad_norm": 1.2022943496704102,
+      "learning_rate": 2.877694412790949e-05,
+      "loss": 0.8506,
+      "step": 39100
+    },
+    {
+      "epoch": 1.5054860442733398,
+      "grad_norm": 1.1159878969192505,
+      "learning_rate": 2.8755720826260492e-05,
+      "loss": 0.7559,
+      "step": 39105
+    },
+    {
+      "epoch": 1.5056785370548604,
+      "grad_norm": 1.8302503824234009,
+      "learning_rate": 2.8734504039428813e-05,
+      "loss": 0.839,
+      "step": 39110
+    },
+    {
+      "epoch": 1.505871029836381,
+      "grad_norm": 1.0273211002349854,
+      "learning_rate": 2.871329376935471e-05,
+      "loss": 0.7604,
+      "step": 39115
+    },
+    {
+      "epoch": 1.506063522617902,
+      "grad_norm": 2.0029349327087402,
+      "learning_rate": 2.8692090017977626e-05,
+      "loss": 0.8672,
+      "step": 39120
+    },
+    {
+      "epoch": 1.5062560153994227,
+      "grad_norm": 1.7396565675735474,
+      "learning_rate": 2.8670892787236536e-05,
+      "loss": 0.8613,
+      "step": 39125
+    },
+    {
+      "epoch": 1.5064485081809433,
+      "grad_norm": 1.0118446350097656,
+      "learning_rate": 2.8649702079069797e-05,
+      "loss": 0.7605,
+      "step": 39130
+    },
+    {
+      "epoch": 1.506641000962464,
+      "grad_norm": 1.3765640258789062,
+      "learning_rate": 2.8628517895415173e-05,
+      "loss": 0.7897,
+      "step": 39135
+    },
+    {
+      "epoch": 1.5068334937439847,
+      "grad_norm": 0.9701485633850098,
+      "learning_rate": 2.8607340238209747e-05,
+      "loss": 0.7784,
+      "step": 39140
+    },
+    {
+      "epoch": 1.5070259865255053,
+      "grad_norm": 1.591834306716919,
+      "learning_rate": 2.858616910939017e-05,
+      "loss": 0.837,
+      "step": 39145
+    },
+    {
+      "epoch": 1.507218479307026,
+      "grad_norm": 1.928481936454773,
+      "learning_rate": 2.8565004510892336e-05,
+      "loss": 0.8828,
+      "step": 39150
+    },
+    {
+      "epoch": 1.5074109720885467,
+      "grad_norm": 1.1933850049972534,
+      "learning_rate": 2.854384644465161e-05,
+      "loss": 0.7229,
+      "step": 39155
+    },
+    {
+      "epoch": 1.5076034648700674,
+      "grad_norm": 1.4722340106964111,
+      "learning_rate": 2.852269491260279e-05,
+      "loss": 0.7789,
+      "step": 39160
+    },
+    {
+      "epoch": 1.507795957651588,
+      "grad_norm": 1.0103994607925415,
+      "learning_rate": 2.8501549916680047e-05,
+      "loss": 0.8624,
+      "step": 39165
+    },
+    {
+      "epoch": 1.5079884504331087,
+      "grad_norm": 1.9095921516418457,
+      "learning_rate": 2.848041145881687e-05,
+      "loss": 0.9677,
+      "step": 39170
+    },
+    {
+      "epoch": 1.5081809432146294,
+      "grad_norm": 0.9631972312927246,
+      "learning_rate": 2.845927954094635e-05,
+      "loss": 0.6126,
+      "step": 39175
+    },
+    {
+      "epoch": 1.50837343599615,
+      "grad_norm": 0.9686654210090637,
+      "learning_rate": 2.8438154165000774e-05,
+      "loss": 0.7955,
+      "step": 39180
+    },
+    {
+      "epoch": 1.5085659287776707,
+      "grad_norm": 1.5519697666168213,
+      "learning_rate": 2.8417035332911946e-05,
+      "loss": 0.8444,
+      "step": 39185
+    },
+    {
+      "epoch": 1.5087584215591914,
+      "grad_norm": 0.9993698596954346,
+      "learning_rate": 2.839592304661107e-05,
+      "loss": 0.8372,
+      "step": 39190
+    },
+    {
+      "epoch": 1.5089509143407123,
+      "grad_norm": 2.2285327911376953,
+      "learning_rate": 2.8374817308028645e-05,
+      "loss": 0.8041,
+      "step": 39195
+    },
+    {
+      "epoch": 1.509143407122233,
+      "grad_norm": 1.6649866104125977,
+      "learning_rate": 2.8353718119094762e-05,
+      "loss": 0.8547,
+      "step": 39200
+    },
+    {
+      "epoch": 1.5093358999037536,
+      "grad_norm": 1.4915626049041748,
+      "learning_rate": 2.8332625481738705e-05,
+      "loss": 0.9186,
+      "step": 39205
+    },
+    {
+      "epoch": 1.5095283926852743,
+      "grad_norm": 0.9645226001739502,
+      "learning_rate": 2.8311539397889307e-05,
+      "loss": 0.6819,
+      "step": 39210
+    },
+    {
+      "epoch": 1.509720885466795,
+      "grad_norm": 1.2806873321533203,
+      "learning_rate": 2.829045986947474e-05,
+      "loss": 0.8868,
+      "step": 39215
+    },
+    {
+      "epoch": 1.5099133782483158,
+      "grad_norm": 1.3241313695907593,
+      "learning_rate": 2.82693868984226e-05,
+      "loss": 0.7592,
+      "step": 39220
+    },
+    {
+      "epoch": 1.5101058710298365,
+      "grad_norm": 0.9667055606842041,
+      "learning_rate": 2.824832048665981e-05,
+      "loss": 0.8439,
+      "step": 39225
+    },
+    {
+      "epoch": 1.5102983638113572,
+      "grad_norm": 1.5384224653244019,
+      "learning_rate": 2.822726063611285e-05,
+      "loss": 0.7319,
+      "step": 39230
+    },
+    {
+      "epoch": 1.5104908565928779,
+      "grad_norm": 1.0840857028961182,
+      "learning_rate": 2.8206207348707435e-05,
+      "loss": 0.7401,
+      "step": 39235
+    },
+    {
+      "epoch": 1.5106833493743985,
+      "grad_norm": 0.7421225905418396,
+      "learning_rate": 2.8185160626368755e-05,
+      "loss": 0.7408,
+      "step": 39240
+    },
+    {
+      "epoch": 1.5108758421559192,
+      "grad_norm": 1.7684946060180664,
+      "learning_rate": 2.8164120471021417e-05,
+      "loss": 0.8469,
+      "step": 39245
+    },
+    {
+      "epoch": 1.5110683349374399,
+      "grad_norm": 1.4627081155776978,
+      "learning_rate": 2.814308688458942e-05,
+      "loss": 0.8621,
+      "step": 39250
+    },
+    {
+      "epoch": 1.5112608277189605,
+      "grad_norm": 0.941478967666626,
+      "learning_rate": 2.8122059868996055e-05,
+      "loss": 0.7465,
+      "step": 39255
+    },
+    {
+      "epoch": 1.5114533205004812,
+      "grad_norm": 1.5532718896865845,
+      "learning_rate": 2.810103942616423e-05,
+      "loss": 0.7821,
+      "step": 39260
+    },
+    {
+      "epoch": 1.5116458132820019,
+      "grad_norm": 0.9734495282173157,
+      "learning_rate": 2.8080025558016033e-05,
+      "loss": 0.8102,
+      "step": 39265
+    },
+    {
+      "epoch": 1.5118383060635225,
+      "grad_norm": 1.1670575141906738,
+      "learning_rate": 2.8059018266473068e-05,
+      "loss": 0.8069,
+      "step": 39270
+    },
+    {
+      "epoch": 1.5120307988450432,
+      "grad_norm": 1.0015208721160889,
+      "learning_rate": 2.803801755345633e-05,
+      "loss": 0.8015,
+      "step": 39275
+    },
+    {
+      "epoch": 1.5122232916265639,
+      "grad_norm": 1.1417772769927979,
+      "learning_rate": 2.8017023420886202e-05,
+      "loss": 0.7996,
+      "step": 39280
+    },
+    {
+      "epoch": 1.5124157844080846,
+      "grad_norm": 1.6754858493804932,
+      "learning_rate": 2.7996035870682402e-05,
+      "loss": 0.7731,
+      "step": 39285
+    },
+    {
+      "epoch": 1.5126082771896054,
+      "grad_norm": 2.357158899307251,
+      "learning_rate": 2.797505490476415e-05,
+      "loss": 0.817,
+      "step": 39290
+    },
+    {
+      "epoch": 1.5128007699711261,
+      "grad_norm": 0.9653787612915039,
+      "learning_rate": 2.7954080525050007e-05,
+      "loss": 0.7779,
+      "step": 39295
+    },
+    {
+      "epoch": 1.5129932627526468,
+      "grad_norm": 0.9732620716094971,
+      "learning_rate": 2.7933112733457933e-05,
+      "loss": 0.8127,
+      "step": 39300
+    },
+    {
+      "epoch": 1.5131857555341675,
+      "grad_norm": 0.9965852499008179,
+      "learning_rate": 2.791215153190533e-05,
+      "loss": 0.9347,
+      "step": 39305
+    },
+    {
+      "epoch": 1.5133782483156881,
+      "grad_norm": 1.4929567575454712,
+      "learning_rate": 2.7891196922308905e-05,
+      "loss": 0.8412,
+      "step": 39310
+    },
+    {
+      "epoch": 1.513570741097209,
+      "grad_norm": 1.590889811515808,
+      "learning_rate": 2.787024890658484e-05,
+      "loss": 0.792,
+      "step": 39315
+    },
+    {
+      "epoch": 1.5137632338787297,
+      "grad_norm": 1.5452625751495361,
+      "learning_rate": 2.7849307486648702e-05,
+      "loss": 0.8094,
+      "step": 39320
+    },
+    {
+      "epoch": 1.5139557266602504,
+      "grad_norm": 1.8608287572860718,
+      "learning_rate": 2.782837266441546e-05,
+      "loss": 0.8371,
+      "step": 39325
+    },
+    {
+      "epoch": 1.514148219441771,
+      "grad_norm": 1.141189455986023,
+      "learning_rate": 2.780744444179939e-05,
+      "loss": 0.8949,
+      "step": 39330
+    },
+    {
+      "epoch": 1.5143407122232917,
+      "grad_norm": 1.3165732622146606,
+      "learning_rate": 2.7786522820714355e-05,
+      "loss": 0.8983,
+      "step": 39335
+    },
+    {
+      "epoch": 1.5145332050048124,
+      "grad_norm": 1.809031367301941,
+      "learning_rate": 2.776560780307341e-05,
+      "loss": 0.7893,
+      "step": 39340
+    },
+    {
+      "epoch": 1.514725697786333,
+      "grad_norm": 0.9109099507331848,
+      "learning_rate": 2.774469939078912e-05,
+      "loss": 0.7076,
+      "step": 39345
+    },
+    {
+      "epoch": 1.5149181905678537,
+      "grad_norm": 1.9604289531707764,
+      "learning_rate": 2.7723797585773426e-05,
+      "loss": 0.916,
+      "step": 39350
+    },
+    {
+      "epoch": 1.5151106833493744,
+      "grad_norm": 1.2984347343444824,
+      "learning_rate": 2.770290238993768e-05,
+      "loss": 0.9168,
+      "step": 39355
+    },
+    {
+      "epoch": 1.515303176130895,
+      "grad_norm": 2.4720041751861572,
+      "learning_rate": 2.768201380519253e-05,
+      "loss": 0.8866,
+      "step": 39360
+    },
+    {
+      "epoch": 1.5154956689124157,
+      "grad_norm": 0.9475586414337158,
+      "learning_rate": 2.7661131833448216e-05,
+      "loss": 0.8357,
+      "step": 39365
+    },
+    {
+      "epoch": 1.5156881616939364,
+      "grad_norm": 1.0659955739974976,
+      "learning_rate": 2.7640256476614157e-05,
+      "loss": 0.8456,
+      "step": 39370
+    },
+    {
+      "epoch": 1.515880654475457,
+      "grad_norm": 1.2105602025985718,
+      "learning_rate": 2.7619387736599312e-05,
+      "loss": 0.7651,
+      "step": 39375
+    },
+    {
+      "epoch": 1.5160731472569777,
+      "grad_norm": 1.759903907775879,
+      "learning_rate": 2.7598525615311975e-05,
+      "loss": 0.7607,
+      "step": 39380
+    },
+    {
+      "epoch": 1.5162656400384984,
+      "grad_norm": 0.6724603176116943,
+      "learning_rate": 2.7577670114659892e-05,
+      "loss": 0.7695,
+      "step": 39385
+    },
+    {
+      "epoch": 1.5164581328200193,
+      "grad_norm": 1.3016246557235718,
+      "learning_rate": 2.755682123655009e-05,
+      "loss": 0.8609,
+      "step": 39390
+    },
+    {
+      "epoch": 1.51665062560154,
+      "grad_norm": 1.1983976364135742,
+      "learning_rate": 2.75359789828891e-05,
+      "loss": 0.8225,
+      "step": 39395
+    },
+    {
+      "epoch": 1.5168431183830606,
+      "grad_norm": 1.0239381790161133,
+      "learning_rate": 2.7515143355582786e-05,
+      "loss": 0.6805,
+      "step": 39400
+    },
+    {
+      "epoch": 1.5170356111645813,
+      "grad_norm": 1.6102731227874756,
+      "learning_rate": 2.7494314356536456e-05,
+      "loss": 0.8922,
+      "step": 39405
+    },
+    {
+      "epoch": 1.517228103946102,
+      "grad_norm": 1.7071436643600464,
+      "learning_rate": 2.7473491987654798e-05,
+      "loss": 0.9025,
+      "step": 39410
+    },
+    {
+      "epoch": 1.5174205967276229,
+      "grad_norm": 1.8176323175430298,
+      "learning_rate": 2.7452676250841824e-05,
+      "loss": 0.9699,
+      "step": 39415
+    },
+    {
+      "epoch": 1.5176130895091435,
+      "grad_norm": 2.1587765216827393,
+      "learning_rate": 2.7431867148001024e-05,
+      "loss": 1.0543,
+      "step": 39420
+    },
+    {
+      "epoch": 1.5178055822906642,
+      "grad_norm": 1.8511626720428467,
+      "learning_rate": 2.741106468103526e-05,
+      "loss": 0.7427,
+      "step": 39425
+    },
+    {
+      "epoch": 1.5179980750721849,
+      "grad_norm": 1.2232860326766968,
+      "learning_rate": 2.739026885184679e-05,
+      "loss": 0.834,
+      "step": 39430
+    },
+    {
+      "epoch": 1.5181905678537055,
+      "grad_norm": 0.9527449011802673,
+      "learning_rate": 2.7369479662337195e-05,
+      "loss": 0.8507,
+      "step": 39435
+    },
+    {
+      "epoch": 1.5183830606352262,
+      "grad_norm": 1.3037556409835815,
+      "learning_rate": 2.7348697114407595e-05,
+      "loss": 0.8492,
+      "step": 39440
+    },
+    {
+      "epoch": 1.5185755534167469,
+      "grad_norm": 0.8813561797142029,
+      "learning_rate": 2.732792120995832e-05,
+      "loss": 0.8644,
+      "step": 39445
+    },
+    {
+      "epoch": 1.5187680461982676,
+      "grad_norm": 1.0147926807403564,
+      "learning_rate": 2.7307151950889308e-05,
+      "loss": 0.7337,
+      "step": 39450
+    },
+    {
+      "epoch": 1.5189605389797882,
+      "grad_norm": 1.2371786832809448,
+      "learning_rate": 2.728638933909966e-05,
+      "loss": 0.7856,
+      "step": 39455
+    },
+    {
+      "epoch": 1.519153031761309,
+      "grad_norm": 1.3520861864089966,
+      "learning_rate": 2.726563337648803e-05,
+      "loss": 0.8224,
+      "step": 39460
+    },
+    {
+      "epoch": 1.5193455245428296,
+      "grad_norm": 1.1333503723144531,
+      "learning_rate": 2.7244884064952414e-05,
+      "loss": 0.7827,
+      "step": 39465
+    },
+    {
+      "epoch": 1.5195380173243502,
+      "grad_norm": 1.4535999298095703,
+      "learning_rate": 2.7224141406390214e-05,
+      "loss": 0.9174,
+      "step": 39470
+    },
+    {
+      "epoch": 1.519730510105871,
+      "grad_norm": 1.9915587902069092,
+      "learning_rate": 2.7203405402698133e-05,
+      "loss": 0.9327,
+      "step": 39475
+    },
+    {
+      "epoch": 1.5199230028873916,
+      "grad_norm": 1.0400042533874512,
+      "learning_rate": 2.7182676055772448e-05,
+      "loss": 0.7821,
+      "step": 39480
+    },
+    {
+      "epoch": 1.5201154956689125,
+      "grad_norm": 1.4389541149139404,
+      "learning_rate": 2.7161953367508643e-05,
+      "loss": 0.826,
+      "step": 39485
+    },
+    {
+      "epoch": 1.5203079884504331,
+      "grad_norm": 1.9297611713409424,
+      "learning_rate": 2.7141237339801685e-05,
+      "loss": 0.9919,
+      "step": 39490
+    },
+    {
+      "epoch": 1.5205004812319538,
+      "grad_norm": 1.1149492263793945,
+      "learning_rate": 2.712052797454594e-05,
+      "loss": 0.7048,
+      "step": 39495
+    },
+    {
+      "epoch": 1.5206929740134745,
+      "grad_norm": 1.5707048177719116,
+      "learning_rate": 2.709982527363515e-05,
+      "loss": 0.8483,
+      "step": 39500
+    },
+    {
+      "epoch": 1.5208854667949951,
+      "grad_norm": 1.289869785308838,
+      "learning_rate": 2.7079129238962386e-05,
+      "loss": 0.6966,
+      "step": 39505
+    },
+    {
+      "epoch": 1.521077959576516,
+      "grad_norm": 1.144384741783142,
+      "learning_rate": 2.7058439872420204e-05,
+      "loss": 0.7713,
+      "step": 39510
+    },
+    {
+      "epoch": 1.5212704523580367,
+      "grad_norm": 0.8240516781806946,
+      "learning_rate": 2.7037757175900503e-05,
+      "loss": 0.7334,
+      "step": 39515
+    },
+    {
+      "epoch": 1.5214629451395574,
+      "grad_norm": 2.318265199661255,
+      "learning_rate": 2.7017081151294576e-05,
+      "loss": 0.8006,
+      "step": 39520
+    },
+    {
+      "epoch": 1.521655437921078,
+      "grad_norm": 1.3979312181472778,
+      "learning_rate": 2.6996411800493137e-05,
+      "loss": 0.7323,
+      "step": 39525
+    },
+    {
+      "epoch": 1.5218479307025987,
+      "grad_norm": 0.9698436856269836,
+      "learning_rate": 2.6975749125386207e-05,
+      "loss": 0.952,
+      "step": 39530
+    },
+    {
+      "epoch": 1.5220404234841194,
+      "grad_norm": 1.6334707736968994,
+      "learning_rate": 2.6955093127863272e-05,
+      "loss": 0.7626,
+      "step": 39535
+    },
+    {
+      "epoch": 1.52223291626564,
+      "grad_norm": 1.1066991090774536,
+      "learning_rate": 2.6934443809813203e-05,
+      "loss": 0.7948,
+      "step": 39540
+    },
+    {
+      "epoch": 1.5224254090471607,
+      "grad_norm": 1.206701636314392,
+      "learning_rate": 2.6913801173124253e-05,
+      "loss": 0.6289,
+      "step": 39545
+    },
+    {
+      "epoch": 1.5226179018286814,
+      "grad_norm": 1.3689944744110107,
+      "learning_rate": 2.689316521968398e-05,
+      "loss": 0.7864,
+      "step": 39550
+    },
+    {
+      "epoch": 1.522810394610202,
+      "grad_norm": 2.103961944580078,
+      "learning_rate": 2.687253595137952e-05,
+      "loss": 0.6763,
+      "step": 39555
+    },
+    {
+      "epoch": 1.5230028873917227,
+      "grad_norm": 1.0298928022384644,
+      "learning_rate": 2.6851913370097194e-05,
+      "loss": 0.7179,
+      "step": 39560
+    },
+    {
+      "epoch": 1.5231953801732434,
+      "grad_norm": 1.2706328630447388,
+      "learning_rate": 2.6831297477722815e-05,
+      "loss": 0.6922,
+      "step": 39565
+    },
+    {
+      "epoch": 1.523387872954764,
+      "grad_norm": 0.9494262933731079,
+      "learning_rate": 2.6810688276141592e-05,
+      "loss": 0.8778,
+      "step": 39570
+    },
+    {
+      "epoch": 1.5235803657362847,
+      "grad_norm": 0.9296857118606567,
+      "learning_rate": 2.6790085767238126e-05,
+      "loss": 0.8307,
+      "step": 39575
+    },
+    {
+      "epoch": 1.5237728585178054,
+      "grad_norm": 1.7679259777069092,
+      "learning_rate": 2.6769489952896275e-05,
+      "loss": 0.8162,
+      "step": 39580
+    },
+    {
+      "epoch": 1.5239653512993263,
+      "grad_norm": 1.1842856407165527,
+      "learning_rate": 2.6748900834999534e-05,
+      "loss": 0.724,
+      "step": 39585
+    },
+    {
+      "epoch": 1.524157844080847,
+      "grad_norm": 0.9693066477775574,
+      "learning_rate": 2.672831841543053e-05,
+      "loss": 0.8326,
+      "step": 39590
+    },
+    {
+      "epoch": 1.5243503368623676,
+      "grad_norm": 1.6108996868133545,
+      "learning_rate": 2.6707742696071425e-05,
+      "loss": 0.7797,
+      "step": 39595
+    },
+    {
+      "epoch": 1.5245428296438883,
+      "grad_norm": 2.4159984588623047,
+      "learning_rate": 2.6687173678803734e-05,
+      "loss": 0.9232,
+      "step": 39600
+    },
+    {
+      "epoch": 1.5247353224254092,
+      "grad_norm": 1.5150139331817627,
+      "learning_rate": 2.6666611365508388e-05,
+      "loss": 0.7605,
+      "step": 39605
+    },
+    {
+      "epoch": 1.5249278152069299,
+      "grad_norm": 1.5143542289733887,
+      "learning_rate": 2.6646055758065613e-05,
+      "loss": 0.9481,
+      "step": 39610
+    },
+    {
+      "epoch": 1.5251203079884506,
+      "grad_norm": 0.9554632306098938,
+      "learning_rate": 2.662550685835511e-05,
+      "loss": 0.9355,
+      "step": 39615
+    },
+    {
+      "epoch": 1.5253128007699712,
+      "grad_norm": 1.936998963356018,
+      "learning_rate": 2.660496466825595e-05,
+      "loss": 0.7781,
+      "step": 39620
+    },
+    {
+      "epoch": 1.525505293551492,
+      "grad_norm": 2.4360251426696777,
+      "learning_rate": 2.6584429189646576e-05,
+      "loss": 0.8181,
+      "step": 39625
+    },
+    {
+      "epoch": 1.5256977863330126,
+      "grad_norm": 1.0304756164550781,
+      "learning_rate": 2.6563900424404843e-05,
+      "loss": 0.6611,
+      "step": 39630
+    },
+    {
+      "epoch": 1.5258902791145332,
+      "grad_norm": 1.5006756782531738,
+      "learning_rate": 2.6543378374407913e-05,
+      "loss": 0.8139,
+      "step": 39635
+    },
+    {
+      "epoch": 1.526082771896054,
+      "grad_norm": 2.475623846054077,
+      "learning_rate": 2.652286304153243e-05,
+      "loss": 0.8565,
+      "step": 39640
+    },
+    {
+      "epoch": 1.5262752646775746,
+      "grad_norm": 1.3540338277816772,
+      "learning_rate": 2.6502354427654375e-05,
+      "loss": 0.9392,
+      "step": 39645
+    },
+    {
+      "epoch": 1.5264677574590952,
+      "grad_norm": 1.5708937644958496,
+      "learning_rate": 2.6481852534649165e-05,
+      "loss": 0.7956,
+      "step": 39650
+    },
+    {
+      "epoch": 1.526660250240616,
+      "grad_norm": 1.0419323444366455,
+      "learning_rate": 2.646135736439147e-05,
+      "loss": 0.7812,
+      "step": 39655
+    },
+    {
+      "epoch": 1.5268527430221366,
+      "grad_norm": 1.3944579362869263,
+      "learning_rate": 2.6440868918755556e-05,
+      "loss": 0.9234,
+      "step": 39660
+    },
+    {
+      "epoch": 1.5270452358036573,
+      "grad_norm": 1.2345746755599976,
+      "learning_rate": 2.642038719961486e-05,
+      "loss": 0.7003,
+      "step": 39665
+    },
+    {
+      "epoch": 1.527237728585178,
+      "grad_norm": 1.7590235471725464,
+      "learning_rate": 2.6399912208842337e-05,
+      "loss": 1.0322,
+      "step": 39670
+    },
+    {
+      "epoch": 1.5274302213666986,
+      "grad_norm": 1.5209788084030151,
+      "learning_rate": 2.637944394831028e-05,
+      "loss": 0.7132,
+      "step": 39675
+    },
+    {
+      "epoch": 1.5276227141482195,
+      "grad_norm": 1.2953201532363892,
+      "learning_rate": 2.635898241989042e-05,
+      "loss": 0.8862,
+      "step": 39680
+    },
+    {
+      "epoch": 1.5278152069297402,
+      "grad_norm": 0.8863399028778076,
+      "learning_rate": 2.6338527625453725e-05,
+      "loss": 0.8335,
+      "step": 39685
+    },
+    {
+      "epoch": 1.5280076997112608,
+      "grad_norm": 0.9998905062675476,
+      "learning_rate": 2.631807956687078e-05,
+      "loss": 0.8523,
+      "step": 39690
+    },
+    {
+      "epoch": 1.5282001924927815,
+      "grad_norm": 1.509939432144165,
+      "learning_rate": 2.62976382460113e-05,
+      "loss": 0.7919,
+      "step": 39695
+    },
+    {
+      "epoch": 1.5283926852743022,
+      "grad_norm": 1.2834575176239014,
+      "learning_rate": 2.627720366474463e-05,
+      "loss": 0.8921,
+      "step": 39700
+    },
+    {
+      "epoch": 1.528585178055823,
+      "grad_norm": 0.9761720299720764,
+      "learning_rate": 2.6256775824939283e-05,
+      "loss": 0.7869,
+      "step": 39705
+    },
+    {
+      "epoch": 1.5287776708373437,
+      "grad_norm": 2.2494795322418213,
+      "learning_rate": 2.6236354728463286e-05,
+      "loss": 0.765,
+      "step": 39710
+    },
+    {
+      "epoch": 1.5289701636188644,
+      "grad_norm": 1.8769516944885254,
+      "learning_rate": 2.6215940377184002e-05,
+      "loss": 0.6916,
+      "step": 39715
+    },
+    {
+      "epoch": 1.529162656400385,
+      "grad_norm": 0.8097474575042725,
+      "learning_rate": 2.6195532772968235e-05,
+      "loss": 0.7983,
+      "step": 39720
+    },
+    {
+      "epoch": 1.5293551491819057,
+      "grad_norm": 1.8454664945602417,
+      "learning_rate": 2.617513191768205e-05,
+      "loss": 0.8183,
+      "step": 39725
+    },
+    {
+      "epoch": 1.5295476419634264,
+      "grad_norm": 1.6850463151931763,
+      "learning_rate": 2.615473781319101e-05,
+      "loss": 1.0079,
+      "step": 39730
+    },
+    {
+      "epoch": 1.529740134744947,
+      "grad_norm": 1.442956566810608,
+      "learning_rate": 2.6134350461360013e-05,
+      "loss": 0.7613,
+      "step": 39735
+    },
+    {
+      "epoch": 1.5299326275264677,
+      "grad_norm": 1.101926326751709,
+      "learning_rate": 2.6113969864053356e-05,
+      "loss": 0.7822,
+      "step": 39740
+    },
+    {
+      "epoch": 1.5301251203079884,
+      "grad_norm": 0.8331781625747681,
+      "learning_rate": 2.6093596023134736e-05,
+      "loss": 0.8167,
+      "step": 39745
+    },
+    {
+      "epoch": 1.530317613089509,
+      "grad_norm": 1.0116935968399048,
+      "learning_rate": 2.6073228940467142e-05,
+      "loss": 0.9155,
+      "step": 39750
+    },
+    {
+      "epoch": 1.5305101058710298,
+      "grad_norm": 1.2244682312011719,
+      "learning_rate": 2.605286861791304e-05,
+      "loss": 0.7738,
+      "step": 39755
+    },
+    {
+      "epoch": 1.5307025986525504,
+      "grad_norm": 1.583099126815796,
+      "learning_rate": 2.603251505733424e-05,
+      "loss": 0.8342,
+      "step": 39760
+    },
+    {
+      "epoch": 1.530895091434071,
+      "grad_norm": 2.7618401050567627,
+      "learning_rate": 2.6012168260591997e-05,
+      "loss": 0.7933,
+      "step": 39765
+    },
+    {
+      "epoch": 1.5310875842155918,
+      "grad_norm": 1.7654789686203003,
+      "learning_rate": 2.5991828229546766e-05,
+      "loss": 0.9115,
+      "step": 39770
+    },
+    {
+      "epoch": 1.5312800769971127,
+      "grad_norm": 1.1165564060211182,
+      "learning_rate": 2.597149496605865e-05,
+      "loss": 0.8808,
+      "step": 39775
+    },
+    {
+      "epoch": 1.5314725697786333,
+      "grad_norm": 2.018768310546875,
+      "learning_rate": 2.5951168471986896e-05,
+      "loss": 0.6884,
+      "step": 39780
+    },
+    {
+      "epoch": 1.531665062560154,
+      "grad_norm": 1.567855954170227,
+      "learning_rate": 2.593084874919025e-05,
+      "loss": 0.7812,
+      "step": 39785
+    },
+    {
+      "epoch": 1.5318575553416747,
+      "grad_norm": 1.3749490976333618,
+      "learning_rate": 2.5910535799526834e-05,
+      "loss": 0.8049,
+      "step": 39790
+    },
+    {
+      "epoch": 1.5320500481231953,
+      "grad_norm": 0.9157447814941406,
+      "learning_rate": 2.5890229624854146e-05,
+      "loss": 0.7877,
+      "step": 39795
+    },
+    {
+      "epoch": 1.5322425409047162,
+      "grad_norm": 1.0359165668487549,
+      "learning_rate": 2.586993022702897e-05,
+      "loss": 0.6694,
+      "step": 39800
+    },
+    {
+      "epoch": 1.532435033686237,
+      "grad_norm": 1.5438929796218872,
+      "learning_rate": 2.584963760790766e-05,
+      "loss": 0.7098,
+      "step": 39805
+    },
+    {
+      "epoch": 1.5326275264677576,
+      "grad_norm": 1.700568675994873,
+      "learning_rate": 2.5829351769345765e-05,
+      "loss": 0.6406,
+      "step": 39810
+    },
+    {
+      "epoch": 1.5328200192492782,
+      "grad_norm": 1.7474377155303955,
+      "learning_rate": 2.5809072713198313e-05,
+      "loss": 0.7675,
+      "step": 39815
+    },
+    {
+      "epoch": 1.533012512030799,
+      "grad_norm": 1.255887746810913,
+      "learning_rate": 2.5788800441319693e-05,
+      "loss": 0.8331,
+      "step": 39820
+    },
+    {
+      "epoch": 1.5332050048123196,
+      "grad_norm": 1.3320515155792236,
+      "learning_rate": 2.5768534955563694e-05,
+      "loss": 0.685,
+      "step": 39825
+    },
+    {
+      "epoch": 1.5333974975938403,
+      "grad_norm": 1.2977993488311768,
+      "learning_rate": 2.574827625778341e-05,
+      "loss": 0.9356,
+      "step": 39830
+    },
+    {
+      "epoch": 1.533589990375361,
+      "grad_norm": 1.2286765575408936,
+      "learning_rate": 2.572802434983139e-05,
+      "loss": 0.7974,
+      "step": 39835
+    },
+    {
+      "epoch": 1.5337824831568816,
+      "grad_norm": 1.5404809713363647,
+      "learning_rate": 2.5707779233559537e-05,
+      "loss": 0.8141,
+      "step": 39840
+    },
+    {
+      "epoch": 1.5339749759384023,
+      "grad_norm": 2.166562795639038,
+      "learning_rate": 2.568754091081913e-05,
+      "loss": 0.7906,
+      "step": 39845
+    },
+    {
+      "epoch": 1.534167468719923,
+      "grad_norm": 2.0202114582061768,
+      "learning_rate": 2.5667309383460857e-05,
+      "loss": 0.8012,
+      "step": 39850
+    },
+    {
+      "epoch": 1.5343599615014436,
+      "grad_norm": 1.2766706943511963,
+      "learning_rate": 2.5647084653334707e-05,
+      "loss": 0.8452,
+      "step": 39855
+    },
+    {
+      "epoch": 1.5345524542829643,
+      "grad_norm": 1.7136098146438599,
+      "learning_rate": 2.562686672229012e-05,
+      "loss": 0.7912,
+      "step": 39860
+    },
+    {
+      "epoch": 1.534744947064485,
+      "grad_norm": 1.3089460134506226,
+      "learning_rate": 2.56066555921759e-05,
+      "loss": 0.7512,
+      "step": 39865
+    },
+    {
+      "epoch": 1.5349374398460056,
+      "grad_norm": 1.317051649093628,
+      "learning_rate": 2.558645126484024e-05,
+      "loss": 0.9469,
+      "step": 39870
+    },
+    {
+      "epoch": 1.5351299326275265,
+      "grad_norm": 1.0458593368530273,
+      "learning_rate": 2.5566253742130608e-05,
+      "loss": 0.8018,
+      "step": 39875
+    },
+    {
+      "epoch": 1.5353224254090472,
+      "grad_norm": 0.9164959788322449,
+      "learning_rate": 2.5546063025894053e-05,
+      "loss": 0.7815,
+      "step": 39880
+    },
+    {
+      "epoch": 1.5355149181905678,
+      "grad_norm": 1.2855099439620972,
+      "learning_rate": 2.5525879117976794e-05,
+      "loss": 0.7955,
+      "step": 39885
+    },
+    {
+      "epoch": 1.5357074109720885,
+      "grad_norm": 1.3438894748687744,
+      "learning_rate": 2.5505702020224542e-05,
+      "loss": 0.724,
+      "step": 39890
+    },
+    {
+      "epoch": 1.5358999037536092,
+      "grad_norm": 1.1192864179611206,
+      "learning_rate": 2.548553173448236e-05,
+      "loss": 0.8516,
+      "step": 39895
+    },
+    {
+      "epoch": 1.53609239653513,
+      "grad_norm": 1.1312519311904907,
+      "learning_rate": 2.5465368262594726e-05,
+      "loss": 0.8709,
+      "step": 39900
+    },
+    {
+      "epoch": 1.5362848893166507,
+      "grad_norm": 1.1640228033065796,
+      "learning_rate": 2.544521160640535e-05,
+      "loss": 0.8806,
+      "step": 39905
+    },
+    {
+      "epoch": 1.5364773820981714,
+      "grad_norm": 0.9546281695365906,
+      "learning_rate": 2.5425061767757553e-05,
+      "loss": 0.7814,
+      "step": 39910
+    },
+    {
+      "epoch": 1.536669874879692,
+      "grad_norm": 1.2589070796966553,
+      "learning_rate": 2.540491874849381e-05,
+      "loss": 0.8007,
+      "step": 39915
+    },
+    {
+      "epoch": 1.5368623676612128,
+      "grad_norm": 1.1835721731185913,
+      "learning_rate": 2.53847825504561e-05,
+      "loss": 0.8729,
+      "step": 39920
+    },
+    {
+      "epoch": 1.5370548604427334,
+      "grad_norm": 1.403292179107666,
+      "learning_rate": 2.5364653175485754e-05,
+      "loss": 0.8297,
+      "step": 39925
+    },
+    {
+      "epoch": 1.537247353224254,
+      "grad_norm": 1.9332057237625122,
+      "learning_rate": 2.534453062542348e-05,
+      "loss": 0.8723,
+      "step": 39930
+    },
+    {
+      "epoch": 1.5374398460057748,
+      "grad_norm": 1.8926912546157837,
+      "learning_rate": 2.5324414902109316e-05,
+      "loss": 0.732,
+      "step": 39935
+    },
+    {
+      "epoch": 1.5376323387872954,
+      "grad_norm": 1.3487472534179688,
+      "learning_rate": 2.5304306007382716e-05,
+      "loss": 0.863,
+      "step": 39940
+    },
+    {
+      "epoch": 1.537824831568816,
+      "grad_norm": 1.4540005922317505,
+      "learning_rate": 2.528420394308253e-05,
+      "loss": 0.6911,
+      "step": 39945
+    },
+    {
+      "epoch": 1.5380173243503368,
+      "grad_norm": 1.1179677248001099,
+      "learning_rate": 2.526410871104694e-05,
+      "loss": 0.779,
+      "step": 39950
+    },
+    {
+      "epoch": 1.5382098171318574,
+      "grad_norm": 1.332623839378357,
+      "learning_rate": 2.5244020313113558e-05,
+      "loss": 0.6558,
+      "step": 39955
+    },
+    {
+      "epoch": 1.5384023099133781,
+      "grad_norm": 2.073150634765625,
+      "learning_rate": 2.5223938751119248e-05,
+      "loss": 0.9246,
+      "step": 39960
+    },
+    {
+      "epoch": 1.5385948026948988,
+      "grad_norm": 1.640775203704834,
+      "learning_rate": 2.5203864026900448e-05,
+      "loss": 0.7962,
+      "step": 39965
+    },
+    {
+      "epoch": 1.5387872954764197,
+      "grad_norm": 1.4655123949050903,
+      "learning_rate": 2.518379614229278e-05,
+      "loss": 0.8892,
+      "step": 39970
+    },
+    {
+      "epoch": 1.5389797882579404,
+      "grad_norm": 0.9323534369468689,
+      "learning_rate": 2.5163735099131336e-05,
+      "loss": 0.7165,
+      "step": 39975
+    },
+    {
+      "epoch": 1.539172281039461,
+      "grad_norm": 1.2594026327133179,
+      "learning_rate": 2.5143680899250577e-05,
+      "loss": 0.6767,
+      "step": 39980
+    },
+    {
+      "epoch": 1.5393647738209817,
+      "grad_norm": 1.3983330726623535,
+      "learning_rate": 2.5123633544484347e-05,
+      "loss": 0.7979,
+      "step": 39985
+    },
+    {
+      "epoch": 1.5395572666025024,
+      "grad_norm": 1.1931166648864746,
+      "learning_rate": 2.510359303666576e-05,
+      "loss": 0.8501,
+      "step": 39990
+    },
+    {
+      "epoch": 1.5397497593840233,
+      "grad_norm": 1.4925061464309692,
+      "learning_rate": 2.5083559377627518e-05,
+      "loss": 0.8053,
+      "step": 39995
+    },
+    {
+      "epoch": 1.539942252165544,
+      "grad_norm": 1.4643807411193848,
+      "learning_rate": 2.5063532569201454e-05,
+      "loss": 0.765,
+      "step": 40000
     }
   ],
   "logging_steps": 5,
@@ -42014,7 +56014,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 10000,
-  "total_flos": 9.370503974649938e+17,
+  "total_flos": 1.2478892695491133e+18,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null