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": 0.6633499170812603,
+  "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 12000,
+  "global_step": 18090,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -16807,6 +16807,8532 @@
       "learning_rate": 1.731097214326322e-05,
       "loss": 4.8989,
       "step": 12000
+    },
+    {
+      "epoch": 0.6636263128800443,
+      "grad_norm": 3.782233953475952,
+      "learning_rate": 1.729675952245594e-05,
+      "loss": 4.989,
+      "step": 12005
+    },
+    {
+      "epoch": 0.6639027086788281,
+      "grad_norm": 3.869248867034912,
+      "learning_rate": 1.7282546901648665e-05,
+      "loss": 4.8152,
+      "step": 12010
+    },
+    {
+      "epoch": 0.664179104477612,
+      "grad_norm": 3.3503940105438232,
+      "learning_rate": 1.726833428084139e-05,
+      "loss": 5.0352,
+      "step": 12015
+    },
+    {
+      "epoch": 0.6644555002763958,
+      "grad_norm": 3.2065019607543945,
+      "learning_rate": 1.7254121660034112e-05,
+      "loss": 4.8756,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6647318960751797,
+      "grad_norm": 3.2806355953216553,
+      "learning_rate": 1.7239909039226833e-05,
+      "loss": 4.9745,
+      "step": 12025
+    },
+    {
+      "epoch": 0.6650082918739635,
+      "grad_norm": 3.366260290145874,
+      "learning_rate": 1.7225696418419557e-05,
+      "loss": 4.9347,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6652846876727474,
+      "grad_norm": 3.2907090187072754,
+      "learning_rate": 1.721148379761228e-05,
+      "loss": 4.9765,
+      "step": 12035
+    },
+    {
+      "epoch": 0.6655610834715312,
+      "grad_norm": 4.094642639160156,
+      "learning_rate": 1.7197271176805005e-05,
+      "loss": 4.8334,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6658374792703151,
+      "grad_norm": 3.0640034675598145,
+      "learning_rate": 1.718305855599773e-05,
+      "loss": 4.6388,
+      "step": 12045
+    },
+    {
+      "epoch": 0.6661138750690989,
+      "grad_norm": 3.662433385848999,
+      "learning_rate": 1.716884593519045e-05,
+      "loss": 4.8606,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6663902708678828,
+      "grad_norm": 4.076939105987549,
+      "learning_rate": 1.7154633314383174e-05,
+      "loss": 4.8626,
+      "step": 12055
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 3.6793127059936523,
+      "learning_rate": 1.7140420693575894e-05,
+      "loss": 4.8432,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6669430624654505,
+      "grad_norm": 3.6468772888183594,
+      "learning_rate": 1.7126208072768618e-05,
+      "loss": 4.7681,
+      "step": 12065
+    },
+    {
+      "epoch": 0.6672194582642343,
+      "grad_norm": 3.1914479732513428,
+      "learning_rate": 1.7111995451961342e-05,
+      "loss": 5.0191,
+      "step": 12070
+    },
+    {
+      "epoch": 0.6674958540630183,
+      "grad_norm": 3.367316484451294,
+      "learning_rate": 1.7097782831154066e-05,
+      "loss": 5.0413,
+      "step": 12075
+    },
+    {
+      "epoch": 0.6677722498618021,
+      "grad_norm": 3.765392780303955,
+      "learning_rate": 1.708357021034679e-05,
+      "loss": 4.8648,
+      "step": 12080
+    },
+    {
+      "epoch": 0.668048645660586,
+      "grad_norm": 3.203326463699341,
+      "learning_rate": 1.706935758953951e-05,
+      "loss": 4.9687,
+      "step": 12085
+    },
+    {
+      "epoch": 0.6683250414593698,
+      "grad_norm": 3.558098554611206,
+      "learning_rate": 1.7055144968732235e-05,
+      "loss": 5.1091,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6686014372581537,
+      "grad_norm": 2.9107744693756104,
+      "learning_rate": 1.704093234792496e-05,
+      "loss": 4.8894,
+      "step": 12095
+    },
+    {
+      "epoch": 0.6688778330569375,
+      "grad_norm": 3.141874074935913,
+      "learning_rate": 1.7026719727117683e-05,
+      "loss": 4.7187,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6691542288557214,
+      "grad_norm": 3.327786684036255,
+      "learning_rate": 1.7012507106310403e-05,
+      "loss": 5.34,
+      "step": 12105
+    },
+    {
+      "epoch": 0.6694306246545052,
+      "grad_norm": 3.6189236640930176,
+      "learning_rate": 1.6998294485503127e-05,
+      "loss": 5.1619,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6697070204532891,
+      "grad_norm": 3.134925365447998,
+      "learning_rate": 1.698408186469585e-05,
+      "loss": 5.4205,
+      "step": 12115
+    },
+    {
+      "epoch": 0.6699834162520729,
+      "grad_norm": 3.9740278720855713,
+      "learning_rate": 1.6969869243888572e-05,
+      "loss": 4.929,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6702598120508568,
+      "grad_norm": 3.459599494934082,
+      "learning_rate": 1.69556566230813e-05,
+      "loss": 5.0055,
+      "step": 12125
+    },
+    {
+      "epoch": 0.6705362078496406,
+      "grad_norm": 3.837351083755493,
+      "learning_rate": 1.694144400227402e-05,
+      "loss": 5.0351,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6708126036484245,
+      "grad_norm": 4.666230201721191,
+      "learning_rate": 1.6927231381466744e-05,
+      "loss": 5.0094,
+      "step": 12135
+    },
+    {
+      "epoch": 0.6710889994472085,
+      "grad_norm": 4.403833866119385,
+      "learning_rate": 1.6913018760659464e-05,
+      "loss": 5.0496,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6713653952459923,
+      "grad_norm": 3.412066698074341,
+      "learning_rate": 1.6898806139852188e-05,
+      "loss": 5.5492,
+      "step": 12145
+    },
+    {
+      "epoch": 0.6716417910447762,
+      "grad_norm": 4.21888542175293,
+      "learning_rate": 1.6884593519044916e-05,
+      "loss": 4.8999,
+      "step": 12150
+    },
+    {
+      "epoch": 0.67191818684356,
+      "grad_norm": 3.7332446575164795,
+      "learning_rate": 1.6870380898237636e-05,
+      "loss": 5.1613,
+      "step": 12155
+    },
+    {
+      "epoch": 0.6721945826423439,
+      "grad_norm": 3.9630868434906006,
+      "learning_rate": 1.685616827743036e-05,
+      "loss": 5.1332,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6724709784411277,
+      "grad_norm": 3.0835654735565186,
+      "learning_rate": 1.684195565662308e-05,
+      "loss": 5.3016,
+      "step": 12165
+    },
+    {
+      "epoch": 0.6727473742399116,
+      "grad_norm": 3.015239715576172,
+      "learning_rate": 1.6827743035815805e-05,
+      "loss": 5.2528,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6730237700386954,
+      "grad_norm": 3.449125051498413,
+      "learning_rate": 1.681353041500853e-05,
+      "loss": 4.8346,
+      "step": 12175
+    },
+    {
+      "epoch": 0.6733001658374793,
+      "grad_norm": 3.250673294067383,
+      "learning_rate": 1.6799317794201253e-05,
+      "loss": 5.0164,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6735765616362631,
+      "grad_norm": 3.8429434299468994,
+      "learning_rate": 1.6785105173393977e-05,
+      "loss": 5.326,
+      "step": 12185
+    },
+    {
+      "epoch": 0.673852957435047,
+      "grad_norm": 3.823730707168579,
+      "learning_rate": 1.6770892552586697e-05,
+      "loss": 5.399,
+      "step": 12190
+    },
+    {
+      "epoch": 0.6741293532338308,
+      "grad_norm": 3.227236747741699,
+      "learning_rate": 1.675667993177942e-05,
+      "loss": 4.8976,
+      "step": 12195
+    },
+    {
+      "epoch": 0.6744057490326147,
+      "grad_norm": 3.5069992542266846,
+      "learning_rate": 1.6742467310972142e-05,
+      "loss": 4.9528,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6746821448313985,
+      "grad_norm": 2.9861481189727783,
+      "learning_rate": 1.672825469016487e-05,
+      "loss": 5.1612,
+      "step": 12205
+    },
+    {
+      "epoch": 0.6749585406301825,
+      "grad_norm": 2.750514507293701,
+      "learning_rate": 1.671404206935759e-05,
+      "loss": 4.7478,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6752349364289663,
+      "grad_norm": 3.113337993621826,
+      "learning_rate": 1.6699829448550314e-05,
+      "loss": 5.0428,
+      "step": 12215
+    },
+    {
+      "epoch": 0.6755113322277502,
+      "grad_norm": 3.744915723800659,
+      "learning_rate": 1.6685616827743038e-05,
+      "loss": 4.9286,
+      "step": 12220
+    },
+    {
+      "epoch": 0.675787728026534,
+      "grad_norm": 2.7885375022888184,
+      "learning_rate": 1.667140420693576e-05,
+      "loss": 5.1185,
+      "step": 12225
+    },
+    {
+      "epoch": 0.6760641238253179,
+      "grad_norm": 3.3463151454925537,
+      "learning_rate": 1.6657191586128482e-05,
+      "loss": 4.815,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6763405196241017,
+      "grad_norm": 3.387303113937378,
+      "learning_rate": 1.6642978965321206e-05,
+      "loss": 5.0025,
+      "step": 12235
+    },
+    {
+      "epoch": 0.6766169154228856,
+      "grad_norm": 3.3395464420318604,
+      "learning_rate": 1.662876634451393e-05,
+      "loss": 5.3222,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6768933112216694,
+      "grad_norm": 3.5981926918029785,
+      "learning_rate": 1.661455372370665e-05,
+      "loss": 5.262,
+      "step": 12245
+    },
+    {
+      "epoch": 0.6771697070204533,
+      "grad_norm": 2.9530062675476074,
+      "learning_rate": 1.6600341102899375e-05,
+      "loss": 5.0093,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6774461028192371,
+      "grad_norm": 2.790177583694458,
+      "learning_rate": 1.65861284820921e-05,
+      "loss": 5.178,
+      "step": 12255
+    },
+    {
+      "epoch": 0.677722498618021,
+      "grad_norm": 3.2655928134918213,
+      "learning_rate": 1.6571915861284823e-05,
+      "loss": 4.9004,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6779988944168048,
+      "grad_norm": 3.1761622428894043,
+      "learning_rate": 1.6557703240477547e-05,
+      "loss": 5.2096,
+      "step": 12265
+    },
+    {
+      "epoch": 0.6782752902155887,
+      "grad_norm": 3.3835296630859375,
+      "learning_rate": 1.6543490619670267e-05,
+      "loss": 4.7797,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6785516860143725,
+      "grad_norm": 3.7050514221191406,
+      "learning_rate": 1.652927799886299e-05,
+      "loss": 5.075,
+      "step": 12275
+    },
+    {
+      "epoch": 0.6788280818131565,
+      "grad_norm": 3.350598096847534,
+      "learning_rate": 1.6515065378055712e-05,
+      "loss": 4.9758,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6791044776119403,
+      "grad_norm": 3.3467860221862793,
+      "learning_rate": 1.6500852757248436e-05,
+      "loss": 4.984,
+      "step": 12285
+    },
+    {
+      "epoch": 0.6793808734107242,
+      "grad_norm": 3.433746814727783,
+      "learning_rate": 1.648664013644116e-05,
+      "loss": 4.8733,
+      "step": 12290
+    },
+    {
+      "epoch": 0.679657269209508,
+      "grad_norm": 3.335195302963257,
+      "learning_rate": 1.6472427515633884e-05,
+      "loss": 4.9734,
+      "step": 12295
+    },
+    {
+      "epoch": 0.6799336650082919,
+      "grad_norm": 3.4817519187927246,
+      "learning_rate": 1.6458214894826608e-05,
+      "loss": 4.9225,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6802100608070757,
+      "grad_norm": 3.4715635776519775,
+      "learning_rate": 1.644400227401933e-05,
+      "loss": 5.0767,
+      "step": 12305
+    },
+    {
+      "epoch": 0.6804864566058596,
+      "grad_norm": 4.440505027770996,
+      "learning_rate": 1.6429789653212052e-05,
+      "loss": 4.6531,
+      "step": 12310
+    },
+    {
+      "epoch": 0.6807628524046434,
+      "grad_norm": 3.239976644515991,
+      "learning_rate": 1.6415577032404776e-05,
+      "loss": 5.0474,
+      "step": 12315
+    },
+    {
+      "epoch": 0.6810392482034273,
+      "grad_norm": 3.019190549850464,
+      "learning_rate": 1.64013644115975e-05,
+      "loss": 4.9771,
+      "step": 12320
+    },
+    {
+      "epoch": 0.6813156440022111,
+      "grad_norm": 3.4960367679595947,
+      "learning_rate": 1.6387151790790224e-05,
+      "loss": 5.1239,
+      "step": 12325
+    },
+    {
+      "epoch": 0.681592039800995,
+      "grad_norm": 4.1001081466674805,
+      "learning_rate": 1.6372939169982945e-05,
+      "loss": 5.0819,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6818684355997788,
+      "grad_norm": 4.748738765716553,
+      "learning_rate": 1.635872654917567e-05,
+      "loss": 4.8874,
+      "step": 12335
+    },
+    {
+      "epoch": 0.6821448313985627,
+      "grad_norm": 3.8815574645996094,
+      "learning_rate": 1.634451392836839e-05,
+      "loss": 5.0887,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6824212271973465,
+      "grad_norm": 3.651639938354492,
+      "learning_rate": 1.6330301307561117e-05,
+      "loss": 4.7795,
+      "step": 12345
+    },
+    {
+      "epoch": 0.6826976229961305,
+      "grad_norm": 3.9696593284606934,
+      "learning_rate": 1.6316088686753837e-05,
+      "loss": 5.243,
+      "step": 12350
+    },
+    {
+      "epoch": 0.6829740187949144,
+      "grad_norm": 3.3984673023223877,
+      "learning_rate": 1.630187606594656e-05,
+      "loss": 5.1969,
+      "step": 12355
+    },
+    {
+      "epoch": 0.6832504145936982,
+      "grad_norm": 3.081726312637329,
+      "learning_rate": 1.6287663445139285e-05,
+      "loss": 5.1571,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6835268103924821,
+      "grad_norm": 3.6568610668182373,
+      "learning_rate": 1.6273450824332006e-05,
+      "loss": 5.2434,
+      "step": 12365
+    },
+    {
+      "epoch": 0.6838032061912659,
+      "grad_norm": 4.056967735290527,
+      "learning_rate": 1.6259238203524733e-05,
+      "loss": 4.993,
+      "step": 12370
+    },
+    {
+      "epoch": 0.6840796019900498,
+      "grad_norm": 3.637042284011841,
+      "learning_rate": 1.6245025582717454e-05,
+      "loss": 5.5494,
+      "step": 12375
+    },
+    {
+      "epoch": 0.6843559977888336,
+      "grad_norm": 4.664302825927734,
+      "learning_rate": 1.6230812961910178e-05,
+      "loss": 4.9485,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6846323935876175,
+      "grad_norm": 3.098987340927124,
+      "learning_rate": 1.62166003411029e-05,
+      "loss": 4.8687,
+      "step": 12385
+    },
+    {
+      "epoch": 0.6849087893864013,
+      "grad_norm": 3.432152509689331,
+      "learning_rate": 1.6202387720295623e-05,
+      "loss": 4.7856,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6851851851851852,
+      "grad_norm": 3.564708709716797,
+      "learning_rate": 1.6188175099488347e-05,
+      "loss": 5.0981,
+      "step": 12395
+    },
+    {
+      "epoch": 0.685461580983969,
+      "grad_norm": 3.6227316856384277,
+      "learning_rate": 1.617396247868107e-05,
+      "loss": 5.0038,
+      "step": 12400
+    },
+    {
+      "epoch": 0.685737976782753,
+      "grad_norm": 3.641265392303467,
+      "learning_rate": 1.6159749857873794e-05,
+      "loss": 4.9751,
+      "step": 12405
+    },
+    {
+      "epoch": 0.6860143725815367,
+      "grad_norm": 3.2415549755096436,
+      "learning_rate": 1.6145537237066515e-05,
+      "loss": 5.1277,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6862907683803207,
+      "grad_norm": 3.59389591217041,
+      "learning_rate": 1.613132461625924e-05,
+      "loss": 5.1952,
+      "step": 12415
+    },
+    {
+      "epoch": 0.6865671641791045,
+      "grad_norm": 3.559694528579712,
+      "learning_rate": 1.611711199545196e-05,
+      "loss": 4.9343,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6868435599778884,
+      "grad_norm": 3.0507447719573975,
+      "learning_rate": 1.6102899374644687e-05,
+      "loss": 5.0918,
+      "step": 12425
+    },
+    {
+      "epoch": 0.6871199557766722,
+      "grad_norm": 3.375110149383545,
+      "learning_rate": 1.6088686753837408e-05,
+      "loss": 4.78,
+      "step": 12430
+    },
+    {
+      "epoch": 0.6873963515754561,
+      "grad_norm": 4.308310031890869,
+      "learning_rate": 1.607447413303013e-05,
+      "loss": 4.9678,
+      "step": 12435
+    },
+    {
+      "epoch": 0.6876727473742399,
+      "grad_norm": 4.061061382293701,
+      "learning_rate": 1.6060261512222856e-05,
+      "loss": 4.9967,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6879491431730238,
+      "grad_norm": 3.5668656826019287,
+      "learning_rate": 1.6046048891415576e-05,
+      "loss": 4.695,
+      "step": 12445
+    },
+    {
+      "epoch": 0.6882255389718076,
+      "grad_norm": 3.55816650390625,
+      "learning_rate": 1.60318362706083e-05,
+      "loss": 5.0108,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6885019347705915,
+      "grad_norm": 3.516140937805176,
+      "learning_rate": 1.6017623649801024e-05,
+      "loss": 5.0607,
+      "step": 12455
+    },
+    {
+      "epoch": 0.6887783305693753,
+      "grad_norm": 3.5040078163146973,
+      "learning_rate": 1.6003411028993748e-05,
+      "loss": 5.2791,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6890547263681592,
+      "grad_norm": 3.3454535007476807,
+      "learning_rate": 1.5989198408186472e-05,
+      "loss": 4.9042,
+      "step": 12465
+    },
+    {
+      "epoch": 0.689331122166943,
+      "grad_norm": 4.622253894805908,
+      "learning_rate": 1.5974985787379193e-05,
+      "loss": 4.8014,
+      "step": 12470
+    },
+    {
+      "epoch": 0.689607517965727,
+      "grad_norm": 4.814120292663574,
+      "learning_rate": 1.5960773166571917e-05,
+      "loss": 5.0986,
+      "step": 12475
+    },
+    {
+      "epoch": 0.6898839137645107,
+      "grad_norm": 3.6000027656555176,
+      "learning_rate": 1.594656054576464e-05,
+      "loss": 5.2295,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6901603095632947,
+      "grad_norm": 3.4990899562835693,
+      "learning_rate": 1.5932347924957365e-05,
+      "loss": 5.2095,
+      "step": 12485
+    },
+    {
+      "epoch": 0.6904367053620785,
+      "grad_norm": 4.27670431137085,
+      "learning_rate": 1.5918135304150085e-05,
+      "loss": 5.3237,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6907131011608624,
+      "grad_norm": 4.3061394691467285,
+      "learning_rate": 1.590392268334281e-05,
+      "loss": 5.0581,
+      "step": 12495
+    },
+    {
+      "epoch": 0.6909894969596462,
+      "grad_norm": 3.4813592433929443,
+      "learning_rate": 1.5889710062535533e-05,
+      "loss": 5.244,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6912658927584301,
+      "grad_norm": 4.107872486114502,
+      "learning_rate": 1.5875497441728254e-05,
+      "loss": 4.9351,
+      "step": 12505
+    },
+    {
+      "epoch": 0.6915422885572139,
+      "grad_norm": 3.8972761631011963,
+      "learning_rate": 1.586128482092098e-05,
+      "loss": 4.954,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6918186843559978,
+      "grad_norm": 3.9529995918273926,
+      "learning_rate": 1.5847072200113702e-05,
+      "loss": 4.553,
+      "step": 12515
+    },
+    {
+      "epoch": 0.6920950801547816,
+      "grad_norm": 3.2394180297851562,
+      "learning_rate": 1.5832859579306426e-05,
+      "loss": 4.8631,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6923714759535655,
+      "grad_norm": 3.906923294067383,
+      "learning_rate": 1.5818646958499146e-05,
+      "loss": 4.655,
+      "step": 12525
+    },
+    {
+      "epoch": 0.6926478717523493,
+      "grad_norm": 3.0384624004364014,
+      "learning_rate": 1.580443433769187e-05,
+      "loss": 4.8143,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6929242675511332,
+      "grad_norm": 3.3072428703308105,
+      "learning_rate": 1.5790221716884594e-05,
+      "loss": 5.1887,
+      "step": 12535
+    },
+    {
+      "epoch": 0.693200663349917,
+      "grad_norm": 3.6409823894500732,
+      "learning_rate": 1.5776009096077318e-05,
+      "loss": 5.3076,
+      "step": 12540
+    },
+    {
+      "epoch": 0.693477059148701,
+      "grad_norm": 4.234898567199707,
+      "learning_rate": 1.5761796475270042e-05,
+      "loss": 5.037,
+      "step": 12545
+    },
+    {
+      "epoch": 0.6937534549474847,
+      "grad_norm": 3.5586225986480713,
+      "learning_rate": 1.5747583854462763e-05,
+      "loss": 5.0659,
+      "step": 12550
+    },
+    {
+      "epoch": 0.6940298507462687,
+      "grad_norm": 3.656480073928833,
+      "learning_rate": 1.5733371233655487e-05,
+      "loss": 5.1345,
+      "step": 12555
+    },
+    {
+      "epoch": 0.6943062465450525,
+      "grad_norm": 3.5738766193389893,
+      "learning_rate": 1.5719158612848207e-05,
+      "loss": 4.8976,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6945826423438364,
+      "grad_norm": 3.6187734603881836,
+      "learning_rate": 1.5704945992040935e-05,
+      "loss": 4.7864,
+      "step": 12565
+    },
+    {
+      "epoch": 0.6948590381426202,
+      "grad_norm": 3.634728193283081,
+      "learning_rate": 1.5690733371233655e-05,
+      "loss": 5.1336,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6951354339414041,
+      "grad_norm": 3.8284621238708496,
+      "learning_rate": 1.567652075042638e-05,
+      "loss": 5.0473,
+      "step": 12575
+    },
+    {
+      "epoch": 0.695411829740188,
+      "grad_norm": 3.628915786743164,
+      "learning_rate": 1.5662308129619103e-05,
+      "loss": 4.8755,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6956882255389718,
+      "grad_norm": 4.438342571258545,
+      "learning_rate": 1.5648095508811824e-05,
+      "loss": 4.7324,
+      "step": 12585
+    },
+    {
+      "epoch": 0.6959646213377557,
+      "grad_norm": 3.574575662612915,
+      "learning_rate": 1.563388288800455e-05,
+      "loss": 5.0009,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6962410171365395,
+      "grad_norm": 3.5201807022094727,
+      "learning_rate": 1.5619670267197272e-05,
+      "loss": 5.0128,
+      "step": 12595
+    },
+    {
+      "epoch": 0.6965174129353234,
+      "grad_norm": 3.0795626640319824,
+      "learning_rate": 1.5605457646389996e-05,
+      "loss": 4.9838,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6967938087341072,
+      "grad_norm": 3.3544068336486816,
+      "learning_rate": 1.5591245025582716e-05,
+      "loss": 5.1475,
+      "step": 12605
+    },
+    {
+      "epoch": 0.6970702045328911,
+      "grad_norm": 3.4172658920288086,
+      "learning_rate": 1.557703240477544e-05,
+      "loss": 4.7901,
+      "step": 12610
+    },
+    {
+      "epoch": 0.697346600331675,
+      "grad_norm": 3.966212511062622,
+      "learning_rate": 1.5562819783968164e-05,
+      "loss": 5.0707,
+      "step": 12615
+    },
+    {
+      "epoch": 0.6976229961304589,
+      "grad_norm": 3.324779748916626,
+      "learning_rate": 1.554860716316089e-05,
+      "loss": 5.0197,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6978993919292427,
+      "grad_norm": 3.247331142425537,
+      "learning_rate": 1.5534394542353612e-05,
+      "loss": 5.3125,
+      "step": 12625
+    },
+    {
+      "epoch": 0.6981757877280266,
+      "grad_norm": 2.768170118331909,
+      "learning_rate": 1.5520181921546333e-05,
+      "loss": 4.8045,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6984521835268104,
+      "grad_norm": 2.8270652294158936,
+      "learning_rate": 1.5505969300739057e-05,
+      "loss": 5.0571,
+      "step": 12635
+    },
+    {
+      "epoch": 0.6987285793255943,
+      "grad_norm": 3.471879005432129,
+      "learning_rate": 1.549175667993178e-05,
+      "loss": 4.9667,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6990049751243781,
+      "grad_norm": 3.826054811477661,
+      "learning_rate": 1.5477544059124505e-05,
+      "loss": 4.9908,
+      "step": 12645
+    },
+    {
+      "epoch": 0.699281370923162,
+      "grad_norm": 3.6149983406066895,
+      "learning_rate": 1.546333143831723e-05,
+      "loss": 4.8681,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6995577667219458,
+      "grad_norm": 4.132063388824463,
+      "learning_rate": 1.544911881750995e-05,
+      "loss": 5.0986,
+      "step": 12655
+    },
+    {
+      "epoch": 0.6998341625207297,
+      "grad_norm": 3.410290002822876,
+      "learning_rate": 1.5434906196702673e-05,
+      "loss": 5.0152,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7001105583195135,
+      "grad_norm": 3.215104103088379,
+      "learning_rate": 1.5420693575895394e-05,
+      "loss": 4.7577,
+      "step": 12665
+    },
+    {
+      "epoch": 0.7003869541182974,
+      "grad_norm": 3.521066427230835,
+      "learning_rate": 1.5406480955088118e-05,
+      "loss": 4.9928,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7006633499170812,
+      "grad_norm": 3.641317367553711,
+      "learning_rate": 1.5392268334280842e-05,
+      "loss": 5.1642,
+      "step": 12675
+    },
+    {
+      "epoch": 0.7009397457158651,
+      "grad_norm": 3.365527629852295,
+      "learning_rate": 1.5378055713473566e-05,
+      "loss": 4.9449,
+      "step": 12680
+    },
+    {
+      "epoch": 0.701216141514649,
+      "grad_norm": 3.430457592010498,
+      "learning_rate": 1.536384309266629e-05,
+      "loss": 5.123,
+      "step": 12685
+    },
+    {
+      "epoch": 0.7014925373134329,
+      "grad_norm": 3.236186981201172,
+      "learning_rate": 1.534963047185901e-05,
+      "loss": 5.1541,
+      "step": 12690
+    },
+    {
+      "epoch": 0.7017689331122167,
+      "grad_norm": 3.792572021484375,
+      "learning_rate": 1.5335417851051734e-05,
+      "loss": 5.1397,
+      "step": 12695
+    },
+    {
+      "epoch": 0.7020453289110006,
+      "grad_norm": 3.7908694744110107,
+      "learning_rate": 1.532120523024446e-05,
+      "loss": 5.0269,
+      "step": 12700
+    },
+    {
+      "epoch": 0.7023217247097844,
+      "grad_norm": 3.2599024772644043,
+      "learning_rate": 1.5306992609437182e-05,
+      "loss": 4.7786,
+      "step": 12705
+    },
+    {
+      "epoch": 0.7025981205085683,
+      "grad_norm": 3.5617377758026123,
+      "learning_rate": 1.5292779988629903e-05,
+      "loss": 5.0488,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7028745163073521,
+      "grad_norm": 3.4966659545898438,
+      "learning_rate": 1.5278567367822627e-05,
+      "loss": 5.0739,
+      "step": 12715
+    },
+    {
+      "epoch": 0.703150912106136,
+      "grad_norm": 3.6931207180023193,
+      "learning_rate": 1.526435474701535e-05,
+      "loss": 4.7299,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7034273079049198,
+      "grad_norm": 3.201852560043335,
+      "learning_rate": 1.5250142126208072e-05,
+      "loss": 5.1792,
+      "step": 12725
+    },
+    {
+      "epoch": 0.7037037037037037,
+      "grad_norm": 3.669928550720215,
+      "learning_rate": 1.5235929505400797e-05,
+      "loss": 4.8751,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7039800995024875,
+      "grad_norm": 3.639188766479492,
+      "learning_rate": 1.5221716884593521e-05,
+      "loss": 4.7316,
+      "step": 12735
+    },
+    {
+      "epoch": 0.7042564953012714,
+      "grad_norm": 3.298006057739258,
+      "learning_rate": 1.5207504263786243e-05,
+      "loss": 4.7265,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7045328911000552,
+      "grad_norm": 4.186183452606201,
+      "learning_rate": 1.5193291642978966e-05,
+      "loss": 4.6542,
+      "step": 12745
+    },
+    {
+      "epoch": 0.7048092868988391,
+      "grad_norm": 3.1352524757385254,
+      "learning_rate": 1.5179079022171688e-05,
+      "loss": 5.1087,
+      "step": 12750
+    },
+    {
+      "epoch": 0.705085682697623,
+      "grad_norm": 3.946307420730591,
+      "learning_rate": 1.516486640136441e-05,
+      "loss": 4.9225,
+      "step": 12755
+    },
+    {
+      "epoch": 0.7053620784964069,
+      "grad_norm": 3.4466960430145264,
+      "learning_rate": 1.5150653780557136e-05,
+      "loss": 5.0924,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7056384742951907,
+      "grad_norm": 3.4541025161743164,
+      "learning_rate": 1.5136441159749858e-05,
+      "loss": 4.7316,
+      "step": 12765
+    },
+    {
+      "epoch": 0.7059148700939746,
+      "grad_norm": 3.1319077014923096,
+      "learning_rate": 1.5122228538942582e-05,
+      "loss": 5.0303,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7061912658927584,
+      "grad_norm": 3.6074001789093018,
+      "learning_rate": 1.5108015918135305e-05,
+      "loss": 4.9892,
+      "step": 12775
+    },
+    {
+      "epoch": 0.7064676616915423,
+      "grad_norm": 3.2462453842163086,
+      "learning_rate": 1.5093803297328027e-05,
+      "loss": 4.9228,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7067440574903261,
+      "grad_norm": 4.183916091918945,
+      "learning_rate": 1.5079590676520753e-05,
+      "loss": 5.0513,
+      "step": 12785
+    },
+    {
+      "epoch": 0.70702045328911,
+      "grad_norm": 4.227228164672852,
+      "learning_rate": 1.5065378055713475e-05,
+      "loss": 5.2663,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7072968490878938,
+      "grad_norm": 4.704550743103027,
+      "learning_rate": 1.5051165434906197e-05,
+      "loss": 5.0818,
+      "step": 12795
+    },
+    {
+      "epoch": 0.7075732448866777,
+      "grad_norm": 3.2340505123138428,
+      "learning_rate": 1.503695281409892e-05,
+      "loss": 5.1363,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7078496406854616,
+      "grad_norm": 4.0538153648376465,
+      "learning_rate": 1.5022740193291643e-05,
+      "loss": 5.6689,
+      "step": 12805
+    },
+    {
+      "epoch": 0.7081260364842454,
+      "grad_norm": 4.062292098999023,
+      "learning_rate": 1.5008527572484366e-05,
+      "loss": 5.157,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7084024322830293,
+      "grad_norm": 3.2612810134887695,
+      "learning_rate": 1.4994314951677091e-05,
+      "loss": 4.9579,
+      "step": 12815
+    },
+    {
+      "epoch": 0.7086788280818132,
+      "grad_norm": 3.235447406768799,
+      "learning_rate": 1.4980102330869814e-05,
+      "loss": 5.1034,
+      "step": 12820
+    },
+    {
+      "epoch": 0.7089552238805971,
+      "grad_norm": 3.9662389755249023,
+      "learning_rate": 1.4965889710062536e-05,
+      "loss": 4.9475,
+      "step": 12825
+    },
+    {
+      "epoch": 0.7092316196793809,
+      "grad_norm": 3.0024068355560303,
+      "learning_rate": 1.4951677089255258e-05,
+      "loss": 4.9537,
+      "step": 12830
+    },
+    {
+      "epoch": 0.7095080154781648,
+      "grad_norm": 3.502521514892578,
+      "learning_rate": 1.4937464468447982e-05,
+      "loss": 4.9855,
+      "step": 12835
+    },
+    {
+      "epoch": 0.7097844112769486,
+      "grad_norm": 3.6317951679229736,
+      "learning_rate": 1.4923251847640706e-05,
+      "loss": 5.1954,
+      "step": 12840
+    },
+    {
+      "epoch": 0.7100608070757325,
+      "grad_norm": 3.3001856803894043,
+      "learning_rate": 1.490903922683343e-05,
+      "loss": 4.9513,
+      "step": 12845
+    },
+    {
+      "epoch": 0.7103372028745163,
+      "grad_norm": 3.484205961227417,
+      "learning_rate": 1.4894826606026152e-05,
+      "loss": 4.6211,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7106135986733002,
+      "grad_norm": 3.8079354763031006,
+      "learning_rate": 1.4880613985218875e-05,
+      "loss": 5.0842,
+      "step": 12855
+    },
+    {
+      "epoch": 0.710889994472084,
+      "grad_norm": 3.5550713539123535,
+      "learning_rate": 1.4866401364411597e-05,
+      "loss": 4.981,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7111663902708679,
+      "grad_norm": 4.202615261077881,
+      "learning_rate": 1.485218874360432e-05,
+      "loss": 5.1481,
+      "step": 12865
+    },
+    {
+      "epoch": 0.7114427860696517,
+      "grad_norm": 3.486539125442505,
+      "learning_rate": 1.4837976122797045e-05,
+      "loss": 5.0235,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7117191818684356,
+      "grad_norm": 3.639721632003784,
+      "learning_rate": 1.4823763501989767e-05,
+      "loss": 4.9426,
+      "step": 12875
+    },
+    {
+      "epoch": 0.7119955776672194,
+      "grad_norm": 3.9784390926361084,
+      "learning_rate": 1.4809550881182491e-05,
+      "loss": 5.0809,
+      "step": 12880
+    },
+    {
+      "epoch": 0.7122719734660033,
+      "grad_norm": 2.89168643951416,
+      "learning_rate": 1.4795338260375213e-05,
+      "loss": 5.0391,
+      "step": 12885
+    },
+    {
+      "epoch": 0.7125483692647872,
+      "grad_norm": 4.273104667663574,
+      "learning_rate": 1.4781125639567936e-05,
+      "loss": 4.8018,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7128247650635711,
+      "grad_norm": 3.9629507064819336,
+      "learning_rate": 1.4766913018760661e-05,
+      "loss": 5.0206,
+      "step": 12895
+    },
+    {
+      "epoch": 0.7131011608623549,
+      "grad_norm": 3.7056665420532227,
+      "learning_rate": 1.4752700397953384e-05,
+      "loss": 5.1737,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7133775566611388,
+      "grad_norm": 3.9631214141845703,
+      "learning_rate": 1.4738487777146106e-05,
+      "loss": 5.4375,
+      "step": 12905
+    },
+    {
+      "epoch": 0.7136539524599226,
+      "grad_norm": 4.301235198974609,
+      "learning_rate": 1.472427515633883e-05,
+      "loss": 5.0811,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7139303482587065,
+      "grad_norm": 3.4924840927124023,
+      "learning_rate": 1.4710062535531552e-05,
+      "loss": 5.0251,
+      "step": 12915
+    },
+    {
+      "epoch": 0.7142067440574903,
+      "grad_norm": 3.845777750015259,
+      "learning_rate": 1.4695849914724275e-05,
+      "loss": 4.9661,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7144831398562742,
+      "grad_norm": 3.8963284492492676,
+      "learning_rate": 1.4681637293917e-05,
+      "loss": 5.0677,
+      "step": 12925
+    },
+    {
+      "epoch": 0.714759535655058,
+      "grad_norm": 3.3300371170043945,
+      "learning_rate": 1.4667424673109723e-05,
+      "loss": 4.565,
+      "step": 12930
+    },
+    {
+      "epoch": 0.7150359314538419,
+      "grad_norm": 3.4866998195648193,
+      "learning_rate": 1.4653212052302445e-05,
+      "loss": 4.9402,
+      "step": 12935
+    },
+    {
+      "epoch": 0.7153123272526257,
+      "grad_norm": 3.7221293449401855,
+      "learning_rate": 1.4638999431495167e-05,
+      "loss": 5.0538,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7155887230514096,
+      "grad_norm": 3.6549923419952393,
+      "learning_rate": 1.4624786810687891e-05,
+      "loss": 5.0269,
+      "step": 12945
+    },
+    {
+      "epoch": 0.7158651188501934,
+      "grad_norm": 3.261334180831909,
+      "learning_rate": 1.4610574189880615e-05,
+      "loss": 5.0511,
+      "step": 12950
+    },
+    {
+      "epoch": 0.7161415146489774,
+      "grad_norm": 3.2248923778533936,
+      "learning_rate": 1.4596361569073339e-05,
+      "loss": 5.0044,
+      "step": 12955
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 3.5878825187683105,
+      "learning_rate": 1.4582148948266061e-05,
+      "loss": 5.0746,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7166943062465451,
+      "grad_norm": 3.1510488986968994,
+      "learning_rate": 1.4567936327458784e-05,
+      "loss": 4.775,
+      "step": 12965
+    },
+    {
+      "epoch": 0.7169707020453289,
+      "grad_norm": 3.1844141483306885,
+      "learning_rate": 1.4553723706651506e-05,
+      "loss": 4.803,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7172470978441128,
+      "grad_norm": 3.417433977127075,
+      "learning_rate": 1.453951108584423e-05,
+      "loss": 4.932,
+      "step": 12975
+    },
+    {
+      "epoch": 0.7175234936428966,
+      "grad_norm": 3.5905191898345947,
+      "learning_rate": 1.4525298465036954e-05,
+      "loss": 5.1135,
+      "step": 12980
+    },
+    {
+      "epoch": 0.7177998894416805,
+      "grad_norm": 3.180964231491089,
+      "learning_rate": 1.4511085844229678e-05,
+      "loss": 4.8405,
+      "step": 12985
+    },
+    {
+      "epoch": 0.7180762852404643,
+      "grad_norm": 3.3651866912841797,
+      "learning_rate": 1.44968732234224e-05,
+      "loss": 4.8142,
+      "step": 12990
+    },
+    {
+      "epoch": 0.7183526810392482,
+      "grad_norm": 3.8147387504577637,
+      "learning_rate": 1.4482660602615122e-05,
+      "loss": 5.152,
+      "step": 12995
+    },
+    {
+      "epoch": 0.718629076838032,
+      "grad_norm": 3.135228157043457,
+      "learning_rate": 1.4468447981807845e-05,
+      "loss": 4.7615,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7189054726368159,
+      "grad_norm": 3.4666500091552734,
+      "learning_rate": 1.445423536100057e-05,
+      "loss": 5.238,
+      "step": 13005
+    },
+    {
+      "epoch": 0.7191818684355997,
+      "grad_norm": 3.4540812969207764,
+      "learning_rate": 1.4440022740193293e-05,
+      "loss": 5.1853,
+      "step": 13010
+    },
+    {
+      "epoch": 0.7194582642343836,
+      "grad_norm": 3.278425455093384,
+      "learning_rate": 1.4425810119386015e-05,
+      "loss": 5.0469,
+      "step": 13015
+    },
+    {
+      "epoch": 0.7197346600331676,
+      "grad_norm": 4.068847179412842,
+      "learning_rate": 1.4411597498578739e-05,
+      "loss": 4.8051,
+      "step": 13020
+    },
+    {
+      "epoch": 0.7200110558319514,
+      "grad_norm": 4.303192615509033,
+      "learning_rate": 1.4397384877771461e-05,
+      "loss": 4.8124,
+      "step": 13025
+    },
+    {
+      "epoch": 0.7202874516307353,
+      "grad_norm": 3.8566951751708984,
+      "learning_rate": 1.4383172256964183e-05,
+      "loss": 4.9964,
+      "step": 13030
+    },
+    {
+      "epoch": 0.7205638474295191,
+      "grad_norm": 3.5754945278167725,
+      "learning_rate": 1.4368959636156909e-05,
+      "loss": 5.055,
+      "step": 13035
+    },
+    {
+      "epoch": 0.720840243228303,
+      "grad_norm": 4.132562637329102,
+      "learning_rate": 1.4354747015349631e-05,
+      "loss": 5.0778,
+      "step": 13040
+    },
+    {
+      "epoch": 0.7211166390270868,
+      "grad_norm": 3.1052451133728027,
+      "learning_rate": 1.4340534394542354e-05,
+      "loss": 4.8714,
+      "step": 13045
+    },
+    {
+      "epoch": 0.7213930348258707,
+      "grad_norm": 3.893859386444092,
+      "learning_rate": 1.4326321773735078e-05,
+      "loss": 5.101,
+      "step": 13050
+    },
+    {
+      "epoch": 0.7216694306246545,
+      "grad_norm": 3.588839054107666,
+      "learning_rate": 1.43121091529278e-05,
+      "loss": 4.6311,
+      "step": 13055
+    },
+    {
+      "epoch": 0.7219458264234384,
+      "grad_norm": 3.061183214187622,
+      "learning_rate": 1.4297896532120526e-05,
+      "loss": 4.9162,
+      "step": 13060
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 3.5575475692749023,
+      "learning_rate": 1.4283683911313248e-05,
+      "loss": 5.3776,
+      "step": 13065
+    },
+    {
+      "epoch": 0.7224986180210061,
+      "grad_norm": 3.373143196105957,
+      "learning_rate": 1.426947129050597e-05,
+      "loss": 4.9563,
+      "step": 13070
+    },
+    {
+      "epoch": 0.7227750138197899,
+      "grad_norm": 4.447892189025879,
+      "learning_rate": 1.4255258669698693e-05,
+      "loss": 4.8872,
+      "step": 13075
+    },
+    {
+      "epoch": 0.7230514096185738,
+      "grad_norm": 4.527839660644531,
+      "learning_rate": 1.4241046048891415e-05,
+      "loss": 4.99,
+      "step": 13080
+    },
+    {
+      "epoch": 0.7233278054173576,
+      "grad_norm": 3.5849502086639404,
+      "learning_rate": 1.4226833428084139e-05,
+      "loss": 4.6041,
+      "step": 13085
+    },
+    {
+      "epoch": 0.7236042012161416,
+      "grad_norm": 3.7297580242156982,
+      "learning_rate": 1.4212620807276863e-05,
+      "loss": 4.948,
+      "step": 13090
+    },
+    {
+      "epoch": 0.7238805970149254,
+      "grad_norm": 3.810117721557617,
+      "learning_rate": 1.4198408186469587e-05,
+      "loss": 4.5659,
+      "step": 13095
+    },
+    {
+      "epoch": 0.7241569928137093,
+      "grad_norm": 3.6611037254333496,
+      "learning_rate": 1.4184195565662309e-05,
+      "loss": 4.7651,
+      "step": 13100
+    },
+    {
+      "epoch": 0.7244333886124931,
+      "grad_norm": 3.81868314743042,
+      "learning_rate": 1.4169982944855031e-05,
+      "loss": 4.8415,
+      "step": 13105
+    },
+    {
+      "epoch": 0.724709784411277,
+      "grad_norm": 3.0699076652526855,
+      "learning_rate": 1.4155770324047754e-05,
+      "loss": 4.8364,
+      "step": 13110
+    },
+    {
+      "epoch": 0.7249861802100608,
+      "grad_norm": 3.566793918609619,
+      "learning_rate": 1.414155770324048e-05,
+      "loss": 4.7472,
+      "step": 13115
+    },
+    {
+      "epoch": 0.7252625760088447,
+      "grad_norm": 3.110363721847534,
+      "learning_rate": 1.4127345082433202e-05,
+      "loss": 4.6858,
+      "step": 13120
+    },
+    {
+      "epoch": 0.7255389718076285,
+      "grad_norm": 3.3906173706054688,
+      "learning_rate": 1.4113132461625926e-05,
+      "loss": 4.9415,
+      "step": 13125
+    },
+    {
+      "epoch": 0.7258153676064124,
+      "grad_norm": 3.598484754562378,
+      "learning_rate": 1.4098919840818648e-05,
+      "loss": 4.9086,
+      "step": 13130
+    },
+    {
+      "epoch": 0.7260917634051962,
+      "grad_norm": 3.2210891246795654,
+      "learning_rate": 1.408470722001137e-05,
+      "loss": 5.1958,
+      "step": 13135
+    },
+    {
+      "epoch": 0.7263681592039801,
+      "grad_norm": 3.400056838989258,
+      "learning_rate": 1.4070494599204092e-05,
+      "loss": 5.2228,
+      "step": 13140
+    },
+    {
+      "epoch": 0.7266445550027639,
+      "grad_norm": 3.5063843727111816,
+      "learning_rate": 1.4056281978396818e-05,
+      "loss": 4.9669,
+      "step": 13145
+    },
+    {
+      "epoch": 0.7269209508015478,
+      "grad_norm": 3.8449692726135254,
+      "learning_rate": 1.404206935758954e-05,
+      "loss": 4.8901,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7271973466003316,
+      "grad_norm": 3.1058237552642822,
+      "learning_rate": 1.4027856736782263e-05,
+      "loss": 5.0105,
+      "step": 13155
+    },
+    {
+      "epoch": 0.7274737423991156,
+      "grad_norm": 3.9088826179504395,
+      "learning_rate": 1.4013644115974987e-05,
+      "loss": 5.0905,
+      "step": 13160
+    },
+    {
+      "epoch": 0.7277501381978994,
+      "grad_norm": 3.3987643718719482,
+      "learning_rate": 1.3999431495167709e-05,
+      "loss": 4.5506,
+      "step": 13165
+    },
+    {
+      "epoch": 0.7280265339966833,
+      "grad_norm": 3.4407613277435303,
+      "learning_rate": 1.3985218874360435e-05,
+      "loss": 4.8973,
+      "step": 13170
+    },
+    {
+      "epoch": 0.7283029297954671,
+      "grad_norm": 3.590932607650757,
+      "learning_rate": 1.3971006253553157e-05,
+      "loss": 5.0722,
+      "step": 13175
+    },
+    {
+      "epoch": 0.728579325594251,
+      "grad_norm": 3.1984472274780273,
+      "learning_rate": 1.3956793632745879e-05,
+      "loss": 4.9033,
+      "step": 13180
+    },
+    {
+      "epoch": 0.7288557213930348,
+      "grad_norm": 3.5542407035827637,
+      "learning_rate": 1.3942581011938601e-05,
+      "loss": 4.8929,
+      "step": 13185
+    },
+    {
+      "epoch": 0.7291321171918187,
+      "grad_norm": 3.841609001159668,
+      "learning_rate": 1.3928368391131324e-05,
+      "loss": 5.0927,
+      "step": 13190
+    },
+    {
+      "epoch": 0.7294085129906025,
+      "grad_norm": 3.5699667930603027,
+      "learning_rate": 1.3914155770324048e-05,
+      "loss": 5.0084,
+      "step": 13195
+    },
+    {
+      "epoch": 0.7296849087893864,
+      "grad_norm": 3.561516046524048,
+      "learning_rate": 1.3899943149516773e-05,
+      "loss": 5.2851,
+      "step": 13200
+    },
+    {
+      "epoch": 0.7299613045881702,
+      "grad_norm": 3.833688497543335,
+      "learning_rate": 1.3885730528709496e-05,
+      "loss": 5.1715,
+      "step": 13205
+    },
+    {
+      "epoch": 0.7302377003869541,
+      "grad_norm": 3.255157232284546,
+      "learning_rate": 1.3871517907902218e-05,
+      "loss": 4.768,
+      "step": 13210
+    },
+    {
+      "epoch": 0.7305140961857379,
+      "grad_norm": 3.3706374168395996,
+      "learning_rate": 1.385730528709494e-05,
+      "loss": 4.8445,
+      "step": 13215
+    },
+    {
+      "epoch": 0.7307904919845218,
+      "grad_norm": 3.0012950897216797,
+      "learning_rate": 1.3843092666287662e-05,
+      "loss": 4.8293,
+      "step": 13220
+    },
+    {
+      "epoch": 0.7310668877833056,
+      "grad_norm": 3.7579739093780518,
+      "learning_rate": 1.3828880045480388e-05,
+      "loss": 5.1837,
+      "step": 13225
+    },
+    {
+      "epoch": 0.7313432835820896,
+      "grad_norm": 3.6330161094665527,
+      "learning_rate": 1.381466742467311e-05,
+      "loss": 4.9532,
+      "step": 13230
+    },
+    {
+      "epoch": 0.7316196793808734,
+      "grad_norm": 4.080511569976807,
+      "learning_rate": 1.3800454803865834e-05,
+      "loss": 5.0781,
+      "step": 13235
+    },
+    {
+      "epoch": 0.7318960751796573,
+      "grad_norm": 3.782904624938965,
+      "learning_rate": 1.3786242183058557e-05,
+      "loss": 4.9824,
+      "step": 13240
+    },
+    {
+      "epoch": 0.7321724709784412,
+      "grad_norm": 3.812208414077759,
+      "learning_rate": 1.3772029562251279e-05,
+      "loss": 5.1982,
+      "step": 13245
+    },
+    {
+      "epoch": 0.732448866777225,
+      "grad_norm": 3.541715145111084,
+      "learning_rate": 1.3757816941444001e-05,
+      "loss": 4.945,
+      "step": 13250
+    },
+    {
+      "epoch": 0.7327252625760089,
+      "grad_norm": 3.311810255050659,
+      "learning_rate": 1.3743604320636727e-05,
+      "loss": 5.0986,
+      "step": 13255
+    },
+    {
+      "epoch": 0.7330016583747927,
+      "grad_norm": 4.002370357513428,
+      "learning_rate": 1.372939169982945e-05,
+      "loss": 4.7093,
+      "step": 13260
+    },
+    {
+      "epoch": 0.7332780541735766,
+      "grad_norm": 3.475691556930542,
+      "learning_rate": 1.3715179079022172e-05,
+      "loss": 5.043,
+      "step": 13265
+    },
+    {
+      "epoch": 0.7335544499723604,
+      "grad_norm": 3.538708448410034,
+      "learning_rate": 1.3700966458214896e-05,
+      "loss": 4.9838,
+      "step": 13270
+    },
+    {
+      "epoch": 0.7338308457711443,
+      "grad_norm": 3.6737594604492188,
+      "learning_rate": 1.3686753837407618e-05,
+      "loss": 4.9001,
+      "step": 13275
+    },
+    {
+      "epoch": 0.7341072415699281,
+      "grad_norm": 3.5421972274780273,
+      "learning_rate": 1.3672541216600343e-05,
+      "loss": 4.9089,
+      "step": 13280
+    },
+    {
+      "epoch": 0.734383637368712,
+      "grad_norm": 3.507235050201416,
+      "learning_rate": 1.3658328595793066e-05,
+      "loss": 5.0143,
+      "step": 13285
+    },
+    {
+      "epoch": 0.7346600331674958,
+      "grad_norm": 2.759843587875366,
+      "learning_rate": 1.3644115974985788e-05,
+      "loss": 4.9846,
+      "step": 13290
+    },
+    {
+      "epoch": 0.7349364289662798,
+      "grad_norm": 3.151954412460327,
+      "learning_rate": 1.362990335417851e-05,
+      "loss": 4.933,
+      "step": 13295
+    },
+    {
+      "epoch": 0.7352128247650636,
+      "grad_norm": 2.9206533432006836,
+      "learning_rate": 1.3615690733371234e-05,
+      "loss": 4.7963,
+      "step": 13300
+    },
+    {
+      "epoch": 0.7354892205638475,
+      "grad_norm": 3.8691372871398926,
+      "learning_rate": 1.3601478112563957e-05,
+      "loss": 5.2145,
+      "step": 13305
+    },
+    {
+      "epoch": 0.7357656163626313,
+      "grad_norm": 3.0782530307769775,
+      "learning_rate": 1.3587265491756682e-05,
+      "loss": 4.6068,
+      "step": 13310
+    },
+    {
+      "epoch": 0.7360420121614152,
+      "grad_norm": 3.9192464351654053,
+      "learning_rate": 1.3573052870949405e-05,
+      "loss": 5.0152,
+      "step": 13315
+    },
+    {
+      "epoch": 0.736318407960199,
+      "grad_norm": 5.0663371086120605,
+      "learning_rate": 1.3558840250142127e-05,
+      "loss": 5.1813,
+      "step": 13320
+    },
+    {
+      "epoch": 0.7365948037589829,
+      "grad_norm": 3.4639241695404053,
+      "learning_rate": 1.3544627629334849e-05,
+      "loss": 5.0714,
+      "step": 13325
+    },
+    {
+      "epoch": 0.7368711995577667,
+      "grad_norm": 3.764646053314209,
+      "learning_rate": 1.3530415008527571e-05,
+      "loss": 4.9129,
+      "step": 13330
+    },
+    {
+      "epoch": 0.7371475953565506,
+      "grad_norm": 3.288600206375122,
+      "learning_rate": 1.3516202387720297e-05,
+      "loss": 4.8982,
+      "step": 13335
+    },
+    {
+      "epoch": 0.7374239911553344,
+      "grad_norm": 3.8759982585906982,
+      "learning_rate": 1.350198976691302e-05,
+      "loss": 5.1059,
+      "step": 13340
+    },
+    {
+      "epoch": 0.7377003869541183,
+      "grad_norm": 3.3489646911621094,
+      "learning_rate": 1.3487777146105743e-05,
+      "loss": 5.1768,
+      "step": 13345
+    },
+    {
+      "epoch": 0.7379767827529021,
+      "grad_norm": 3.005626916885376,
+      "learning_rate": 1.3473564525298466e-05,
+      "loss": 5.0332,
+      "step": 13350
+    },
+    {
+      "epoch": 0.738253178551686,
+      "grad_norm": 3.416574239730835,
+      "learning_rate": 1.3459351904491188e-05,
+      "loss": 5.0185,
+      "step": 13355
+    },
+    {
+      "epoch": 0.7385295743504698,
+      "grad_norm": 3.5634119510650635,
+      "learning_rate": 1.344513928368391e-05,
+      "loss": 4.7656,
+      "step": 13360
+    },
+    {
+      "epoch": 0.7388059701492538,
+      "grad_norm": 3.0939366817474365,
+      "learning_rate": 1.3430926662876636e-05,
+      "loss": 4.8285,
+      "step": 13365
+    },
+    {
+      "epoch": 0.7390823659480376,
+      "grad_norm": 3.3977231979370117,
+      "learning_rate": 1.3416714042069358e-05,
+      "loss": 4.9093,
+      "step": 13370
+    },
+    {
+      "epoch": 0.7393587617468215,
+      "grad_norm": 3.6465513706207275,
+      "learning_rate": 1.3402501421262082e-05,
+      "loss": 5.0178,
+      "step": 13375
+    },
+    {
+      "epoch": 0.7396351575456053,
+      "grad_norm": 3.207108974456787,
+      "learning_rate": 1.3388288800454804e-05,
+      "loss": 5.2148,
+      "step": 13380
+    },
+    {
+      "epoch": 0.7399115533443892,
+      "grad_norm": 3.185753107070923,
+      "learning_rate": 1.3374076179647527e-05,
+      "loss": 5.0929,
+      "step": 13385
+    },
+    {
+      "epoch": 0.740187949143173,
+      "grad_norm": 4.584200859069824,
+      "learning_rate": 1.3359863558840252e-05,
+      "loss": 4.651,
+      "step": 13390
+    },
+    {
+      "epoch": 0.7404643449419569,
+      "grad_norm": 3.067399024963379,
+      "learning_rate": 1.3345650938032975e-05,
+      "loss": 5.0875,
+      "step": 13395
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 3.611589193344116,
+      "learning_rate": 1.3331438317225697e-05,
+      "loss": 4.725,
+      "step": 13400
+    },
+    {
+      "epoch": 0.7410171365395246,
+      "grad_norm": 3.321446418762207,
+      "learning_rate": 1.331722569641842e-05,
+      "loss": 4.8858,
+      "step": 13405
+    },
+    {
+      "epoch": 0.7412935323383084,
+      "grad_norm": 3.519653797149658,
+      "learning_rate": 1.3303013075611143e-05,
+      "loss": 4.645,
+      "step": 13410
+    },
+    {
+      "epoch": 0.7415699281370923,
+      "grad_norm": 3.325186014175415,
+      "learning_rate": 1.3288800454803865e-05,
+      "loss": 4.8125,
+      "step": 13415
+    },
+    {
+      "epoch": 0.7418463239358761,
+      "grad_norm": 3.7790181636810303,
+      "learning_rate": 1.3274587833996591e-05,
+      "loss": 4.9224,
+      "step": 13420
+    },
+    {
+      "epoch": 0.74212271973466,
+      "grad_norm": 3.3539559841156006,
+      "learning_rate": 1.3260375213189313e-05,
+      "loss": 4.5994,
+      "step": 13425
+    },
+    {
+      "epoch": 0.7423991155334438,
+      "grad_norm": 3.002809762954712,
+      "learning_rate": 1.3246162592382036e-05,
+      "loss": 5.2474,
+      "step": 13430
+    },
+    {
+      "epoch": 0.7426755113322278,
+      "grad_norm": 3.0804951190948486,
+      "learning_rate": 1.3231949971574758e-05,
+      "loss": 4.6779,
+      "step": 13435
+    },
+    {
+      "epoch": 0.7429519071310116,
+      "grad_norm": 4.401519775390625,
+      "learning_rate": 1.3217737350767482e-05,
+      "loss": 4.8666,
+      "step": 13440
+    },
+    {
+      "epoch": 0.7432283029297955,
+      "grad_norm": 3.0811357498168945,
+      "learning_rate": 1.3203524729960206e-05,
+      "loss": 4.909,
+      "step": 13445
+    },
+    {
+      "epoch": 0.7435046987285793,
+      "grad_norm": 4.676823139190674,
+      "learning_rate": 1.318931210915293e-05,
+      "loss": 5.3051,
+      "step": 13450
+    },
+    {
+      "epoch": 0.7437810945273632,
+      "grad_norm": 3.572274684906006,
+      "learning_rate": 1.3175099488345652e-05,
+      "loss": 4.8484,
+      "step": 13455
+    },
+    {
+      "epoch": 0.7440574903261471,
+      "grad_norm": 3.933598041534424,
+      "learning_rate": 1.3160886867538375e-05,
+      "loss": 4.9078,
+      "step": 13460
+    },
+    {
+      "epoch": 0.7443338861249309,
+      "grad_norm": 4.784248352050781,
+      "learning_rate": 1.3146674246731097e-05,
+      "loss": 4.9795,
+      "step": 13465
+    },
+    {
+      "epoch": 0.7446102819237148,
+      "grad_norm": 3.54833984375,
+      "learning_rate": 1.3132461625923819e-05,
+      "loss": 5.0677,
+      "step": 13470
+    },
+    {
+      "epoch": 0.7448866777224986,
+      "grad_norm": 3.983644962310791,
+      "learning_rate": 1.3118249005116545e-05,
+      "loss": 4.8627,
+      "step": 13475
+    },
+    {
+      "epoch": 0.7451630735212825,
+      "grad_norm": 3.837332010269165,
+      "learning_rate": 1.3104036384309267e-05,
+      "loss": 4.7729,
+      "step": 13480
+    },
+    {
+      "epoch": 0.7454394693200663,
+      "grad_norm": 4.445468425750732,
+      "learning_rate": 1.3089823763501991e-05,
+      "loss": 5.0954,
+      "step": 13485
+    },
+    {
+      "epoch": 0.7457158651188502,
+      "grad_norm": 3.518786668777466,
+      "learning_rate": 1.3075611142694713e-05,
+      "loss": 4.855,
+      "step": 13490
+    },
+    {
+      "epoch": 0.745992260917634,
+      "grad_norm": 3.9900107383728027,
+      "learning_rate": 1.3061398521887436e-05,
+      "loss": 4.7721,
+      "step": 13495
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 3.3050553798675537,
+      "learning_rate": 1.3047185901080161e-05,
+      "loss": 4.9431,
+      "step": 13500
+    },
+    {
+      "epoch": 0.7465450525152018,
+      "grad_norm": 3.213809013366699,
+      "learning_rate": 1.3032973280272884e-05,
+      "loss": 5.172,
+      "step": 13505
+    },
+    {
+      "epoch": 0.7468214483139857,
+      "grad_norm": 3.7360849380493164,
+      "learning_rate": 1.3018760659465606e-05,
+      "loss": 5.0361,
+      "step": 13510
+    },
+    {
+      "epoch": 0.7470978441127695,
+      "grad_norm": 4.024673938751221,
+      "learning_rate": 1.300454803865833e-05,
+      "loss": 4.8585,
+      "step": 13515
+    },
+    {
+      "epoch": 0.7473742399115534,
+      "grad_norm": 3.3230862617492676,
+      "learning_rate": 1.2990335417851052e-05,
+      "loss": 5.1286,
+      "step": 13520
+    },
+    {
+      "epoch": 0.7476506357103372,
+      "grad_norm": 4.014267921447754,
+      "learning_rate": 1.2976122797043774e-05,
+      "loss": 5.0204,
+      "step": 13525
+    },
+    {
+      "epoch": 0.7479270315091211,
+      "grad_norm": 3.804960250854492,
+      "learning_rate": 1.29619101762365e-05,
+      "loss": 4.6567,
+      "step": 13530
+    },
+    {
+      "epoch": 0.7482034273079049,
+      "grad_norm": 3.149284601211548,
+      "learning_rate": 1.2947697555429222e-05,
+      "loss": 4.857,
+      "step": 13535
+    },
+    {
+      "epoch": 0.7484798231066888,
+      "grad_norm": 3.5097615718841553,
+      "learning_rate": 1.2933484934621945e-05,
+      "loss": 4.5143,
+      "step": 13540
+    },
+    {
+      "epoch": 0.7487562189054726,
+      "grad_norm": 3.3128249645233154,
+      "learning_rate": 1.2919272313814667e-05,
+      "loss": 5.1963,
+      "step": 13545
+    },
+    {
+      "epoch": 0.7490326147042565,
+      "grad_norm": 3.3906478881835938,
+      "learning_rate": 1.2905059693007391e-05,
+      "loss": 4.741,
+      "step": 13550
+    },
+    {
+      "epoch": 0.7493090105030403,
+      "grad_norm": 3.369476556777954,
+      "learning_rate": 1.2890847072200115e-05,
+      "loss": 5.0067,
+      "step": 13555
+    },
+    {
+      "epoch": 0.7495854063018242,
+      "grad_norm": 4.018293857574463,
+      "learning_rate": 1.2876634451392839e-05,
+      "loss": 5.2337,
+      "step": 13560
+    },
+    {
+      "epoch": 0.749861802100608,
+      "grad_norm": 4.163745880126953,
+      "learning_rate": 1.2862421830585561e-05,
+      "loss": 4.9956,
+      "step": 13565
+    },
+    {
+      "epoch": 0.750138197899392,
+      "grad_norm": 3.6494219303131104,
+      "learning_rate": 1.2848209209778283e-05,
+      "loss": 4.6203,
+      "step": 13570
+    },
+    {
+      "epoch": 0.7504145936981758,
+      "grad_norm": 3.210977077484131,
+      "learning_rate": 1.2833996588971006e-05,
+      "loss": 4.9079,
+      "step": 13575
+    },
+    {
+      "epoch": 0.7506909894969597,
+      "grad_norm": 3.0875606536865234,
+      "learning_rate": 1.2819783968163728e-05,
+      "loss": 5.0435,
+      "step": 13580
+    },
+    {
+      "epoch": 0.7509673852957435,
+      "grad_norm": 3.371025323867798,
+      "learning_rate": 1.2805571347356454e-05,
+      "loss": 5.0249,
+      "step": 13585
+    },
+    {
+      "epoch": 0.7512437810945274,
+      "grad_norm": 3.431417226791382,
+      "learning_rate": 1.2791358726549178e-05,
+      "loss": 4.8918,
+      "step": 13590
+    },
+    {
+      "epoch": 0.7515201768933112,
+      "grad_norm": 4.608222007751465,
+      "learning_rate": 1.27771461057419e-05,
+      "loss": 4.8191,
+      "step": 13595
+    },
+    {
+      "epoch": 0.7517965726920951,
+      "grad_norm": 4.438573360443115,
+      "learning_rate": 1.2762933484934622e-05,
+      "loss": 5.1095,
+      "step": 13600
+    },
+    {
+      "epoch": 0.7520729684908789,
+      "grad_norm": 3.5704104900360107,
+      "learning_rate": 1.2748720864127345e-05,
+      "loss": 4.8193,
+      "step": 13605
+    },
+    {
+      "epoch": 0.7523493642896628,
+      "grad_norm": 3.665001630783081,
+      "learning_rate": 1.273450824332007e-05,
+      "loss": 4.9377,
+      "step": 13610
+    },
+    {
+      "epoch": 0.7526257600884466,
+      "grad_norm": 4.091150760650635,
+      "learning_rate": 1.2720295622512792e-05,
+      "loss": 4.758,
+      "step": 13615
+    },
+    {
+      "epoch": 0.7529021558872305,
+      "grad_norm": 3.687591791152954,
+      "learning_rate": 1.2706083001705515e-05,
+      "loss": 4.9703,
+      "step": 13620
+    },
+    {
+      "epoch": 0.7531785516860143,
+      "grad_norm": 3.9441850185394287,
+      "learning_rate": 1.2691870380898239e-05,
+      "loss": 5.0439,
+      "step": 13625
+    },
+    {
+      "epoch": 0.7534549474847982,
+      "grad_norm": 3.240487575531006,
+      "learning_rate": 1.2677657760090961e-05,
+      "loss": 4.6799,
+      "step": 13630
+    },
+    {
+      "epoch": 0.753731343283582,
+      "grad_norm": 3.4183409214019775,
+      "learning_rate": 1.2663445139283683e-05,
+      "loss": 5.1698,
+      "step": 13635
+    },
+    {
+      "epoch": 0.754007739082366,
+      "grad_norm": 3.549414873123169,
+      "learning_rate": 1.2649232518476409e-05,
+      "loss": 4.9135,
+      "step": 13640
+    },
+    {
+      "epoch": 0.7542841348811498,
+      "grad_norm": 3.5168375968933105,
+      "learning_rate": 1.2635019897669131e-05,
+      "loss": 5.078,
+      "step": 13645
+    },
+    {
+      "epoch": 0.7545605306799337,
+      "grad_norm": 2.9269330501556396,
+      "learning_rate": 1.2620807276861854e-05,
+      "loss": 4.9437,
+      "step": 13650
+    },
+    {
+      "epoch": 0.7548369264787175,
+      "grad_norm": 3.613555908203125,
+      "learning_rate": 1.2606594656054576e-05,
+      "loss": 5.1616,
+      "step": 13655
+    },
+    {
+      "epoch": 0.7551133222775014,
+      "grad_norm": 3.630427122116089,
+      "learning_rate": 1.25923820352473e-05,
+      "loss": 5.0425,
+      "step": 13660
+    },
+    {
+      "epoch": 0.7553897180762852,
+      "grad_norm": 3.5098865032196045,
+      "learning_rate": 1.2578169414440025e-05,
+      "loss": 4.9677,
+      "step": 13665
+    },
+    {
+      "epoch": 0.7556661138750691,
+      "grad_norm": 3.6350200176239014,
+      "learning_rate": 1.2563956793632748e-05,
+      "loss": 5.1123,
+      "step": 13670
+    },
+    {
+      "epoch": 0.7559425096738529,
+      "grad_norm": 3.2419891357421875,
+      "learning_rate": 1.254974417282547e-05,
+      "loss": 5.0052,
+      "step": 13675
+    },
+    {
+      "epoch": 0.7562189054726368,
+      "grad_norm": 3.6156861782073975,
+      "learning_rate": 1.2535531552018192e-05,
+      "loss": 4.6917,
+      "step": 13680
+    },
+    {
+      "epoch": 0.7564953012714207,
+      "grad_norm": 3.5388574600219727,
+      "learning_rate": 1.2521318931210915e-05,
+      "loss": 4.9765,
+      "step": 13685
+    },
+    {
+      "epoch": 0.7567716970702045,
+      "grad_norm": 3.379070997238159,
+      "learning_rate": 1.2507106310403639e-05,
+      "loss": 4.8777,
+      "step": 13690
+    },
+    {
+      "epoch": 0.7570480928689884,
+      "grad_norm": 3.054107904434204,
+      "learning_rate": 1.2492893689596363e-05,
+      "loss": 4.6504,
+      "step": 13695
+    },
+    {
+      "epoch": 0.7573244886677722,
+      "grad_norm": 4.416358947753906,
+      "learning_rate": 1.2478681068789087e-05,
+      "loss": 4.9261,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7576008844665562,
+      "grad_norm": 3.9947686195373535,
+      "learning_rate": 1.2464468447981809e-05,
+      "loss": 5.0292,
+      "step": 13705
+    },
+    {
+      "epoch": 0.75787728026534,
+      "grad_norm": 3.1346933841705322,
+      "learning_rate": 1.2450255827174531e-05,
+      "loss": 4.9894,
+      "step": 13710
+    },
+    {
+      "epoch": 0.7581536760641239,
+      "grad_norm": 3.530553102493286,
+      "learning_rate": 1.2436043206367255e-05,
+      "loss": 5.091,
+      "step": 13715
+    },
+    {
+      "epoch": 0.7584300718629077,
+      "grad_norm": 3.7621891498565674,
+      "learning_rate": 1.2421830585559977e-05,
+      "loss": 4.74,
+      "step": 13720
+    },
+    {
+      "epoch": 0.7587064676616916,
+      "grad_norm": 3.6707799434661865,
+      "learning_rate": 1.24076179647527e-05,
+      "loss": 5.1945,
+      "step": 13725
+    },
+    {
+      "epoch": 0.7589828634604754,
+      "grad_norm": 3.413144111633301,
+      "learning_rate": 1.2393405343945425e-05,
+      "loss": 4.9481,
+      "step": 13730
+    },
+    {
+      "epoch": 0.7592592592592593,
+      "grad_norm": 3.7922909259796143,
+      "learning_rate": 1.2379192723138148e-05,
+      "loss": 5.0266,
+      "step": 13735
+    },
+    {
+      "epoch": 0.7595356550580431,
+      "grad_norm": 3.63356876373291,
+      "learning_rate": 1.236498010233087e-05,
+      "loss": 5.0376,
+      "step": 13740
+    },
+    {
+      "epoch": 0.759812050856827,
+      "grad_norm": 3.9122843742370605,
+      "learning_rate": 1.2350767481523594e-05,
+      "loss": 4.7132,
+      "step": 13745
+    },
+    {
+      "epoch": 0.7600884466556108,
+      "grad_norm": 4.200931072235107,
+      "learning_rate": 1.2336554860716316e-05,
+      "loss": 4.8281,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7603648424543947,
+      "grad_norm": 3.4712514877319336,
+      "learning_rate": 1.232234223990904e-05,
+      "loss": 4.9505,
+      "step": 13755
+    },
+    {
+      "epoch": 0.7606412382531785,
+      "grad_norm": 3.5167579650878906,
+      "learning_rate": 1.2308129619101762e-05,
+      "loss": 4.8998,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7609176340519624,
+      "grad_norm": 2.8712246417999268,
+      "learning_rate": 1.2293916998294486e-05,
+      "loss": 4.9693,
+      "step": 13765
+    },
+    {
+      "epoch": 0.7611940298507462,
+      "grad_norm": 3.5918309688568115,
+      "learning_rate": 1.227970437748721e-05,
+      "loss": 5.3341,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7614704256495302,
+      "grad_norm": 3.4594647884368896,
+      "learning_rate": 1.2265491756679933e-05,
+      "loss": 4.7239,
+      "step": 13775
+    },
+    {
+      "epoch": 0.761746821448314,
+      "grad_norm": 3.583055257797241,
+      "learning_rate": 1.2251279135872655e-05,
+      "loss": 4.8022,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7620232172470979,
+      "grad_norm": 3.9287872314453125,
+      "learning_rate": 1.2237066515065379e-05,
+      "loss": 4.6748,
+      "step": 13785
+    },
+    {
+      "epoch": 0.7622996130458817,
+      "grad_norm": 4.147637367248535,
+      "learning_rate": 1.2222853894258101e-05,
+      "loss": 4.9476,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7625760088446656,
+      "grad_norm": 3.5439789295196533,
+      "learning_rate": 1.2208641273450824e-05,
+      "loss": 4.9426,
+      "step": 13795
+    },
+    {
+      "epoch": 0.7628524046434494,
+      "grad_norm": 3.4278995990753174,
+      "learning_rate": 1.2194428652643548e-05,
+      "loss": 5.0638,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7631288004422333,
+      "grad_norm": 3.199315071105957,
+      "learning_rate": 1.2180216031836271e-05,
+      "loss": 4.9484,
+      "step": 13805
+    },
+    {
+      "epoch": 0.7634051962410171,
+      "grad_norm": 3.8618929386138916,
+      "learning_rate": 1.2166003411028995e-05,
+      "loss": 4.7998,
+      "step": 13810
+    },
+    {
+      "epoch": 0.763681592039801,
+      "grad_norm": 3.2143681049346924,
+      "learning_rate": 1.2151790790221718e-05,
+      "loss": 4.9469,
+      "step": 13815
+    },
+    {
+      "epoch": 0.7639579878385848,
+      "grad_norm": 3.6355247497558594,
+      "learning_rate": 1.213757816941444e-05,
+      "loss": 4.7807,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7642343836373687,
+      "grad_norm": 4.1098246574401855,
+      "learning_rate": 1.2123365548607164e-05,
+      "loss": 4.9628,
+      "step": 13825
+    },
+    {
+      "epoch": 0.7645107794361525,
+      "grad_norm": 3.3956034183502197,
+      "learning_rate": 1.2109152927799886e-05,
+      "loss": 5.1022,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7647871752349364,
+      "grad_norm": 3.51596736907959,
+      "learning_rate": 1.209494030699261e-05,
+      "loss": 4.7867,
+      "step": 13835
+    },
+    {
+      "epoch": 0.7650635710337202,
+      "grad_norm": 3.4072954654693604,
+      "learning_rate": 1.2080727686185334e-05,
+      "loss": 5.0985,
+      "step": 13840
+    },
+    {
+      "epoch": 0.7653399668325042,
+      "grad_norm": 2.994511842727661,
+      "learning_rate": 1.2066515065378057e-05,
+      "loss": 4.9816,
+      "step": 13845
+    },
+    {
+      "epoch": 0.765616362631288,
+      "grad_norm": 3.5091991424560547,
+      "learning_rate": 1.2052302444570779e-05,
+      "loss": 5.0666,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7658927584300719,
+      "grad_norm": 3.4935755729675293,
+      "learning_rate": 1.2038089823763503e-05,
+      "loss": 4.8671,
+      "step": 13855
+    },
+    {
+      "epoch": 0.7661691542288557,
+      "grad_norm": 3.766550302505493,
+      "learning_rate": 1.2023877202956225e-05,
+      "loss": 4.9597,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7664455500276396,
+      "grad_norm": 3.4647650718688965,
+      "learning_rate": 1.2009664582148949e-05,
+      "loss": 5.0242,
+      "step": 13865
+    },
+    {
+      "epoch": 0.7667219458264234,
+      "grad_norm": 3.2583017349243164,
+      "learning_rate": 1.1995451961341671e-05,
+      "loss": 5.1654,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7669983416252073,
+      "grad_norm": 4.087442874908447,
+      "learning_rate": 1.1981239340534395e-05,
+      "loss": 5.029,
+      "step": 13875
+    },
+    {
+      "epoch": 0.7672747374239911,
+      "grad_norm": 3.532439947128296,
+      "learning_rate": 1.196702671972712e-05,
+      "loss": 4.7802,
+      "step": 13880
+    },
+    {
+      "epoch": 0.767551133222775,
+      "grad_norm": 2.9457197189331055,
+      "learning_rate": 1.1952814098919842e-05,
+      "loss": 4.8096,
+      "step": 13885
+    },
+    {
+      "epoch": 0.7678275290215588,
+      "grad_norm": 3.6090781688690186,
+      "learning_rate": 1.1938601478112564e-05,
+      "loss": 4.9288,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7681039248203427,
+      "grad_norm": 4.044902801513672,
+      "learning_rate": 1.1924388857305288e-05,
+      "loss": 4.8477,
+      "step": 13895
+    },
+    {
+      "epoch": 0.7683803206191266,
+      "grad_norm": 4.630081653594971,
+      "learning_rate": 1.191017623649801e-05,
+      "loss": 4.9998,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7686567164179104,
+      "grad_norm": 4.869181156158447,
+      "learning_rate": 1.1895963615690734e-05,
+      "loss": 4.8786,
+      "step": 13905
+    },
+    {
+      "epoch": 0.7689331122166944,
+      "grad_norm": 3.5769691467285156,
+      "learning_rate": 1.1881750994883458e-05,
+      "loss": 4.9105,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7692095080154782,
+      "grad_norm": 2.8430140018463135,
+      "learning_rate": 1.186753837407618e-05,
+      "loss": 4.9828,
+      "step": 13915
+    },
+    {
+      "epoch": 0.7694859038142621,
+      "grad_norm": 3.4973304271698,
+      "learning_rate": 1.1853325753268904e-05,
+      "loss": 4.7579,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7697622996130459,
+      "grad_norm": 3.7342896461486816,
+      "learning_rate": 1.1839113132461627e-05,
+      "loss": 5.0691,
+      "step": 13925
+    },
+    {
+      "epoch": 0.7700386954118298,
+      "grad_norm": 3.4098775386810303,
+      "learning_rate": 1.1824900511654349e-05,
+      "loss": 4.8514,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7703150912106136,
+      "grad_norm": 3.3334388732910156,
+      "learning_rate": 1.1810687890847073e-05,
+      "loss": 5.1646,
+      "step": 13935
+    },
+    {
+      "epoch": 0.7705914870093975,
+      "grad_norm": 4.081144332885742,
+      "learning_rate": 1.1796475270039795e-05,
+      "loss": 5.1261,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7708678828081813,
+      "grad_norm": 3.5878841876983643,
+      "learning_rate": 1.178226264923252e-05,
+      "loss": 4.9911,
+      "step": 13945
+    },
+    {
+      "epoch": 0.7711442786069652,
+      "grad_norm": 3.167708158493042,
+      "learning_rate": 1.1768050028425243e-05,
+      "loss": 4.8064,
+      "step": 13950
+    },
+    {
+      "epoch": 0.771420674405749,
+      "grad_norm": 4.6932806968688965,
+      "learning_rate": 1.1753837407617965e-05,
+      "loss": 5.0513,
+      "step": 13955
+    },
+    {
+      "epoch": 0.7716970702045329,
+      "grad_norm": 3.659235954284668,
+      "learning_rate": 1.1739624786810688e-05,
+      "loss": 4.88,
+      "step": 13960
+    },
+    {
+      "epoch": 0.7719734660033167,
+      "grad_norm": 3.2741990089416504,
+      "learning_rate": 1.1725412166003412e-05,
+      "loss": 4.7233,
+      "step": 13965
+    },
+    {
+      "epoch": 0.7722498618021006,
+      "grad_norm": 3.8330390453338623,
+      "learning_rate": 1.1711199545196134e-05,
+      "loss": 4.9704,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7725262576008844,
+      "grad_norm": 3.3870649337768555,
+      "learning_rate": 1.1696986924388858e-05,
+      "loss": 4.7279,
+      "step": 13975
+    },
+    {
+      "epoch": 0.7728026533996684,
+      "grad_norm": 3.542768716812134,
+      "learning_rate": 1.1682774303581582e-05,
+      "loss": 5.0573,
+      "step": 13980
+    },
+    {
+      "epoch": 0.7730790491984522,
+      "grad_norm": 4.090266227722168,
+      "learning_rate": 1.1668561682774304e-05,
+      "loss": 5.2034,
+      "step": 13985
+    },
+    {
+      "epoch": 0.7733554449972361,
+      "grad_norm": 3.7428195476531982,
+      "learning_rate": 1.1654349061967028e-05,
+      "loss": 5.0463,
+      "step": 13990
+    },
+    {
+      "epoch": 0.7736318407960199,
+      "grad_norm": 4.728503227233887,
+      "learning_rate": 1.164013644115975e-05,
+      "loss": 4.7835,
+      "step": 13995
+    },
+    {
+      "epoch": 0.7739082365948038,
+      "grad_norm": 3.689528226852417,
+      "learning_rate": 1.1625923820352473e-05,
+      "loss": 5.2645,
+      "step": 14000
+    },
+    {
+      "epoch": 0.7741846323935876,
+      "grad_norm": 4.224091529846191,
+      "learning_rate": 1.1611711199545197e-05,
+      "loss": 4.7057,
+      "step": 14005
+    },
+    {
+      "epoch": 0.7744610281923715,
+      "grad_norm": 3.54939866065979,
+      "learning_rate": 1.1597498578737919e-05,
+      "loss": 5.2695,
+      "step": 14010
+    },
+    {
+      "epoch": 0.7747374239911553,
+      "grad_norm": 3.281078338623047,
+      "learning_rate": 1.1583285957930643e-05,
+      "loss": 5.0044,
+      "step": 14015
+    },
+    {
+      "epoch": 0.7750138197899392,
+      "grad_norm": 5.049131870269775,
+      "learning_rate": 1.1569073337123367e-05,
+      "loss": 5.2212,
+      "step": 14020
+    },
+    {
+      "epoch": 0.775290215588723,
+      "grad_norm": 3.362013339996338,
+      "learning_rate": 1.155486071631609e-05,
+      "loss": 4.8976,
+      "step": 14025
+    },
+    {
+      "epoch": 0.7755666113875069,
+      "grad_norm": 5.27224063873291,
+      "learning_rate": 1.1540648095508812e-05,
+      "loss": 4.898,
+      "step": 14030
+    },
+    {
+      "epoch": 0.7758430071862907,
+      "grad_norm": 3.802133560180664,
+      "learning_rate": 1.1526435474701536e-05,
+      "loss": 4.8342,
+      "step": 14035
+    },
+    {
+      "epoch": 0.7761194029850746,
+      "grad_norm": 3.688088893890381,
+      "learning_rate": 1.1512222853894258e-05,
+      "loss": 4.8441,
+      "step": 14040
+    },
+    {
+      "epoch": 0.7763957987838584,
+      "grad_norm": 3.9619669914245605,
+      "learning_rate": 1.1498010233086982e-05,
+      "loss": 5.2455,
+      "step": 14045
+    },
+    {
+      "epoch": 0.7766721945826424,
+      "grad_norm": 3.759188413619995,
+      "learning_rate": 1.1483797612279706e-05,
+      "loss": 5.1307,
+      "step": 14050
+    },
+    {
+      "epoch": 0.7769485903814262,
+      "grad_norm": 3.8243587017059326,
+      "learning_rate": 1.1469584991472428e-05,
+      "loss": 4.9645,
+      "step": 14055
+    },
+    {
+      "epoch": 0.7772249861802101,
+      "grad_norm": 3.4312751293182373,
+      "learning_rate": 1.1455372370665152e-05,
+      "loss": 4.9075,
+      "step": 14060
+    },
+    {
+      "epoch": 0.7775013819789939,
+      "grad_norm": 3.4000275135040283,
+      "learning_rate": 1.1441159749857874e-05,
+      "loss": 4.8989,
+      "step": 14065
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 3.78006911277771,
+      "learning_rate": 1.1426947129050597e-05,
+      "loss": 5.2333,
+      "step": 14070
+    },
+    {
+      "epoch": 0.7780541735765616,
+      "grad_norm": 3.4480984210968018,
+      "learning_rate": 1.141273450824332e-05,
+      "loss": 4.9613,
+      "step": 14075
+    },
+    {
+      "epoch": 0.7783305693753455,
+      "grad_norm": 3.377744674682617,
+      "learning_rate": 1.1398521887436043e-05,
+      "loss": 5.0646,
+      "step": 14080
+    },
+    {
+      "epoch": 0.7786069651741293,
+      "grad_norm": 3.1980888843536377,
+      "learning_rate": 1.1384309266628767e-05,
+      "loss": 4.9803,
+      "step": 14085
+    },
+    {
+      "epoch": 0.7788833609729132,
+      "grad_norm": 3.6203091144561768,
+      "learning_rate": 1.1370096645821491e-05,
+      "loss": 4.8729,
+      "step": 14090
+    },
+    {
+      "epoch": 0.779159756771697,
+      "grad_norm": 3.161928176879883,
+      "learning_rate": 1.1355884025014213e-05,
+      "loss": 5.1262,
+      "step": 14095
+    },
+    {
+      "epoch": 0.7794361525704809,
+      "grad_norm": 3.8327505588531494,
+      "learning_rate": 1.1341671404206937e-05,
+      "loss": 5.0207,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7797125483692647,
+      "grad_norm": 3.690469980239868,
+      "learning_rate": 1.132745878339966e-05,
+      "loss": 5.0829,
+      "step": 14105
+    },
+    {
+      "epoch": 0.7799889441680486,
+      "grad_norm": 3.06017804145813,
+      "learning_rate": 1.1313246162592382e-05,
+      "loss": 5.2518,
+      "step": 14110
+    },
+    {
+      "epoch": 0.7802653399668324,
+      "grad_norm": 4.095527172088623,
+      "learning_rate": 1.1299033541785106e-05,
+      "loss": 5.1885,
+      "step": 14115
+    },
+    {
+      "epoch": 0.7805417357656164,
+      "grad_norm": 3.2654521465301514,
+      "learning_rate": 1.128482092097783e-05,
+      "loss": 4.9736,
+      "step": 14120
+    },
+    {
+      "epoch": 0.7808181315644003,
+      "grad_norm": 3.9866836071014404,
+      "learning_rate": 1.1270608300170552e-05,
+      "loss": 5.0144,
+      "step": 14125
+    },
+    {
+      "epoch": 0.7810945273631841,
+      "grad_norm": 4.128415584564209,
+      "learning_rate": 1.1256395679363276e-05,
+      "loss": 5.0449,
+      "step": 14130
+    },
+    {
+      "epoch": 0.781370923161968,
+      "grad_norm": 3.3916735649108887,
+      "learning_rate": 1.1242183058555998e-05,
+      "loss": 5.0108,
+      "step": 14135
+    },
+    {
+      "epoch": 0.7816473189607518,
+      "grad_norm": 3.543978452682495,
+      "learning_rate": 1.122797043774872e-05,
+      "loss": 4.8539,
+      "step": 14140
+    },
+    {
+      "epoch": 0.7819237147595357,
+      "grad_norm": 3.8419699668884277,
+      "learning_rate": 1.1213757816941444e-05,
+      "loss": 4.722,
+      "step": 14145
+    },
+    {
+      "epoch": 0.7822001105583195,
+      "grad_norm": 3.7718849182128906,
+      "learning_rate": 1.1199545196134167e-05,
+      "loss": 4.9846,
+      "step": 14150
+    },
+    {
+      "epoch": 0.7824765063571034,
+      "grad_norm": 3.22829532623291,
+      "learning_rate": 1.118533257532689e-05,
+      "loss": 4.6035,
+      "step": 14155
+    },
+    {
+      "epoch": 0.7827529021558872,
+      "grad_norm": 3.4191575050354004,
+      "learning_rate": 1.1171119954519615e-05,
+      "loss": 5.0507,
+      "step": 14160
+    },
+    {
+      "epoch": 0.7830292979546711,
+      "grad_norm": 4.302538871765137,
+      "learning_rate": 1.1156907333712337e-05,
+      "loss": 4.8843,
+      "step": 14165
+    },
+    {
+      "epoch": 0.7833056937534549,
+      "grad_norm": 4.03009557723999,
+      "learning_rate": 1.1142694712905061e-05,
+      "loss": 4.8532,
+      "step": 14170
+    },
+    {
+      "epoch": 0.7835820895522388,
+      "grad_norm": 3.964759349822998,
+      "learning_rate": 1.1128482092097783e-05,
+      "loss": 5.3011,
+      "step": 14175
+    },
+    {
+      "epoch": 0.7838584853510226,
+      "grad_norm": 3.5552785396575928,
+      "learning_rate": 1.1114269471290506e-05,
+      "loss": 5.0725,
+      "step": 14180
+    },
+    {
+      "epoch": 0.7841348811498066,
+      "grad_norm": 3.765488862991333,
+      "learning_rate": 1.110005685048323e-05,
+      "loss": 4.8646,
+      "step": 14185
+    },
+    {
+      "epoch": 0.7844112769485904,
+      "grad_norm": 3.883822441101074,
+      "learning_rate": 1.1085844229675952e-05,
+      "loss": 4.9223,
+      "step": 14190
+    },
+    {
+      "epoch": 0.7846876727473743,
+      "grad_norm": 3.70638370513916,
+      "learning_rate": 1.1071631608868676e-05,
+      "loss": 5.1099,
+      "step": 14195
+    },
+    {
+      "epoch": 0.7849640685461581,
+      "grad_norm": 4.200986385345459,
+      "learning_rate": 1.10574189880614e-05,
+      "loss": 4.722,
+      "step": 14200
+    },
+    {
+      "epoch": 0.785240464344942,
+      "grad_norm": 3.2610623836517334,
+      "learning_rate": 1.1043206367254122e-05,
+      "loss": 4.8481,
+      "step": 14205
+    },
+    {
+      "epoch": 0.7855168601437258,
+      "grad_norm": 3.607994556427002,
+      "learning_rate": 1.1028993746446846e-05,
+      "loss": 4.6585,
+      "step": 14210
+    },
+    {
+      "epoch": 0.7857932559425097,
+      "grad_norm": 3.5993244647979736,
+      "learning_rate": 1.1014781125639568e-05,
+      "loss": 4.8042,
+      "step": 14215
+    },
+    {
+      "epoch": 0.7860696517412935,
+      "grad_norm": 3.5972378253936768,
+      "learning_rate": 1.100056850483229e-05,
+      "loss": 4.8298,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7863460475400774,
+      "grad_norm": 3.5555312633514404,
+      "learning_rate": 1.0986355884025015e-05,
+      "loss": 5.0829,
+      "step": 14225
+    },
+    {
+      "epoch": 0.7866224433388612,
+      "grad_norm": 3.3311171531677246,
+      "learning_rate": 1.0972143263217739e-05,
+      "loss": 5.0186,
+      "step": 14230
+    },
+    {
+      "epoch": 0.7868988391376451,
+      "grad_norm": 3.552480936050415,
+      "learning_rate": 1.0957930642410461e-05,
+      "loss": 4.6854,
+      "step": 14235
+    },
+    {
+      "epoch": 0.7871752349364289,
+      "grad_norm": 3.2884206771850586,
+      "learning_rate": 1.0943718021603185e-05,
+      "loss": 5.0206,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7874516307352128,
+      "grad_norm": 3.535804033279419,
+      "learning_rate": 1.0929505400795907e-05,
+      "loss": 4.8758,
+      "step": 14245
+    },
+    {
+      "epoch": 0.7877280265339967,
+      "grad_norm": 4.292071342468262,
+      "learning_rate": 1.091529277998863e-05,
+      "loss": 5.2122,
+      "step": 14250
+    },
+    {
+      "epoch": 0.7880044223327806,
+      "grad_norm": 3.89021372795105,
+      "learning_rate": 1.0901080159181353e-05,
+      "loss": 4.775,
+      "step": 14255
+    },
+    {
+      "epoch": 0.7882808181315644,
+      "grad_norm": 3.114368438720703,
+      "learning_rate": 1.0886867538374076e-05,
+      "loss": 5.1345,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7885572139303483,
+      "grad_norm": 3.6420657634735107,
+      "learning_rate": 1.08726549175668e-05,
+      "loss": 4.7937,
+      "step": 14265
+    },
+    {
+      "epoch": 0.7888336097291321,
+      "grad_norm": 3.4523563385009766,
+      "learning_rate": 1.0858442296759524e-05,
+      "loss": 4.8285,
+      "step": 14270
+    },
+    {
+      "epoch": 0.789110005527916,
+      "grad_norm": 3.896808624267578,
+      "learning_rate": 1.0844229675952246e-05,
+      "loss": 5.1532,
+      "step": 14275
+    },
+    {
+      "epoch": 0.7893864013266998,
+      "grad_norm": 3.9850831031799316,
+      "learning_rate": 1.083001705514497e-05,
+      "loss": 5.117,
+      "step": 14280
+    },
+    {
+      "epoch": 0.7896627971254837,
+      "grad_norm": 3.7816109657287598,
+      "learning_rate": 1.0815804434337692e-05,
+      "loss": 4.8639,
+      "step": 14285
+    },
+    {
+      "epoch": 0.7899391929242675,
+      "grad_norm": 3.2434422969818115,
+      "learning_rate": 1.0801591813530414e-05,
+      "loss": 4.716,
+      "step": 14290
+    },
+    {
+      "epoch": 0.7902155887230514,
+      "grad_norm": 4.700549125671387,
+      "learning_rate": 1.0787379192723138e-05,
+      "loss": 5.1435,
+      "step": 14295
+    },
+    {
+      "epoch": 0.7904919845218352,
+      "grad_norm": 4.932585716247559,
+      "learning_rate": 1.0773166571915862e-05,
+      "loss": 5.0201,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7907683803206191,
+      "grad_norm": 4.089287281036377,
+      "learning_rate": 1.0758953951108585e-05,
+      "loss": 4.8033,
+      "step": 14305
+    },
+    {
+      "epoch": 0.7910447761194029,
+      "grad_norm": 4.705197811126709,
+      "learning_rate": 1.0744741330301309e-05,
+      "loss": 4.6571,
+      "step": 14310
+    },
+    {
+      "epoch": 0.7913211719181868,
+      "grad_norm": 4.048484802246094,
+      "learning_rate": 1.0730528709494031e-05,
+      "loss": 4.8813,
+      "step": 14315
+    },
+    {
+      "epoch": 0.7915975677169707,
+      "grad_norm": 4.197638988494873,
+      "learning_rate": 1.0716316088686755e-05,
+      "loss": 4.991,
+      "step": 14320
+    },
+    {
+      "epoch": 0.7918739635157546,
+      "grad_norm": 3.081589698791504,
+      "learning_rate": 1.0702103467879477e-05,
+      "loss": 4.7785,
+      "step": 14325
+    },
+    {
+      "epoch": 0.7921503593145384,
+      "grad_norm": 3.1407415866851807,
+      "learning_rate": 1.06878908470722e-05,
+      "loss": 4.7891,
+      "step": 14330
+    },
+    {
+      "epoch": 0.7924267551133223,
+      "grad_norm": 3.6896438598632812,
+      "learning_rate": 1.0673678226264924e-05,
+      "loss": 4.7142,
+      "step": 14335
+    },
+    {
+      "epoch": 0.7927031509121062,
+      "grad_norm": 3.5066559314727783,
+      "learning_rate": 1.0659465605457647e-05,
+      "loss": 5.1038,
+      "step": 14340
+    },
+    {
+      "epoch": 0.79297954671089,
+      "grad_norm": 4.803554058074951,
+      "learning_rate": 1.064525298465037e-05,
+      "loss": 5.1086,
+      "step": 14345
+    },
+    {
+      "epoch": 0.7932559425096739,
+      "grad_norm": 3.1235287189483643,
+      "learning_rate": 1.0631040363843094e-05,
+      "loss": 4.952,
+      "step": 14350
+    },
+    {
+      "epoch": 0.7935323383084577,
+      "grad_norm": 4.96946382522583,
+      "learning_rate": 1.0616827743035816e-05,
+      "loss": 5.2588,
+      "step": 14355
+    },
+    {
+      "epoch": 0.7938087341072416,
+      "grad_norm": 3.8352441787719727,
+      "learning_rate": 1.0602615122228538e-05,
+      "loss": 4.871,
+      "step": 14360
+    },
+    {
+      "epoch": 0.7940851299060254,
+      "grad_norm": 3.559532642364502,
+      "learning_rate": 1.0588402501421262e-05,
+      "loss": 4.5453,
+      "step": 14365
+    },
+    {
+      "epoch": 0.7943615257048093,
+      "grad_norm": 4.492837905883789,
+      "learning_rate": 1.0574189880613986e-05,
+      "loss": 5.0764,
+      "step": 14370
+    },
+    {
+      "epoch": 0.7946379215035931,
+      "grad_norm": 3.7767648696899414,
+      "learning_rate": 1.055997725980671e-05,
+      "loss": 5.1033,
+      "step": 14375
+    },
+    {
+      "epoch": 0.794914317302377,
+      "grad_norm": 3.4074301719665527,
+      "learning_rate": 1.0545764638999433e-05,
+      "loss": 4.6381,
+      "step": 14380
+    },
+    {
+      "epoch": 0.7951907131011609,
+      "grad_norm": 4.1058783531188965,
+      "learning_rate": 1.0531552018192155e-05,
+      "loss": 4.7906,
+      "step": 14385
+    },
+    {
+      "epoch": 0.7954671088999448,
+      "grad_norm": 3.570598840713501,
+      "learning_rate": 1.0517339397384879e-05,
+      "loss": 4.7018,
+      "step": 14390
+    },
+    {
+      "epoch": 0.7957435046987286,
+      "grad_norm": 4.346248626708984,
+      "learning_rate": 1.0503126776577601e-05,
+      "loss": 5.2179,
+      "step": 14395
+    },
+    {
+      "epoch": 0.7960199004975125,
+      "grad_norm": 4.531290054321289,
+      "learning_rate": 1.0488914155770323e-05,
+      "loss": 5.124,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7962962962962963,
+      "grad_norm": 4.11277961730957,
+      "learning_rate": 1.0474701534963047e-05,
+      "loss": 4.9454,
+      "step": 14405
+    },
+    {
+      "epoch": 0.7965726920950802,
+      "grad_norm": 4.518270015716553,
+      "learning_rate": 1.0460488914155771e-05,
+      "loss": 5.0708,
+      "step": 14410
+    },
+    {
+      "epoch": 0.796849087893864,
+      "grad_norm": 3.6835479736328125,
+      "learning_rate": 1.0446276293348494e-05,
+      "loss": 4.9844,
+      "step": 14415
+    },
+    {
+      "epoch": 0.7971254836926479,
+      "grad_norm": 4.1077880859375,
+      "learning_rate": 1.0432063672541218e-05,
+      "loss": 4.8259,
+      "step": 14420
+    },
+    {
+      "epoch": 0.7974018794914317,
+      "grad_norm": 3.0021631717681885,
+      "learning_rate": 1.041785105173394e-05,
+      "loss": 5.0534,
+      "step": 14425
+    },
+    {
+      "epoch": 0.7976782752902156,
+      "grad_norm": 3.630094289779663,
+      "learning_rate": 1.0403638430926662e-05,
+      "loss": 4.9099,
+      "step": 14430
+    },
+    {
+      "epoch": 0.7979546710889994,
+      "grad_norm": 3.7591092586517334,
+      "learning_rate": 1.0389425810119386e-05,
+      "loss": 5.1555,
+      "step": 14435
+    },
+    {
+      "epoch": 0.7982310668877833,
+      "grad_norm": 4.0143656730651855,
+      "learning_rate": 1.037521318931211e-05,
+      "loss": 4.811,
+      "step": 14440
+    },
+    {
+      "epoch": 0.7985074626865671,
+      "grad_norm": 4.8033976554870605,
+      "learning_rate": 1.0361000568504834e-05,
+      "loss": 5.0588,
+      "step": 14445
+    },
+    {
+      "epoch": 0.798783858485351,
+      "grad_norm": 4.348042964935303,
+      "learning_rate": 1.0346787947697556e-05,
+      "loss": 4.6481,
+      "step": 14450
+    },
+    {
+      "epoch": 0.7990602542841349,
+      "grad_norm": 3.549741506576538,
+      "learning_rate": 1.0332575326890279e-05,
+      "loss": 4.9424,
+      "step": 14455
+    },
+    {
+      "epoch": 0.7993366500829188,
+      "grad_norm": 3.3581111431121826,
+      "learning_rate": 1.0318362706083003e-05,
+      "loss": 5.1743,
+      "step": 14460
+    },
+    {
+      "epoch": 0.7996130458817026,
+      "grad_norm": 4.2459540367126465,
+      "learning_rate": 1.0304150085275725e-05,
+      "loss": 5.1007,
+      "step": 14465
+    },
+    {
+      "epoch": 0.7998894416804865,
+      "grad_norm": 5.176113128662109,
+      "learning_rate": 1.0289937464468447e-05,
+      "loss": 5.1488,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8001658374792703,
+      "grad_norm": 3.7594547271728516,
+      "learning_rate": 1.0275724843661171e-05,
+      "loss": 5.0909,
+      "step": 14475
+    },
+    {
+      "epoch": 0.8004422332780542,
+      "grad_norm": 4.071329593658447,
+      "learning_rate": 1.0261512222853895e-05,
+      "loss": 5.1695,
+      "step": 14480
+    },
+    {
+      "epoch": 0.800718629076838,
+      "grad_norm": 2.798461437225342,
+      "learning_rate": 1.0247299602046617e-05,
+      "loss": 4.7519,
+      "step": 14485
+    },
+    {
+      "epoch": 0.8009950248756219,
+      "grad_norm": 3.2360336780548096,
+      "learning_rate": 1.0233086981239341e-05,
+      "loss": 5.04,
+      "step": 14490
+    },
+    {
+      "epoch": 0.8012714206744057,
+      "grad_norm": 3.919640064239502,
+      "learning_rate": 1.0218874360432064e-05,
+      "loss": 4.8624,
+      "step": 14495
+    },
+    {
+      "epoch": 0.8015478164731896,
+      "grad_norm": 3.5546929836273193,
+      "learning_rate": 1.0204661739624788e-05,
+      "loss": 4.8529,
+      "step": 14500
+    },
+    {
+      "epoch": 0.8018242122719734,
+      "grad_norm": 4.039145469665527,
+      "learning_rate": 1.019044911881751e-05,
+      "loss": 5.2288,
+      "step": 14505
+    },
+    {
+      "epoch": 0.8021006080707573,
+      "grad_norm": 2.8271827697753906,
+      "learning_rate": 1.0176236498010234e-05,
+      "loss": 4.7984,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8023770038695411,
+      "grad_norm": 3.8265321254730225,
+      "learning_rate": 1.0162023877202958e-05,
+      "loss": 4.8457,
+      "step": 14515
+    },
+    {
+      "epoch": 0.802653399668325,
+      "grad_norm": 3.9370837211608887,
+      "learning_rate": 1.014781125639568e-05,
+      "loss": 5.279,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8029297954671089,
+      "grad_norm": 3.1449708938598633,
+      "learning_rate": 1.0133598635588403e-05,
+      "loss": 4.7764,
+      "step": 14525
+    },
+    {
+      "epoch": 0.8032061912658928,
+      "grad_norm": 3.421999454498291,
+      "learning_rate": 1.0119386014781127e-05,
+      "loss": 4.887,
+      "step": 14530
+    },
+    {
+      "epoch": 0.8034825870646766,
+      "grad_norm": 3.242093324661255,
+      "learning_rate": 1.0105173393973849e-05,
+      "loss": 4.9132,
+      "step": 14535
+    },
+    {
+      "epoch": 0.8037589828634605,
+      "grad_norm": 3.8317198753356934,
+      "learning_rate": 1.0090960773166571e-05,
+      "loss": 4.9708,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8040353786622443,
+      "grad_norm": 2.940936326980591,
+      "learning_rate": 1.0076748152359295e-05,
+      "loss": 5.1001,
+      "step": 14545
+    },
+    {
+      "epoch": 0.8043117744610282,
+      "grad_norm": 3.3959786891937256,
+      "learning_rate": 1.0062535531552019e-05,
+      "loss": 4.722,
+      "step": 14550
+    },
+    {
+      "epoch": 0.804588170259812,
+      "grad_norm": 3.077038049697876,
+      "learning_rate": 1.0048322910744743e-05,
+      "loss": 4.8792,
+      "step": 14555
+    },
+    {
+      "epoch": 0.8048645660585959,
+      "grad_norm": 3.9798264503479004,
+      "learning_rate": 1.0034110289937465e-05,
+      "loss": 4.864,
+      "step": 14560
+    },
+    {
+      "epoch": 0.8051409618573798,
+      "grad_norm": 3.65219783782959,
+      "learning_rate": 1.0019897669130188e-05,
+      "loss": 5.1996,
+      "step": 14565
+    },
+    {
+      "epoch": 0.8054173576561636,
+      "grad_norm": 3.0886595249176025,
+      "learning_rate": 1.0005685048322912e-05,
+      "loss": 4.9891,
+      "step": 14570
+    },
+    {
+      "epoch": 0.8056937534549475,
+      "grad_norm": 3.2638444900512695,
+      "learning_rate": 9.991472427515634e-06,
+      "loss": 4.4643,
+      "step": 14575
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 3.3993797302246094,
+      "learning_rate": 9.977259806708356e-06,
+      "loss": 4.5518,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8062465450525153,
+      "grad_norm": 3.588103771209717,
+      "learning_rate": 9.963047185901082e-06,
+      "loss": 4.8706,
+      "step": 14585
+    },
+    {
+      "epoch": 0.806522940851299,
+      "grad_norm": 3.355370044708252,
+      "learning_rate": 9.948834565093804e-06,
+      "loss": 5.0936,
+      "step": 14590
+    },
+    {
+      "epoch": 0.806799336650083,
+      "grad_norm": 3.478992462158203,
+      "learning_rate": 9.934621944286526e-06,
+      "loss": 4.8008,
+      "step": 14595
+    },
+    {
+      "epoch": 0.8070757324488668,
+      "grad_norm": 3.620059013366699,
+      "learning_rate": 9.92040932347925e-06,
+      "loss": 4.8624,
+      "step": 14600
+    },
+    {
+      "epoch": 0.8073521282476507,
+      "grad_norm": 3.095531463623047,
+      "learning_rate": 9.906196702671973e-06,
+      "loss": 4.7169,
+      "step": 14605
+    },
+    {
+      "epoch": 0.8076285240464345,
+      "grad_norm": 3.2527894973754883,
+      "learning_rate": 9.891984081864697e-06,
+      "loss": 4.9315,
+      "step": 14610
+    },
+    {
+      "epoch": 0.8079049198452184,
+      "grad_norm": 3.8432343006134033,
+      "learning_rate": 9.877771461057419e-06,
+      "loss": 5.0862,
+      "step": 14615
+    },
+    {
+      "epoch": 0.8081813156440022,
+      "grad_norm": 4.290609836578369,
+      "learning_rate": 9.863558840250143e-06,
+      "loss": 5.1276,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8084577114427861,
+      "grad_norm": 3.555579662322998,
+      "learning_rate": 9.849346219442867e-06,
+      "loss": 5.0188,
+      "step": 14625
+    },
+    {
+      "epoch": 0.8087341072415699,
+      "grad_norm": 3.1154568195343018,
+      "learning_rate": 9.835133598635589e-06,
+      "loss": 5.083,
+      "step": 14630
+    },
+    {
+      "epoch": 0.8090105030403538,
+      "grad_norm": 3.5717074871063232,
+      "learning_rate": 9.820920977828311e-06,
+      "loss": 5.0745,
+      "step": 14635
+    },
+    {
+      "epoch": 0.8092868988391376,
+      "grad_norm": 3.1403820514678955,
+      "learning_rate": 9.806708357021035e-06,
+      "loss": 4.8718,
+      "step": 14640
+    },
+    {
+      "epoch": 0.8095632946379215,
+      "grad_norm": 3.6060197353363037,
+      "learning_rate": 9.792495736213758e-06,
+      "loss": 5.1306,
+      "step": 14645
+    },
+    {
+      "epoch": 0.8098396904367053,
+      "grad_norm": 3.7077431678771973,
+      "learning_rate": 9.77828311540648e-06,
+      "loss": 4.6169,
+      "step": 14650
+    },
+    {
+      "epoch": 0.8101160862354893,
+      "grad_norm": 4.237201690673828,
+      "learning_rate": 9.764070494599204e-06,
+      "loss": 4.6571,
+      "step": 14655
+    },
+    {
+      "epoch": 0.810392482034273,
+      "grad_norm": 3.263676166534424,
+      "learning_rate": 9.749857873791928e-06,
+      "loss": 5.0076,
+      "step": 14660
+    },
+    {
+      "epoch": 0.810668877833057,
+      "grad_norm": 3.406616687774658,
+      "learning_rate": 9.735645252984652e-06,
+      "loss": 4.6793,
+      "step": 14665
+    },
+    {
+      "epoch": 0.8109452736318408,
+      "grad_norm": 3.664701461791992,
+      "learning_rate": 9.721432632177374e-06,
+      "loss": 4.7769,
+      "step": 14670
+    },
+    {
+      "epoch": 0.8112216694306247,
+      "grad_norm": 3.6615958213806152,
+      "learning_rate": 9.707220011370096e-06,
+      "loss": 5.1376,
+      "step": 14675
+    },
+    {
+      "epoch": 0.8114980652294085,
+      "grad_norm": 3.585599660873413,
+      "learning_rate": 9.69300739056282e-06,
+      "loss": 4.8062,
+      "step": 14680
+    },
+    {
+      "epoch": 0.8117744610281924,
+      "grad_norm": 3.3951401710510254,
+      "learning_rate": 9.678794769755543e-06,
+      "loss": 5.0233,
+      "step": 14685
+    },
+    {
+      "epoch": 0.8120508568269762,
+      "grad_norm": 3.261725664138794,
+      "learning_rate": 9.664582148948267e-06,
+      "loss": 4.957,
+      "step": 14690
+    },
+    {
+      "epoch": 0.8123272526257601,
+      "grad_norm": 3.23929762840271,
+      "learning_rate": 9.65036952814099e-06,
+      "loss": 4.885,
+      "step": 14695
+    },
+    {
+      "epoch": 0.8126036484245439,
+      "grad_norm": 3.207082748413086,
+      "learning_rate": 9.636156907333713e-06,
+      "loss": 4.5523,
+      "step": 14700
+    },
+    {
+      "epoch": 0.8128800442233278,
+      "grad_norm": 3.7475905418395996,
+      "learning_rate": 9.621944286526435e-06,
+      "loss": 5.0544,
+      "step": 14705
+    },
+    {
+      "epoch": 0.8131564400221116,
+      "grad_norm": 3.0875489711761475,
+      "learning_rate": 9.60773166571916e-06,
+      "loss": 5.127,
+      "step": 14710
+    },
+    {
+      "epoch": 0.8134328358208955,
+      "grad_norm": 3.5430004596710205,
+      "learning_rate": 9.593519044911882e-06,
+      "loss": 4.9775,
+      "step": 14715
+    },
+    {
+      "epoch": 0.8137092316196793,
+      "grad_norm": 3.482557535171509,
+      "learning_rate": 9.579306424104606e-06,
+      "loss": 4.8944,
+      "step": 14720
+    },
+    {
+      "epoch": 0.8139856274184633,
+      "grad_norm": 3.779650926589966,
+      "learning_rate": 9.565093803297328e-06,
+      "loss": 5.2317,
+      "step": 14725
+    },
+    {
+      "epoch": 0.814262023217247,
+      "grad_norm": 4.0570244789123535,
+      "learning_rate": 9.550881182490052e-06,
+      "loss": 4.9351,
+      "step": 14730
+    },
+    {
+      "epoch": 0.814538419016031,
+      "grad_norm": 3.294402599334717,
+      "learning_rate": 9.536668561682776e-06,
+      "loss": 4.63,
+      "step": 14735
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 3.926581859588623,
+      "learning_rate": 9.522455940875498e-06,
+      "loss": 4.9833,
+      "step": 14740
+    },
+    {
+      "epoch": 0.8150912106135987,
+      "grad_norm": 3.623554229736328,
+      "learning_rate": 9.50824332006822e-06,
+      "loss": 4.9796,
+      "step": 14745
+    },
+    {
+      "epoch": 0.8153676064123825,
+      "grad_norm": 4.15550422668457,
+      "learning_rate": 9.494030699260944e-06,
+      "loss": 4.9854,
+      "step": 14750
+    },
+    {
+      "epoch": 0.8156440022111664,
+      "grad_norm": 4.647388935089111,
+      "learning_rate": 9.479818078453667e-06,
+      "loss": 5.336,
+      "step": 14755
+    },
+    {
+      "epoch": 0.8159203980099502,
+      "grad_norm": 4.139171600341797,
+      "learning_rate": 9.46560545764639e-06,
+      "loss": 4.9827,
+      "step": 14760
+    },
+    {
+      "epoch": 0.8161967938087341,
+      "grad_norm": 3.5119783878326416,
+      "learning_rate": 9.451392836839115e-06,
+      "loss": 4.7924,
+      "step": 14765
+    },
+    {
+      "epoch": 0.8164731896075179,
+      "grad_norm": 3.4265787601470947,
+      "learning_rate": 9.437180216031837e-06,
+      "loss": 5.0749,
+      "step": 14770
+    },
+    {
+      "epoch": 0.8167495854063018,
+      "grad_norm": 3.3151018619537354,
+      "learning_rate": 9.42296759522456e-06,
+      "loss": 4.7777,
+      "step": 14775
+    },
+    {
+      "epoch": 0.8170259812050856,
+      "grad_norm": 3.0593369007110596,
+      "learning_rate": 9.408754974417283e-06,
+      "loss": 4.9151,
+      "step": 14780
+    },
+    {
+      "epoch": 0.8173023770038695,
+      "grad_norm": 3.627476930618286,
+      "learning_rate": 9.394542353610005e-06,
+      "loss": 5.2975,
+      "step": 14785
+    },
+    {
+      "epoch": 0.8175787728026535,
+      "grad_norm": 3.161733388900757,
+      "learning_rate": 9.38032973280273e-06,
+      "loss": 5.0477,
+      "step": 14790
+    },
+    {
+      "epoch": 0.8178551686014373,
+      "grad_norm": 3.5496115684509277,
+      "learning_rate": 9.366117111995452e-06,
+      "loss": 5.0122,
+      "step": 14795
+    },
+    {
+      "epoch": 0.8181315644002212,
+      "grad_norm": 3.702653169631958,
+      "learning_rate": 9.351904491188176e-06,
+      "loss": 4.7585,
+      "step": 14800
+    },
+    {
+      "epoch": 0.818407960199005,
+      "grad_norm": 3.9237051010131836,
+      "learning_rate": 9.3376918703809e-06,
+      "loss": 4.6184,
+      "step": 14805
+    },
+    {
+      "epoch": 0.8186843559977889,
+      "grad_norm": 3.077275276184082,
+      "learning_rate": 9.323479249573622e-06,
+      "loss": 4.6302,
+      "step": 14810
+    },
+    {
+      "epoch": 0.8189607517965727,
+      "grad_norm": 3.5210278034210205,
+      "learning_rate": 9.309266628766344e-06,
+      "loss": 5.0052,
+      "step": 14815
+    },
+    {
+      "epoch": 0.8192371475953566,
+      "grad_norm": 3.1332850456237793,
+      "learning_rate": 9.295054007959068e-06,
+      "loss": 4.9236,
+      "step": 14820
+    },
+    {
+      "epoch": 0.8195135433941404,
+      "grad_norm": 3.8537850379943848,
+      "learning_rate": 9.28084138715179e-06,
+      "loss": 5.0165,
+      "step": 14825
+    },
+    {
+      "epoch": 0.8197899391929243,
+      "grad_norm": 3.768019199371338,
+      "learning_rate": 9.266628766344514e-06,
+      "loss": 4.7222,
+      "step": 14830
+    },
+    {
+      "epoch": 0.8200663349917081,
+      "grad_norm": 4.06325101852417,
+      "learning_rate": 9.252416145537238e-06,
+      "loss": 4.942,
+      "step": 14835
+    },
+    {
+      "epoch": 0.820342730790492,
+      "grad_norm": 3.555703639984131,
+      "learning_rate": 9.23820352472996e-06,
+      "loss": 5.0981,
+      "step": 14840
+    },
+    {
+      "epoch": 0.8206191265892758,
+      "grad_norm": 3.216472625732422,
+      "learning_rate": 9.223990903922685e-06,
+      "loss": 5.1655,
+      "step": 14845
+    },
+    {
+      "epoch": 0.8208955223880597,
+      "grad_norm": 2.9533238410949707,
+      "learning_rate": 9.209778283115407e-06,
+      "loss": 4.868,
+      "step": 14850
+    },
+    {
+      "epoch": 0.8211719181868435,
+      "grad_norm": 3.6127288341522217,
+      "learning_rate": 9.19556566230813e-06,
+      "loss": 5.1768,
+      "step": 14855
+    },
+    {
+      "epoch": 0.8214483139856275,
+      "grad_norm": 3.685889720916748,
+      "learning_rate": 9.181353041500853e-06,
+      "loss": 4.8842,
+      "step": 14860
+    },
+    {
+      "epoch": 0.8217247097844113,
+      "grad_norm": 3.554274559020996,
+      "learning_rate": 9.167140420693576e-06,
+      "loss": 4.7565,
+      "step": 14865
+    },
+    {
+      "epoch": 0.8220011055831952,
+      "grad_norm": 4.190557956695557,
+      "learning_rate": 9.1529277998863e-06,
+      "loss": 4.8703,
+      "step": 14870
+    },
+    {
+      "epoch": 0.822277501381979,
+      "grad_norm": 3.2858259677886963,
+      "learning_rate": 9.138715179079023e-06,
+      "loss": 4.8811,
+      "step": 14875
+    },
+    {
+      "epoch": 0.8225538971807629,
+      "grad_norm": 4.575525760650635,
+      "learning_rate": 9.124502558271746e-06,
+      "loss": 5.1692,
+      "step": 14880
+    },
+    {
+      "epoch": 0.8228302929795467,
+      "grad_norm": 3.1636838912963867,
+      "learning_rate": 9.11028993746447e-06,
+      "loss": 4.8458,
+      "step": 14885
+    },
+    {
+      "epoch": 0.8231066887783306,
+      "grad_norm": 3.5616745948791504,
+      "learning_rate": 9.096077316657192e-06,
+      "loss": 4.9314,
+      "step": 14890
+    },
+    {
+      "epoch": 0.8233830845771144,
+      "grad_norm": 4.337357044219971,
+      "learning_rate": 9.081864695849914e-06,
+      "loss": 5.0281,
+      "step": 14895
+    },
+    {
+      "epoch": 0.8236594803758983,
+      "grad_norm": 3.906620502471924,
+      "learning_rate": 9.067652075042638e-06,
+      "loss": 4.895,
+      "step": 14900
+    },
+    {
+      "epoch": 0.8239358761746821,
+      "grad_norm": 3.766569137573242,
+      "learning_rate": 9.053439454235362e-06,
+      "loss": 4.7679,
+      "step": 14905
+    },
+    {
+      "epoch": 0.824212271973466,
+      "grad_norm": 3.7413177490234375,
+      "learning_rate": 9.039226833428085e-06,
+      "loss": 4.8692,
+      "step": 14910
+    },
+    {
+      "epoch": 0.8244886677722498,
+      "grad_norm": 3.4053595066070557,
+      "learning_rate": 9.025014212620809e-06,
+      "loss": 4.625,
+      "step": 14915
+    },
+    {
+      "epoch": 0.8247650635710337,
+      "grad_norm": 4.579228401184082,
+      "learning_rate": 9.01080159181353e-06,
+      "loss": 4.6284,
+      "step": 14920
+    },
+    {
+      "epoch": 0.8250414593698175,
+      "grad_norm": 4.081772327423096,
+      "learning_rate": 8.996588971006253e-06,
+      "loss": 4.9273,
+      "step": 14925
+    },
+    {
+      "epoch": 0.8253178551686015,
+      "grad_norm": 3.704084634780884,
+      "learning_rate": 8.982376350198977e-06,
+      "loss": 4.8167,
+      "step": 14930
+    },
+    {
+      "epoch": 0.8255942509673853,
+      "grad_norm": 4.213784217834473,
+      "learning_rate": 8.9681637293917e-06,
+      "loss": 4.8445,
+      "step": 14935
+    },
+    {
+      "epoch": 0.8258706467661692,
+      "grad_norm": 3.3844199180603027,
+      "learning_rate": 8.953951108584423e-06,
+      "loss": 4.9753,
+      "step": 14940
+    },
+    {
+      "epoch": 0.826147042564953,
+      "grad_norm": 4.304365158081055,
+      "learning_rate": 8.939738487777147e-06,
+      "loss": 4.9407,
+      "step": 14945
+    },
+    {
+      "epoch": 0.8264234383637369,
+      "grad_norm": 5.198899269104004,
+      "learning_rate": 8.92552586696987e-06,
+      "loss": 5.0407,
+      "step": 14950
+    },
+    {
+      "epoch": 0.8266998341625207,
+      "grad_norm": 3.2233057022094727,
+      "learning_rate": 8.911313246162594e-06,
+      "loss": 4.6535,
+      "step": 14955
+    },
+    {
+      "epoch": 0.8269762299613046,
+      "grad_norm": 3.325000286102295,
+      "learning_rate": 8.897100625355316e-06,
+      "loss": 4.9527,
+      "step": 14960
+    },
+    {
+      "epoch": 0.8272526257600884,
+      "grad_norm": 4.237645626068115,
+      "learning_rate": 8.882888004548038e-06,
+      "loss": 4.831,
+      "step": 14965
+    },
+    {
+      "epoch": 0.8275290215588723,
+      "grad_norm": 3.883349895477295,
+      "learning_rate": 8.868675383740762e-06,
+      "loss": 5.2111,
+      "step": 14970
+    },
+    {
+      "epoch": 0.8278054173576561,
+      "grad_norm": 4.307447910308838,
+      "learning_rate": 8.854462762933486e-06,
+      "loss": 4.8234,
+      "step": 14975
+    },
+    {
+      "epoch": 0.82808181315644,
+      "grad_norm": 3.646482229232788,
+      "learning_rate": 8.840250142126208e-06,
+      "loss": 4.6802,
+      "step": 14980
+    },
+    {
+      "epoch": 0.8283582089552238,
+      "grad_norm": 4.150384902954102,
+      "learning_rate": 8.826037521318932e-06,
+      "loss": 4.939,
+      "step": 14985
+    },
+    {
+      "epoch": 0.8286346047540077,
+      "grad_norm": 3.2273035049438477,
+      "learning_rate": 8.811824900511655e-06,
+      "loss": 4.758,
+      "step": 14990
+    },
+    {
+      "epoch": 0.8289110005527915,
+      "grad_norm": 3.5394163131713867,
+      "learning_rate": 8.797612279704377e-06,
+      "loss": 5.039,
+      "step": 14995
+    },
+    {
+      "epoch": 0.8291873963515755,
+      "grad_norm": 3.0084025859832764,
+      "learning_rate": 8.783399658897101e-06,
+      "loss": 5.2781,
+      "step": 15000
+    },
+    {
+      "epoch": 0.8294637921503594,
+      "grad_norm": 3.8212339878082275,
+      "learning_rate": 8.769187038089823e-06,
+      "loss": 4.8291,
+      "step": 15005
+    },
+    {
+      "epoch": 0.8297401879491432,
+      "grad_norm": 3.359372854232788,
+      "learning_rate": 8.754974417282547e-06,
+      "loss": 4.9139,
+      "step": 15010
+    },
+    {
+      "epoch": 0.8300165837479271,
+      "grad_norm": 3.9898805618286133,
+      "learning_rate": 8.740761796475271e-06,
+      "loss": 4.738,
+      "step": 15015
+    },
+    {
+      "epoch": 0.8302929795467109,
+      "grad_norm": 3.4556021690368652,
+      "learning_rate": 8.726549175667993e-06,
+      "loss": 4.6281,
+      "step": 15020
+    },
+    {
+      "epoch": 0.8305693753454948,
+      "grad_norm": 4.344873905181885,
+      "learning_rate": 8.712336554860717e-06,
+      "loss": 5.041,
+      "step": 15025
+    },
+    {
+      "epoch": 0.8308457711442786,
+      "grad_norm": 4.770878314971924,
+      "learning_rate": 8.69812393405344e-06,
+      "loss": 4.8658,
+      "step": 15030
+    },
+    {
+      "epoch": 0.8311221669430625,
+      "grad_norm": 3.748448610305786,
+      "learning_rate": 8.683911313246162e-06,
+      "loss": 4.969,
+      "step": 15035
+    },
+    {
+      "epoch": 0.8313985627418463,
+      "grad_norm": 4.457357883453369,
+      "learning_rate": 8.669698692438886e-06,
+      "loss": 5.2523,
+      "step": 15040
+    },
+    {
+      "epoch": 0.8316749585406302,
+      "grad_norm": 4.152907371520996,
+      "learning_rate": 8.65548607163161e-06,
+      "loss": 4.8646,
+      "step": 15045
+    },
+    {
+      "epoch": 0.831951354339414,
+      "grad_norm": 3.441721200942993,
+      "learning_rate": 8.641273450824332e-06,
+      "loss": 4.8011,
+      "step": 15050
+    },
+    {
+      "epoch": 0.8322277501381979,
+      "grad_norm": 4.53822660446167,
+      "learning_rate": 8.627060830017056e-06,
+      "loss": 5.184,
+      "step": 15055
+    },
+    {
+      "epoch": 0.8325041459369817,
+      "grad_norm": 3.868192195892334,
+      "learning_rate": 8.612848209209779e-06,
+      "loss": 4.8637,
+      "step": 15060
+    },
+    {
+      "epoch": 0.8327805417357657,
+      "grad_norm": 3.9942774772644043,
+      "learning_rate": 8.598635588402502e-06,
+      "loss": 4.8782,
+      "step": 15065
+    },
+    {
+      "epoch": 0.8330569375345495,
+      "grad_norm": 2.853957176208496,
+      "learning_rate": 8.584422967595225e-06,
+      "loss": 4.9473,
+      "step": 15070
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 3.749110460281372,
+      "learning_rate": 8.570210346787947e-06,
+      "loss": 5.012,
+      "step": 15075
+    },
+    {
+      "epoch": 0.8336097291321172,
+      "grad_norm": 4.649420261383057,
+      "learning_rate": 8.555997725980671e-06,
+      "loss": 4.7911,
+      "step": 15080
+    },
+    {
+      "epoch": 0.8338861249309011,
+      "grad_norm": 3.9291510581970215,
+      "learning_rate": 8.541785105173395e-06,
+      "loss": 5.1986,
+      "step": 15085
+    },
+    {
+      "epoch": 0.8341625207296849,
+      "grad_norm": 3.576134443283081,
+      "learning_rate": 8.527572484366117e-06,
+      "loss": 5.0059,
+      "step": 15090
+    },
+    {
+      "epoch": 0.8344389165284688,
+      "grad_norm": 3.902698516845703,
+      "learning_rate": 8.513359863558841e-06,
+      "loss": 4.8758,
+      "step": 15095
+    },
+    {
+      "epoch": 0.8347153123272526,
+      "grad_norm": 4.259060859680176,
+      "learning_rate": 8.499147242751564e-06,
+      "loss": 4.8227,
+      "step": 15100
+    },
+    {
+      "epoch": 0.8349917081260365,
+      "grad_norm": 3.539156436920166,
+      "learning_rate": 8.484934621944286e-06,
+      "loss": 5.0189,
+      "step": 15105
+    },
+    {
+      "epoch": 0.8352681039248203,
+      "grad_norm": 4.712406635284424,
+      "learning_rate": 8.47072200113701e-06,
+      "loss": 5.0471,
+      "step": 15110
+    },
+    {
+      "epoch": 0.8355444997236042,
+      "grad_norm": 3.5508053302764893,
+      "learning_rate": 8.456509380329732e-06,
+      "loss": 4.8707,
+      "step": 15115
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 3.3851168155670166,
+      "learning_rate": 8.442296759522458e-06,
+      "loss": 4.8683,
+      "step": 15120
+    },
+    {
+      "epoch": 0.8360972913211719,
+      "grad_norm": 3.6790895462036133,
+      "learning_rate": 8.42808413871518e-06,
+      "loss": 4.8464,
+      "step": 15125
+    },
+    {
+      "epoch": 0.8363736871199557,
+      "grad_norm": 3.053698778152466,
+      "learning_rate": 8.413871517907902e-06,
+      "loss": 4.9889,
+      "step": 15130
+    },
+    {
+      "epoch": 0.8366500829187397,
+      "grad_norm": 4.331393718719482,
+      "learning_rate": 8.399658897100626e-06,
+      "loss": 5.4066,
+      "step": 15135
+    },
+    {
+      "epoch": 0.8369264787175235,
+      "grad_norm": 3.729410171508789,
+      "learning_rate": 8.385446276293349e-06,
+      "loss": 5.0611,
+      "step": 15140
+    },
+    {
+      "epoch": 0.8372028745163074,
+      "grad_norm": 3.60453724861145,
+      "learning_rate": 8.371233655486071e-06,
+      "loss": 4.8469,
+      "step": 15145
+    },
+    {
+      "epoch": 0.8374792703150912,
+      "grad_norm": 4.959109306335449,
+      "learning_rate": 8.357021034678795e-06,
+      "loss": 4.938,
+      "step": 15150
+    },
+    {
+      "epoch": 0.8377556661138751,
+      "grad_norm": 4.496500015258789,
+      "learning_rate": 8.342808413871519e-06,
+      "loss": 4.7706,
+      "step": 15155
+    },
+    {
+      "epoch": 0.8380320619126589,
+      "grad_norm": 4.093534469604492,
+      "learning_rate": 8.328595793064241e-06,
+      "loss": 5.1374,
+      "step": 15160
+    },
+    {
+      "epoch": 0.8383084577114428,
+      "grad_norm": 3.628263473510742,
+      "learning_rate": 8.314383172256965e-06,
+      "loss": 4.7076,
+      "step": 15165
+    },
+    {
+      "epoch": 0.8385848535102266,
+      "grad_norm": 3.8275880813598633,
+      "learning_rate": 8.300170551449687e-06,
+      "loss": 4.5418,
+      "step": 15170
+    },
+    {
+      "epoch": 0.8388612493090105,
+      "grad_norm": 4.351752758026123,
+      "learning_rate": 8.285957930642411e-06,
+      "loss": 4.903,
+      "step": 15175
+    },
+    {
+      "epoch": 0.8391376451077943,
+      "grad_norm": 3.621277093887329,
+      "learning_rate": 8.271745309835134e-06,
+      "loss": 4.9818,
+      "step": 15180
+    },
+    {
+      "epoch": 0.8394140409065782,
+      "grad_norm": 3.3908252716064453,
+      "learning_rate": 8.257532689027856e-06,
+      "loss": 5.084,
+      "step": 15185
+    },
+    {
+      "epoch": 0.839690436705362,
+      "grad_norm": 3.7233057022094727,
+      "learning_rate": 8.24332006822058e-06,
+      "loss": 5.2591,
+      "step": 15190
+    },
+    {
+      "epoch": 0.8399668325041459,
+      "grad_norm": 3.7437429428100586,
+      "learning_rate": 8.229107447413304e-06,
+      "loss": 4.817,
+      "step": 15195
+    },
+    {
+      "epoch": 0.8402432283029297,
+      "grad_norm": 3.7888224124908447,
+      "learning_rate": 8.214894826606026e-06,
+      "loss": 4.598,
+      "step": 15200
+    },
+    {
+      "epoch": 0.8405196241017137,
+      "grad_norm": 3.3755719661712646,
+      "learning_rate": 8.20068220579875e-06,
+      "loss": 5.1114,
+      "step": 15205
+    },
+    {
+      "epoch": 0.8407960199004975,
+      "grad_norm": 3.7054946422576904,
+      "learning_rate": 8.186469584991472e-06,
+      "loss": 4.6614,
+      "step": 15210
+    },
+    {
+      "epoch": 0.8410724156992814,
+      "grad_norm": 3.747762441635132,
+      "learning_rate": 8.172256964184195e-06,
+      "loss": 4.7052,
+      "step": 15215
+    },
+    {
+      "epoch": 0.8413488114980652,
+      "grad_norm": 3.4834938049316406,
+      "learning_rate": 8.158044343376919e-06,
+      "loss": 5.1221,
+      "step": 15220
+    },
+    {
+      "epoch": 0.8416252072968491,
+      "grad_norm": 3.943834066390991,
+      "learning_rate": 8.143831722569643e-06,
+      "loss": 4.9175,
+      "step": 15225
+    },
+    {
+      "epoch": 0.841901603095633,
+      "grad_norm": 3.3898112773895264,
+      "learning_rate": 8.129619101762367e-06,
+      "loss": 5.0414,
+      "step": 15230
+    },
+    {
+      "epoch": 0.8421779988944168,
+      "grad_norm": 3.7850241661071777,
+      "learning_rate": 8.115406480955089e-06,
+      "loss": 4.9141,
+      "step": 15235
+    },
+    {
+      "epoch": 0.8424543946932007,
+      "grad_norm": 3.2577099800109863,
+      "learning_rate": 8.101193860147811e-06,
+      "loss": 4.7304,
+      "step": 15240
+    },
+    {
+      "epoch": 0.8427307904919845,
+      "grad_norm": 3.3998703956604004,
+      "learning_rate": 8.086981239340535e-06,
+      "loss": 4.9697,
+      "step": 15245
+    },
+    {
+      "epoch": 0.8430071862907684,
+      "grad_norm": 4.133670330047607,
+      "learning_rate": 8.072768618533258e-06,
+      "loss": 4.9938,
+      "step": 15250
+    },
+    {
+      "epoch": 0.8432835820895522,
+      "grad_norm": 4.74376916885376,
+      "learning_rate": 8.05855599772598e-06,
+      "loss": 4.8981,
+      "step": 15255
+    },
+    {
+      "epoch": 0.8435599778883361,
+      "grad_norm": 3.422651529312134,
+      "learning_rate": 8.044343376918704e-06,
+      "loss": 4.9323,
+      "step": 15260
+    },
+    {
+      "epoch": 0.84383637368712,
+      "grad_norm": 4.132688999176025,
+      "learning_rate": 8.030130756111428e-06,
+      "loss": 4.8445,
+      "step": 15265
+    },
+    {
+      "epoch": 0.8441127694859039,
+      "grad_norm": 4.657894134521484,
+      "learning_rate": 8.01591813530415e-06,
+      "loss": 4.7014,
+      "step": 15270
+    },
+    {
+      "epoch": 0.8443891652846877,
+      "grad_norm": 4.646186828613281,
+      "learning_rate": 8.001705514496874e-06,
+      "loss": 5.2851,
+      "step": 15275
+    },
+    {
+      "epoch": 0.8446655610834716,
+      "grad_norm": 4.24678897857666,
+      "learning_rate": 7.987492893689596e-06,
+      "loss": 4.6824,
+      "step": 15280
+    },
+    {
+      "epoch": 0.8449419568822554,
+      "grad_norm": 3.599647045135498,
+      "learning_rate": 7.97328027288232e-06,
+      "loss": 4.8003,
+      "step": 15285
+    },
+    {
+      "epoch": 0.8452183526810393,
+      "grad_norm": 3.2087178230285645,
+      "learning_rate": 7.959067652075043e-06,
+      "loss": 4.6724,
+      "step": 15290
+    },
+    {
+      "epoch": 0.8454947484798231,
+      "grad_norm": 3.3129324913024902,
+      "learning_rate": 7.944855031267767e-06,
+      "loss": 4.9888,
+      "step": 15295
+    },
+    {
+      "epoch": 0.845771144278607,
+      "grad_norm": 4.16626501083374,
+      "learning_rate": 7.93064241046049e-06,
+      "loss": 4.9972,
+      "step": 15300
+    },
+    {
+      "epoch": 0.8460475400773908,
+      "grad_norm": 3.8757193088531494,
+      "learning_rate": 7.916429789653213e-06,
+      "loss": 4.9566,
+      "step": 15305
+    },
+    {
+      "epoch": 0.8463239358761747,
+      "grad_norm": 3.9059760570526123,
+      "learning_rate": 7.902217168845935e-06,
+      "loss": 4.7056,
+      "step": 15310
+    },
+    {
+      "epoch": 0.8466003316749585,
+      "grad_norm": 3.2571303844451904,
+      "learning_rate": 7.888004548038659e-06,
+      "loss": 5.0104,
+      "step": 15315
+    },
+    {
+      "epoch": 0.8468767274737424,
+      "grad_norm": 3.6211488246917725,
+      "learning_rate": 7.873791927231381e-06,
+      "loss": 4.9973,
+      "step": 15320
+    },
+    {
+      "epoch": 0.8471531232725262,
+      "grad_norm": 4.511015892028809,
+      "learning_rate": 7.859579306424104e-06,
+      "loss": 4.8835,
+      "step": 15325
+    },
+    {
+      "epoch": 0.8474295190713101,
+      "grad_norm": 3.3514885902404785,
+      "learning_rate": 7.845366685616828e-06,
+      "loss": 5.1082,
+      "step": 15330
+    },
+    {
+      "epoch": 0.847705914870094,
+      "grad_norm": 4.153277397155762,
+      "learning_rate": 7.831154064809552e-06,
+      "loss": 5.1117,
+      "step": 15335
+    },
+    {
+      "epoch": 0.8479823106688779,
+      "grad_norm": 3.661910057067871,
+      "learning_rate": 7.816941444002276e-06,
+      "loss": 4.785,
+      "step": 15340
+    },
+    {
+      "epoch": 0.8482587064676617,
+      "grad_norm": 3.4714293479919434,
+      "learning_rate": 7.802728823194998e-06,
+      "loss": 4.9615,
+      "step": 15345
+    },
+    {
+      "epoch": 0.8485351022664456,
+      "grad_norm": 3.412801504135132,
+      "learning_rate": 7.78851620238772e-06,
+      "loss": 4.7827,
+      "step": 15350
+    },
+    {
+      "epoch": 0.8488114980652294,
+      "grad_norm": 4.46990966796875,
+      "learning_rate": 7.774303581580444e-06,
+      "loss": 4.8436,
+      "step": 15355
+    },
+    {
+      "epoch": 0.8490878938640133,
+      "grad_norm": 4.014577388763428,
+      "learning_rate": 7.760090960773166e-06,
+      "loss": 4.9757,
+      "step": 15360
+    },
+    {
+      "epoch": 0.8493642896627971,
+      "grad_norm": 3.89868426322937,
+      "learning_rate": 7.74587833996589e-06,
+      "loss": 4.8749,
+      "step": 15365
+    },
+    {
+      "epoch": 0.849640685461581,
+      "grad_norm": 2.8966128826141357,
+      "learning_rate": 7.731665719158614e-06,
+      "loss": 4.7,
+      "step": 15370
+    },
+    {
+      "epoch": 0.8499170812603648,
+      "grad_norm": 4.55120849609375,
+      "learning_rate": 7.717453098351337e-06,
+      "loss": 4.7576,
+      "step": 15375
+    },
+    {
+      "epoch": 0.8501934770591487,
+      "grad_norm": 4.043015003204346,
+      "learning_rate": 7.703240477544059e-06,
+      "loss": 4.7656,
+      "step": 15380
+    },
+    {
+      "epoch": 0.8504698728579325,
+      "grad_norm": 4.398715019226074,
+      "learning_rate": 7.689027856736783e-06,
+      "loss": 4.7285,
+      "step": 15385
+    },
+    {
+      "epoch": 0.8507462686567164,
+      "grad_norm": 4.311310291290283,
+      "learning_rate": 7.674815235929505e-06,
+      "loss": 4.8688,
+      "step": 15390
+    },
+    {
+      "epoch": 0.8510226644555002,
+      "grad_norm": 3.4514060020446777,
+      "learning_rate": 7.66060261512223e-06,
+      "loss": 4.7912,
+      "step": 15395
+    },
+    {
+      "epoch": 0.8512990602542841,
+      "grad_norm": 3.418095588684082,
+      "learning_rate": 7.646389994314952e-06,
+      "loss": 4.8534,
+      "step": 15400
+    },
+    {
+      "epoch": 0.851575456053068,
+      "grad_norm": 4.0209503173828125,
+      "learning_rate": 7.632177373507675e-06,
+      "loss": 5.2002,
+      "step": 15405
+    },
+    {
+      "epoch": 0.8518518518518519,
+      "grad_norm": 2.8665504455566406,
+      "learning_rate": 7.617964752700399e-06,
+      "loss": 5.142,
+      "step": 15410
+    },
+    {
+      "epoch": 0.8521282476506357,
+      "grad_norm": 3.491189956665039,
+      "learning_rate": 7.603752131893122e-06,
+      "loss": 4.6841,
+      "step": 15415
+    },
+    {
+      "epoch": 0.8524046434494196,
+      "grad_norm": 3.8220739364624023,
+      "learning_rate": 7.589539511085844e-06,
+      "loss": 4.929,
+      "step": 15420
+    },
+    {
+      "epoch": 0.8526810392482034,
+      "grad_norm": 3.5828657150268555,
+      "learning_rate": 7.575326890278568e-06,
+      "loss": 5.0016,
+      "step": 15425
+    },
+    {
+      "epoch": 0.8529574350469873,
+      "grad_norm": 4.3814921379089355,
+      "learning_rate": 7.561114269471291e-06,
+      "loss": 4.8032,
+      "step": 15430
+    },
+    {
+      "epoch": 0.8532338308457711,
+      "grad_norm": 3.4918086528778076,
+      "learning_rate": 7.5469016486640134e-06,
+      "loss": 4.846,
+      "step": 15435
+    },
+    {
+      "epoch": 0.853510226644555,
+      "grad_norm": 3.771287679672241,
+      "learning_rate": 7.532689027856737e-06,
+      "loss": 4.791,
+      "step": 15440
+    },
+    {
+      "epoch": 0.8537866224433389,
+      "grad_norm": 3.5989866256713867,
+      "learning_rate": 7.51847640704946e-06,
+      "loss": 4.9933,
+      "step": 15445
+    },
+    {
+      "epoch": 0.8540630182421227,
+      "grad_norm": 3.3803904056549072,
+      "learning_rate": 7.504263786242183e-06,
+      "loss": 4.8067,
+      "step": 15450
+    },
+    {
+      "epoch": 0.8543394140409066,
+      "grad_norm": 4.378183841705322,
+      "learning_rate": 7.490051165434907e-06,
+      "loss": 5.3456,
+      "step": 15455
+    },
+    {
+      "epoch": 0.8546158098396904,
+      "grad_norm": 4.145111560821533,
+      "learning_rate": 7.475838544627629e-06,
+      "loss": 4.833,
+      "step": 15460
+    },
+    {
+      "epoch": 0.8548922056384743,
+      "grad_norm": 3.1699306964874268,
+      "learning_rate": 7.461625923820353e-06,
+      "loss": 4.7371,
+      "step": 15465
+    },
+    {
+      "epoch": 0.8551686014372581,
+      "grad_norm": 3.570518732070923,
+      "learning_rate": 7.447413303013076e-06,
+      "loss": 5.0778,
+      "step": 15470
+    },
+    {
+      "epoch": 0.8554449972360421,
+      "grad_norm": 3.922854423522949,
+      "learning_rate": 7.4332006822057985e-06,
+      "loss": 4.9517,
+      "step": 15475
+    },
+    {
+      "epoch": 0.8557213930348259,
+      "grad_norm": 4.147318363189697,
+      "learning_rate": 7.4189880613985225e-06,
+      "loss": 4.9921,
+      "step": 15480
+    },
+    {
+      "epoch": 0.8559977888336098,
+      "grad_norm": 3.7411043643951416,
+      "learning_rate": 7.404775440591246e-06,
+      "loss": 5.2079,
+      "step": 15485
+    },
+    {
+      "epoch": 0.8562741846323936,
+      "grad_norm": 3.129054069519043,
+      "learning_rate": 7.390562819783968e-06,
+      "loss": 4.8514,
+      "step": 15490
+    },
+    {
+      "epoch": 0.8565505804311775,
+      "grad_norm": 3.2814624309539795,
+      "learning_rate": 7.376350198976692e-06,
+      "loss": 4.9005,
+      "step": 15495
+    },
+    {
+      "epoch": 0.8568269762299613,
+      "grad_norm": 4.446510314941406,
+      "learning_rate": 7.362137578169415e-06,
+      "loss": 4.6255,
+      "step": 15500
+    },
+    {
+      "epoch": 0.8571033720287452,
+      "grad_norm": 4.147850036621094,
+      "learning_rate": 7.347924957362137e-06,
+      "loss": 4.7585,
+      "step": 15505
+    },
+    {
+      "epoch": 0.857379767827529,
+      "grad_norm": 3.926903247833252,
+      "learning_rate": 7.333712336554861e-06,
+      "loss": 5.0353,
+      "step": 15510
+    },
+    {
+      "epoch": 0.8576561636263129,
+      "grad_norm": 4.027170658111572,
+      "learning_rate": 7.3194997157475835e-06,
+      "loss": 4.9129,
+      "step": 15515
+    },
+    {
+      "epoch": 0.8579325594250967,
+      "grad_norm": 3.412487030029297,
+      "learning_rate": 7.3052870949403075e-06,
+      "loss": 4.8461,
+      "step": 15520
+    },
+    {
+      "epoch": 0.8582089552238806,
+      "grad_norm": 4.255180358886719,
+      "learning_rate": 7.291074474133031e-06,
+      "loss": 5.133,
+      "step": 15525
+    },
+    {
+      "epoch": 0.8584853510226644,
+      "grad_norm": 3.505549192428589,
+      "learning_rate": 7.276861853325753e-06,
+      "loss": 5.1632,
+      "step": 15530
+    },
+    {
+      "epoch": 0.8587617468214483,
+      "grad_norm": 3.490243434906006,
+      "learning_rate": 7.262649232518477e-06,
+      "loss": 4.8109,
+      "step": 15535
+    },
+    {
+      "epoch": 0.8590381426202321,
+      "grad_norm": 3.9529364109039307,
+      "learning_rate": 7.2484366117112e-06,
+      "loss": 4.7609,
+      "step": 15540
+    },
+    {
+      "epoch": 0.8593145384190161,
+      "grad_norm": 4.20280122756958,
+      "learning_rate": 7.234223990903922e-06,
+      "loss": 4.8306,
+      "step": 15545
+    },
+    {
+      "epoch": 0.8595909342177999,
+      "grad_norm": 3.574753761291504,
+      "learning_rate": 7.220011370096646e-06,
+      "loss": 5.051,
+      "step": 15550
+    },
+    {
+      "epoch": 0.8598673300165838,
+      "grad_norm": 3.6286284923553467,
+      "learning_rate": 7.2057987492893694e-06,
+      "loss": 4.7709,
+      "step": 15555
+    },
+    {
+      "epoch": 0.8601437258153676,
+      "grad_norm": 5.038646697998047,
+      "learning_rate": 7.191586128482092e-06,
+      "loss": 4.9982,
+      "step": 15560
+    },
+    {
+      "epoch": 0.8604201216141515,
+      "grad_norm": 3.0044443607330322,
+      "learning_rate": 7.177373507674816e-06,
+      "loss": 4.8929,
+      "step": 15565
+    },
+    {
+      "epoch": 0.8606965174129353,
+      "grad_norm": 3.639540672302246,
+      "learning_rate": 7.163160886867539e-06,
+      "loss": 4.9908,
+      "step": 15570
+    },
+    {
+      "epoch": 0.8609729132117192,
+      "grad_norm": 3.9962198734283447,
+      "learning_rate": 7.148948266060263e-06,
+      "loss": 4.6839,
+      "step": 15575
+    },
+    {
+      "epoch": 0.861249309010503,
+      "grad_norm": 3.4947597980499268,
+      "learning_rate": 7.134735645252985e-06,
+      "loss": 4.9089,
+      "step": 15580
+    },
+    {
+      "epoch": 0.8615257048092869,
+      "grad_norm": 3.2274651527404785,
+      "learning_rate": 7.120523024445707e-06,
+      "loss": 4.6201,
+      "step": 15585
+    },
+    {
+      "epoch": 0.8618021006080707,
+      "grad_norm": 4.237607955932617,
+      "learning_rate": 7.106310403638431e-06,
+      "loss": 5.0777,
+      "step": 15590
+    },
+    {
+      "epoch": 0.8620784964068546,
+      "grad_norm": 2.7979462146759033,
+      "learning_rate": 7.0920977828311545e-06,
+      "loss": 4.7732,
+      "step": 15595
+    },
+    {
+      "epoch": 0.8623548922056384,
+      "grad_norm": 3.392643928527832,
+      "learning_rate": 7.077885162023877e-06,
+      "loss": 4.8292,
+      "step": 15600
+    },
+    {
+      "epoch": 0.8626312880044223,
+      "grad_norm": 4.662559509277344,
+      "learning_rate": 7.063672541216601e-06,
+      "loss": 4.7918,
+      "step": 15605
+    },
+    {
+      "epoch": 0.8629076838032061,
+      "grad_norm": 3.801340341567993,
+      "learning_rate": 7.049459920409324e-06,
+      "loss": 5.173,
+      "step": 15610
+    },
+    {
+      "epoch": 0.8631840796019901,
+      "grad_norm": 3.8719255924224854,
+      "learning_rate": 7.035247299602046e-06,
+      "loss": 4.9831,
+      "step": 15615
+    },
+    {
+      "epoch": 0.8634604754007739,
+      "grad_norm": 3.9387316703796387,
+      "learning_rate": 7.02103467879477e-06,
+      "loss": 4.8149,
+      "step": 15620
+    },
+    {
+      "epoch": 0.8637368711995578,
+      "grad_norm": 3.568291664123535,
+      "learning_rate": 7.006822057987493e-06,
+      "loss": 4.8419,
+      "step": 15625
+    },
+    {
+      "epoch": 0.8640132669983416,
+      "grad_norm": 4.031901836395264,
+      "learning_rate": 6.992609437180217e-06,
+      "loss": 4.6986,
+      "step": 15630
+    },
+    {
+      "epoch": 0.8642896627971255,
+      "grad_norm": 4.286612033843994,
+      "learning_rate": 6.9783968163729396e-06,
+      "loss": 4.9399,
+      "step": 15635
+    },
+    {
+      "epoch": 0.8645660585959093,
+      "grad_norm": 3.5853421688079834,
+      "learning_rate": 6.964184195565662e-06,
+      "loss": 5.2017,
+      "step": 15640
+    },
+    {
+      "epoch": 0.8648424543946932,
+      "grad_norm": 4.815415382385254,
+      "learning_rate": 6.949971574758387e-06,
+      "loss": 5.2114,
+      "step": 15645
+    },
+    {
+      "epoch": 0.865118850193477,
+      "grad_norm": 3.401326894760132,
+      "learning_rate": 6.935758953951109e-06,
+      "loss": 4.821,
+      "step": 15650
+    },
+    {
+      "epoch": 0.8653952459922609,
+      "grad_norm": 3.949343681335449,
+      "learning_rate": 6.921546333143831e-06,
+      "loss": 5.2086,
+      "step": 15655
+    },
+    {
+      "epoch": 0.8656716417910447,
+      "grad_norm": 4.092352867126465,
+      "learning_rate": 6.907333712336555e-06,
+      "loss": 4.8635,
+      "step": 15660
+    },
+    {
+      "epoch": 0.8659480375898286,
+      "grad_norm": 4.154953956604004,
+      "learning_rate": 6.893121091529278e-06,
+      "loss": 4.7274,
+      "step": 15665
+    },
+    {
+      "epoch": 0.8662244333886125,
+      "grad_norm": 3.2108614444732666,
+      "learning_rate": 6.878908470722001e-06,
+      "loss": 5.0751,
+      "step": 15670
+    },
+    {
+      "epoch": 0.8665008291873963,
+      "grad_norm": 3.6406795978546143,
+      "learning_rate": 6.864695849914725e-06,
+      "loss": 4.935,
+      "step": 15675
+    },
+    {
+      "epoch": 0.8667772249861803,
+      "grad_norm": 3.644953727722168,
+      "learning_rate": 6.850483229107448e-06,
+      "loss": 5.0036,
+      "step": 15680
+    },
+    {
+      "epoch": 0.8670536207849641,
+      "grad_norm": 3.906993865966797,
+      "learning_rate": 6.836270608300172e-06,
+      "loss": 4.7136,
+      "step": 15685
+    },
+    {
+      "epoch": 0.867330016583748,
+      "grad_norm": 3.677238702774048,
+      "learning_rate": 6.822057987492894e-06,
+      "loss": 4.8841,
+      "step": 15690
+    },
+    {
+      "epoch": 0.8676064123825318,
+      "grad_norm": 3.396554946899414,
+      "learning_rate": 6.807845366685617e-06,
+      "loss": 4.9053,
+      "step": 15695
+    },
+    {
+      "epoch": 0.8678828081813157,
+      "grad_norm": 4.678235054016113,
+      "learning_rate": 6.793632745878341e-06,
+      "loss": 4.7476,
+      "step": 15700
+    },
+    {
+      "epoch": 0.8681592039800995,
+      "grad_norm": 3.4147768020629883,
+      "learning_rate": 6.779420125071063e-06,
+      "loss": 5.0731,
+      "step": 15705
+    },
+    {
+      "epoch": 0.8684355997788834,
+      "grad_norm": 4.460427761077881,
+      "learning_rate": 6.765207504263786e-06,
+      "loss": 4.9435,
+      "step": 15710
+    },
+    {
+      "epoch": 0.8687119955776672,
+      "grad_norm": 4.205367565155029,
+      "learning_rate": 6.75099488345651e-06,
+      "loss": 4.7848,
+      "step": 15715
+    },
+    {
+      "epoch": 0.8689883913764511,
+      "grad_norm": 3.436741828918457,
+      "learning_rate": 6.736782262649233e-06,
+      "loss": 4.7811,
+      "step": 15720
+    },
+    {
+      "epoch": 0.8692647871752349,
+      "grad_norm": 3.21610164642334,
+      "learning_rate": 6.722569641841955e-06,
+      "loss": 4.8657,
+      "step": 15725
+    },
+    {
+      "epoch": 0.8695411829740188,
+      "grad_norm": 3.93279767036438,
+      "learning_rate": 6.708357021034679e-06,
+      "loss": 4.8837,
+      "step": 15730
+    },
+    {
+      "epoch": 0.8698175787728026,
+      "grad_norm": 3.3323113918304443,
+      "learning_rate": 6.694144400227402e-06,
+      "loss": 5.143,
+      "step": 15735
+    },
+    {
+      "epoch": 0.8700939745715865,
+      "grad_norm": 3.971737861633301,
+      "learning_rate": 6.679931779420126e-06,
+      "loss": 4.9652,
+      "step": 15740
+    },
+    {
+      "epoch": 0.8703703703703703,
+      "grad_norm": 3.596555471420288,
+      "learning_rate": 6.6657191586128485e-06,
+      "loss": 4.9259,
+      "step": 15745
+    },
+    {
+      "epoch": 0.8706467661691543,
+      "grad_norm": 3.582368850708008,
+      "learning_rate": 6.651506537805572e-06,
+      "loss": 5.0243,
+      "step": 15750
+    },
+    {
+      "epoch": 0.8709231619679381,
+      "grad_norm": 3.9322404861450195,
+      "learning_rate": 6.637293916998296e-06,
+      "loss": 5.2276,
+      "step": 15755
+    },
+    {
+      "epoch": 0.871199557766722,
+      "grad_norm": 4.058863162994385,
+      "learning_rate": 6.623081296191018e-06,
+      "loss": 5.2218,
+      "step": 15760
+    },
+    {
+      "epoch": 0.8714759535655058,
+      "grad_norm": 3.4161527156829834,
+      "learning_rate": 6.608868675383741e-06,
+      "loss": 4.6876,
+      "step": 15765
+    },
+    {
+      "epoch": 0.8717523493642897,
+      "grad_norm": 2.900460958480835,
+      "learning_rate": 6.594656054576465e-06,
+      "loss": 4.7734,
+      "step": 15770
+    },
+    {
+      "epoch": 0.8720287451630735,
+      "grad_norm": 3.5918712615966797,
+      "learning_rate": 6.580443433769187e-06,
+      "loss": 4.7184,
+      "step": 15775
+    },
+    {
+      "epoch": 0.8723051409618574,
+      "grad_norm": 3.992398500442505,
+      "learning_rate": 6.5662308129619095e-06,
+      "loss": 4.966,
+      "step": 15780
+    },
+    {
+      "epoch": 0.8725815367606412,
+      "grad_norm": 3.817514419555664,
+      "learning_rate": 6.5520181921546335e-06,
+      "loss": 5.0287,
+      "step": 15785
+    },
+    {
+      "epoch": 0.8728579325594251,
+      "grad_norm": 3.5824077129364014,
+      "learning_rate": 6.537805571347357e-06,
+      "loss": 4.9329,
+      "step": 15790
+    },
+    {
+      "epoch": 0.8731343283582089,
+      "grad_norm": 3.7401037216186523,
+      "learning_rate": 6.523592950540081e-06,
+      "loss": 4.9287,
+      "step": 15795
+    },
+    {
+      "epoch": 0.8734107241569928,
+      "grad_norm": 3.4796454906463623,
+      "learning_rate": 6.509380329732803e-06,
+      "loss": 4.9727,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8736871199557766,
+      "grad_norm": 5.003295421600342,
+      "learning_rate": 6.495167708925526e-06,
+      "loss": 4.7314,
+      "step": 15805
+    },
+    {
+      "epoch": 0.8739635157545605,
+      "grad_norm": 4.505870819091797,
+      "learning_rate": 6.48095508811825e-06,
+      "loss": 5.0809,
+      "step": 15810
+    },
+    {
+      "epoch": 0.8742399115533444,
+      "grad_norm": 3.863006114959717,
+      "learning_rate": 6.466742467310972e-06,
+      "loss": 4.6385,
+      "step": 15815
+    },
+    {
+      "epoch": 0.8745163073521283,
+      "grad_norm": 3.4922070503234863,
+      "learning_rate": 6.4525298465036955e-06,
+      "loss": 4.7127,
+      "step": 15820
+    },
+    {
+      "epoch": 0.8747927031509121,
+      "grad_norm": 3.6463775634765625,
+      "learning_rate": 6.4383172256964194e-06,
+      "loss": 4.938,
+      "step": 15825
+    },
+    {
+      "epoch": 0.875069098949696,
+      "grad_norm": 3.5146822929382324,
+      "learning_rate": 6.424104604889142e-06,
+      "loss": 4.8737,
+      "step": 15830
+    },
+    {
+      "epoch": 0.8753454947484798,
+      "grad_norm": 2.9423630237579346,
+      "learning_rate": 6.409891984081864e-06,
+      "loss": 4.7986,
+      "step": 15835
+    },
+    {
+      "epoch": 0.8756218905472637,
+      "grad_norm": 4.098637580871582,
+      "learning_rate": 6.395679363274589e-06,
+      "loss": 4.7571,
+      "step": 15840
+    },
+    {
+      "epoch": 0.8758982863460475,
+      "grad_norm": 2.9004056453704834,
+      "learning_rate": 6.381466742467311e-06,
+      "loss": 4.716,
+      "step": 15845
+    },
+    {
+      "epoch": 0.8761746821448314,
+      "grad_norm": 3.797128915786743,
+      "learning_rate": 6.367254121660035e-06,
+      "loss": 5.0502,
+      "step": 15850
+    },
+    {
+      "epoch": 0.8764510779436152,
+      "grad_norm": 3.994654417037964,
+      "learning_rate": 6.353041500852757e-06,
+      "loss": 4.898,
+      "step": 15855
+    },
+    {
+      "epoch": 0.8767274737423991,
+      "grad_norm": 3.2940444946289062,
+      "learning_rate": 6.3388288800454805e-06,
+      "loss": 4.8664,
+      "step": 15860
+    },
+    {
+      "epoch": 0.8770038695411829,
+      "grad_norm": 4.2146124839782715,
+      "learning_rate": 6.3246162592382045e-06,
+      "loss": 4.8941,
+      "step": 15865
+    },
+    {
+      "epoch": 0.8772802653399668,
+      "grad_norm": 3.3862946033477783,
+      "learning_rate": 6.310403638430927e-06,
+      "loss": 4.8247,
+      "step": 15870
+    },
+    {
+      "epoch": 0.8775566611387506,
+      "grad_norm": 3.480405807495117,
+      "learning_rate": 6.29619101762365e-06,
+      "loss": 5.1447,
+      "step": 15875
+    },
+    {
+      "epoch": 0.8778330569375346,
+      "grad_norm": 3.4733119010925293,
+      "learning_rate": 6.281978396816374e-06,
+      "loss": 4.6635,
+      "step": 15880
+    },
+    {
+      "epoch": 0.8781094527363185,
+      "grad_norm": 3.853881359100342,
+      "learning_rate": 6.267765776009096e-06,
+      "loss": 5.0099,
+      "step": 15885
+    },
+    {
+      "epoch": 0.8783858485351023,
+      "grad_norm": 3.801287889480591,
+      "learning_rate": 6.253553155201819e-06,
+      "loss": 4.6277,
+      "step": 15890
+    },
+    {
+      "epoch": 0.8786622443338862,
+      "grad_norm": 3.785771131515503,
+      "learning_rate": 6.239340534394543e-06,
+      "loss": 4.7045,
+      "step": 15895
+    },
+    {
+      "epoch": 0.87893864013267,
+      "grad_norm": 4.128140449523926,
+      "learning_rate": 6.2251279135872656e-06,
+      "loss": 5.0229,
+      "step": 15900
+    },
+    {
+      "epoch": 0.8792150359314539,
+      "grad_norm": 3.9363701343536377,
+      "learning_rate": 6.210915292779989e-06,
+      "loss": 5.1252,
+      "step": 15905
+    },
+    {
+      "epoch": 0.8794914317302377,
+      "grad_norm": 3.420142889022827,
+      "learning_rate": 6.196702671972713e-06,
+      "loss": 5.1304,
+      "step": 15910
+    },
+    {
+      "epoch": 0.8797678275290216,
+      "grad_norm": 4.293994426727295,
+      "learning_rate": 6.182490051165435e-06,
+      "loss": 5.1745,
+      "step": 15915
+    },
+    {
+      "epoch": 0.8800442233278054,
+      "grad_norm": 4.5453314781188965,
+      "learning_rate": 6.168277430358158e-06,
+      "loss": 4.7134,
+      "step": 15920
+    },
+    {
+      "epoch": 0.8803206191265893,
+      "grad_norm": 3.731394052505493,
+      "learning_rate": 6.154064809550881e-06,
+      "loss": 4.7987,
+      "step": 15925
+    },
+    {
+      "epoch": 0.8805970149253731,
+      "grad_norm": 3.9155333042144775,
+      "learning_rate": 6.139852188743605e-06,
+      "loss": 4.8764,
+      "step": 15930
+    },
+    {
+      "epoch": 0.880873410724157,
+      "grad_norm": 3.251206636428833,
+      "learning_rate": 6.1256395679363275e-06,
+      "loss": 4.9742,
+      "step": 15935
+    },
+    {
+      "epoch": 0.8811498065229408,
+      "grad_norm": 3.428173542022705,
+      "learning_rate": 6.111426947129051e-06,
+      "loss": 5.0398,
+      "step": 15940
+    },
+    {
+      "epoch": 0.8814262023217247,
+      "grad_norm": 4.512082576751709,
+      "learning_rate": 6.097214326321774e-06,
+      "loss": 4.9136,
+      "step": 15945
+    },
+    {
+      "epoch": 0.8817025981205086,
+      "grad_norm": 3.4978713989257812,
+      "learning_rate": 6.083001705514498e-06,
+      "loss": 4.732,
+      "step": 15950
+    },
+    {
+      "epoch": 0.8819789939192925,
+      "grad_norm": 4.127050876617432,
+      "learning_rate": 6.06878908470722e-06,
+      "loss": 4.9104,
+      "step": 15955
+    },
+    {
+      "epoch": 0.8822553897180763,
+      "grad_norm": 3.6599040031433105,
+      "learning_rate": 6.054576463899943e-06,
+      "loss": 5.2423,
+      "step": 15960
+    },
+    {
+      "epoch": 0.8825317855168602,
+      "grad_norm": 3.895284652709961,
+      "learning_rate": 6.040363843092667e-06,
+      "loss": 4.9198,
+      "step": 15965
+    },
+    {
+      "epoch": 0.882808181315644,
+      "grad_norm": 3.3798611164093018,
+      "learning_rate": 6.026151222285389e-06,
+      "loss": 4.7896,
+      "step": 15970
+    },
+    {
+      "epoch": 0.8830845771144279,
+      "grad_norm": 3.8356308937072754,
+      "learning_rate": 6.0119386014781126e-06,
+      "loss": 5.1028,
+      "step": 15975
+    },
+    {
+      "epoch": 0.8833609729132117,
+      "grad_norm": 3.9979400634765625,
+      "learning_rate": 5.997725980670836e-06,
+      "loss": 4.9204,
+      "step": 15980
+    },
+    {
+      "epoch": 0.8836373687119956,
+      "grad_norm": 3.993461847305298,
+      "learning_rate": 5.98351335986356e-06,
+      "loss": 4.7276,
+      "step": 15985
+    },
+    {
+      "epoch": 0.8839137645107794,
+      "grad_norm": 3.9294278621673584,
+      "learning_rate": 5.969300739056282e-06,
+      "loss": 4.666,
+      "step": 15990
+    },
+    {
+      "epoch": 0.8841901603095633,
+      "grad_norm": 5.005484104156494,
+      "learning_rate": 5.955088118249005e-06,
+      "loss": 4.9111,
+      "step": 15995
+    },
+    {
+      "epoch": 0.8844665561083471,
+      "grad_norm": 3.216432809829712,
+      "learning_rate": 5.940875497441729e-06,
+      "loss": 4.6319,
+      "step": 16000
+    },
+    {
+      "epoch": 0.884742951907131,
+      "grad_norm": 3.6470518112182617,
+      "learning_rate": 5.926662876634452e-06,
+      "loss": 4.7934,
+      "step": 16005
+    },
+    {
+      "epoch": 0.8850193477059148,
+      "grad_norm": 3.7931535243988037,
+      "learning_rate": 5.9124502558271745e-06,
+      "loss": 4.9271,
+      "step": 16010
+    },
+    {
+      "epoch": 0.8852957435046988,
+      "grad_norm": 3.077409267425537,
+      "learning_rate": 5.898237635019898e-06,
+      "loss": 4.7351,
+      "step": 16015
+    },
+    {
+      "epoch": 0.8855721393034826,
+      "grad_norm": 3.6445906162261963,
+      "learning_rate": 5.884025014212622e-06,
+      "loss": 4.8911,
+      "step": 16020
+    },
+    {
+      "epoch": 0.8858485351022665,
+      "grad_norm": 4.014408111572266,
+      "learning_rate": 5.869812393405344e-06,
+      "loss": 4.9776,
+      "step": 16025
+    },
+    {
+      "epoch": 0.8861249309010503,
+      "grad_norm": 3.4272587299346924,
+      "learning_rate": 5.855599772598067e-06,
+      "loss": 4.9316,
+      "step": 16030
+    },
+    {
+      "epoch": 0.8864013266998342,
+      "grad_norm": 3.4246294498443604,
+      "learning_rate": 5.841387151790791e-06,
+      "loss": 4.7851,
+      "step": 16035
+    },
+    {
+      "epoch": 0.886677722498618,
+      "grad_norm": 3.426506996154785,
+      "learning_rate": 5.827174530983514e-06,
+      "loss": 5.0848,
+      "step": 16040
+    },
+    {
+      "epoch": 0.8869541182974019,
+      "grad_norm": 4.0810956954956055,
+      "learning_rate": 5.812961910176236e-06,
+      "loss": 4.5215,
+      "step": 16045
+    },
+    {
+      "epoch": 0.8872305140961857,
+      "grad_norm": 2.9384403228759766,
+      "learning_rate": 5.7987492893689595e-06,
+      "loss": 4.7552,
+      "step": 16050
+    },
+    {
+      "epoch": 0.8875069098949696,
+      "grad_norm": 3.227759838104248,
+      "learning_rate": 5.7845366685616835e-06,
+      "loss": 4.7817,
+      "step": 16055
+    },
+    {
+      "epoch": 0.8877833056937534,
+      "grad_norm": 4.257302761077881,
+      "learning_rate": 5.770324047754406e-06,
+      "loss": 4.7766,
+      "step": 16060
+    },
+    {
+      "epoch": 0.8880597014925373,
+      "grad_norm": 4.3714919090271,
+      "learning_rate": 5.756111426947129e-06,
+      "loss": 5.0406,
+      "step": 16065
+    },
+    {
+      "epoch": 0.8883360972913211,
+      "grad_norm": 3.33453369140625,
+      "learning_rate": 5.741898806139853e-06,
+      "loss": 5.4695,
+      "step": 16070
+    },
+    {
+      "epoch": 0.888612493090105,
+      "grad_norm": 3.6694791316986084,
+      "learning_rate": 5.727686185332576e-06,
+      "loss": 4.7901,
+      "step": 16075
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 3.3092100620269775,
+      "learning_rate": 5.713473564525298e-06,
+      "loss": 5.0055,
+      "step": 16080
+    },
+    {
+      "epoch": 0.8891652846876728,
+      "grad_norm": 3.254375457763672,
+      "learning_rate": 5.6992609437180215e-06,
+      "loss": 4.8603,
+      "step": 16085
+    },
+    {
+      "epoch": 0.8894416804864566,
+      "grad_norm": 3.5926568508148193,
+      "learning_rate": 5.6850483229107454e-06,
+      "loss": 4.9635,
+      "step": 16090
+    },
+    {
+      "epoch": 0.8897180762852405,
+      "grad_norm": 3.4119133949279785,
+      "learning_rate": 5.6708357021034686e-06,
+      "loss": 5.0395,
+      "step": 16095
+    },
+    {
+      "epoch": 0.8899944720840243,
+      "grad_norm": 3.3834893703460693,
+      "learning_rate": 5.656623081296191e-06,
+      "loss": 5.145,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8902708678828082,
+      "grad_norm": 3.796090602874756,
+      "learning_rate": 5.642410460488915e-06,
+      "loss": 4.8246,
+      "step": 16105
+    },
+    {
+      "epoch": 0.8905472636815921,
+      "grad_norm": 3.5259482860565186,
+      "learning_rate": 5.628197839681638e-06,
+      "loss": 4.8869,
+      "step": 16110
+    },
+    {
+      "epoch": 0.8908236594803759,
+      "grad_norm": 4.100551605224609,
+      "learning_rate": 5.61398521887436e-06,
+      "loss": 4.674,
+      "step": 16115
+    },
+    {
+      "epoch": 0.8911000552791598,
+      "grad_norm": 4.206357479095459,
+      "learning_rate": 5.599772598067083e-06,
+      "loss": 4.7143,
+      "step": 16120
+    },
+    {
+      "epoch": 0.8913764510779436,
+      "grad_norm": 3.823300838470459,
+      "learning_rate": 5.585559977259807e-06,
+      "loss": 4.8694,
+      "step": 16125
+    },
+    {
+      "epoch": 0.8916528468767275,
+      "grad_norm": 4.283262729644775,
+      "learning_rate": 5.5713473564525305e-06,
+      "loss": 5.0432,
+      "step": 16130
+    },
+    {
+      "epoch": 0.8919292426755113,
+      "grad_norm": 4.096394062042236,
+      "learning_rate": 5.557134735645253e-06,
+      "loss": 4.9027,
+      "step": 16135
+    },
+    {
+      "epoch": 0.8922056384742952,
+      "grad_norm": 3.4035484790802,
+      "learning_rate": 5.542922114837976e-06,
+      "loss": 4.8498,
+      "step": 16140
+    },
+    {
+      "epoch": 0.892482034273079,
+      "grad_norm": 3.974168539047241,
+      "learning_rate": 5.5287094940307e-06,
+      "loss": 4.9876,
+      "step": 16145
+    },
+    {
+      "epoch": 0.892758430071863,
+      "grad_norm": 4.489919662475586,
+      "learning_rate": 5.514496873223423e-06,
+      "loss": 4.757,
+      "step": 16150
+    },
+    {
+      "epoch": 0.8930348258706468,
+      "grad_norm": 3.5792441368103027,
+      "learning_rate": 5.500284252416145e-06,
+      "loss": 4.821,
+      "step": 16155
+    },
+    {
+      "epoch": 0.8933112216694307,
+      "grad_norm": 4.130000591278076,
+      "learning_rate": 5.486071631608869e-06,
+      "loss": 4.5594,
+      "step": 16160
+    },
+    {
+      "epoch": 0.8935876174682145,
+      "grad_norm": 4.009551048278809,
+      "learning_rate": 5.471859010801592e-06,
+      "loss": 5.024,
+      "step": 16165
+    },
+    {
+      "epoch": 0.8938640132669984,
+      "grad_norm": 4.058741569519043,
+      "learning_rate": 5.457646389994315e-06,
+      "loss": 5.031,
+      "step": 16170
+    },
+    {
+      "epoch": 0.8941404090657822,
+      "grad_norm": 3.8095996379852295,
+      "learning_rate": 5.443433769187038e-06,
+      "loss": 4.7272,
+      "step": 16175
+    },
+    {
+      "epoch": 0.8944168048645661,
+      "grad_norm": 3.7943928241729736,
+      "learning_rate": 5.429221148379762e-06,
+      "loss": 4.8543,
+      "step": 16180
+    },
+    {
+      "epoch": 0.8946932006633499,
+      "grad_norm": 3.153428316116333,
+      "learning_rate": 5.415008527572485e-06,
+      "loss": 4.6109,
+      "step": 16185
+    },
+    {
+      "epoch": 0.8949695964621338,
+      "grad_norm": 3.685887575149536,
+      "learning_rate": 5.400795906765207e-06,
+      "loss": 4.676,
+      "step": 16190
+    },
+    {
+      "epoch": 0.8952459922609176,
+      "grad_norm": 3.823957920074463,
+      "learning_rate": 5.386583285957931e-06,
+      "loss": 5.0588,
+      "step": 16195
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 3.5146028995513916,
+      "learning_rate": 5.372370665150654e-06,
+      "loss": 4.5432,
+      "step": 16200
+    },
+    {
+      "epoch": 0.8957987838584853,
+      "grad_norm": 3.8643455505371094,
+      "learning_rate": 5.3581580443433775e-06,
+      "loss": 5.029,
+      "step": 16205
+    },
+    {
+      "epoch": 0.8960751796572692,
+      "grad_norm": 4.048851490020752,
+      "learning_rate": 5.3439454235361e-06,
+      "loss": 4.7677,
+      "step": 16210
+    },
+    {
+      "epoch": 0.896351575456053,
+      "grad_norm": 3.798962354660034,
+      "learning_rate": 5.329732802728824e-06,
+      "loss": 5.0452,
+      "step": 16215
+    },
+    {
+      "epoch": 0.896627971254837,
+      "grad_norm": 3.9125924110412598,
+      "learning_rate": 5.315520181921547e-06,
+      "loss": 4.8263,
+      "step": 16220
+    },
+    {
+      "epoch": 0.8969043670536208,
+      "grad_norm": 5.244776248931885,
+      "learning_rate": 5.301307561114269e-06,
+      "loss": 5.1466,
+      "step": 16225
+    },
+    {
+      "epoch": 0.8971807628524047,
+      "grad_norm": 4.3974528312683105,
+      "learning_rate": 5.287094940306993e-06,
+      "loss": 4.6972,
+      "step": 16230
+    },
+    {
+      "epoch": 0.8974571586511885,
+      "grad_norm": 3.663856267929077,
+      "learning_rate": 5.272882319499716e-06,
+      "loss": 4.8124,
+      "step": 16235
+    },
+    {
+      "epoch": 0.8977335544499724,
+      "grad_norm": 3.177274465560913,
+      "learning_rate": 5.258669698692439e-06,
+      "loss": 4.7413,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8980099502487562,
+      "grad_norm": 4.527139663696289,
+      "learning_rate": 5.244457077885162e-06,
+      "loss": 4.6418,
+      "step": 16245
+    },
+    {
+      "epoch": 0.8982863460475401,
+      "grad_norm": 3.90124773979187,
+      "learning_rate": 5.230244457077886e-06,
+      "loss": 4.985,
+      "step": 16250
+    },
+    {
+      "epoch": 0.8985627418463239,
+      "grad_norm": 4.100987911224365,
+      "learning_rate": 5.216031836270609e-06,
+      "loss": 4.9227,
+      "step": 16255
+    },
+    {
+      "epoch": 0.8988391376451078,
+      "grad_norm": 3.7627832889556885,
+      "learning_rate": 5.201819215463331e-06,
+      "loss": 4.658,
+      "step": 16260
+    },
+    {
+      "epoch": 0.8991155334438916,
+      "grad_norm": 3.8064067363739014,
+      "learning_rate": 5.187606594656055e-06,
+      "loss": 4.7242,
+      "step": 16265
+    },
+    {
+      "epoch": 0.8993919292426755,
+      "grad_norm": 4.380611419677734,
+      "learning_rate": 5.173393973848778e-06,
+      "loss": 4.957,
+      "step": 16270
+    },
+    {
+      "epoch": 0.8996683250414593,
+      "grad_norm": 4.4961676597595215,
+      "learning_rate": 5.159181353041501e-06,
+      "loss": 4.7186,
+      "step": 16275
+    },
+    {
+      "epoch": 0.8999447208402432,
+      "grad_norm": 4.06073522567749,
+      "learning_rate": 5.144968732234224e-06,
+      "loss": 5.024,
+      "step": 16280
+    },
+    {
+      "epoch": 0.900221116639027,
+      "grad_norm": 3.396531581878662,
+      "learning_rate": 5.130756111426948e-06,
+      "loss": 4.8516,
+      "step": 16285
+    },
+    {
+      "epoch": 0.900497512437811,
+      "grad_norm": 3.7098519802093506,
+      "learning_rate": 5.116543490619671e-06,
+      "loss": 4.719,
+      "step": 16290
+    },
+    {
+      "epoch": 0.9007739082365948,
+      "grad_norm": 3.9446892738342285,
+      "learning_rate": 5.102330869812394e-06,
+      "loss": 5.1406,
+      "step": 16295
+    },
+    {
+      "epoch": 0.9010503040353787,
+      "grad_norm": 3.210261344909668,
+      "learning_rate": 5.088118249005117e-06,
+      "loss": 4.6729,
+      "step": 16300
+    },
+    {
+      "epoch": 0.9013266998341625,
+      "grad_norm": 4.007756233215332,
+      "learning_rate": 5.07390562819784e-06,
+      "loss": 4.7955,
+      "step": 16305
+    },
+    {
+      "epoch": 0.9016030956329464,
+      "grad_norm": 3.227644205093384,
+      "learning_rate": 5.059693007390563e-06,
+      "loss": 4.7183,
+      "step": 16310
+    },
+    {
+      "epoch": 0.9018794914317302,
+      "grad_norm": 4.634718894958496,
+      "learning_rate": 5.0454803865832855e-06,
+      "loss": 5.101,
+      "step": 16315
+    },
+    {
+      "epoch": 0.9021558872305141,
+      "grad_norm": 4.174224853515625,
+      "learning_rate": 5.0312677657760095e-06,
+      "loss": 4.9855,
+      "step": 16320
+    },
+    {
+      "epoch": 0.9024322830292979,
+      "grad_norm": 3.7757952213287354,
+      "learning_rate": 5.017055144968733e-06,
+      "loss": 5.0235,
+      "step": 16325
+    },
+    {
+      "epoch": 0.9027086788280818,
+      "grad_norm": 3.770765781402588,
+      "learning_rate": 5.002842524161456e-06,
+      "loss": 5.1194,
+      "step": 16330
+    },
+    {
+      "epoch": 0.9029850746268657,
+      "grad_norm": 3.6249911785125732,
+      "learning_rate": 4.988629903354178e-06,
+      "loss": 4.7802,
+      "step": 16335
+    },
+    {
+      "epoch": 0.9032614704256495,
+      "grad_norm": 4.837128162384033,
+      "learning_rate": 4.974417282546902e-06,
+      "loss": 4.599,
+      "step": 16340
+    },
+    {
+      "epoch": 0.9035378662244334,
+      "grad_norm": 4.374870777130127,
+      "learning_rate": 4.960204661739625e-06,
+      "loss": 4.8379,
+      "step": 16345
+    },
+    {
+      "epoch": 0.9038142620232172,
+      "grad_norm": 4.0973734855651855,
+      "learning_rate": 4.945992040932348e-06,
+      "loss": 5.0881,
+      "step": 16350
+    },
+    {
+      "epoch": 0.9040906578220012,
+      "grad_norm": 4.1243977546691895,
+      "learning_rate": 4.9317794201250714e-06,
+      "loss": 5.1151,
+      "step": 16355
+    },
+    {
+      "epoch": 0.904367053620785,
+      "grad_norm": 3.1973705291748047,
+      "learning_rate": 4.9175667993177946e-06,
+      "loss": 5.0274,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9046434494195689,
+      "grad_norm": 3.458981990814209,
+      "learning_rate": 4.903354178510518e-06,
+      "loss": 4.8251,
+      "step": 16365
+    },
+    {
+      "epoch": 0.9049198452183527,
+      "grad_norm": 3.640015125274658,
+      "learning_rate": 4.88914155770324e-06,
+      "loss": 4.9673,
+      "step": 16370
+    },
+    {
+      "epoch": 0.9051962410171366,
+      "grad_norm": 3.2531189918518066,
+      "learning_rate": 4.874928936895964e-06,
+      "loss": 4.5497,
+      "step": 16375
+    },
+    {
+      "epoch": 0.9054726368159204,
+      "grad_norm": 3.353008508682251,
+      "learning_rate": 4.860716316088687e-06,
+      "loss": 4.8228,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9057490326147043,
+      "grad_norm": 4.130444049835205,
+      "learning_rate": 4.84650369528141e-06,
+      "loss": 4.7594,
+      "step": 16385
+    },
+    {
+      "epoch": 0.9060254284134881,
+      "grad_norm": 3.1232221126556396,
+      "learning_rate": 4.832291074474133e-06,
+      "loss": 4.9349,
+      "step": 16390
+    },
+    {
+      "epoch": 0.906301824212272,
+      "grad_norm": 3.7138428688049316,
+      "learning_rate": 4.8180784536668565e-06,
+      "loss": 4.7187,
+      "step": 16395
+    },
+    {
+      "epoch": 0.9065782200110558,
+      "grad_norm": 3.310006618499756,
+      "learning_rate": 4.80386583285958e-06,
+      "loss": 5.0194,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9068546158098397,
+      "grad_norm": 5.019920349121094,
+      "learning_rate": 4.789653212052303e-06,
+      "loss": 4.9197,
+      "step": 16405
+    },
+    {
+      "epoch": 0.9071310116086235,
+      "grad_norm": 3.792968273162842,
+      "learning_rate": 4.775440591245026e-06,
+      "loss": 4.9581,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9074074074074074,
+      "grad_norm": 4.522578716278076,
+      "learning_rate": 4.761227970437749e-06,
+      "loss": 4.7076,
+      "step": 16415
+    },
+    {
+      "epoch": 0.9076838032061912,
+      "grad_norm": 3.9118542671203613,
+      "learning_rate": 4.747015349630472e-06,
+      "loss": 4.9707,
+      "step": 16420
+    },
+    {
+      "epoch": 0.9079601990049752,
+      "grad_norm": 4.168303489685059,
+      "learning_rate": 4.732802728823195e-06,
+      "loss": 4.603,
+      "step": 16425
+    },
+    {
+      "epoch": 0.908236594803759,
+      "grad_norm": 3.641812324523926,
+      "learning_rate": 4.7185901080159184e-06,
+      "loss": 4.8405,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9085129906025429,
+      "grad_norm": 4.432767868041992,
+      "learning_rate": 4.7043774872086416e-06,
+      "loss": 5.1405,
+      "step": 16435
+    },
+    {
+      "epoch": 0.9087893864013267,
+      "grad_norm": 3.7158989906311035,
+      "learning_rate": 4.690164866401365e-06,
+      "loss": 4.8914,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9090657822001106,
+      "grad_norm": 3.8096296787261963,
+      "learning_rate": 4.675952245594088e-06,
+      "loss": 4.8564,
+      "step": 16445
+    },
+    {
+      "epoch": 0.9093421779988944,
+      "grad_norm": 3.3491950035095215,
+      "learning_rate": 4.661739624786811e-06,
+      "loss": 5.1232,
+      "step": 16450
+    },
+    {
+      "epoch": 0.9096185737976783,
+      "grad_norm": 4.407762050628662,
+      "learning_rate": 4.647527003979534e-06,
+      "loss": 5.2513,
+      "step": 16455
+    },
+    {
+      "epoch": 0.9098949695964621,
+      "grad_norm": 4.356717109680176,
+      "learning_rate": 4.633314383172257e-06,
+      "loss": 5.0331,
+      "step": 16460
+    },
+    {
+      "epoch": 0.910171365395246,
+      "grad_norm": 3.2973270416259766,
+      "learning_rate": 4.61910176236498e-06,
+      "loss": 4.8999,
+      "step": 16465
+    },
+    {
+      "epoch": 0.9104477611940298,
+      "grad_norm": 3.944094181060791,
+      "learning_rate": 4.6048891415577035e-06,
+      "loss": 4.7392,
+      "step": 16470
+    },
+    {
+      "epoch": 0.9107241569928137,
+      "grad_norm": 3.541595220565796,
+      "learning_rate": 4.590676520750427e-06,
+      "loss": 5.0021,
+      "step": 16475
+    },
+    {
+      "epoch": 0.9110005527915975,
+      "grad_norm": 4.237240314483643,
+      "learning_rate": 4.57646389994315e-06,
+      "loss": 4.586,
+      "step": 16480
+    },
+    {
+      "epoch": 0.9112769485903814,
+      "grad_norm": 3.664316177368164,
+      "learning_rate": 4.562251279135873e-06,
+      "loss": 4.9774,
+      "step": 16485
+    },
+    {
+      "epoch": 0.9115533443891652,
+      "grad_norm": 3.5280942916870117,
+      "learning_rate": 4.548038658328596e-06,
+      "loss": 4.9678,
+      "step": 16490
+    },
+    {
+      "epoch": 0.9118297401879492,
+      "grad_norm": 3.860072135925293,
+      "learning_rate": 4.533826037521319e-06,
+      "loss": 5.1321,
+      "step": 16495
+    },
+    {
+      "epoch": 0.912106135986733,
+      "grad_norm": 3.7712130546569824,
+      "learning_rate": 4.519613416714042e-06,
+      "loss": 4.9678,
+      "step": 16500
+    },
+    {
+      "epoch": 0.9123825317855169,
+      "grad_norm": 3.747433662414551,
+      "learning_rate": 4.505400795906765e-06,
+      "loss": 4.9893,
+      "step": 16505
+    },
+    {
+      "epoch": 0.9126589275843007,
+      "grad_norm": 6.0445170402526855,
+      "learning_rate": 4.4911881750994885e-06,
+      "loss": 4.8376,
+      "step": 16510
+    },
+    {
+      "epoch": 0.9129353233830846,
+      "grad_norm": 3.5901873111724854,
+      "learning_rate": 4.476975554292212e-06,
+      "loss": 4.8362,
+      "step": 16515
+    },
+    {
+      "epoch": 0.9132117191818684,
+      "grad_norm": 3.6722826957702637,
+      "learning_rate": 4.462762933484935e-06,
+      "loss": 5.1034,
+      "step": 16520
+    },
+    {
+      "epoch": 0.9134881149806523,
+      "grad_norm": 3.5389833450317383,
+      "learning_rate": 4.448550312677658e-06,
+      "loss": 4.6817,
+      "step": 16525
+    },
+    {
+      "epoch": 0.9137645107794361,
+      "grad_norm": 3.544811248779297,
+      "learning_rate": 4.434337691870381e-06,
+      "loss": 5.0695,
+      "step": 16530
+    },
+    {
+      "epoch": 0.91404090657822,
+      "grad_norm": 4.132248401641846,
+      "learning_rate": 4.420125071063104e-06,
+      "loss": 4.9197,
+      "step": 16535
+    },
+    {
+      "epoch": 0.9143173023770038,
+      "grad_norm": 3.680356979370117,
+      "learning_rate": 4.405912450255827e-06,
+      "loss": 4.5352,
+      "step": 16540
+    },
+    {
+      "epoch": 0.9145936981757877,
+      "grad_norm": 4.035909175872803,
+      "learning_rate": 4.3916998294485505e-06,
+      "loss": 4.8975,
+      "step": 16545
+    },
+    {
+      "epoch": 0.9148700939745716,
+      "grad_norm": 3.8951597213745117,
+      "learning_rate": 4.377487208641274e-06,
+      "loss": 4.817,
+      "step": 16550
+    },
+    {
+      "epoch": 0.9151464897733554,
+      "grad_norm": 3.9107933044433594,
+      "learning_rate": 4.363274587833997e-06,
+      "loss": 4.7651,
+      "step": 16555
+    },
+    {
+      "epoch": 0.9154228855721394,
+      "grad_norm": 3.4493396282196045,
+      "learning_rate": 4.34906196702672e-06,
+      "loss": 5.076,
+      "step": 16560
+    },
+    {
+      "epoch": 0.9156992813709232,
+      "grad_norm": 4.296689987182617,
+      "learning_rate": 4.334849346219443e-06,
+      "loss": 4.9039,
+      "step": 16565
+    },
+    {
+      "epoch": 0.9159756771697071,
+      "grad_norm": 3.8431396484375,
+      "learning_rate": 4.320636725412166e-06,
+      "loss": 4.6727,
+      "step": 16570
+    },
+    {
+      "epoch": 0.9162520729684909,
+      "grad_norm": 4.500297546386719,
+      "learning_rate": 4.306424104604889e-06,
+      "loss": 5.3912,
+      "step": 16575
+    },
+    {
+      "epoch": 0.9165284687672748,
+      "grad_norm": 4.726571083068848,
+      "learning_rate": 4.292211483797612e-06,
+      "loss": 5.2984,
+      "step": 16580
+    },
+    {
+      "epoch": 0.9168048645660586,
+      "grad_norm": 3.416921615600586,
+      "learning_rate": 4.2779988629903355e-06,
+      "loss": 4.6175,
+      "step": 16585
+    },
+    {
+      "epoch": 0.9170812603648425,
+      "grad_norm": 3.1644046306610107,
+      "learning_rate": 4.263786242183059e-06,
+      "loss": 4.6776,
+      "step": 16590
+    },
+    {
+      "epoch": 0.9173576561636263,
+      "grad_norm": 3.82157826423645,
+      "learning_rate": 4.249573621375782e-06,
+      "loss": 4.881,
+      "step": 16595
+    },
+    {
+      "epoch": 0.9176340519624102,
+      "grad_norm": 3.8752880096435547,
+      "learning_rate": 4.235361000568505e-06,
+      "loss": 4.749,
+      "step": 16600
+    },
+    {
+      "epoch": 0.917910447761194,
+      "grad_norm": 3.842712879180908,
+      "learning_rate": 4.221148379761229e-06,
+      "loss": 4.7253,
+      "step": 16605
+    },
+    {
+      "epoch": 0.9181868435599779,
+      "grad_norm": 4.872412204742432,
+      "learning_rate": 4.206935758953951e-06,
+      "loss": 4.7417,
+      "step": 16610
+    },
+    {
+      "epoch": 0.9184632393587617,
+      "grad_norm": 3.414172649383545,
+      "learning_rate": 4.192723138146674e-06,
+      "loss": 5.1289,
+      "step": 16615
+    },
+    {
+      "epoch": 0.9187396351575456,
+      "grad_norm": 4.3799543380737305,
+      "learning_rate": 4.1785105173393974e-06,
+      "loss": 4.9575,
+      "step": 16620
+    },
+    {
+      "epoch": 0.9190160309563294,
+      "grad_norm": 3.8672125339508057,
+      "learning_rate": 4.164297896532121e-06,
+      "loss": 4.7991,
+      "step": 16625
+    },
+    {
+      "epoch": 0.9192924267551134,
+      "grad_norm": 3.75191593170166,
+      "learning_rate": 4.150085275724844e-06,
+      "loss": 4.8514,
+      "step": 16630
+    },
+    {
+      "epoch": 0.9195688225538972,
+      "grad_norm": 3.8761277198791504,
+      "learning_rate": 4.135872654917567e-06,
+      "loss": 4.9027,
+      "step": 16635
+    },
+    {
+      "epoch": 0.9198452183526811,
+      "grad_norm": 3.8116533756256104,
+      "learning_rate": 4.12166003411029e-06,
+      "loss": 4.7569,
+      "step": 16640
+    },
+    {
+      "epoch": 0.9201216141514649,
+      "grad_norm": 3.882634162902832,
+      "learning_rate": 4.107447413303013e-06,
+      "loss": 4.6937,
+      "step": 16645
+    },
+    {
+      "epoch": 0.9203980099502488,
+      "grad_norm": 3.3166961669921875,
+      "learning_rate": 4.093234792495736e-06,
+      "loss": 4.8617,
+      "step": 16650
+    },
+    {
+      "epoch": 0.9206744057490326,
+      "grad_norm": 4.038979530334473,
+      "learning_rate": 4.079022171688459e-06,
+      "loss": 4.9206,
+      "step": 16655
+    },
+    {
+      "epoch": 0.9209508015478165,
+      "grad_norm": 3.362423896789551,
+      "learning_rate": 4.064809550881183e-06,
+      "loss": 4.9875,
+      "step": 16660
+    },
+    {
+      "epoch": 0.9212271973466003,
+      "grad_norm": 3.3738491535186768,
+      "learning_rate": 4.050596930073906e-06,
+      "loss": 4.7933,
+      "step": 16665
+    },
+    {
+      "epoch": 0.9215035931453842,
+      "grad_norm": 3.503601551055908,
+      "learning_rate": 4.036384309266629e-06,
+      "loss": 5.0732,
+      "step": 16670
+    },
+    {
+      "epoch": 0.921779988944168,
+      "grad_norm": 3.1450142860412598,
+      "learning_rate": 4.022171688459352e-06,
+      "loss": 4.7811,
+      "step": 16675
+    },
+    {
+      "epoch": 0.9220563847429519,
+      "grad_norm": 3.4202752113342285,
+      "learning_rate": 4.007959067652075e-06,
+      "loss": 5.0072,
+      "step": 16680
+    },
+    {
+      "epoch": 0.9223327805417357,
+      "grad_norm": 3.7823784351348877,
+      "learning_rate": 3.993746446844798e-06,
+      "loss": 5.1254,
+      "step": 16685
+    },
+    {
+      "epoch": 0.9226091763405196,
+      "grad_norm": 4.299469947814941,
+      "learning_rate": 3.979533826037521e-06,
+      "loss": 4.9729,
+      "step": 16690
+    },
+    {
+      "epoch": 0.9228855721393034,
+      "grad_norm": 3.722261905670166,
+      "learning_rate": 3.965321205230245e-06,
+      "loss": 4.9265,
+      "step": 16695
+    },
+    {
+      "epoch": 0.9231619679380874,
+      "grad_norm": 4.488846778869629,
+      "learning_rate": 3.9511085844229676e-06,
+      "loss": 4.8332,
+      "step": 16700
+    },
+    {
+      "epoch": 0.9234383637368712,
+      "grad_norm": 3.4476404190063477,
+      "learning_rate": 3.936895963615691e-06,
+      "loss": 4.8358,
+      "step": 16705
+    },
+    {
+      "epoch": 0.9237147595356551,
+      "grad_norm": 3.9329166412353516,
+      "learning_rate": 3.922683342808414e-06,
+      "loss": 4.721,
+      "step": 16710
+    },
+    {
+      "epoch": 0.9239911553344389,
+      "grad_norm": 4.05313777923584,
+      "learning_rate": 3.908470722001138e-06,
+      "loss": 4.9849,
+      "step": 16715
+    },
+    {
+      "epoch": 0.9242675511332228,
+      "grad_norm": 3.66392183303833,
+      "learning_rate": 3.89425810119386e-06,
+      "loss": 4.8486,
+      "step": 16720
+    },
+    {
+      "epoch": 0.9245439469320066,
+      "grad_norm": 4.335777759552002,
+      "learning_rate": 3.880045480386583e-06,
+      "loss": 4.9046,
+      "step": 16725
+    },
+    {
+      "epoch": 0.9248203427307905,
+      "grad_norm": 4.8382954597473145,
+      "learning_rate": 3.865832859579307e-06,
+      "loss": 4.8476,
+      "step": 16730
+    },
+    {
+      "epoch": 0.9250967385295743,
+      "grad_norm": 3.354020357131958,
+      "learning_rate": 3.8516202387720295e-06,
+      "loss": 5.12,
+      "step": 16735
+    },
+    {
+      "epoch": 0.9253731343283582,
+      "grad_norm": 3.5212149620056152,
+      "learning_rate": 3.837407617964753e-06,
+      "loss": 4.6406,
+      "step": 16740
+    },
+    {
+      "epoch": 0.925649530127142,
+      "grad_norm": 4.447071552276611,
+      "learning_rate": 3.823194997157476e-06,
+      "loss": 5.1303,
+      "step": 16745
+    },
+    {
+      "epoch": 0.9259259259259259,
+      "grad_norm": 4.714621067047119,
+      "learning_rate": 3.8089823763501993e-06,
+      "loss": 5.0047,
+      "step": 16750
+    },
+    {
+      "epoch": 0.9262023217247097,
+      "grad_norm": 4.3438568115234375,
+      "learning_rate": 3.794769755542922e-06,
+      "loss": 4.8138,
+      "step": 16755
+    },
+    {
+      "epoch": 0.9264787175234936,
+      "grad_norm": 3.487215280532837,
+      "learning_rate": 3.7805571347356456e-06,
+      "loss": 5.0629,
+      "step": 16760
+    },
+    {
+      "epoch": 0.9267551133222774,
+      "grad_norm": 3.3207781314849854,
+      "learning_rate": 3.7663445139283687e-06,
+      "loss": 4.6624,
+      "step": 16765
+    },
+    {
+      "epoch": 0.9270315091210614,
+      "grad_norm": 3.6874585151672363,
+      "learning_rate": 3.7521318931210914e-06,
+      "loss": 4.9265,
+      "step": 16770
+    },
+    {
+      "epoch": 0.9273079049198453,
+      "grad_norm": 3.417487859725952,
+      "learning_rate": 3.7379192723138145e-06,
+      "loss": 4.8298,
+      "step": 16775
+    },
+    {
+      "epoch": 0.9275843007186291,
+      "grad_norm": 3.985056161880493,
+      "learning_rate": 3.723706651506538e-06,
+      "loss": 5.0497,
+      "step": 16780
+    },
+    {
+      "epoch": 0.927860696517413,
+      "grad_norm": 3.5535130500793457,
+      "learning_rate": 3.7094940306992612e-06,
+      "loss": 4.9327,
+      "step": 16785
+    },
+    {
+      "epoch": 0.9281370923161968,
+      "grad_norm": 3.7775490283966064,
+      "learning_rate": 3.695281409891984e-06,
+      "loss": 4.7599,
+      "step": 16790
+    },
+    {
+      "epoch": 0.9284134881149807,
+      "grad_norm": 3.6912879943847656,
+      "learning_rate": 3.6810687890847075e-06,
+      "loss": 4.73,
+      "step": 16795
+    },
+    {
+      "epoch": 0.9286898839137645,
+      "grad_norm": 3.8959012031555176,
+      "learning_rate": 3.6668561682774306e-06,
+      "loss": 4.6683,
+      "step": 16800
+    },
+    {
+      "epoch": 0.9289662797125484,
+      "grad_norm": 3.763211250305176,
+      "learning_rate": 3.6526435474701538e-06,
+      "loss": 4.9404,
+      "step": 16805
+    },
+    {
+      "epoch": 0.9292426755113322,
+      "grad_norm": 2.960988998413086,
+      "learning_rate": 3.6384309266628765e-06,
+      "loss": 4.7549,
+      "step": 16810
+    },
+    {
+      "epoch": 0.9295190713101161,
+      "grad_norm": 3.7784411907196045,
+      "learning_rate": 3.6242183058556e-06,
+      "loss": 5.0343,
+      "step": 16815
+    },
+    {
+      "epoch": 0.9297954671088999,
+      "grad_norm": 3.5923070907592773,
+      "learning_rate": 3.610005685048323e-06,
+      "loss": 4.9945,
+      "step": 16820
+    },
+    {
+      "epoch": 0.9300718629076838,
+      "grad_norm": 4.040002822875977,
+      "learning_rate": 3.595793064241046e-06,
+      "loss": 4.9812,
+      "step": 16825
+    },
+    {
+      "epoch": 0.9303482587064676,
+      "grad_norm": 4.1851806640625,
+      "learning_rate": 3.5815804434337694e-06,
+      "loss": 5.0198,
+      "step": 16830
+    },
+    {
+      "epoch": 0.9306246545052516,
+      "grad_norm": 3.2891855239868164,
+      "learning_rate": 3.5673678226264926e-06,
+      "loss": 4.9279,
+      "step": 16835
+    },
+    {
+      "epoch": 0.9309010503040354,
+      "grad_norm": 4.148435115814209,
+      "learning_rate": 3.5531552018192157e-06,
+      "loss": 4.7171,
+      "step": 16840
+    },
+    {
+      "epoch": 0.9311774461028193,
+      "grad_norm": 3.529670238494873,
+      "learning_rate": 3.5389425810119384e-06,
+      "loss": 5.2854,
+      "step": 16845
+    },
+    {
+      "epoch": 0.9314538419016031,
+      "grad_norm": 4.003045082092285,
+      "learning_rate": 3.524729960204662e-06,
+      "loss": 4.829,
+      "step": 16850
+    },
+    {
+      "epoch": 0.931730237700387,
+      "grad_norm": 4.24057674407959,
+      "learning_rate": 3.510517339397385e-06,
+      "loss": 4.6103,
+      "step": 16855
+    },
+    {
+      "epoch": 0.9320066334991708,
+      "grad_norm": 4.488296985626221,
+      "learning_rate": 3.4963047185901086e-06,
+      "loss": 4.9311,
+      "step": 16860
+    },
+    {
+      "epoch": 0.9322830292979547,
+      "grad_norm": 3.4120020866394043,
+      "learning_rate": 3.482092097782831e-06,
+      "loss": 4.8596,
+      "step": 16865
+    },
+    {
+      "epoch": 0.9325594250967385,
+      "grad_norm": 3.683335781097412,
+      "learning_rate": 3.4678794769755545e-06,
+      "loss": 4.5523,
+      "step": 16870
+    },
+    {
+      "epoch": 0.9328358208955224,
+      "grad_norm": 3.4997408390045166,
+      "learning_rate": 3.4536668561682776e-06,
+      "loss": 4.8484,
+      "step": 16875
+    },
+    {
+      "epoch": 0.9331122166943062,
+      "grad_norm": 4.172585964202881,
+      "learning_rate": 3.4394542353610003e-06,
+      "loss": 4.8826,
+      "step": 16880
+    },
+    {
+      "epoch": 0.9333886124930901,
+      "grad_norm": 3.2831268310546875,
+      "learning_rate": 3.425241614553724e-06,
+      "loss": 4.8996,
+      "step": 16885
+    },
+    {
+      "epoch": 0.9336650082918739,
+      "grad_norm": 4.524475574493408,
+      "learning_rate": 3.411028993746447e-06,
+      "loss": 5.0182,
+      "step": 16890
+    },
+    {
+      "epoch": 0.9339414040906578,
+      "grad_norm": 3.8012876510620117,
+      "learning_rate": 3.3968163729391706e-06,
+      "loss": 4.9137,
+      "step": 16895
+    },
+    {
+      "epoch": 0.9342177998894416,
+      "grad_norm": 4.124208450317383,
+      "learning_rate": 3.382603752131893e-06,
+      "loss": 4.7937,
+      "step": 16900
+    },
+    {
+      "epoch": 0.9344941956882256,
+      "grad_norm": 3.331831932067871,
+      "learning_rate": 3.3683911313246164e-06,
+      "loss": 4.7215,
+      "step": 16905
+    },
+    {
+      "epoch": 0.9347705914870094,
+      "grad_norm": 4.77909517288208,
+      "learning_rate": 3.3541785105173395e-06,
+      "loss": 4.811,
+      "step": 16910
+    },
+    {
+      "epoch": 0.9350469872857933,
+      "grad_norm": 4.471525192260742,
+      "learning_rate": 3.339965889710063e-06,
+      "loss": 4.704,
+      "step": 16915
+    },
+    {
+      "epoch": 0.9353233830845771,
+      "grad_norm": 3.200888156890869,
+      "learning_rate": 3.325753268902786e-06,
+      "loss": 4.5741,
+      "step": 16920
+    },
+    {
+      "epoch": 0.935599778883361,
+      "grad_norm": 3.5812439918518066,
+      "learning_rate": 3.311540648095509e-06,
+      "loss": 4.9324,
+      "step": 16925
+    },
+    {
+      "epoch": 0.9358761746821448,
+      "grad_norm": 4.601278305053711,
+      "learning_rate": 3.2973280272882325e-06,
+      "loss": 4.908,
+      "step": 16930
+    },
+    {
+      "epoch": 0.9361525704809287,
+      "grad_norm": 3.4622044563293457,
+      "learning_rate": 3.2831154064809548e-06,
+      "loss": 4.5155,
+      "step": 16935
+    },
+    {
+      "epoch": 0.9364289662797125,
+      "grad_norm": 4.049286842346191,
+      "learning_rate": 3.2689027856736783e-06,
+      "loss": 4.7219,
+      "step": 16940
+    },
+    {
+      "epoch": 0.9367053620784964,
+      "grad_norm": 4.182099342346191,
+      "learning_rate": 3.2546901648664015e-06,
+      "loss": 4.8228,
+      "step": 16945
+    },
+    {
+      "epoch": 0.9369817578772802,
+      "grad_norm": 3.8937876224517822,
+      "learning_rate": 3.240477544059125e-06,
+      "loss": 4.8272,
+      "step": 16950
+    },
+    {
+      "epoch": 0.9372581536760641,
+      "grad_norm": 4.102065086364746,
+      "learning_rate": 3.2262649232518477e-06,
+      "loss": 5.0221,
+      "step": 16955
+    },
+    {
+      "epoch": 0.9375345494748479,
+      "grad_norm": 3.8191025257110596,
+      "learning_rate": 3.212052302444571e-06,
+      "loss": 4.9027,
+      "step": 16960
+    },
+    {
+      "epoch": 0.9378109452736318,
+      "grad_norm": 3.0223190784454346,
+      "learning_rate": 3.1978396816372944e-06,
+      "loss": 4.7848,
+      "step": 16965
+    },
+    {
+      "epoch": 0.9380873410724156,
+      "grad_norm": 3.0020852088928223,
+      "learning_rate": 3.1836270608300175e-06,
+      "loss": 4.8542,
+      "step": 16970
+    },
+    {
+      "epoch": 0.9383637368711996,
+      "grad_norm": 3.872197389602661,
+      "learning_rate": 3.1694144400227403e-06,
+      "loss": 4.4044,
+      "step": 16975
+    },
+    {
+      "epoch": 0.9386401326699834,
+      "grad_norm": 4.418979167938232,
+      "learning_rate": 3.1552018192154634e-06,
+      "loss": 4.732,
+      "step": 16980
+    },
+    {
+      "epoch": 0.9389165284687673,
+      "grad_norm": 3.6312499046325684,
+      "learning_rate": 3.140989198408187e-06,
+      "loss": 5.0736,
+      "step": 16985
+    },
+    {
+      "epoch": 0.9391929242675512,
+      "grad_norm": 4.047171592712402,
+      "learning_rate": 3.1267765776009097e-06,
+      "loss": 4.8834,
+      "step": 16990
+    },
+    {
+      "epoch": 0.939469320066335,
+      "grad_norm": 3.8443045616149902,
+      "learning_rate": 3.1125639567936328e-06,
+      "loss": 5.0899,
+      "step": 16995
+    },
+    {
+      "epoch": 0.9397457158651189,
+      "grad_norm": 3.2416129112243652,
+      "learning_rate": 3.0983513359863563e-06,
+      "loss": 4.6055,
+      "step": 17000
+    },
+    {
+      "epoch": 0.9400221116639027,
+      "grad_norm": 4.050750255584717,
+      "learning_rate": 3.084138715179079e-06,
+      "loss": 4.6478,
+      "step": 17005
+    },
+    {
+      "epoch": 0.9402985074626866,
+      "grad_norm": 3.5441677570343018,
+      "learning_rate": 3.0699260943718026e-06,
+      "loss": 4.7012,
+      "step": 17010
+    },
+    {
+      "epoch": 0.9405749032614704,
+      "grad_norm": 3.7427573204040527,
+      "learning_rate": 3.0557134735645253e-06,
+      "loss": 4.8786,
+      "step": 17015
+    },
+    {
+      "epoch": 0.9408512990602543,
+      "grad_norm": 3.412027359008789,
+      "learning_rate": 3.041500852757249e-06,
+      "loss": 5.0255,
+      "step": 17020
+    },
+    {
+      "epoch": 0.9411276948590381,
+      "grad_norm": 4.9815497398376465,
+      "learning_rate": 3.0272882319499716e-06,
+      "loss": 4.6702,
+      "step": 17025
+    },
+    {
+      "epoch": 0.941404090657822,
+      "grad_norm": 4.5840044021606445,
+      "learning_rate": 3.0130756111426947e-06,
+      "loss": 4.7252,
+      "step": 17030
+    },
+    {
+      "epoch": 0.9416804864566058,
+      "grad_norm": 3.795989751815796,
+      "learning_rate": 2.998862990335418e-06,
+      "loss": 5.215,
+      "step": 17035
+    },
+    {
+      "epoch": 0.9419568822553898,
+      "grad_norm": 3.6358675956726074,
+      "learning_rate": 2.984650369528141e-06,
+      "loss": 4.6967,
+      "step": 17040
+    },
+    {
+      "epoch": 0.9422332780541736,
+      "grad_norm": 3.8705015182495117,
+      "learning_rate": 2.9704377487208645e-06,
+      "loss": 4.806,
+      "step": 17045
+    },
+    {
+      "epoch": 0.9425096738529575,
+      "grad_norm": 3.705080270767212,
+      "learning_rate": 2.9562251279135872e-06,
+      "loss": 4.7858,
+      "step": 17050
+    },
+    {
+      "epoch": 0.9427860696517413,
+      "grad_norm": 3.397587299346924,
+      "learning_rate": 2.942012507106311e-06,
+      "loss": 4.4487,
+      "step": 17055
+    },
+    {
+      "epoch": 0.9430624654505252,
+      "grad_norm": 4.504406929016113,
+      "learning_rate": 2.9277998862990335e-06,
+      "loss": 5.0349,
+      "step": 17060
+    },
+    {
+      "epoch": 0.943338861249309,
+      "grad_norm": 4.826185703277588,
+      "learning_rate": 2.913587265491757e-06,
+      "loss": 5.0815,
+      "step": 17065
+    },
+    {
+      "epoch": 0.9436152570480929,
+      "grad_norm": 3.7354743480682373,
+      "learning_rate": 2.8993746446844798e-06,
+      "loss": 4.5416,
+      "step": 17070
+    },
+    {
+      "epoch": 0.9438916528468767,
+      "grad_norm": 3.69808030128479,
+      "learning_rate": 2.885162023877203e-06,
+      "loss": 5.134,
+      "step": 17075
+    },
+    {
+      "epoch": 0.9441680486456606,
+      "grad_norm": 3.6593971252441406,
+      "learning_rate": 2.8709494030699265e-06,
+      "loss": 4.9477,
+      "step": 17080
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 3.8395798206329346,
+      "learning_rate": 2.856736782262649e-06,
+      "loss": 4.4886,
+      "step": 17085
+    },
+    {
+      "epoch": 0.9447208402432283,
+      "grad_norm": 4.302114009857178,
+      "learning_rate": 2.8425241614553727e-06,
+      "loss": 4.9254,
+      "step": 17090
+    },
+    {
+      "epoch": 0.9449972360420121,
+      "grad_norm": 3.074225664138794,
+      "learning_rate": 2.8283115406480954e-06,
+      "loss": 4.586,
+      "step": 17095
+    },
+    {
+      "epoch": 0.945273631840796,
+      "grad_norm": 4.352227687835693,
+      "learning_rate": 2.814098919840819e-06,
+      "loss": 4.9557,
+      "step": 17100
+    },
+    {
+      "epoch": 0.9455500276395798,
+      "grad_norm": 4.241294860839844,
+      "learning_rate": 2.7998862990335417e-06,
+      "loss": 5.0972,
+      "step": 17105
+    },
+    {
+      "epoch": 0.9458264234383638,
+      "grad_norm": 3.86039662361145,
+      "learning_rate": 2.7856736782262652e-06,
+      "loss": 4.9244,
+      "step": 17110
+    },
+    {
+      "epoch": 0.9461028192371476,
+      "grad_norm": 3.4481401443481445,
+      "learning_rate": 2.771461057418988e-06,
+      "loss": 4.6124,
+      "step": 17115
+    },
+    {
+      "epoch": 0.9463792150359315,
+      "grad_norm": 3.4227750301361084,
+      "learning_rate": 2.7572484366117115e-06,
+      "loss": 4.7664,
+      "step": 17120
+    },
+    {
+      "epoch": 0.9466556108347153,
+      "grad_norm": 4.316078186035156,
+      "learning_rate": 2.7430358158044346e-06,
+      "loss": 4.8977,
+      "step": 17125
+    },
+    {
+      "epoch": 0.9469320066334992,
+      "grad_norm": 4.136010646820068,
+      "learning_rate": 2.7288231949971574e-06,
+      "loss": 5.169,
+      "step": 17130
+    },
+    {
+      "epoch": 0.947208402432283,
+      "grad_norm": 3.5034444332122803,
+      "learning_rate": 2.714610574189881e-06,
+      "loss": 4.9702,
+      "step": 17135
+    },
+    {
+      "epoch": 0.9474847982310669,
+      "grad_norm": 3.8330748081207275,
+      "learning_rate": 2.7003979533826036e-06,
+      "loss": 5.0539,
+      "step": 17140
+    },
+    {
+      "epoch": 0.9477611940298507,
+      "grad_norm": 3.778444290161133,
+      "learning_rate": 2.686185332575327e-06,
+      "loss": 4.9856,
+      "step": 17145
+    },
+    {
+      "epoch": 0.9480375898286346,
+      "grad_norm": 4.075733184814453,
+      "learning_rate": 2.67197271176805e-06,
+      "loss": 4.4412,
+      "step": 17150
+    },
+    {
+      "epoch": 0.9483139856274184,
+      "grad_norm": 4.561306953430176,
+      "learning_rate": 2.6577600909607734e-06,
+      "loss": 4.5683,
+      "step": 17155
+    },
+    {
+      "epoch": 0.9485903814262023,
+      "grad_norm": 4.139663219451904,
+      "learning_rate": 2.6435474701534966e-06,
+      "loss": 5.1347,
+      "step": 17160
+    },
+    {
+      "epoch": 0.9488667772249861,
+      "grad_norm": 4.152304172515869,
+      "learning_rate": 2.6293348493462197e-06,
+      "loss": 4.851,
+      "step": 17165
+    },
+    {
+      "epoch": 0.94914317302377,
+      "grad_norm": 3.967447280883789,
+      "learning_rate": 2.615122228538943e-06,
+      "loss": 4.9876,
+      "step": 17170
+    },
+    {
+      "epoch": 0.9494195688225538,
+      "grad_norm": 3.313063144683838,
+      "learning_rate": 2.6009096077316655e-06,
+      "loss": 4.8804,
+      "step": 17175
+    },
+    {
+      "epoch": 0.9496959646213378,
+      "grad_norm": 3.359733819961548,
+      "learning_rate": 2.586696986924389e-06,
+      "loss": 4.673,
+      "step": 17180
+    },
+    {
+      "epoch": 0.9499723604201216,
+      "grad_norm": 3.648066520690918,
+      "learning_rate": 2.572484366117112e-06,
+      "loss": 4.9545,
+      "step": 17185
+    },
+    {
+      "epoch": 0.9502487562189055,
+      "grad_norm": 3.5169296264648438,
+      "learning_rate": 2.5582717453098354e-06,
+      "loss": 4.7484,
+      "step": 17190
+    },
+    {
+      "epoch": 0.9505251520176893,
+      "grad_norm": 3.443011522293091,
+      "learning_rate": 2.5440591245025585e-06,
+      "loss": 4.6217,
+      "step": 17195
+    },
+    {
+      "epoch": 0.9508015478164732,
+      "grad_norm": 3.1068191528320312,
+      "learning_rate": 2.5298465036952816e-06,
+      "loss": 4.6914,
+      "step": 17200
+    },
+    {
+      "epoch": 0.951077943615257,
+      "grad_norm": 3.6582255363464355,
+      "learning_rate": 2.5156338828880048e-06,
+      "loss": 4.7874,
+      "step": 17205
+    },
+    {
+      "epoch": 0.9513543394140409,
+      "grad_norm": 4.528609752655029,
+      "learning_rate": 2.501421262080728e-06,
+      "loss": 4.8206,
+      "step": 17210
+    },
+    {
+      "epoch": 0.9516307352128248,
+      "grad_norm": 4.048614025115967,
+      "learning_rate": 2.487208641273451e-06,
+      "loss": 5.1836,
+      "step": 17215
+    },
+    {
+      "epoch": 0.9519071310116086,
+      "grad_norm": 3.5472965240478516,
+      "learning_rate": 2.472996020466174e-06,
+      "loss": 4.8558,
+      "step": 17220
+    },
+    {
+      "epoch": 0.9521835268103925,
+      "grad_norm": 3.3577089309692383,
+      "learning_rate": 2.4587833996588973e-06,
+      "loss": 5.1221,
+      "step": 17225
+    },
+    {
+      "epoch": 0.9524599226091763,
+      "grad_norm": 3.5389177799224854,
+      "learning_rate": 2.44457077885162e-06,
+      "loss": 5.0266,
+      "step": 17230
+    },
+    {
+      "epoch": 0.9527363184079602,
+      "grad_norm": 4.607699871063232,
+      "learning_rate": 2.4303581580443436e-06,
+      "loss": 5.0358,
+      "step": 17235
+    },
+    {
+      "epoch": 0.953012714206744,
+      "grad_norm": 3.5849125385284424,
+      "learning_rate": 2.4161455372370667e-06,
+      "loss": 4.8641,
+      "step": 17240
+    },
+    {
+      "epoch": 0.953289110005528,
+      "grad_norm": 4.045363426208496,
+      "learning_rate": 2.40193291642979e-06,
+      "loss": 4.5956,
+      "step": 17245
+    },
+    {
+      "epoch": 0.9535655058043118,
+      "grad_norm": 3.181807041168213,
+      "learning_rate": 2.387720295622513e-06,
+      "loss": 4.868,
+      "step": 17250
+    },
+    {
+      "epoch": 0.9538419016030957,
+      "grad_norm": 3.7292158603668213,
+      "learning_rate": 2.373507674815236e-06,
+      "loss": 4.9381,
+      "step": 17255
+    },
+    {
+      "epoch": 0.9541182974018795,
+      "grad_norm": 3.4840378761291504,
+      "learning_rate": 2.3592950540079592e-06,
+      "loss": 4.6714,
+      "step": 17260
+    },
+    {
+      "epoch": 0.9543946932006634,
+      "grad_norm": 4.143215179443359,
+      "learning_rate": 2.3450824332006823e-06,
+      "loss": 4.5491,
+      "step": 17265
+    },
+    {
+      "epoch": 0.9546710889994472,
+      "grad_norm": 4.7000555992126465,
+      "learning_rate": 2.3308698123934055e-06,
+      "loss": 4.7206,
+      "step": 17270
+    },
+    {
+      "epoch": 0.9549474847982311,
+      "grad_norm": 3.5755834579467773,
+      "learning_rate": 2.3166571915861286e-06,
+      "loss": 4.6387,
+      "step": 17275
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 3.0749599933624268,
+      "learning_rate": 2.3024445707788517e-06,
+      "loss": 4.6379,
+      "step": 17280
+    },
+    {
+      "epoch": 0.9555002763957988,
+      "grad_norm": 4.241413593292236,
+      "learning_rate": 2.288231949971575e-06,
+      "loss": 4.7813,
+      "step": 17285
+    },
+    {
+      "epoch": 0.9557766721945826,
+      "grad_norm": 4.150501251220703,
+      "learning_rate": 2.274019329164298e-06,
+      "loss": 5.048,
+      "step": 17290
+    },
+    {
+      "epoch": 0.9560530679933665,
+      "grad_norm": 3.893385887145996,
+      "learning_rate": 2.259806708357021e-06,
+      "loss": 5.1972,
+      "step": 17295
+    },
+    {
+      "epoch": 0.9563294637921503,
+      "grad_norm": 3.0259926319122314,
+      "learning_rate": 2.2455940875497443e-06,
+      "loss": 4.9307,
+      "step": 17300
+    },
+    {
+      "epoch": 0.9566058595909342,
+      "grad_norm": 4.032433986663818,
+      "learning_rate": 2.2313814667424674e-06,
+      "loss": 4.8233,
+      "step": 17305
+    },
+    {
+      "epoch": 0.956882255389718,
+      "grad_norm": 3.581953525543213,
+      "learning_rate": 2.2171688459351905e-06,
+      "loss": 4.994,
+      "step": 17310
+    },
+    {
+      "epoch": 0.957158651188502,
+      "grad_norm": 3.022449493408203,
+      "learning_rate": 2.2029562251279137e-06,
+      "loss": 4.8858,
+      "step": 17315
+    },
+    {
+      "epoch": 0.9574350469872858,
+      "grad_norm": 3.664137125015259,
+      "learning_rate": 2.188743604320637e-06,
+      "loss": 4.8783,
+      "step": 17320
+    },
+    {
+      "epoch": 0.9577114427860697,
+      "grad_norm": 3.227354049682617,
+      "learning_rate": 2.17453098351336e-06,
+      "loss": 4.923,
+      "step": 17325
+    },
+    {
+      "epoch": 0.9579878385848535,
+      "grad_norm": 3.1704261302948,
+      "learning_rate": 2.160318362706083e-06,
+      "loss": 4.8075,
+      "step": 17330
+    },
+    {
+      "epoch": 0.9582642343836374,
+      "grad_norm": 3.537228584289551,
+      "learning_rate": 2.146105741898806e-06,
+      "loss": 4.7424,
+      "step": 17335
+    },
+    {
+      "epoch": 0.9585406301824212,
+      "grad_norm": 4.334897518157959,
+      "learning_rate": 2.1318931210915293e-06,
+      "loss": 4.9451,
+      "step": 17340
+    },
+    {
+      "epoch": 0.9588170259812051,
+      "grad_norm": 3.4837393760681152,
+      "learning_rate": 2.1176805002842525e-06,
+      "loss": 4.8,
+      "step": 17345
+    },
+    {
+      "epoch": 0.9590934217799889,
+      "grad_norm": 3.615074396133423,
+      "learning_rate": 2.1034678794769756e-06,
+      "loss": 4.9119,
+      "step": 17350
+    },
+    {
+      "epoch": 0.9593698175787728,
+      "grad_norm": 3.716219425201416,
+      "learning_rate": 2.0892552586696987e-06,
+      "loss": 5.1262,
+      "step": 17355
+    },
+    {
+      "epoch": 0.9596462133775566,
+      "grad_norm": 3.8585875034332275,
+      "learning_rate": 2.075042637862422e-06,
+      "loss": 4.9943,
+      "step": 17360
+    },
+    {
+      "epoch": 0.9599226091763405,
+      "grad_norm": 3.1968586444854736,
+      "learning_rate": 2.060830017055145e-06,
+      "loss": 4.8486,
+      "step": 17365
+    },
+    {
+      "epoch": 0.9601990049751243,
+      "grad_norm": 5.157547950744629,
+      "learning_rate": 2.046617396247868e-06,
+      "loss": 5.1108,
+      "step": 17370
+    },
+    {
+      "epoch": 0.9604754007739082,
+      "grad_norm": 4.623779296875,
+      "learning_rate": 2.0324047754405917e-06,
+      "loss": 4.89,
+      "step": 17375
+    },
+    {
+      "epoch": 0.960751796572692,
+      "grad_norm": 3.5606627464294434,
+      "learning_rate": 2.0181921546333144e-06,
+      "loss": 4.8832,
+      "step": 17380
+    },
+    {
+      "epoch": 0.961028192371476,
+      "grad_norm": 3.792771577835083,
+      "learning_rate": 2.0039795338260375e-06,
+      "loss": 4.8013,
+      "step": 17385
+    },
+    {
+      "epoch": 0.9613045881702598,
+      "grad_norm": 3.4608917236328125,
+      "learning_rate": 1.9897669130187606e-06,
+      "loss": 4.888,
+      "step": 17390
+    },
+    {
+      "epoch": 0.9615809839690437,
+      "grad_norm": 4.744699001312256,
+      "learning_rate": 1.9755542922114838e-06,
+      "loss": 5.3272,
+      "step": 17395
+    },
+    {
+      "epoch": 0.9618573797678275,
+      "grad_norm": 3.0279312133789062,
+      "learning_rate": 1.961341671404207e-06,
+      "loss": 4.6455,
+      "step": 17400
+    },
+    {
+      "epoch": 0.9621337755666114,
+      "grad_norm": 3.420703649520874,
+      "learning_rate": 1.94712905059693e-06,
+      "loss": 4.8824,
+      "step": 17405
+    },
+    {
+      "epoch": 0.9624101713653952,
+      "grad_norm": 3.330737590789795,
+      "learning_rate": 1.9329164297896536e-06,
+      "loss": 4.7754,
+      "step": 17410
+    },
+    {
+      "epoch": 0.9626865671641791,
+      "grad_norm": 3.5434346199035645,
+      "learning_rate": 1.9187038089823763e-06,
+      "loss": 4.6785,
+      "step": 17415
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 4.19118595123291,
+      "learning_rate": 1.9044911881750997e-06,
+      "loss": 4.8247,
+      "step": 17420
+    },
+    {
+      "epoch": 0.9632393587617468,
+      "grad_norm": 4.7213454246521,
+      "learning_rate": 1.8902785673678228e-06,
+      "loss": 4.8312,
+      "step": 17425
+    },
+    {
+      "epoch": 0.9635157545605307,
+      "grad_norm": 3.6587038040161133,
+      "learning_rate": 1.8760659465605457e-06,
+      "loss": 5.1918,
+      "step": 17430
+    },
+    {
+      "epoch": 0.9637921503593145,
+      "grad_norm": 3.242537021636963,
+      "learning_rate": 1.861853325753269e-06,
+      "loss": 4.9542,
+      "step": 17435
+    },
+    {
+      "epoch": 0.9640685461580984,
+      "grad_norm": 3.7462520599365234,
+      "learning_rate": 1.847640704945992e-06,
+      "loss": 4.9511,
+      "step": 17440
+    },
+    {
+      "epoch": 0.9643449419568823,
+      "grad_norm": 3.6050407886505127,
+      "learning_rate": 1.8334280841387153e-06,
+      "loss": 4.8826,
+      "step": 17445
+    },
+    {
+      "epoch": 0.9646213377556662,
+      "grad_norm": 3.979299783706665,
+      "learning_rate": 1.8192154633314382e-06,
+      "loss": 4.879,
+      "step": 17450
+    },
+    {
+      "epoch": 0.96489773355445,
+      "grad_norm": 4.422558784484863,
+      "learning_rate": 1.8050028425241616e-06,
+      "loss": 5.2462,
+      "step": 17455
+    },
+    {
+      "epoch": 0.9651741293532339,
+      "grad_norm": 3.984792947769165,
+      "learning_rate": 1.7907902217168847e-06,
+      "loss": 4.9413,
+      "step": 17460
+    },
+    {
+      "epoch": 0.9654505251520177,
+      "grad_norm": 3.3358352184295654,
+      "learning_rate": 1.7765776009096078e-06,
+      "loss": 4.8726,
+      "step": 17465
+    },
+    {
+      "epoch": 0.9657269209508016,
+      "grad_norm": 3.0157341957092285,
+      "learning_rate": 1.762364980102331e-06,
+      "loss": 4.9832,
+      "step": 17470
+    },
+    {
+      "epoch": 0.9660033167495854,
+      "grad_norm": 3.5615592002868652,
+      "learning_rate": 1.7481523592950543e-06,
+      "loss": 4.6077,
+      "step": 17475
+    },
+    {
+      "epoch": 0.9662797125483693,
+      "grad_norm": 3.1383845806121826,
+      "learning_rate": 1.7339397384877772e-06,
+      "loss": 4.7842,
+      "step": 17480
+    },
+    {
+      "epoch": 0.9665561083471531,
+      "grad_norm": 4.033356189727783,
+      "learning_rate": 1.7197271176805002e-06,
+      "loss": 4.9063,
+      "step": 17485
+    },
+    {
+      "epoch": 0.966832504145937,
+      "grad_norm": 3.651676654815674,
+      "learning_rate": 1.7055144968732235e-06,
+      "loss": 4.7036,
+      "step": 17490
+    },
+    {
+      "epoch": 0.9671088999447208,
+      "grad_norm": 3.519984245300293,
+      "learning_rate": 1.6913018760659464e-06,
+      "loss": 4.9829,
+      "step": 17495
+    },
+    {
+      "epoch": 0.9673852957435047,
+      "grad_norm": 2.9878344535827637,
+      "learning_rate": 1.6770892552586698e-06,
+      "loss": 4.6443,
+      "step": 17500
+    },
+    {
+      "epoch": 0.9676616915422885,
+      "grad_norm": 3.7152388095855713,
+      "learning_rate": 1.662876634451393e-06,
+      "loss": 4.8157,
+      "step": 17505
+    },
+    {
+      "epoch": 0.9679380873410725,
+      "grad_norm": 3.1234991550445557,
+      "learning_rate": 1.6486640136441162e-06,
+      "loss": 5.0816,
+      "step": 17510
+    },
+    {
+      "epoch": 0.9682144831398563,
+      "grad_norm": 3.695985794067383,
+      "learning_rate": 1.6344513928368392e-06,
+      "loss": 5.1458,
+      "step": 17515
+    },
+    {
+      "epoch": 0.9684908789386402,
+      "grad_norm": 3.487590789794922,
+      "learning_rate": 1.6202387720295625e-06,
+      "loss": 4.5166,
+      "step": 17520
+    },
+    {
+      "epoch": 0.968767274737424,
+      "grad_norm": 3.8516764640808105,
+      "learning_rate": 1.6060261512222854e-06,
+      "loss": 5.0152,
+      "step": 17525
+    },
+    {
+      "epoch": 0.9690436705362079,
+      "grad_norm": 3.8771095275878906,
+      "learning_rate": 1.5918135304150088e-06,
+      "loss": 4.7823,
+      "step": 17530
+    },
+    {
+      "epoch": 0.9693200663349917,
+      "grad_norm": 4.3737287521362305,
+      "learning_rate": 1.5776009096077317e-06,
+      "loss": 4.905,
+      "step": 17535
+    },
+    {
+      "epoch": 0.9695964621337756,
+      "grad_norm": 4.190290451049805,
+      "learning_rate": 1.5633882888004548e-06,
+      "loss": 4.7303,
+      "step": 17540
+    },
+    {
+      "epoch": 0.9698728579325594,
+      "grad_norm": 4.203909397125244,
+      "learning_rate": 1.5491756679931782e-06,
+      "loss": 4.6031,
+      "step": 17545
+    },
+    {
+      "epoch": 0.9701492537313433,
+      "grad_norm": 3.6529483795166016,
+      "learning_rate": 1.5349630471859013e-06,
+      "loss": 4.9744,
+      "step": 17550
+    },
+    {
+      "epoch": 0.9704256495301271,
+      "grad_norm": 3.739197015762329,
+      "learning_rate": 1.5207504263786244e-06,
+      "loss": 4.8951,
+      "step": 17555
+    },
+    {
+      "epoch": 0.970702045328911,
+      "grad_norm": 4.100447177886963,
+      "learning_rate": 1.5065378055713474e-06,
+      "loss": 5.0774,
+      "step": 17560
+    },
+    {
+      "epoch": 0.9709784411276948,
+      "grad_norm": 4.293310165405273,
+      "learning_rate": 1.4923251847640705e-06,
+      "loss": 4.7931,
+      "step": 17565
+    },
+    {
+      "epoch": 0.9712548369264787,
+      "grad_norm": 3.8114206790924072,
+      "learning_rate": 1.4781125639567936e-06,
+      "loss": 4.6042,
+      "step": 17570
+    },
+    {
+      "epoch": 0.9715312327252625,
+      "grad_norm": 3.7172446250915527,
+      "learning_rate": 1.4638999431495168e-06,
+      "loss": 4.7856,
+      "step": 17575
+    },
+    {
+      "epoch": 0.9718076285240465,
+      "grad_norm": 3.316227912902832,
+      "learning_rate": 1.4496873223422399e-06,
+      "loss": 4.8669,
+      "step": 17580
+    },
+    {
+      "epoch": 0.9720840243228303,
+      "grad_norm": 4.058900833129883,
+      "learning_rate": 1.4354747015349632e-06,
+      "loss": 4.5917,
+      "step": 17585
+    },
+    {
+      "epoch": 0.9723604201216142,
+      "grad_norm": 3.633906841278076,
+      "learning_rate": 1.4212620807276864e-06,
+      "loss": 4.7329,
+      "step": 17590
+    },
+    {
+      "epoch": 0.972636815920398,
+      "grad_norm": 4.460212230682373,
+      "learning_rate": 1.4070494599204095e-06,
+      "loss": 4.7991,
+      "step": 17595
+    },
+    {
+      "epoch": 0.9729132117191819,
+      "grad_norm": 3.0964694023132324,
+      "learning_rate": 1.3928368391131326e-06,
+      "loss": 4.6162,
+      "step": 17600
+    },
+    {
+      "epoch": 0.9731896075179657,
+      "grad_norm": 3.333630323410034,
+      "learning_rate": 1.3786242183058558e-06,
+      "loss": 4.5838,
+      "step": 17605
+    },
+    {
+      "epoch": 0.9734660033167496,
+      "grad_norm": 3.469555139541626,
+      "learning_rate": 1.3644115974985787e-06,
+      "loss": 5.2188,
+      "step": 17610
+    },
+    {
+      "epoch": 0.9737423991155334,
+      "grad_norm": 3.8169009685516357,
+      "learning_rate": 1.3501989766913018e-06,
+      "loss": 5.018,
+      "step": 17615
+    },
+    {
+      "epoch": 0.9740187949143173,
+      "grad_norm": 4.118598461151123,
+      "learning_rate": 1.335986355884025e-06,
+      "loss": 5.0628,
+      "step": 17620
+    },
+    {
+      "epoch": 0.9742951907131011,
+      "grad_norm": 3.506039619445801,
+      "learning_rate": 1.3217737350767483e-06,
+      "loss": 4.6609,
+      "step": 17625
+    },
+    {
+      "epoch": 0.974571586511885,
+      "grad_norm": 3.59202241897583,
+      "learning_rate": 1.3075611142694714e-06,
+      "loss": 4.9948,
+      "step": 17630
+    },
+    {
+      "epoch": 0.9748479823106688,
+      "grad_norm": 3.3904547691345215,
+      "learning_rate": 1.2933484934621945e-06,
+      "loss": 5.0171,
+      "step": 17635
+    },
+    {
+      "epoch": 0.9751243781094527,
+      "grad_norm": 3.5725204944610596,
+      "learning_rate": 1.2791358726549177e-06,
+      "loss": 4.719,
+      "step": 17640
+    },
+    {
+      "epoch": 0.9754007739082365,
+      "grad_norm": 4.067680835723877,
+      "learning_rate": 1.2649232518476408e-06,
+      "loss": 5.1472,
+      "step": 17645
+    },
+    {
+      "epoch": 0.9756771697070205,
+      "grad_norm": 3.5419161319732666,
+      "learning_rate": 1.250710631040364e-06,
+      "loss": 4.9008,
+      "step": 17650
+    },
+    {
+      "epoch": 0.9759535655058044,
+      "grad_norm": 4.060971260070801,
+      "learning_rate": 1.236498010233087e-06,
+      "loss": 4.8635,
+      "step": 17655
+    },
+    {
+      "epoch": 0.9762299613045882,
+      "grad_norm": 3.4809699058532715,
+      "learning_rate": 1.22228538942581e-06,
+      "loss": 4.7176,
+      "step": 17660
+    },
+    {
+      "epoch": 0.9765063571033721,
+      "grad_norm": 3.121370792388916,
+      "learning_rate": 1.2080727686185333e-06,
+      "loss": 4.7872,
+      "step": 17665
+    },
+    {
+      "epoch": 0.9767827529021559,
+      "grad_norm": 3.8386788368225098,
+      "learning_rate": 1.1938601478112565e-06,
+      "loss": 4.6917,
+      "step": 17670
+    },
+    {
+      "epoch": 0.9770591487009398,
+      "grad_norm": 3.4671096801757812,
+      "learning_rate": 1.1796475270039796e-06,
+      "loss": 5.0222,
+      "step": 17675
+    },
+    {
+      "epoch": 0.9773355444997236,
+      "grad_norm": 4.064565658569336,
+      "learning_rate": 1.1654349061967027e-06,
+      "loss": 4.9298,
+      "step": 17680
+    },
+    {
+      "epoch": 0.9776119402985075,
+      "grad_norm": 3.155139923095703,
+      "learning_rate": 1.1512222853894259e-06,
+      "loss": 4.7341,
+      "step": 17685
+    },
+    {
+      "epoch": 0.9778883360972913,
+      "grad_norm": 3.358921527862549,
+      "learning_rate": 1.137009664582149e-06,
+      "loss": 4.8701,
+      "step": 17690
+    },
+    {
+      "epoch": 0.9781647318960752,
+      "grad_norm": 3.9727041721343994,
+      "learning_rate": 1.1227970437748721e-06,
+      "loss": 4.9399,
+      "step": 17695
+    },
+    {
+      "epoch": 0.978441127694859,
+      "grad_norm": 3.551064968109131,
+      "learning_rate": 1.1085844229675953e-06,
+      "loss": 4.4855,
+      "step": 17700
+    },
+    {
+      "epoch": 0.9787175234936429,
+      "grad_norm": 3.285154104232788,
+      "learning_rate": 1.0943718021603184e-06,
+      "loss": 4.6904,
+      "step": 17705
+    },
+    {
+      "epoch": 0.9789939192924267,
+      "grad_norm": 3.77357816696167,
+      "learning_rate": 1.0801591813530415e-06,
+      "loss": 4.7048,
+      "step": 17710
+    },
+    {
+      "epoch": 0.9792703150912107,
+      "grad_norm": 3.7652862071990967,
+      "learning_rate": 1.0659465605457647e-06,
+      "loss": 4.8469,
+      "step": 17715
+    },
+    {
+      "epoch": 0.9795467108899945,
+      "grad_norm": 4.785614013671875,
+      "learning_rate": 1.0517339397384878e-06,
+      "loss": 4.6331,
+      "step": 17720
+    },
+    {
+      "epoch": 0.9798231066887784,
+      "grad_norm": 3.4421780109405518,
+      "learning_rate": 1.037521318931211e-06,
+      "loss": 4.8027,
+      "step": 17725
+    },
+    {
+      "epoch": 0.9800995024875622,
+      "grad_norm": 3.414224863052368,
+      "learning_rate": 1.023308698123934e-06,
+      "loss": 4.6746,
+      "step": 17730
+    },
+    {
+      "epoch": 0.9803758982863461,
+      "grad_norm": 3.6963014602661133,
+      "learning_rate": 1.0090960773166572e-06,
+      "loss": 4.9759,
+      "step": 17735
+    },
+    {
+      "epoch": 0.9806522940851299,
+      "grad_norm": 4.396233558654785,
+      "learning_rate": 9.948834565093803e-07,
+      "loss": 4.9611,
+      "step": 17740
+    },
+    {
+      "epoch": 0.9809286898839138,
+      "grad_norm": 3.8200385570526123,
+      "learning_rate": 9.806708357021035e-07,
+      "loss": 4.7393,
+      "step": 17745
+    },
+    {
+      "epoch": 0.9812050856826976,
+      "grad_norm": 3.469111680984497,
+      "learning_rate": 9.664582148948268e-07,
+      "loss": 4.9948,
+      "step": 17750
+    },
+    {
+      "epoch": 0.9814814814814815,
+      "grad_norm": 3.5124690532684326,
+      "learning_rate": 9.522455940875498e-07,
+      "loss": 4.9789,
+      "step": 17755
+    },
+    {
+      "epoch": 0.9817578772802653,
+      "grad_norm": 3.9192700386047363,
+      "learning_rate": 9.380329732802729e-07,
+      "loss": 4.7527,
+      "step": 17760
+    },
+    {
+      "epoch": 0.9820342730790492,
+      "grad_norm": 3.603703260421753,
+      "learning_rate": 9.23820352472996e-07,
+      "loss": 4.8427,
+      "step": 17765
+    },
+    {
+      "epoch": 0.982310668877833,
+      "grad_norm": 3.82257342338562,
+      "learning_rate": 9.096077316657191e-07,
+      "loss": 4.6548,
+      "step": 17770
+    },
+    {
+      "epoch": 0.9825870646766169,
+      "grad_norm": 3.6828839778900146,
+      "learning_rate": 8.953951108584424e-07,
+      "loss": 4.7383,
+      "step": 17775
+    },
+    {
+      "epoch": 0.9828634604754007,
+      "grad_norm": 4.09072732925415,
+      "learning_rate": 8.811824900511655e-07,
+      "loss": 4.6382,
+      "step": 17780
+    },
+    {
+      "epoch": 0.9831398562741847,
+      "grad_norm": 3.557931900024414,
+      "learning_rate": 8.669698692438886e-07,
+      "loss": 4.7966,
+      "step": 17785
+    },
+    {
+      "epoch": 0.9834162520729685,
+      "grad_norm": 3.4907984733581543,
+      "learning_rate": 8.527572484366118e-07,
+      "loss": 5.0257,
+      "step": 17790
+    },
+    {
+      "epoch": 0.9836926478717524,
+      "grad_norm": 4.357024192810059,
+      "learning_rate": 8.385446276293349e-07,
+      "loss": 5.2176,
+      "step": 17795
+    },
+    {
+      "epoch": 0.9839690436705362,
+      "grad_norm": 4.168322563171387,
+      "learning_rate": 8.243320068220581e-07,
+      "loss": 4.9219,
+      "step": 17800
+    },
+    {
+      "epoch": 0.9842454394693201,
+      "grad_norm": 4.186100006103516,
+      "learning_rate": 8.101193860147813e-07,
+      "loss": 5.1336,
+      "step": 17805
+    },
+    {
+      "epoch": 0.9845218352681039,
+      "grad_norm": 4.123225212097168,
+      "learning_rate": 7.959067652075044e-07,
+      "loss": 4.7761,
+      "step": 17810
+    },
+    {
+      "epoch": 0.9847982310668878,
+      "grad_norm": 4.013276100158691,
+      "learning_rate": 7.816941444002274e-07,
+      "loss": 4.8168,
+      "step": 17815
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 3.063643217086792,
+      "learning_rate": 7.674815235929507e-07,
+      "loss": 4.8048,
+      "step": 17820
+    },
+    {
+      "epoch": 0.9853510226644555,
+      "grad_norm": 4.011346817016602,
+      "learning_rate": 7.532689027856737e-07,
+      "loss": 4.3031,
+      "step": 17825
+    },
+    {
+      "epoch": 0.9856274184632393,
+      "grad_norm": 4.323795318603516,
+      "learning_rate": 7.390562819783968e-07,
+      "loss": 4.7946,
+      "step": 17830
+    },
+    {
+      "epoch": 0.9859038142620232,
+      "grad_norm": 4.777589321136475,
+      "learning_rate": 7.248436611711199e-07,
+      "loss": 4.6589,
+      "step": 17835
+    },
+    {
+      "epoch": 0.986180210060807,
+      "grad_norm": 4.321505546569824,
+      "learning_rate": 7.106310403638432e-07,
+      "loss": 4.9267,
+      "step": 17840
+    },
+    {
+      "epoch": 0.9864566058595909,
+      "grad_norm": 3.9251480102539062,
+      "learning_rate": 6.964184195565663e-07,
+      "loss": 4.6038,
+      "step": 17845
+    },
+    {
+      "epoch": 0.9867330016583747,
+      "grad_norm": 3.0840792655944824,
+      "learning_rate": 6.822057987492893e-07,
+      "loss": 4.5904,
+      "step": 17850
+    },
+    {
+      "epoch": 0.9870093974571587,
+      "grad_norm": 4.160499572753906,
+      "learning_rate": 6.679931779420125e-07,
+      "loss": 4.8037,
+      "step": 17855
+    },
+    {
+      "epoch": 0.9872857932559425,
+      "grad_norm": 4.162237167358398,
+      "learning_rate": 6.537805571347357e-07,
+      "loss": 5.0783,
+      "step": 17860
+    },
+    {
+      "epoch": 0.9875621890547264,
+      "grad_norm": 3.218111515045166,
+      "learning_rate": 6.395679363274588e-07,
+      "loss": 4.8852,
+      "step": 17865
+    },
+    {
+      "epoch": 0.9878385848535103,
+      "grad_norm": 3.7877068519592285,
+      "learning_rate": 6.25355315520182e-07,
+      "loss": 5.0094,
+      "step": 17870
+    },
+    {
+      "epoch": 0.9881149806522941,
+      "grad_norm": 3.7480013370513916,
+      "learning_rate": 6.11142694712905e-07,
+      "loss": 4.9167,
+      "step": 17875
+    },
+    {
+      "epoch": 0.988391376451078,
+      "grad_norm": 3.9518656730651855,
+      "learning_rate": 5.969300739056282e-07,
+      "loss": 4.9065,
+      "step": 17880
+    },
+    {
+      "epoch": 0.9886677722498618,
+      "grad_norm": 4.303936958312988,
+      "learning_rate": 5.827174530983514e-07,
+      "loss": 5.0365,
+      "step": 17885
+    },
+    {
+      "epoch": 0.9889441680486457,
+      "grad_norm": 3.944624423980713,
+      "learning_rate": 5.685048322910745e-07,
+      "loss": 4.9906,
+      "step": 17890
+    },
+    {
+      "epoch": 0.9892205638474295,
+      "grad_norm": 3.4610772132873535,
+      "learning_rate": 5.542922114837976e-07,
+      "loss": 4.6898,
+      "step": 17895
+    },
+    {
+      "epoch": 0.9894969596462134,
+      "grad_norm": 3.5662105083465576,
+      "learning_rate": 5.400795906765208e-07,
+      "loss": 5.1878,
+      "step": 17900
+    },
+    {
+      "epoch": 0.9897733554449972,
+      "grad_norm": 3.8425076007843018,
+      "learning_rate": 5.258669698692439e-07,
+      "loss": 4.482,
+      "step": 17905
+    },
+    {
+      "epoch": 0.9900497512437811,
+      "grad_norm": 3.6111109256744385,
+      "learning_rate": 5.11654349061967e-07,
+      "loss": 4.8749,
+      "step": 17910
+    },
+    {
+      "epoch": 0.9903261470425649,
+      "grad_norm": 3.664844274520874,
+      "learning_rate": 4.974417282546902e-07,
+      "loss": 5.1134,
+      "step": 17915
+    },
+    {
+      "epoch": 0.9906025428413489,
+      "grad_norm": 3.341689109802246,
+      "learning_rate": 4.832291074474134e-07,
+      "loss": 5.1036,
+      "step": 17920
+    },
+    {
+      "epoch": 0.9908789386401327,
+      "grad_norm": 3.0677828788757324,
+      "learning_rate": 4.6901648664013643e-07,
+      "loss": 5.0625,
+      "step": 17925
+    },
+    {
+      "epoch": 0.9911553344389166,
+      "grad_norm": 4.164408206939697,
+      "learning_rate": 4.5480386583285956e-07,
+      "loss": 4.8023,
+      "step": 17930
+    },
+    {
+      "epoch": 0.9914317302377004,
+      "grad_norm": 3.8659586906433105,
+      "learning_rate": 4.4059124502558274e-07,
+      "loss": 4.9043,
+      "step": 17935
+    },
+    {
+      "epoch": 0.9917081260364843,
+      "grad_norm": 4.336862087249756,
+      "learning_rate": 4.263786242183059e-07,
+      "loss": 5.1321,
+      "step": 17940
+    },
+    {
+      "epoch": 0.9919845218352681,
+      "grad_norm": 3.780893325805664,
+      "learning_rate": 4.1216600341102906e-07,
+      "loss": 5.1098,
+      "step": 17945
+    },
+    {
+      "epoch": 0.992260917634052,
+      "grad_norm": 3.7630019187927246,
+      "learning_rate": 3.979533826037522e-07,
+      "loss": 4.7298,
+      "step": 17950
+    },
+    {
+      "epoch": 0.9925373134328358,
+      "grad_norm": 4.459870338439941,
+      "learning_rate": 3.837407617964753e-07,
+      "loss": 5.0754,
+      "step": 17955
+    },
+    {
+      "epoch": 0.9928137092316197,
+      "grad_norm": 3.592924118041992,
+      "learning_rate": 3.695281409891984e-07,
+      "loss": 4.9391,
+      "step": 17960
+    },
+    {
+      "epoch": 0.9930901050304035,
+      "grad_norm": 3.273549795150757,
+      "learning_rate": 3.553155201819216e-07,
+      "loss": 4.6995,
+      "step": 17965
+    },
+    {
+      "epoch": 0.9933665008291874,
+      "grad_norm": 3.976041555404663,
+      "learning_rate": 3.4110289937464467e-07,
+      "loss": 5.0032,
+      "step": 17970
+    },
+    {
+      "epoch": 0.9936428966279712,
+      "grad_norm": 3.7148942947387695,
+      "learning_rate": 3.2689027856736785e-07,
+      "loss": 4.9053,
+      "step": 17975
+    },
+    {
+      "epoch": 0.9939192924267551,
+      "grad_norm": 4.372527122497559,
+      "learning_rate": 3.12677657760091e-07,
+      "loss": 4.6127,
+      "step": 17980
+    },
+    {
+      "epoch": 0.9941956882255389,
+      "grad_norm": 4.019134521484375,
+      "learning_rate": 2.984650369528141e-07,
+      "loss": 4.9816,
+      "step": 17985
+    },
+    {
+      "epoch": 0.9944720840243229,
+      "grad_norm": 4.684123992919922,
+      "learning_rate": 2.8425241614553725e-07,
+      "loss": 4.9317,
+      "step": 17990
+    },
+    {
+      "epoch": 0.9947484798231067,
+      "grad_norm": 4.50031852722168,
+      "learning_rate": 2.700397953382604e-07,
+      "loss": 4.8163,
+      "step": 17995
+    },
+    {
+      "epoch": 0.9950248756218906,
+      "grad_norm": 3.7687060832977295,
+      "learning_rate": 2.558271745309835e-07,
+      "loss": 4.8997,
+      "step": 18000
+    },
+    {
+      "epoch": 0.9953012714206744,
+      "grad_norm": 5.449435234069824,
+      "learning_rate": 2.416145537237067e-07,
+      "loss": 4.8225,
+      "step": 18005
+    },
+    {
+      "epoch": 0.9955776672194583,
+      "grad_norm": 3.540816068649292,
+      "learning_rate": 2.2740193291642978e-07,
+      "loss": 4.8916,
+      "step": 18010
+    },
+    {
+      "epoch": 0.9958540630182421,
+      "grad_norm": 3.5293331146240234,
+      "learning_rate": 2.1318931210915294e-07,
+      "loss": 4.8202,
+      "step": 18015
+    },
+    {
+      "epoch": 0.996130458817026,
+      "grad_norm": 4.026026248931885,
+      "learning_rate": 1.989766913018761e-07,
+      "loss": 5.0024,
+      "step": 18020
+    },
+    {
+      "epoch": 0.9964068546158098,
+      "grad_norm": 4.405704021453857,
+      "learning_rate": 1.847640704945992e-07,
+      "loss": 4.9342,
+      "step": 18025
+    },
+    {
+      "epoch": 0.9966832504145937,
+      "grad_norm": 3.3129539489746094,
+      "learning_rate": 1.7055144968732233e-07,
+      "loss": 4.8739,
+      "step": 18030
+    },
+    {
+      "epoch": 0.9969596462133775,
+      "grad_norm": 4.042524814605713,
+      "learning_rate": 1.563388288800455e-07,
+      "loss": 4.9343,
+      "step": 18035
+    },
+    {
+      "epoch": 0.9972360420121614,
+      "grad_norm": 3.105717658996582,
+      "learning_rate": 1.4212620807276863e-07,
+      "loss": 4.463,
+      "step": 18040
+    },
+    {
+      "epoch": 0.9975124378109452,
+      "grad_norm": 3.765864133834839,
+      "learning_rate": 1.2791358726549176e-07,
+      "loss": 4.9118,
+      "step": 18045
+    },
+    {
+      "epoch": 0.9977888336097291,
+      "grad_norm": 4.456231594085693,
+      "learning_rate": 1.1370096645821489e-07,
+      "loss": 4.8341,
+      "step": 18050
+    },
+    {
+      "epoch": 0.9980652294085129,
+      "grad_norm": 3.701779365539551,
+      "learning_rate": 9.948834565093805e-08,
+      "loss": 4.9091,
+      "step": 18055
+    },
+    {
+      "epoch": 0.9983416252072969,
+      "grad_norm": 3.981372833251953,
+      "learning_rate": 8.527572484366117e-08,
+      "loss": 4.9035,
+      "step": 18060
+    },
+    {
+      "epoch": 0.9986180210060807,
+      "grad_norm": 4.189070224761963,
+      "learning_rate": 7.106310403638431e-08,
+      "loss": 4.8013,
+      "step": 18065
+    },
+    {
+      "epoch": 0.9988944168048646,
+      "grad_norm": 3.8687007427215576,
+      "learning_rate": 5.6850483229107445e-08,
+      "loss": 4.7042,
+      "step": 18070
+    },
+    {
+      "epoch": 0.9991708126036484,
+      "grad_norm": 4.483328342437744,
+      "learning_rate": 4.2637862421830584e-08,
+      "loss": 4.6875,
+      "step": 18075
+    },
+    {
+      "epoch": 0.9994472084024323,
+      "grad_norm": 4.180428981781006,
+      "learning_rate": 2.8425241614553722e-08,
+      "loss": 4.816,
+      "step": 18080
+    },
+    {
+      "epoch": 0.9997236042012161,
+      "grad_norm": 3.824141025543213,
+      "learning_rate": 1.4212620807276861e-08,
+      "loss": 4.8663,
+      "step": 18085
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4549672603607178,
+      "learning_rate": 0.0,
+      "loss": 4.3516,
+      "step": 18090
     }
   ],
   "logging_steps": 5,
@@ -16821,12 +25347,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.1151619342811136e+16,
+  "total_flos": 1.6815556451321856e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null