atsuki-yamaguchi's picture
Upload folder using huggingface_hub
a652a12 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.698944636348022,
"eval_steps": 500,
"global_step": 27468,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007888191250469157,
"grad_norm": 19.37966537475586,
"learning_rate": 1.0157273918741808e-06,
"loss": 8.7613,
"step": 31
},
{
"epoch": 0.0015776382500938314,
"grad_norm": 13.922501564025879,
"learning_rate": 2.0314547837483616e-06,
"loss": 7.7444,
"step": 62
},
{
"epoch": 0.002366457375140747,
"grad_norm": 11.746194839477539,
"learning_rate": 3.0471821756225426e-06,
"loss": 6.5314,
"step": 93
},
{
"epoch": 0.003155276500187663,
"grad_norm": 16.49347686767578,
"learning_rate": 4.062909567496723e-06,
"loss": 5.3965,
"step": 124
},
{
"epoch": 0.003944095625234578,
"grad_norm": 14.946576118469238,
"learning_rate": 5.078636959370905e-06,
"loss": 4.8936,
"step": 155
},
{
"epoch": 0.004732914750281494,
"grad_norm": 16.985593795776367,
"learning_rate": 6.094364351245085e-06,
"loss": 4.5739,
"step": 186
},
{
"epoch": 0.00552173387532841,
"grad_norm": 17.912887573242188,
"learning_rate": 7.110091743119267e-06,
"loss": 4.3466,
"step": 217
},
{
"epoch": 0.006310553000375326,
"grad_norm": 15.746530532836914,
"learning_rate": 8.125819134993446e-06,
"loss": 4.1902,
"step": 248
},
{
"epoch": 0.007099372125422241,
"grad_norm": 24.640979766845703,
"learning_rate": 9.141546526867629e-06,
"loss": 4.0289,
"step": 279
},
{
"epoch": 0.007888191250469157,
"grad_norm": 14.921712875366211,
"learning_rate": 1.015727391874181e-05,
"loss": 3.8831,
"step": 310
},
{
"epoch": 0.008677010375516072,
"grad_norm": 13.711713790893555,
"learning_rate": 1.117300131061599e-05,
"loss": 3.7542,
"step": 341
},
{
"epoch": 0.009465829500562987,
"grad_norm": 12.796462059020996,
"learning_rate": 1.218872870249017e-05,
"loss": 3.6361,
"step": 372
},
{
"epoch": 0.010254648625609904,
"grad_norm": 13.015061378479004,
"learning_rate": 1.3204456094364351e-05,
"loss": 3.5447,
"step": 403
},
{
"epoch": 0.01104346775065682,
"grad_norm": 11.99329662322998,
"learning_rate": 1.4220183486238533e-05,
"loss": 3.4755,
"step": 434
},
{
"epoch": 0.011832286875703734,
"grad_norm": 14.673354148864746,
"learning_rate": 1.5235910878112714e-05,
"loss": 3.384,
"step": 465
},
{
"epoch": 0.012621106000750651,
"grad_norm": 14.919001579284668,
"learning_rate": 1.6251638269986893e-05,
"loss": 3.3096,
"step": 496
},
{
"epoch": 0.013409925125797566,
"grad_norm": 9.896846771240234,
"learning_rate": 1.7267365661861077e-05,
"loss": 3.2666,
"step": 527
},
{
"epoch": 0.014198744250844482,
"grad_norm": 10.563855171203613,
"learning_rate": 1.8283093053735257e-05,
"loss": 3.1977,
"step": 558
},
{
"epoch": 0.014987563375891397,
"grad_norm": 8.470562934875488,
"learning_rate": 1.9298820445609438e-05,
"loss": 3.1561,
"step": 589
},
{
"epoch": 0.015776382500938314,
"grad_norm": 7.783871173858643,
"learning_rate": 2.031454783748362e-05,
"loss": 3.1059,
"step": 620
},
{
"epoch": 0.016565201625985227,
"grad_norm": 7.2594194412231445,
"learning_rate": 2.13302752293578e-05,
"loss": 3.0697,
"step": 651
},
{
"epoch": 0.017354020751032144,
"grad_norm": 6.209031581878662,
"learning_rate": 2.234600262123198e-05,
"loss": 3.0443,
"step": 682
},
{
"epoch": 0.01814283987607906,
"grad_norm": 8.110218048095703,
"learning_rate": 2.336173001310616e-05,
"loss": 2.9893,
"step": 713
},
{
"epoch": 0.018931659001125974,
"grad_norm": 6.372434616088867,
"learning_rate": 2.437745740498034e-05,
"loss": 2.9522,
"step": 744
},
{
"epoch": 0.01972047812617289,
"grad_norm": 6.396059989929199,
"learning_rate": 2.5393184796854525e-05,
"loss": 2.9286,
"step": 775
},
{
"epoch": 0.020509297251219808,
"grad_norm": 5.576442718505859,
"learning_rate": 2.6408912188728702e-05,
"loss": 2.8986,
"step": 806
},
{
"epoch": 0.02129811637626672,
"grad_norm": 7.455265045166016,
"learning_rate": 2.7424639580602886e-05,
"loss": 2.8346,
"step": 837
},
{
"epoch": 0.02208693550131364,
"grad_norm": 5.682501792907715,
"learning_rate": 2.8440366972477066e-05,
"loss": 2.8033,
"step": 868
},
{
"epoch": 0.022875754626360555,
"grad_norm": 6.197375297546387,
"learning_rate": 2.9456094364351244e-05,
"loss": 2.8146,
"step": 899
},
{
"epoch": 0.02366457375140747,
"grad_norm": 6.124543190002441,
"learning_rate": 3.0471821756225428e-05,
"loss": 2.7633,
"step": 930
},
{
"epoch": 0.024453392876454386,
"grad_norm": 5.52219820022583,
"learning_rate": 3.148754914809961e-05,
"loss": 2.7328,
"step": 961
},
{
"epoch": 0.025242212001501303,
"grad_norm": 5.2266950607299805,
"learning_rate": 3.2503276539973785e-05,
"loss": 2.7086,
"step": 992
},
{
"epoch": 0.026031031126548216,
"grad_norm": 4.860825538635254,
"learning_rate": 3.351900393184797e-05,
"loss": 2.6655,
"step": 1023
},
{
"epoch": 0.026819850251595133,
"grad_norm": 4.891534805297852,
"learning_rate": 3.453473132372215e-05,
"loss": 2.6643,
"step": 1054
},
{
"epoch": 0.027608669376642046,
"grad_norm": 4.9125566482543945,
"learning_rate": 3.555045871559633e-05,
"loss": 2.6084,
"step": 1085
},
{
"epoch": 0.028397488501688963,
"grad_norm": 4.432997703552246,
"learning_rate": 3.6566186107470514e-05,
"loss": 2.5912,
"step": 1116
},
{
"epoch": 0.02918630762673588,
"grad_norm": 4.265169143676758,
"learning_rate": 3.7581913499344695e-05,
"loss": 2.574,
"step": 1147
},
{
"epoch": 0.029975126751782794,
"grad_norm": 4.731688022613525,
"learning_rate": 3.8597640891218876e-05,
"loss": 2.548,
"step": 1178
},
{
"epoch": 0.03076394587682971,
"grad_norm": 4.037458419799805,
"learning_rate": 3.9613368283093056e-05,
"loss": 2.5411,
"step": 1209
},
{
"epoch": 0.03155276500187663,
"grad_norm": 4.526784896850586,
"learning_rate": 4.062909567496724e-05,
"loss": 2.5111,
"step": 1240
},
{
"epoch": 0.03234158412692354,
"grad_norm": 4.06175422668457,
"learning_rate": 4.164482306684142e-05,
"loss": 2.519,
"step": 1271
},
{
"epoch": 0.033130403251970454,
"grad_norm": 4.599717617034912,
"learning_rate": 4.26605504587156e-05,
"loss": 2.4858,
"step": 1302
},
{
"epoch": 0.033919222377017375,
"grad_norm": 3.97731876373291,
"learning_rate": 4.367627785058978e-05,
"loss": 2.4902,
"step": 1333
},
{
"epoch": 0.03470804150206429,
"grad_norm": 3.732372522354126,
"learning_rate": 4.469200524246396e-05,
"loss": 2.4718,
"step": 1364
},
{
"epoch": 0.0354968606271112,
"grad_norm": 3.583721160888672,
"learning_rate": 4.570773263433814e-05,
"loss": 2.423,
"step": 1395
},
{
"epoch": 0.03628567975215812,
"grad_norm": 3.5558745861053467,
"learning_rate": 4.672346002621232e-05,
"loss": 2.4137,
"step": 1426
},
{
"epoch": 0.037074498877205035,
"grad_norm": 3.757810115814209,
"learning_rate": 4.77391874180865e-05,
"loss": 2.4368,
"step": 1457
},
{
"epoch": 0.03786331800225195,
"grad_norm": 3.6044111251831055,
"learning_rate": 4.875491480996068e-05,
"loss": 2.4129,
"step": 1488
},
{
"epoch": 0.03865213712729887,
"grad_norm": 3.366729736328125,
"learning_rate": 4.977064220183487e-05,
"loss": 2.3889,
"step": 1519
},
{
"epoch": 0.03944095625234578,
"grad_norm": 3.2331082820892334,
"learning_rate": 4.9999915451558777e-05,
"loss": 2.3745,
"step": 1550
},
{
"epoch": 0.040229775377392696,
"grad_norm": 3.150716781616211,
"learning_rate": 4.999955597496219e-05,
"loss": 2.3588,
"step": 1581
},
{
"epoch": 0.041018594502439616,
"grad_norm": 3.4237565994262695,
"learning_rate": 4.9998914381774255e-05,
"loss": 2.3639,
"step": 1612
},
{
"epoch": 0.04180741362748653,
"grad_norm": 2.9677655696868896,
"learning_rate": 4.999799067923527e-05,
"loss": 2.3587,
"step": 1643
},
{
"epoch": 0.04259623275253344,
"grad_norm": 3.1596009731292725,
"learning_rate": 4.999678487776908e-05,
"loss": 2.3365,
"step": 1674
},
{
"epoch": 0.04338505187758036,
"grad_norm": 3.4485344886779785,
"learning_rate": 4.9995296990983006e-05,
"loss": 2.3353,
"step": 1705
},
{
"epoch": 0.04417387100262728,
"grad_norm": 3.016425848007202,
"learning_rate": 4.999352703566763e-05,
"loss": 2.3172,
"step": 1736
},
{
"epoch": 0.04496269012767419,
"grad_norm": 2.8494279384613037,
"learning_rate": 4.999147503179668e-05,
"loss": 2.3241,
"step": 1767
},
{
"epoch": 0.04575150925272111,
"grad_norm": 2.7547197341918945,
"learning_rate": 4.998914100252672e-05,
"loss": 2.301,
"step": 1798
},
{
"epoch": 0.046540328377768024,
"grad_norm": 2.7102956771850586,
"learning_rate": 4.998652497419696e-05,
"loss": 2.3046,
"step": 1829
},
{
"epoch": 0.04732914750281494,
"grad_norm": 2.5211455821990967,
"learning_rate": 4.9983626976328927e-05,
"loss": 2.2794,
"step": 1860
},
{
"epoch": 0.04811796662786186,
"grad_norm": 2.7271173000335693,
"learning_rate": 4.998044704162613e-05,
"loss": 2.2554,
"step": 1891
},
{
"epoch": 0.04890678575290877,
"grad_norm": 2.850342273712158,
"learning_rate": 4.9976985205973705e-05,
"loss": 2.2722,
"step": 1922
},
{
"epoch": 0.049695604877955685,
"grad_norm": 2.6712722778320312,
"learning_rate": 4.997324150843799e-05,
"loss": 2.2412,
"step": 1953
},
{
"epoch": 0.050484424003002605,
"grad_norm": 2.6086056232452393,
"learning_rate": 4.99692159912661e-05,
"loss": 2.2724,
"step": 1984
},
{
"epoch": 0.05127324312804952,
"grad_norm": 2.495509147644043,
"learning_rate": 4.996490869988546e-05,
"loss": 2.2588,
"step": 2015
},
{
"epoch": 0.05206206225309643,
"grad_norm": 2.671813488006592,
"learning_rate": 4.996031968290326e-05,
"loss": 2.2334,
"step": 2046
},
{
"epoch": 0.052850881378143345,
"grad_norm": 2.6381373405456543,
"learning_rate": 4.995544899210594e-05,
"loss": 2.2361,
"step": 2077
},
{
"epoch": 0.053639700503190266,
"grad_norm": 2.54028058052063,
"learning_rate": 4.9950296682458583e-05,
"loss": 2.2216,
"step": 2108
},
{
"epoch": 0.05442851962823718,
"grad_norm": 2.67765212059021,
"learning_rate": 4.994486281210429e-05,
"loss": 2.2064,
"step": 2139
},
{
"epoch": 0.05521733875328409,
"grad_norm": 2.570333957672119,
"learning_rate": 4.9939147442363566e-05,
"loss": 2.2334,
"step": 2170
},
{
"epoch": 0.05600615787833101,
"grad_norm": 2.4661099910736084,
"learning_rate": 4.9933150637733574e-05,
"loss": 2.1975,
"step": 2201
},
{
"epoch": 0.056794977003377926,
"grad_norm": 2.5485525131225586,
"learning_rate": 4.992687246588743e-05,
"loss": 2.1883,
"step": 2232
},
{
"epoch": 0.05758379612842484,
"grad_norm": 2.471689462661743,
"learning_rate": 4.992031299767347e-05,
"loss": 2.1976,
"step": 2263
},
{
"epoch": 0.05837261525347176,
"grad_norm": 2.4743897914886475,
"learning_rate": 4.9913472307114386e-05,
"loss": 2.175,
"step": 2294
},
{
"epoch": 0.059161434378518674,
"grad_norm": 2.445451259613037,
"learning_rate": 4.9906350471406446e-05,
"loss": 2.1683,
"step": 2325
},
{
"epoch": 0.05995025350356559,
"grad_norm": 2.406235456466675,
"learning_rate": 4.989894757091861e-05,
"loss": 2.1685,
"step": 2356
},
{
"epoch": 0.06073907262861251,
"grad_norm": 2.2473762035369873,
"learning_rate": 4.989126368919158e-05,
"loss": 2.191,
"step": 2387
},
{
"epoch": 0.06152789175365942,
"grad_norm": 2.3035123348236084,
"learning_rate": 4.988329891293693e-05,
"loss": 2.1702,
"step": 2418
},
{
"epoch": 0.062316710878706334,
"grad_norm": 2.3511545658111572,
"learning_rate": 4.987505333203608e-05,
"loss": 2.1565,
"step": 2449
},
{
"epoch": 0.06310553000375325,
"grad_norm": 2.2300775051116943,
"learning_rate": 4.9866527039539276e-05,
"loss": 2.1558,
"step": 2480
},
{
"epoch": 0.06389434912880017,
"grad_norm": 2.3471133708953857,
"learning_rate": 4.9857720131664594e-05,
"loss": 2.1516,
"step": 2511
},
{
"epoch": 0.06468316825384708,
"grad_norm": 2.3770556449890137,
"learning_rate": 4.9848632707796773e-05,
"loss": 2.1384,
"step": 2542
},
{
"epoch": 0.065471987378894,
"grad_norm": 2.235302209854126,
"learning_rate": 4.9839264870486155e-05,
"loss": 2.1287,
"step": 2573
},
{
"epoch": 0.06626080650394091,
"grad_norm": 2.3938326835632324,
"learning_rate": 4.9829616725447526e-05,
"loss": 2.1387,
"step": 2604
},
{
"epoch": 0.06704962562898784,
"grad_norm": 2.177912712097168,
"learning_rate": 4.981968838155888e-05,
"loss": 2.1126,
"step": 2635
},
{
"epoch": 0.06783844475403475,
"grad_norm": 2.2535035610198975,
"learning_rate": 4.980947995086024e-05,
"loss": 2.126,
"step": 2666
},
{
"epoch": 0.06862726387908166,
"grad_norm": 2.14459228515625,
"learning_rate": 4.979899154855234e-05,
"loss": 2.1334,
"step": 2697
},
{
"epoch": 0.06941608300412858,
"grad_norm": 2.156930446624756,
"learning_rate": 4.9788223292995386e-05,
"loss": 2.1121,
"step": 2728
},
{
"epoch": 0.07020490212917549,
"grad_norm": 2.344521999359131,
"learning_rate": 4.977717530570768e-05,
"loss": 2.123,
"step": 2759
},
{
"epoch": 0.0709937212542224,
"grad_norm": 2.1548428535461426,
"learning_rate": 4.976584771136425e-05,
"loss": 2.113,
"step": 2790
},
{
"epoch": 0.07178254037926933,
"grad_norm": 2.2467427253723145,
"learning_rate": 4.975424063779547e-05,
"loss": 2.1048,
"step": 2821
},
{
"epoch": 0.07257135950431624,
"grad_norm": 2.601376533508301,
"learning_rate": 4.974235421598557e-05,
"loss": 2.1144,
"step": 2852
},
{
"epoch": 0.07336017862936316,
"grad_norm": 2.143150568008423,
"learning_rate": 4.973018858007122e-05,
"loss": 2.0793,
"step": 2883
},
{
"epoch": 0.07414899775441007,
"grad_norm": 2.1503472328186035,
"learning_rate": 4.9717743867339963e-05,
"loss": 2.0996,
"step": 2914
},
{
"epoch": 0.07493781687945698,
"grad_norm": 2.1850991249084473,
"learning_rate": 4.9705020218228695e-05,
"loss": 2.0912,
"step": 2945
},
{
"epoch": 0.0757266360045039,
"grad_norm": 2.207058906555176,
"learning_rate": 4.969201777632205e-05,
"loss": 2.104,
"step": 2976
},
{
"epoch": 0.07651545512955082,
"grad_norm": 2.1132214069366455,
"learning_rate": 4.9678736688350846e-05,
"loss": 2.0857,
"step": 3007
},
{
"epoch": 0.07730427425459774,
"grad_norm": 4.404266834259033,
"learning_rate": 4.966517710419033e-05,
"loss": 2.0888,
"step": 3038
},
{
"epoch": 0.07809309337964465,
"grad_norm": 2.0443875789642334,
"learning_rate": 4.965133917685858e-05,
"loss": 2.0688,
"step": 3069
},
{
"epoch": 0.07888191250469156,
"grad_norm": 1.9510438442230225,
"learning_rate": 4.9637223062514714e-05,
"loss": 2.0891,
"step": 3100
},
{
"epoch": 0.07967073162973848,
"grad_norm": 2.089725971221924,
"learning_rate": 4.962282892045718e-05,
"loss": 2.0949,
"step": 3131
},
{
"epoch": 0.08045955075478539,
"grad_norm": 2.1167550086975098,
"learning_rate": 4.9608156913121904e-05,
"loss": 2.0911,
"step": 3162
},
{
"epoch": 0.08124836987983232,
"grad_norm": 2.2896945476531982,
"learning_rate": 4.959320720608049e-05,
"loss": 2.0709,
"step": 3193
},
{
"epoch": 0.08203718900487923,
"grad_norm": 2.0464134216308594,
"learning_rate": 4.9577979968038354e-05,
"loss": 2.0743,
"step": 3224
},
{
"epoch": 0.08282600812992615,
"grad_norm": 2.0745983123779297,
"learning_rate": 4.956247537083282e-05,
"loss": 2.0644,
"step": 3255
},
{
"epoch": 0.08361482725497306,
"grad_norm": 2.1493799686431885,
"learning_rate": 4.9546693589431145e-05,
"loss": 2.0651,
"step": 3286
},
{
"epoch": 0.08440364638001997,
"grad_norm": 2.07473087310791,
"learning_rate": 4.9530634801928595e-05,
"loss": 2.0354,
"step": 3317
},
{
"epoch": 0.08519246550506689,
"grad_norm": 1.962320327758789,
"learning_rate": 4.9514299189546395e-05,
"loss": 2.0664,
"step": 3348
},
{
"epoch": 0.0859812846301138,
"grad_norm": 1.9907017946243286,
"learning_rate": 4.949768693662973e-05,
"loss": 2.0772,
"step": 3379
},
{
"epoch": 0.08677010375516073,
"grad_norm": 2.2022948265075684,
"learning_rate": 4.948079823064559e-05,
"loss": 2.0629,
"step": 3410
},
{
"epoch": 0.08755892288020764,
"grad_norm": 1.9094113111495972,
"learning_rate": 4.946363326218074e-05,
"loss": 2.0516,
"step": 3441
},
{
"epoch": 0.08834774200525455,
"grad_norm": 2.16923189163208,
"learning_rate": 4.9446192224939525e-05,
"loss": 2.0712,
"step": 3472
},
{
"epoch": 0.08913656113030147,
"grad_norm": 2.2181923389434814,
"learning_rate": 4.942847531574167e-05,
"loss": 2.0476,
"step": 3503
},
{
"epoch": 0.08992538025534838,
"grad_norm": 2.0701868534088135,
"learning_rate": 4.941048273452008e-05,
"loss": 2.0662,
"step": 3534
},
{
"epoch": 0.0907141993803953,
"grad_norm": 2.1713969707489014,
"learning_rate": 4.9392214684318605e-05,
"loss": 2.0316,
"step": 3565
},
{
"epoch": 0.09150301850544222,
"grad_norm": 1.8713674545288086,
"learning_rate": 4.93736713712897e-05,
"loss": 2.0276,
"step": 3596
},
{
"epoch": 0.09229183763048913,
"grad_norm": 1.957153081893921,
"learning_rate": 4.9354853004692124e-05,
"loss": 2.0423,
"step": 3627
},
{
"epoch": 0.09308065675553605,
"grad_norm": 2.025723457336426,
"learning_rate": 4.93357597968886e-05,
"loss": 2.0361,
"step": 3658
},
{
"epoch": 0.09386947588058296,
"grad_norm": 1.9253807067871094,
"learning_rate": 4.931639196334338e-05,
"loss": 2.0254,
"step": 3689
},
{
"epoch": 0.09465829500562987,
"grad_norm": 2.1453866958618164,
"learning_rate": 4.9296749722619826e-05,
"loss": 2.0434,
"step": 3720
},
{
"epoch": 0.09544711413067679,
"grad_norm": 2.116746187210083,
"learning_rate": 4.9276833296377966e-05,
"loss": 2.0535,
"step": 3751
},
{
"epoch": 0.09623593325572372,
"grad_norm": 2.060053825378418,
"learning_rate": 4.925664290937196e-05,
"loss": 2.0162,
"step": 3782
},
{
"epoch": 0.09702475238077063,
"grad_norm": 2.033005952835083,
"learning_rate": 4.9236178789447576e-05,
"loss": 2.0377,
"step": 3813
},
{
"epoch": 0.09781357150581754,
"grad_norm": 1.930992841720581,
"learning_rate": 4.921544116753962e-05,
"loss": 2.0091,
"step": 3844
},
{
"epoch": 0.09860239063086446,
"grad_norm": 1.987481951713562,
"learning_rate": 4.919443027766935e-05,
"loss": 2.0191,
"step": 3875
},
{
"epoch": 0.09939120975591137,
"grad_norm": 2.1110246181488037,
"learning_rate": 4.91731463569418e-05,
"loss": 2.015,
"step": 3906
},
{
"epoch": 0.10018002888095828,
"grad_norm": 2.009760618209839,
"learning_rate": 4.915158964554312e-05,
"loss": 2.0196,
"step": 3937
},
{
"epoch": 0.10096884800600521,
"grad_norm": 1.9978469610214233,
"learning_rate": 4.912976038673786e-05,
"loss": 2.0095,
"step": 3968
},
{
"epoch": 0.10175766713105212,
"grad_norm": 1.9885358810424805,
"learning_rate": 4.9107658826866254e-05,
"loss": 2.0148,
"step": 3999
},
{
"epoch": 0.10254648625609904,
"grad_norm": 1.9853264093399048,
"learning_rate": 4.908528521534139e-05,
"loss": 2.014,
"step": 4030
},
{
"epoch": 0.10333530538114595,
"grad_norm": 1.9636098146438599,
"learning_rate": 4.906263980464644e-05,
"loss": 2.0197,
"step": 4061
},
{
"epoch": 0.10412412450619286,
"grad_norm": 1.9954477548599243,
"learning_rate": 4.903972285033178e-05,
"loss": 1.9974,
"step": 4092
},
{
"epoch": 0.10491294363123978,
"grad_norm": 1.9644993543624878,
"learning_rate": 4.901653461101213e-05,
"loss": 2.0035,
"step": 4123
},
{
"epoch": 0.10570176275628669,
"grad_norm": 1.9775594472885132,
"learning_rate": 4.8993075348363626e-05,
"loss": 2.0108,
"step": 4154
},
{
"epoch": 0.10649058188133362,
"grad_norm": 1.9478306770324707,
"learning_rate": 4.896934532712084e-05,
"loss": 1.992,
"step": 4185
},
{
"epoch": 0.10727940100638053,
"grad_norm": 2.2822272777557373,
"learning_rate": 4.8945344815073846e-05,
"loss": 2.0038,
"step": 4216
},
{
"epoch": 0.10806822013142744,
"grad_norm": 1.9325686693191528,
"learning_rate": 4.892107408306516e-05,
"loss": 1.9914,
"step": 4247
},
{
"epoch": 0.10885703925647436,
"grad_norm": 1.9367748498916626,
"learning_rate": 4.889653340498669e-05,
"loss": 2.0012,
"step": 4278
},
{
"epoch": 0.10964585838152127,
"grad_norm": 1.9141064882278442,
"learning_rate": 4.8871723057776664e-05,
"loss": 2.0192,
"step": 4309
},
{
"epoch": 0.11043467750656819,
"grad_norm": 2.1826133728027344,
"learning_rate": 4.8846643321416476e-05,
"loss": 2.0014,
"step": 4340
},
{
"epoch": 0.11122349663161511,
"grad_norm": 1.9631683826446533,
"learning_rate": 4.882129447892753e-05,
"loss": 1.9848,
"step": 4371
},
{
"epoch": 0.11201231575666203,
"grad_norm": 1.969335913658142,
"learning_rate": 4.8795676816368076e-05,
"loss": 1.9832,
"step": 4402
},
{
"epoch": 0.11280113488170894,
"grad_norm": 1.8996473550796509,
"learning_rate": 4.876979062282995e-05,
"loss": 1.9866,
"step": 4433
},
{
"epoch": 0.11358995400675585,
"grad_norm": 1.9162741899490356,
"learning_rate": 4.8743636190435325e-05,
"loss": 1.9985,
"step": 4464
},
{
"epoch": 0.11437877313180277,
"grad_norm": 1.9941459894180298,
"learning_rate": 4.871721381433344e-05,
"loss": 1.9831,
"step": 4495
},
{
"epoch": 0.11516759225684968,
"grad_norm": 1.7893972396850586,
"learning_rate": 4.869052379269719e-05,
"loss": 1.9965,
"step": 4526
},
{
"epoch": 0.1159564113818966,
"grad_norm": 1.8150454759597778,
"learning_rate": 4.866356642671985e-05,
"loss": 1.9814,
"step": 4557
},
{
"epoch": 0.11674523050694352,
"grad_norm": 1.9319466352462769,
"learning_rate": 4.8636342020611634e-05,
"loss": 1.9761,
"step": 4588
},
{
"epoch": 0.11753404963199043,
"grad_norm": 1.7858814001083374,
"learning_rate": 4.860885088159626e-05,
"loss": 1.9912,
"step": 4619
},
{
"epoch": 0.11832286875703735,
"grad_norm": 1.8554847240447998,
"learning_rate": 4.858109331990751e-05,
"loss": 1.9709,
"step": 4650
},
{
"epoch": 0.11911168788208426,
"grad_norm": 1.8804956674575806,
"learning_rate": 4.855306964878567e-05,
"loss": 2.0065,
"step": 4681
},
{
"epoch": 0.11990050700713117,
"grad_norm": 1.8334413766860962,
"learning_rate": 4.8524780184474084e-05,
"loss": 1.9883,
"step": 4712
},
{
"epoch": 0.1206893261321781,
"grad_norm": 2.0563790798187256,
"learning_rate": 4.8496225246215496e-05,
"loss": 1.9805,
"step": 4743
},
{
"epoch": 0.12147814525722501,
"grad_norm": 1.7843685150146484,
"learning_rate": 4.8467405156248505e-05,
"loss": 1.9732,
"step": 4774
},
{
"epoch": 0.12226696438227193,
"grad_norm": 1.8490878343582153,
"learning_rate": 4.843832023980392e-05,
"loss": 1.9955,
"step": 4805
},
{
"epoch": 0.12305578350731884,
"grad_norm": 1.6674724817276,
"learning_rate": 4.840897082510106e-05,
"loss": 1.9837,
"step": 4836
},
{
"epoch": 0.12384460263236576,
"grad_norm": 1.7055600881576538,
"learning_rate": 4.8379357243344084e-05,
"loss": 1.9782,
"step": 4867
},
{
"epoch": 0.12463342175741267,
"grad_norm": 1.7836189270019531,
"learning_rate": 4.8349479828718236e-05,
"loss": 1.9696,
"step": 4898
},
{
"epoch": 0.1254222408824596,
"grad_norm": 1.8213683366775513,
"learning_rate": 4.8319338918386075e-05,
"loss": 1.9689,
"step": 4929
},
{
"epoch": 0.1262110600075065,
"grad_norm": 1.7552939653396606,
"learning_rate": 4.828893485248369e-05,
"loss": 2.0059,
"step": 4960
},
{
"epoch": 0.12699987913255342,
"grad_norm": 1.7021842002868652,
"learning_rate": 4.825826797411682e-05,
"loss": 1.966,
"step": 4991
},
{
"epoch": 0.12778869825760034,
"grad_norm": 2.696760416030884,
"learning_rate": 4.822733862935702e-05,
"loss": 1.9909,
"step": 5022
},
{
"epoch": 0.12857751738264725,
"grad_norm": 1.8321341276168823,
"learning_rate": 4.819614716723775e-05,
"loss": 1.967,
"step": 5053
},
{
"epoch": 0.12936633650769416,
"grad_norm": 2.1158053874969482,
"learning_rate": 4.8164693939750425e-05,
"loss": 1.9692,
"step": 5084
},
{
"epoch": 0.13015515563274108,
"grad_norm": 1.9149221181869507,
"learning_rate": 4.813297930184042e-05,
"loss": 1.9676,
"step": 5115
},
{
"epoch": 0.130943974757788,
"grad_norm": 1.7447597980499268,
"learning_rate": 4.810100361140314e-05,
"loss": 1.9538,
"step": 5146
},
{
"epoch": 0.1317327938828349,
"grad_norm": 1.7692095041275024,
"learning_rate": 4.8068767229279885e-05,
"loss": 1.9612,
"step": 5177
},
{
"epoch": 0.13252161300788182,
"grad_norm": 1.8180122375488281,
"learning_rate": 4.8036270519253854e-05,
"loss": 1.9679,
"step": 5208
},
{
"epoch": 0.13331043213292876,
"grad_norm": 1.7397531270980835,
"learning_rate": 4.8003513848046e-05,
"loss": 1.9483,
"step": 5239
},
{
"epoch": 0.13409925125797567,
"grad_norm": 1.773038387298584,
"learning_rate": 4.79704975853109e-05,
"loss": 1.9706,
"step": 5270
},
{
"epoch": 0.13488807038302258,
"grad_norm": 1.846821904182434,
"learning_rate": 4.793722210363262e-05,
"loss": 1.9557,
"step": 5301
},
{
"epoch": 0.1356768895080695,
"grad_norm": 1.8578094244003296,
"learning_rate": 4.7903687778520414e-05,
"loss": 1.9613,
"step": 5332
},
{
"epoch": 0.1364657086331164,
"grad_norm": 1.6778779029846191,
"learning_rate": 4.7869894988404593e-05,
"loss": 1.9481,
"step": 5363
},
{
"epoch": 0.13725452775816332,
"grad_norm": 1.8821789026260376,
"learning_rate": 4.783584411463221e-05,
"loss": 1.9475,
"step": 5394
},
{
"epoch": 0.13804334688321024,
"grad_norm": 1.7182306051254272,
"learning_rate": 4.780153554146274e-05,
"loss": 1.9541,
"step": 5425
},
{
"epoch": 0.13883216600825715,
"grad_norm": 1.9063465595245361,
"learning_rate": 4.7766969656063766e-05,
"loss": 1.9475,
"step": 5456
},
{
"epoch": 0.13962098513330407,
"grad_norm": 1.8275973796844482,
"learning_rate": 4.773214684850662e-05,
"loss": 1.9672,
"step": 5487
},
{
"epoch": 0.14040980425835098,
"grad_norm": 1.798607349395752,
"learning_rate": 4.769706751176193e-05,
"loss": 1.9603,
"step": 5518
},
{
"epoch": 0.1411986233833979,
"grad_norm": 1.9662116765975952,
"learning_rate": 4.7661732041695264e-05,
"loss": 1.959,
"step": 5549
},
{
"epoch": 0.1419874425084448,
"grad_norm": 1.9336684942245483,
"learning_rate": 4.762614083706258e-05,
"loss": 1.9427,
"step": 5580
},
{
"epoch": 0.14277626163349175,
"grad_norm": 1.68619704246521,
"learning_rate": 4.759029429950581e-05,
"loss": 1.9459,
"step": 5611
},
{
"epoch": 0.14356508075853866,
"grad_norm": 1.9060170650482178,
"learning_rate": 4.7554192833548235e-05,
"loss": 1.9522,
"step": 5642
},
{
"epoch": 0.14435389988358557,
"grad_norm": 1.7184319496154785,
"learning_rate": 4.751783684659e-05,
"loss": 1.9253,
"step": 5673
},
{
"epoch": 0.1451427190086325,
"grad_norm": 1.7833524942398071,
"learning_rate": 4.748122674890348e-05,
"loss": 1.9484,
"step": 5704
},
{
"epoch": 0.1459315381336794,
"grad_norm": 1.8405711650848389,
"learning_rate": 4.7444362953628654e-05,
"loss": 1.9576,
"step": 5735
},
{
"epoch": 0.1467203572587263,
"grad_norm": 1.7746304273605347,
"learning_rate": 4.7407245876768424e-05,
"loss": 1.9549,
"step": 5766
},
{
"epoch": 0.14750917638377323,
"grad_norm": 1.6882883310317993,
"learning_rate": 4.736987593718397e-05,
"loss": 1.9407,
"step": 5797
},
{
"epoch": 0.14829799550882014,
"grad_norm": 1.6760326623916626,
"learning_rate": 4.733225355658999e-05,
"loss": 1.9358,
"step": 5828
},
{
"epoch": 0.14908681463386705,
"grad_norm": 1.6274425983428955,
"learning_rate": 4.7294379159549926e-05,
"loss": 1.929,
"step": 5859
},
{
"epoch": 0.14987563375891397,
"grad_norm": 1.8570448160171509,
"learning_rate": 4.725625317347119e-05,
"loss": 1.926,
"step": 5890
},
{
"epoch": 0.15066445288396088,
"grad_norm": 1.754744052886963,
"learning_rate": 4.7217876028600374e-05,
"loss": 1.948,
"step": 5921
},
{
"epoch": 0.1514532720090078,
"grad_norm": 1.7100951671600342,
"learning_rate": 4.717924815801832e-05,
"loss": 1.9527,
"step": 5952
},
{
"epoch": 0.1522420911340547,
"grad_norm": 1.731489658355713,
"learning_rate": 4.714036999763532e-05,
"loss": 1.933,
"step": 5983
},
{
"epoch": 0.15303091025910165,
"grad_norm": 1.7440745830535889,
"learning_rate": 4.7101241986186116e-05,
"loss": 1.9214,
"step": 6014
},
{
"epoch": 0.15381972938414856,
"grad_norm": 1.7062227725982666,
"learning_rate": 4.7061864565225e-05,
"loss": 1.9381,
"step": 6045
},
{
"epoch": 0.15460854850919548,
"grad_norm": 1.6533842086791992,
"learning_rate": 4.702223817912081e-05,
"loss": 1.9097,
"step": 6076
},
{
"epoch": 0.1553973676342424,
"grad_norm": 1.8380335569381714,
"learning_rate": 4.698236327505195e-05,
"loss": 1.9357,
"step": 6107
},
{
"epoch": 0.1561861867592893,
"grad_norm": 1.7769367694854736,
"learning_rate": 4.694224030300127e-05,
"loss": 1.9176,
"step": 6138
},
{
"epoch": 0.15697500588433622,
"grad_norm": 1.697293758392334,
"learning_rate": 4.690186971575107e-05,
"loss": 1.9242,
"step": 6169
},
{
"epoch": 0.15776382500938313,
"grad_norm": 1.6948884725570679,
"learning_rate": 4.6861251968877916e-05,
"loss": 1.9277,
"step": 6200
},
{
"epoch": 0.15855264413443004,
"grad_norm": 1.8506945371627808,
"learning_rate": 4.68203875207476e-05,
"loss": 1.9413,
"step": 6231
},
{
"epoch": 0.15934146325947696,
"grad_norm": 1.6183403730392456,
"learning_rate": 4.677927683250983e-05,
"loss": 1.929,
"step": 6262
},
{
"epoch": 0.16013028238452387,
"grad_norm": 1.7411370277404785,
"learning_rate": 4.6737920368093156e-05,
"loss": 1.9272,
"step": 6293
},
{
"epoch": 0.16091910150957078,
"grad_norm": 1.6963975429534912,
"learning_rate": 4.669631859419965e-05,
"loss": 1.9183,
"step": 6324
},
{
"epoch": 0.1617079206346177,
"grad_norm": 1.664981484413147,
"learning_rate": 4.6654471980299676e-05,
"loss": 1.9358,
"step": 6355
},
{
"epoch": 0.16249673975966464,
"grad_norm": 1.8351246118545532,
"learning_rate": 4.661238099862658e-05,
"loss": 1.9126,
"step": 6386
},
{
"epoch": 0.16328555888471155,
"grad_norm": 1.8973945379257202,
"learning_rate": 4.657004612417138e-05,
"loss": 1.9171,
"step": 6417
},
{
"epoch": 0.16407437800975846,
"grad_norm": 1.7961071729660034,
"learning_rate": 4.6527467834677374e-05,
"loss": 1.9095,
"step": 6448
},
{
"epoch": 0.16486319713480538,
"grad_norm": 1.73550283908844,
"learning_rate": 4.648464661063478e-05,
"loss": 1.924,
"step": 6479
},
{
"epoch": 0.1656520162598523,
"grad_norm": 1.77008056640625,
"learning_rate": 4.6441582935275264e-05,
"loss": 1.9338,
"step": 6510
},
{
"epoch": 0.1664408353848992,
"grad_norm": 1.7121846675872803,
"learning_rate": 4.6398277294566586e-05,
"loss": 1.9094,
"step": 6541
},
{
"epoch": 0.16722965450994612,
"grad_norm": 1.6560605764389038,
"learning_rate": 4.6354730177207e-05,
"loss": 1.9195,
"step": 6572
},
{
"epoch": 0.16801847363499303,
"grad_norm": 1.7044239044189453,
"learning_rate": 4.6310942074619787e-05,
"loss": 1.9197,
"step": 6603
},
{
"epoch": 0.16880729276003995,
"grad_norm": 1.6876695156097412,
"learning_rate": 4.626691348094777e-05,
"loss": 1.9078,
"step": 6634
},
{
"epoch": 0.16959611188508686,
"grad_norm": 1.7991163730621338,
"learning_rate": 4.622264489304762e-05,
"loss": 1.9147,
"step": 6665
},
{
"epoch": 0.17038493101013377,
"grad_norm": 1.662376046180725,
"learning_rate": 4.617813681048434e-05,
"loss": 1.9263,
"step": 6696
},
{
"epoch": 0.17117375013518069,
"grad_norm": 1.6112306118011475,
"learning_rate": 4.61333897355256e-05,
"loss": 1.9102,
"step": 6727
},
{
"epoch": 0.1719625692602276,
"grad_norm": 1.757332444190979,
"learning_rate": 4.608840417313604e-05,
"loss": 1.921,
"step": 6758
},
{
"epoch": 0.17275138838527454,
"grad_norm": 1.6604098081588745,
"learning_rate": 4.6043180630971646e-05,
"loss": 1.905,
"step": 6789
},
{
"epoch": 0.17354020751032145,
"grad_norm": 1.6792216300964355,
"learning_rate": 4.599771961937391e-05,
"loss": 1.906,
"step": 6820
},
{
"epoch": 0.17432902663536837,
"grad_norm": 1.635213851928711,
"learning_rate": 4.5952021651364204e-05,
"loss": 1.9038,
"step": 6851
},
{
"epoch": 0.17511784576041528,
"grad_norm": 1.6034295558929443,
"learning_rate": 4.590608724263786e-05,
"loss": 1.9214,
"step": 6882
},
{
"epoch": 0.1759066648854622,
"grad_norm": 1.730025053024292,
"learning_rate": 4.585991691155845e-05,
"loss": 1.9023,
"step": 6913
},
{
"epoch": 0.1766954840105091,
"grad_norm": 1.6403334140777588,
"learning_rate": 4.581351117915188e-05,
"loss": 1.8944,
"step": 6944
},
{
"epoch": 0.17748430313555602,
"grad_norm": 1.7030408382415771,
"learning_rate": 4.5766870569100534e-05,
"loss": 1.9082,
"step": 6975
},
{
"epoch": 0.17827312226060293,
"grad_norm": 1.9924838542938232,
"learning_rate": 4.571999560773736e-05,
"loss": 1.9033,
"step": 7006
},
{
"epoch": 0.17906194138564985,
"grad_norm": 1.688637137413025,
"learning_rate": 4.5672886824039915e-05,
"loss": 1.9087,
"step": 7037
},
{
"epoch": 0.17985076051069676,
"grad_norm": 1.7744380235671997,
"learning_rate": 4.5625544749624435e-05,
"loss": 1.8911,
"step": 7068
},
{
"epoch": 0.18063957963574367,
"grad_norm": 1.6221119165420532,
"learning_rate": 4.5577969918739794e-05,
"loss": 1.9018,
"step": 7099
},
{
"epoch": 0.1814283987607906,
"grad_norm": 1.6945770978927612,
"learning_rate": 4.5530162868261486e-05,
"loss": 1.9042,
"step": 7130
},
{
"epoch": 0.18221721788583753,
"grad_norm": 1.6812230348587036,
"learning_rate": 4.548212413768558e-05,
"loss": 1.8992,
"step": 7161
},
{
"epoch": 0.18300603701088444,
"grad_norm": 1.5889744758605957,
"learning_rate": 4.543385426912261e-05,
"loss": 1.9132,
"step": 7192
},
{
"epoch": 0.18379485613593136,
"grad_norm": 1.9001593589782715,
"learning_rate": 4.53853538072915e-05,
"loss": 1.9053,
"step": 7223
},
{
"epoch": 0.18458367526097827,
"grad_norm": 1.647740125656128,
"learning_rate": 4.533662329951336e-05,
"loss": 1.9191,
"step": 7254
},
{
"epoch": 0.18537249438602518,
"grad_norm": 1.6723840236663818,
"learning_rate": 4.528766329570536e-05,
"loss": 1.8782,
"step": 7285
},
{
"epoch": 0.1861613135110721,
"grad_norm": 1.668317437171936,
"learning_rate": 4.523847434837447e-05,
"loss": 1.9026,
"step": 7316
},
{
"epoch": 0.186950132636119,
"grad_norm": 1.640189528465271,
"learning_rate": 4.518905701261128e-05,
"loss": 1.9087,
"step": 7347
},
{
"epoch": 0.18773895176116592,
"grad_norm": 1.5775387287139893,
"learning_rate": 4.5139411846083715e-05,
"loss": 1.9129,
"step": 7378
},
{
"epoch": 0.18852777088621284,
"grad_norm": 1.5601975917816162,
"learning_rate": 4.508953940903073e-05,
"loss": 1.9088,
"step": 7409
},
{
"epoch": 0.18931659001125975,
"grad_norm": 1.648223876953125,
"learning_rate": 4.5039440264255994e-05,
"loss": 1.8977,
"step": 7440
},
{
"epoch": 0.19010540913630666,
"grad_norm": 1.6973642110824585,
"learning_rate": 4.498911497712155e-05,
"loss": 1.8849,
"step": 7471
},
{
"epoch": 0.19089422826135358,
"grad_norm": 1.6180808544158936,
"learning_rate": 4.493856411554142e-05,
"loss": 1.8859,
"step": 7502
},
{
"epoch": 0.1916830473864005,
"grad_norm": 1.6768759489059448,
"learning_rate": 4.4887788249975206e-05,
"loss": 1.9005,
"step": 7533
},
{
"epoch": 0.19247186651144743,
"grad_norm": 1.6389315128326416,
"learning_rate": 4.4836787953421656e-05,
"loss": 1.9025,
"step": 7564
},
{
"epoch": 0.19326068563649434,
"grad_norm": 1.6695667505264282,
"learning_rate": 4.478556380141218e-05,
"loss": 1.9015,
"step": 7595
},
{
"epoch": 0.19404950476154126,
"grad_norm": 1.634464979171753,
"learning_rate": 4.4734116372004375e-05,
"loss": 1.8848,
"step": 7626
},
{
"epoch": 0.19483832388658817,
"grad_norm": 1.6077677011489868,
"learning_rate": 4.4682446245775477e-05,
"loss": 1.8944,
"step": 7657
},
{
"epoch": 0.19562714301163509,
"grad_norm": 1.7401186227798462,
"learning_rate": 4.463055400581586e-05,
"loss": 1.8841,
"step": 7688
},
{
"epoch": 0.196415962136682,
"grad_norm": 1.588080644607544,
"learning_rate": 4.4578440237722374e-05,
"loss": 1.8988,
"step": 7719
},
{
"epoch": 0.1972047812617289,
"grad_norm": 1.5799649953842163,
"learning_rate": 4.452610552959183e-05,
"loss": 1.8828,
"step": 7750
},
{
"epoch": 0.19799360038677583,
"grad_norm": 1.6081531047821045,
"learning_rate": 4.447355047201428e-05,
"loss": 1.8967,
"step": 7781
},
{
"epoch": 0.19878241951182274,
"grad_norm": 1.6133239269256592,
"learning_rate": 4.4420775658066414e-05,
"loss": 1.8898,
"step": 7812
},
{
"epoch": 0.19957123863686965,
"grad_norm": 1.612506628036499,
"learning_rate": 4.436778168330484e-05,
"loss": 1.8962,
"step": 7843
},
{
"epoch": 0.20036005776191657,
"grad_norm": 1.6066638231277466,
"learning_rate": 4.4314569145759353e-05,
"loss": 1.8885,
"step": 7874
},
{
"epoch": 0.20114887688696348,
"grad_norm": 1.6667569875717163,
"learning_rate": 4.42611386459262e-05,
"loss": 1.8836,
"step": 7905
},
{
"epoch": 0.20193769601201042,
"grad_norm": 1.7658061981201172,
"learning_rate": 4.420749078676133e-05,
"loss": 1.8627,
"step": 7936
},
{
"epoch": 0.20272651513705733,
"grad_norm": 1.5527122020721436,
"learning_rate": 4.4153626173673516e-05,
"loss": 1.8682,
"step": 7967
},
{
"epoch": 0.20351533426210425,
"grad_norm": 1.6022471189498901,
"learning_rate": 4.409954541451762e-05,
"loss": 1.8892,
"step": 7998
},
{
"epoch": 0.20430415338715116,
"grad_norm": 1.6246200799942017,
"learning_rate": 4.404524911958764e-05,
"loss": 1.8703,
"step": 8029
},
{
"epoch": 0.20509297251219807,
"grad_norm": 1.5030767917633057,
"learning_rate": 4.399073790160989e-05,
"loss": 1.8897,
"step": 8060
},
{
"epoch": 0.205881791637245,
"grad_norm": 1.603633165359497,
"learning_rate": 4.393601237573607e-05,
"loss": 1.8655,
"step": 8091
},
{
"epoch": 0.2066706107622919,
"grad_norm": 1.5709718465805054,
"learning_rate": 4.388107315953628e-05,
"loss": 1.8715,
"step": 8122
},
{
"epoch": 0.20745942988733881,
"grad_norm": 1.529145359992981,
"learning_rate": 4.382592087299212e-05,
"loss": 1.8628,
"step": 8153
},
{
"epoch": 0.20824824901238573,
"grad_norm": 1.6858514547348022,
"learning_rate": 4.377055613848964e-05,
"loss": 1.8871,
"step": 8184
},
{
"epoch": 0.20903706813743264,
"grad_norm": 1.7125133275985718,
"learning_rate": 4.3714979580812355e-05,
"loss": 1.8768,
"step": 8215
},
{
"epoch": 0.20982588726247955,
"grad_norm": 1.5397855043411255,
"learning_rate": 4.365919182713416e-05,
"loss": 1.8807,
"step": 8246
},
{
"epoch": 0.21061470638752647,
"grad_norm": 1.6475356817245483,
"learning_rate": 4.360319350701226e-05,
"loss": 1.885,
"step": 8277
},
{
"epoch": 0.21140352551257338,
"grad_norm": 1.5485234260559082,
"learning_rate": 4.3546985252380115e-05,
"loss": 1.8777,
"step": 8308
},
{
"epoch": 0.21219234463762032,
"grad_norm": 1.5416793823242188,
"learning_rate": 4.349056769754021e-05,
"loss": 1.8577,
"step": 8339
},
{
"epoch": 0.21298116376266724,
"grad_norm": 1.6852402687072754,
"learning_rate": 4.3433941479156994e-05,
"loss": 1.8807,
"step": 8370
},
{
"epoch": 0.21376998288771415,
"grad_norm": 1.5214046239852905,
"learning_rate": 4.3377107236249647e-05,
"loss": 1.8631,
"step": 8401
},
{
"epoch": 0.21455880201276106,
"grad_norm": 1.6051750183105469,
"learning_rate": 4.332006561018488e-05,
"loss": 1.8684,
"step": 8432
},
{
"epoch": 0.21534762113780798,
"grad_norm": 1.5419507026672363,
"learning_rate": 4.3262817244669683e-05,
"loss": 1.9002,
"step": 8463
},
{
"epoch": 0.2161364402628549,
"grad_norm": 1.6802319288253784,
"learning_rate": 4.3205362785744083e-05,
"loss": 1.8577,
"step": 8494
},
{
"epoch": 0.2169252593879018,
"grad_norm": 1.5763416290283203,
"learning_rate": 4.314770288177384e-05,
"loss": 1.865,
"step": 8525
},
{
"epoch": 0.21771407851294872,
"grad_norm": 1.5372315645217896,
"learning_rate": 4.308983818344313e-05,
"loss": 1.8634,
"step": 8556
},
{
"epoch": 0.21850289763799563,
"grad_norm": 1.6838102340698242,
"learning_rate": 4.3031769343747206e-05,
"loss": 1.8685,
"step": 8587
},
{
"epoch": 0.21929171676304254,
"grad_norm": 1.6022460460662842,
"learning_rate": 4.297349701798505e-05,
"loss": 1.8593,
"step": 8618
},
{
"epoch": 0.22008053588808946,
"grad_norm": 1.6018834114074707,
"learning_rate": 4.2915021863751916e-05,
"loss": 1.8852,
"step": 8649
},
{
"epoch": 0.22086935501313637,
"grad_norm": 1.5209150314331055,
"learning_rate": 4.285634454093198e-05,
"loss": 1.8522,
"step": 8680
},
{
"epoch": 0.2216581741381833,
"grad_norm": 1.6187845468521118,
"learning_rate": 4.279746571169086e-05,
"loss": 1.8497,
"step": 8711
},
{
"epoch": 0.22244699326323022,
"grad_norm": 1.6452182531356812,
"learning_rate": 4.2738386040468136e-05,
"loss": 1.8551,
"step": 8742
},
{
"epoch": 0.22323581238827714,
"grad_norm": 1.613244652748108,
"learning_rate": 4.2679106193969866e-05,
"loss": 1.8629,
"step": 8773
},
{
"epoch": 0.22402463151332405,
"grad_norm": 1.5586017370224,
"learning_rate": 4.261962684116106e-05,
"loss": 1.8674,
"step": 8804
},
{
"epoch": 0.22481345063837097,
"grad_norm": 1.5552334785461426,
"learning_rate": 4.2559948653258145e-05,
"loss": 1.8923,
"step": 8835
},
{
"epoch": 0.22560226976341788,
"grad_norm": 1.609717607498169,
"learning_rate": 4.250007230372134e-05,
"loss": 1.8854,
"step": 8866
},
{
"epoch": 0.2263910888884648,
"grad_norm": 1.617981195449829,
"learning_rate": 4.2439998468247126e-05,
"loss": 1.8794,
"step": 8897
},
{
"epoch": 0.2271799080135117,
"grad_norm": 1.5466769933700562,
"learning_rate": 4.2379727824760566e-05,
"loss": 1.8594,
"step": 8928
},
{
"epoch": 0.22796872713855862,
"grad_norm": 1.5078591108322144,
"learning_rate": 4.231926105340768e-05,
"loss": 1.8543,
"step": 8959
},
{
"epoch": 0.22875754626360553,
"grad_norm": 1.5065971612930298,
"learning_rate": 4.225859883654776e-05,
"loss": 1.8506,
"step": 8990
},
{
"epoch": 0.22954636538865245,
"grad_norm": 1.583526849746704,
"learning_rate": 4.219774185874569e-05,
"loss": 1.8741,
"step": 9021
},
{
"epoch": 0.23033518451369936,
"grad_norm": 1.5880491733551025,
"learning_rate": 4.213669080676418e-05,
"loss": 1.8531,
"step": 9052
},
{
"epoch": 0.2311240036387463,
"grad_norm": 1.6649373769760132,
"learning_rate": 4.2075446369556056e-05,
"loss": 1.8524,
"step": 9083
},
{
"epoch": 0.2319128227637932,
"grad_norm": 1.556809425354004,
"learning_rate": 4.201400923825648e-05,
"loss": 1.8581,
"step": 9114
},
{
"epoch": 0.23270164188884013,
"grad_norm": 1.5572203397750854,
"learning_rate": 4.195238010617511e-05,
"loss": 1.8546,
"step": 9145
},
{
"epoch": 0.23349046101388704,
"grad_norm": 1.5201469659805298,
"learning_rate": 4.1890559668788344e-05,
"loss": 1.8644,
"step": 9176
},
{
"epoch": 0.23427928013893395,
"grad_norm": 1.5816295146942139,
"learning_rate": 4.1828548623731405e-05,
"loss": 1.8556,
"step": 9207
},
{
"epoch": 0.23506809926398087,
"grad_norm": 1.5696643590927124,
"learning_rate": 4.1766347670790506e-05,
"loss": 1.8491,
"step": 9238
},
{
"epoch": 0.23585691838902778,
"grad_norm": 1.5091686248779297,
"learning_rate": 4.170395751189495e-05,
"loss": 1.8493,
"step": 9269
},
{
"epoch": 0.2366457375140747,
"grad_norm": 1.5627410411834717,
"learning_rate": 4.164137885110921e-05,
"loss": 1.8502,
"step": 9300
},
{
"epoch": 0.2374345566391216,
"grad_norm": 1.5759685039520264,
"learning_rate": 4.157861239462495e-05,
"loss": 1.8427,
"step": 9331
},
{
"epoch": 0.23822337576416852,
"grad_norm": 1.533496379852295,
"learning_rate": 4.1515658850753114e-05,
"loss": 1.8533,
"step": 9362
},
{
"epoch": 0.23901219488921543,
"grad_norm": 1.5628371238708496,
"learning_rate": 4.145251892991588e-05,
"loss": 1.8439,
"step": 9393
},
{
"epoch": 0.23980101401426235,
"grad_norm": 1.568591594696045,
"learning_rate": 4.138919334463868e-05,
"loss": 1.8533,
"step": 9424
},
{
"epoch": 0.24058983313930926,
"grad_norm": 1.5815969705581665,
"learning_rate": 4.1325682809542124e-05,
"loss": 1.8511,
"step": 9455
},
{
"epoch": 0.2413786522643562,
"grad_norm": 1.5089821815490723,
"learning_rate": 4.126198804133398e-05,
"loss": 1.8609,
"step": 9486
},
{
"epoch": 0.24216747138940312,
"grad_norm": 1.5676932334899902,
"learning_rate": 4.1198109758801055e-05,
"loss": 1.8565,
"step": 9517
},
{
"epoch": 0.24295629051445003,
"grad_norm": 1.5041563510894775,
"learning_rate": 4.113404868280107e-05,
"loss": 1.8536,
"step": 9548
},
{
"epoch": 0.24374510963949694,
"grad_norm": 1.5710649490356445,
"learning_rate": 4.106980553625457e-05,
"loss": 1.8547,
"step": 9579
},
{
"epoch": 0.24453392876454386,
"grad_norm": 1.5798616409301758,
"learning_rate": 4.100538104413674e-05,
"loss": 1.8572,
"step": 9610
},
{
"epoch": 0.24532274788959077,
"grad_norm": 2.4222729206085205,
"learning_rate": 4.09407759334692e-05,
"loss": 1.8645,
"step": 9641
},
{
"epoch": 0.24611156701463768,
"grad_norm": 1.4646201133728027,
"learning_rate": 4.087599093331186e-05,
"loss": 1.8496,
"step": 9672
},
{
"epoch": 0.2469003861396846,
"grad_norm": 1.606465458869934,
"learning_rate": 4.081102677475462e-05,
"loss": 1.8459,
"step": 9703
},
{
"epoch": 0.2476892052647315,
"grad_norm": 1.5241799354553223,
"learning_rate": 4.0745884190909194e-05,
"loss": 1.8511,
"step": 9734
},
{
"epoch": 0.24847802438977842,
"grad_norm": 1.6257604360580444,
"learning_rate": 4.0680563916900796e-05,
"loss": 1.8416,
"step": 9765
},
{
"epoch": 0.24926684351482534,
"grad_norm": 1.5233662128448486,
"learning_rate": 4.0615066689859815e-05,
"loss": 1.8465,
"step": 9796
},
{
"epoch": 0.2500556626398723,
"grad_norm": 1.4877644777297974,
"learning_rate": 4.0549393248913584e-05,
"loss": 1.8467,
"step": 9827
},
{
"epoch": 0.2508444817649192,
"grad_norm": 1.530216932296753,
"learning_rate": 4.048354433517794e-05,
"loss": 1.8304,
"step": 9858
},
{
"epoch": 0.2516333008899661,
"grad_norm": 1.4812994003295898,
"learning_rate": 4.0417520691748916e-05,
"loss": 1.8324,
"step": 9889
},
{
"epoch": 0.252422120015013,
"grad_norm": 1.5022152662277222,
"learning_rate": 4.035132306369438e-05,
"loss": 1.8521,
"step": 9920
},
{
"epoch": 0.25321093914005993,
"grad_norm": 1.5792648792266846,
"learning_rate": 4.028495219804555e-05,
"loss": 1.8275,
"step": 9951
},
{
"epoch": 0.25399975826510685,
"grad_norm": 1.4738425016403198,
"learning_rate": 4.021840884378864e-05,
"loss": 1.8484,
"step": 9982
},
{
"epoch": 0.25478857739015376,
"grad_norm": 1.4558053016662598,
"learning_rate": 4.015169375185633e-05,
"loss": 1.842,
"step": 10013
},
{
"epoch": 0.25557739651520067,
"grad_norm": 1.6721614599227905,
"learning_rate": 4.0084807675119396e-05,
"loss": 1.8582,
"step": 10044
},
{
"epoch": 0.2563662156402476,
"grad_norm": 1.643314003944397,
"learning_rate": 4.0017751368378106e-05,
"loss": 1.8467,
"step": 10075
},
{
"epoch": 0.2571550347652945,
"grad_norm": 1.600417137145996,
"learning_rate": 3.995052558835377e-05,
"loss": 1.8466,
"step": 10106
},
{
"epoch": 0.2579438538903414,
"grad_norm": 1.5018486976623535,
"learning_rate": 3.988313109368017e-05,
"loss": 1.8299,
"step": 10137
},
{
"epoch": 0.2587326730153883,
"grad_norm": 1.4656468629837036,
"learning_rate": 3.981556864489504e-05,
"loss": 1.8254,
"step": 10168
},
{
"epoch": 0.25952149214043524,
"grad_norm": 1.573634386062622,
"learning_rate": 3.974783900443142e-05,
"loss": 1.8387,
"step": 10199
},
{
"epoch": 0.26031031126548215,
"grad_norm": 1.4818580150604248,
"learning_rate": 3.9679942936609095e-05,
"loss": 1.8364,
"step": 10230
},
{
"epoch": 0.26109913039052907,
"grad_norm": 1.560797095298767,
"learning_rate": 3.961188120762596e-05,
"loss": 1.8504,
"step": 10261
},
{
"epoch": 0.261887949515576,
"grad_norm": 1.535651683807373,
"learning_rate": 3.954365458554938e-05,
"loss": 1.8408,
"step": 10292
},
{
"epoch": 0.2626767686406229,
"grad_norm": 1.5009502172470093,
"learning_rate": 3.947526384030751e-05,
"loss": 1.8612,
"step": 10323
},
{
"epoch": 0.2634655877656698,
"grad_norm": 1.5589120388031006,
"learning_rate": 3.9406709743680624e-05,
"loss": 1.843,
"step": 10354
},
{
"epoch": 0.2642544068907167,
"grad_norm": 1.6979308128356934,
"learning_rate": 3.9337993069292366e-05,
"loss": 1.8445,
"step": 10385
},
{
"epoch": 0.26504322601576363,
"grad_norm": 1.4456043243408203,
"learning_rate": 3.926911459260109e-05,
"loss": 1.8455,
"step": 10416
},
{
"epoch": 0.2658320451408106,
"grad_norm": 1.5096760988235474,
"learning_rate": 3.920007509089102e-05,
"loss": 1.8243,
"step": 10447
},
{
"epoch": 0.2666208642658575,
"grad_norm": 1.534406065940857,
"learning_rate": 3.913087534326357e-05,
"loss": 1.828,
"step": 10478
},
{
"epoch": 0.26740968339090443,
"grad_norm": 1.7499176263809204,
"learning_rate": 3.9061516130628475e-05,
"loss": 1.8372,
"step": 10509
},
{
"epoch": 0.26819850251595134,
"grad_norm": 1.5080372095108032,
"learning_rate": 3.8991998235695025e-05,
"loss": 1.8237,
"step": 10540
},
{
"epoch": 0.26898732164099826,
"grad_norm": 1.495322346687317,
"learning_rate": 3.8922322442963224e-05,
"loss": 1.856,
"step": 10571
},
{
"epoch": 0.26977614076604517,
"grad_norm": 1.548775315284729,
"learning_rate": 3.885248953871491e-05,
"loss": 1.8305,
"step": 10602
},
{
"epoch": 0.2705649598910921,
"grad_norm": 1.655852198600769,
"learning_rate": 3.8782500311004915e-05,
"loss": 1.826,
"step": 10633
},
{
"epoch": 0.271353779016139,
"grad_norm": 1.4688091278076172,
"learning_rate": 3.871235554965218e-05,
"loss": 1.8277,
"step": 10664
},
{
"epoch": 0.2721425981411859,
"grad_norm": 1.4753823280334473,
"learning_rate": 3.864205604623078e-05,
"loss": 1.8343,
"step": 10695
},
{
"epoch": 0.2729314172662328,
"grad_norm": 1.5295604467391968,
"learning_rate": 3.857160259406107e-05,
"loss": 1.8212,
"step": 10726
},
{
"epoch": 0.27372023639127974,
"grad_norm": 1.5936615467071533,
"learning_rate": 3.8500995988200674e-05,
"loss": 1.8369,
"step": 10757
},
{
"epoch": 0.27450905551632665,
"grad_norm": 1.61776864528656,
"learning_rate": 3.843023702543556e-05,
"loss": 1.8075,
"step": 10788
},
{
"epoch": 0.27529787464137356,
"grad_norm": 1.5151159763336182,
"learning_rate": 3.8359326504270984e-05,
"loss": 1.824,
"step": 10819
},
{
"epoch": 0.2760866937664205,
"grad_norm": 1.6150144338607788,
"learning_rate": 3.828826522492255e-05,
"loss": 1.8392,
"step": 10850
},
{
"epoch": 0.2768755128914674,
"grad_norm": 1.4708991050720215,
"learning_rate": 3.821705398930713e-05,
"loss": 1.8428,
"step": 10881
},
{
"epoch": 0.2776643320165143,
"grad_norm": 1.4772177934646606,
"learning_rate": 3.814569360103385e-05,
"loss": 1.816,
"step": 10912
},
{
"epoch": 0.2784531511415612,
"grad_norm": 1.5404503345489502,
"learning_rate": 3.807418486539499e-05,
"loss": 1.8148,
"step": 10943
},
{
"epoch": 0.27924197026660813,
"grad_norm": 1.5612235069274902,
"learning_rate": 3.80025285893569e-05,
"loss": 1.8345,
"step": 10974
},
{
"epoch": 0.28003078939165504,
"grad_norm": 1.5301347970962524,
"learning_rate": 3.793072558155093e-05,
"loss": 1.8248,
"step": 11005
},
{
"epoch": 0.28081960851670196,
"grad_norm": 1.4737629890441895,
"learning_rate": 3.785877665226426e-05,
"loss": 1.8198,
"step": 11036
},
{
"epoch": 0.28160842764174887,
"grad_norm": 1.4503047466278076,
"learning_rate": 3.778668261343079e-05,
"loss": 1.8371,
"step": 11067
},
{
"epoch": 0.2823972467667958,
"grad_norm": 1.5823520421981812,
"learning_rate": 3.771444427862192e-05,
"loss": 1.8325,
"step": 11098
},
{
"epoch": 0.2831860658918427,
"grad_norm": 1.6144663095474243,
"learning_rate": 3.7642062463037465e-05,
"loss": 1.8209,
"step": 11129
},
{
"epoch": 0.2839748850168896,
"grad_norm": 1.6444216966629028,
"learning_rate": 3.7569537983496373e-05,
"loss": 1.825,
"step": 11160
},
{
"epoch": 0.2847637041419365,
"grad_norm": 1.503098487854004,
"learning_rate": 3.749687165842753e-05,
"loss": 1.8392,
"step": 11191
},
{
"epoch": 0.2855525232669835,
"grad_norm": 1.83393394947052,
"learning_rate": 3.7424064307860536e-05,
"loss": 1.8252,
"step": 11222
},
{
"epoch": 0.2863413423920304,
"grad_norm": 1.4748836755752563,
"learning_rate": 3.735111675341645e-05,
"loss": 1.806,
"step": 11253
},
{
"epoch": 0.2871301615170773,
"grad_norm": 1.912840485572815,
"learning_rate": 3.7278029818298524e-05,
"loss": 1.8095,
"step": 11284
},
{
"epoch": 0.28791898064212423,
"grad_norm": 1.5209791660308838,
"learning_rate": 3.720480432728287e-05,
"loss": 1.8162,
"step": 11315
},
{
"epoch": 0.28870779976717115,
"grad_norm": 1.5430744886398315,
"learning_rate": 3.71314411067092e-05,
"loss": 1.833,
"step": 11346
},
{
"epoch": 0.28949661889221806,
"grad_norm": 1.592753529548645,
"learning_rate": 3.70579409844715e-05,
"loss": 1.8402,
"step": 11377
},
{
"epoch": 0.290285438017265,
"grad_norm": 1.5640573501586914,
"learning_rate": 3.698430479000865e-05,
"loss": 1.8275,
"step": 11408
},
{
"epoch": 0.2910742571423119,
"grad_norm": 1.4649739265441895,
"learning_rate": 3.691053335429509e-05,
"loss": 1.8092,
"step": 11439
},
{
"epoch": 0.2918630762673588,
"grad_norm": 1.5399479866027832,
"learning_rate": 3.683662750983147e-05,
"loss": 1.8131,
"step": 11470
},
{
"epoch": 0.2926518953924057,
"grad_norm": 1.5420011281967163,
"learning_rate": 3.676258809063518e-05,
"loss": 1.8287,
"step": 11501
},
{
"epoch": 0.2934407145174526,
"grad_norm": 1.4778993129730225,
"learning_rate": 3.6688415932231004e-05,
"loss": 1.8202,
"step": 11532
},
{
"epoch": 0.29422953364249954,
"grad_norm": 1.5598480701446533,
"learning_rate": 3.661411187164166e-05,
"loss": 1.8425,
"step": 11563
},
{
"epoch": 0.29501835276754645,
"grad_norm": 1.6161003112792969,
"learning_rate": 3.65396767473784e-05,
"loss": 1.8402,
"step": 11594
},
{
"epoch": 0.29580717189259337,
"grad_norm": 1.4399648904800415,
"learning_rate": 3.6465111399431465e-05,
"loss": 1.8083,
"step": 11625
},
{
"epoch": 0.2965959910176403,
"grad_norm": 1.5157605409622192,
"learning_rate": 3.6390416669260674e-05,
"loss": 1.8041,
"step": 11656
},
{
"epoch": 0.2973848101426872,
"grad_norm": 1.5234498977661133,
"learning_rate": 3.63155933997859e-05,
"loss": 1.8104,
"step": 11687
},
{
"epoch": 0.2981736292677341,
"grad_norm": 1.4988269805908203,
"learning_rate": 3.624064243537758e-05,
"loss": 1.8192,
"step": 11718
},
{
"epoch": 0.298962448392781,
"grad_norm": 1.5283297300338745,
"learning_rate": 3.616556462184716e-05,
"loss": 1.8123,
"step": 11749
},
{
"epoch": 0.29975126751782794,
"grad_norm": 1.523959994316101,
"learning_rate": 3.609036080643755e-05,
"loss": 1.8391,
"step": 11780
},
{
"epoch": 0.30054008664287485,
"grad_norm": 1.4363136291503906,
"learning_rate": 3.60150318378136e-05,
"loss": 1.8176,
"step": 11811
},
{
"epoch": 0.30132890576792176,
"grad_norm": 1.521101474761963,
"learning_rate": 3.5939578566052465e-05,
"loss": 1.8164,
"step": 11842
},
{
"epoch": 0.3021177248929687,
"grad_norm": 1.4613672494888306,
"learning_rate": 3.586400184263408e-05,
"loss": 1.8204,
"step": 11873
},
{
"epoch": 0.3029065440180156,
"grad_norm": 1.4992350339889526,
"learning_rate": 3.578830252043148e-05,
"loss": 1.8122,
"step": 11904
},
{
"epoch": 0.3036953631430625,
"grad_norm": 1.659454584121704,
"learning_rate": 3.571248145370125e-05,
"loss": 1.8243,
"step": 11935
},
{
"epoch": 0.3044841822681094,
"grad_norm": 1.429630160331726,
"learning_rate": 3.5636539498073794e-05,
"loss": 1.8079,
"step": 11966
},
{
"epoch": 0.3052730013931564,
"grad_norm": 1.4403787851333618,
"learning_rate": 3.556047751054378e-05,
"loss": 1.8079,
"step": 11997
},
{
"epoch": 0.3060618205182033,
"grad_norm": 1.4795056581497192,
"learning_rate": 3.548429634946039e-05,
"loss": 1.7949,
"step": 12028
},
{
"epoch": 0.3068506396432502,
"grad_norm": 1.5466026067733765,
"learning_rate": 3.540799687451768e-05,
"loss": 1.8039,
"step": 12059
},
{
"epoch": 0.3076394587682971,
"grad_norm": 1.5872678756713867,
"learning_rate": 3.533157994674485e-05,
"loss": 1.8174,
"step": 12090
},
{
"epoch": 0.30842827789334404,
"grad_norm": 1.4539406299591064,
"learning_rate": 3.5255046428496546e-05,
"loss": 1.8212,
"step": 12121
},
{
"epoch": 0.30921709701839095,
"grad_norm": 1.443831443786621,
"learning_rate": 3.517839718344311e-05,
"loss": 1.8387,
"step": 12152
},
{
"epoch": 0.31000591614343787,
"grad_norm": 1.4756397008895874,
"learning_rate": 3.510163307656086e-05,
"loss": 1.827,
"step": 12183
},
{
"epoch": 0.3107947352684848,
"grad_norm": 1.5675030946731567,
"learning_rate": 3.5024754974122324e-05,
"loss": 1.8198,
"step": 12214
},
{
"epoch": 0.3115835543935317,
"grad_norm": 1.4247853755950928,
"learning_rate": 3.494776374368643e-05,
"loss": 1.8072,
"step": 12245
},
{
"epoch": 0.3123723735185786,
"grad_norm": 1.567158579826355,
"learning_rate": 3.4870660254088724e-05,
"loss": 1.8158,
"step": 12276
},
{
"epoch": 0.3131611926436255,
"grad_norm": 1.4549590349197388,
"learning_rate": 3.479344537543164e-05,
"loss": 1.8044,
"step": 12307
},
{
"epoch": 0.31395001176867243,
"grad_norm": 1.4478166103363037,
"learning_rate": 3.4716119979074565e-05,
"loss": 1.8099,
"step": 12338
},
{
"epoch": 0.31473883089371935,
"grad_norm": 1.4160298109054565,
"learning_rate": 3.463868493762412e-05,
"loss": 1.8179,
"step": 12369
},
{
"epoch": 0.31552765001876626,
"grad_norm": 1.5618009567260742,
"learning_rate": 3.456114112492418e-05,
"loss": 1.801,
"step": 12400
},
{
"epoch": 0.3163164691438132,
"grad_norm": 1.3777692317962646,
"learning_rate": 3.4483489416046164e-05,
"loss": 1.8117,
"step": 12431
},
{
"epoch": 0.3171052882688601,
"grad_norm": 1.5304317474365234,
"learning_rate": 3.440573068727905e-05,
"loss": 1.8085,
"step": 12462
},
{
"epoch": 0.317894107393907,
"grad_norm": 1.5578155517578125,
"learning_rate": 3.4327865816119495e-05,
"loss": 1.8106,
"step": 12493
},
{
"epoch": 0.3186829265189539,
"grad_norm": 1.51682710647583,
"learning_rate": 3.4249895681262025e-05,
"loss": 1.8199,
"step": 12524
},
{
"epoch": 0.3194717456440008,
"grad_norm": 1.4429659843444824,
"learning_rate": 3.417182116258899e-05,
"loss": 1.7975,
"step": 12555
},
{
"epoch": 0.32026056476904774,
"grad_norm": 1.4837095737457275,
"learning_rate": 3.409364314116074e-05,
"loss": 1.8039,
"step": 12586
},
{
"epoch": 0.32104938389409465,
"grad_norm": 1.4250850677490234,
"learning_rate": 3.401536249920559e-05,
"loss": 1.7886,
"step": 12617
},
{
"epoch": 0.32183820301914157,
"grad_norm": 1.429291009902954,
"learning_rate": 3.393698012010998e-05,
"loss": 1.7938,
"step": 12648
},
{
"epoch": 0.3226270221441885,
"grad_norm": 1.4839123487472534,
"learning_rate": 3.385849688840839e-05,
"loss": 1.811,
"step": 12679
},
{
"epoch": 0.3234158412692354,
"grad_norm": 1.5528687238693237,
"learning_rate": 3.3779913689773414e-05,
"loss": 1.8004,
"step": 12710
},
{
"epoch": 0.3242046603942823,
"grad_norm": 1.4324339628219604,
"learning_rate": 3.370123141100578e-05,
"loss": 1.8048,
"step": 12741
},
{
"epoch": 0.3249934795193293,
"grad_norm": 1.5204551219940186,
"learning_rate": 3.3622450940024305e-05,
"loss": 1.7944,
"step": 12772
},
{
"epoch": 0.3257822986443762,
"grad_norm": 1.3667051792144775,
"learning_rate": 3.35435731658559e-05,
"loss": 1.8086,
"step": 12803
},
{
"epoch": 0.3265711177694231,
"grad_norm": 1.4720345735549927,
"learning_rate": 3.346459897862552e-05,
"loss": 1.7982,
"step": 12834
},
{
"epoch": 0.32735993689447,
"grad_norm": 1.4867823123931885,
"learning_rate": 3.338552926954613e-05,
"loss": 1.7855,
"step": 12865
},
{
"epoch": 0.32814875601951693,
"grad_norm": 1.468201994895935,
"learning_rate": 3.330636493090868e-05,
"loss": 1.7909,
"step": 12896
},
{
"epoch": 0.32893757514456384,
"grad_norm": 1.4950802326202393,
"learning_rate": 3.322710685607193e-05,
"loss": 1.78,
"step": 12927
},
{
"epoch": 0.32972639426961076,
"grad_norm": 1.6039625406265259,
"learning_rate": 3.314775593945251e-05,
"loss": 1.8094,
"step": 12958
},
{
"epoch": 0.33051521339465767,
"grad_norm": 1.5084600448608398,
"learning_rate": 3.3068313076514714e-05,
"loss": 1.7908,
"step": 12989
},
{
"epoch": 0.3313040325197046,
"grad_norm": 1.4338363409042358,
"learning_rate": 3.298877916376047e-05,
"loss": 1.8092,
"step": 13020
},
{
"epoch": 0.3320928516447515,
"grad_norm": 1.4125412702560425,
"learning_rate": 3.290915509871915e-05,
"loss": 1.7756,
"step": 13051
},
{
"epoch": 0.3328816707697984,
"grad_norm": 1.4515947103500366,
"learning_rate": 3.282944177993753e-05,
"loss": 1.7845,
"step": 13082
},
{
"epoch": 0.3336704898948453,
"grad_norm": 1.586795449256897,
"learning_rate": 3.274964010696957e-05,
"loss": 1.8047,
"step": 13113
},
{
"epoch": 0.33445930901989224,
"grad_norm": 1.486333966255188,
"learning_rate": 3.266975098036629e-05,
"loss": 1.7913,
"step": 13144
},
{
"epoch": 0.33524812814493915,
"grad_norm": 1.4172276258468628,
"learning_rate": 3.258977530166562e-05,
"loss": 1.7962,
"step": 13175
},
{
"epoch": 0.33603694726998606,
"grad_norm": 1.4856761693954468,
"learning_rate": 3.250971397338227e-05,
"loss": 1.8232,
"step": 13206
},
{
"epoch": 0.336825766395033,
"grad_norm": 1.496001124382019,
"learning_rate": 3.2429567898997404e-05,
"loss": 1.7968,
"step": 13237
},
{
"epoch": 0.3376145855200799,
"grad_norm": 1.5100946426391602,
"learning_rate": 3.234933798294859e-05,
"loss": 1.806,
"step": 13268
},
{
"epoch": 0.3384034046451268,
"grad_norm": 1.4353514909744263,
"learning_rate": 3.2269025130619535e-05,
"loss": 1.8088,
"step": 13299
},
{
"epoch": 0.3391922237701737,
"grad_norm": 1.393201231956482,
"learning_rate": 3.218863024832985e-05,
"loss": 1.778,
"step": 13330
},
{
"epoch": 0.33998104289522063,
"grad_norm": 1.4197418689727783,
"learning_rate": 3.2108154243324864e-05,
"loss": 1.7986,
"step": 13361
},
{
"epoch": 0.34076986202026754,
"grad_norm": 1.5117108821868896,
"learning_rate": 3.2027598023765345e-05,
"loss": 1.812,
"step": 13392
},
{
"epoch": 0.34155868114531446,
"grad_norm": 1.4815988540649414,
"learning_rate": 3.194696249871729e-05,
"loss": 1.7971,
"step": 13423
},
{
"epoch": 0.34234750027036137,
"grad_norm": 1.4436742067337036,
"learning_rate": 3.186624857814164e-05,
"loss": 1.7946,
"step": 13454
},
{
"epoch": 0.3431363193954083,
"grad_norm": 1.6074799299240112,
"learning_rate": 3.178545717288401e-05,
"loss": 1.8018,
"step": 13485
},
{
"epoch": 0.3439251385204552,
"grad_norm": 1.5564550161361694,
"learning_rate": 3.170458919466444e-05,
"loss": 1.7816,
"step": 13516
},
{
"epoch": 0.34471395764550217,
"grad_norm": 1.4918630123138428,
"learning_rate": 3.1623645556067063e-05,
"loss": 1.7759,
"step": 13547
},
{
"epoch": 0.3455027767705491,
"grad_norm": 1.537247896194458,
"learning_rate": 3.154262717052985e-05,
"loss": 1.7959,
"step": 13578
},
{
"epoch": 0.346291595895596,
"grad_norm": 1.4561980962753296,
"learning_rate": 3.146153495233426e-05,
"loss": 1.7923,
"step": 13609
},
{
"epoch": 0.3470804150206429,
"grad_norm": 1.4964359998703003,
"learning_rate": 3.1380369816594944e-05,
"loss": 1.8033,
"step": 13640
},
{
"epoch": 0.3478692341456898,
"grad_norm": 1.463797688484192,
"learning_rate": 3.129913267924946e-05,
"loss": 1.7794,
"step": 13671
},
{
"epoch": 0.34865805327073673,
"grad_norm": 1.4784711599349976,
"learning_rate": 3.121782445704782e-05,
"loss": 1.8002,
"step": 13702
},
{
"epoch": 0.34944687239578365,
"grad_norm": 1.4917761087417603,
"learning_rate": 3.11364460675423e-05,
"loss": 1.7964,
"step": 13733
},
{
"epoch": 0.35023569152083056,
"grad_norm": 1.4284688234329224,
"learning_rate": 3.1054998429076934e-05,
"loss": 1.7981,
"step": 13764
},
{
"epoch": 0.3510245106458775,
"grad_norm": 1.4910475015640259,
"learning_rate": 3.097348246077728e-05,
"loss": 1.7952,
"step": 13795
},
{
"epoch": 0.3518133297709244,
"grad_norm": 1.4870178699493408,
"learning_rate": 3.0891899082539924e-05,
"loss": 1.7876,
"step": 13826
},
{
"epoch": 0.3526021488959713,
"grad_norm": 1.5134365558624268,
"learning_rate": 3.0810249215022233e-05,
"loss": 1.7961,
"step": 13857
},
{
"epoch": 0.3533909680210182,
"grad_norm": 1.595760464668274,
"learning_rate": 3.0728533779631865e-05,
"loss": 1.8069,
"step": 13888
},
{
"epoch": 0.35417978714606513,
"grad_norm": 1.5597907304763794,
"learning_rate": 3.064675369851637e-05,
"loss": 1.787,
"step": 13919
},
{
"epoch": 0.35496860627111204,
"grad_norm": 1.4652432203292847,
"learning_rate": 3.056490989455289e-05,
"loss": 1.7936,
"step": 13950
},
{
"epoch": 0.35575742539615895,
"grad_norm": 1.5195232629776,
"learning_rate": 3.0483003291337596e-05,
"loss": 1.7988,
"step": 13981
},
{
"epoch": 0.35654624452120587,
"grad_norm": 1.5883373022079468,
"learning_rate": 3.040103481317539e-05,
"loss": 1.7752,
"step": 14012
},
{
"epoch": 0.3573350636462528,
"grad_norm": 1.4016722440719604,
"learning_rate": 3.03190053850694e-05,
"loss": 1.7793,
"step": 14043
},
{
"epoch": 0.3581238827712997,
"grad_norm": 1.4025518894195557,
"learning_rate": 3.0236915932710573e-05,
"loss": 1.7788,
"step": 14074
},
{
"epoch": 0.3589127018963466,
"grad_norm": 1.3901499509811401,
"learning_rate": 3.0154767382467232e-05,
"loss": 1.7766,
"step": 14105
},
{
"epoch": 0.3597015210213935,
"grad_norm": 1.4077001810073853,
"learning_rate": 3.0072560661374582e-05,
"loss": 1.7903,
"step": 14136
},
{
"epoch": 0.36049034014644044,
"grad_norm": 1.4476062059402466,
"learning_rate": 2.999029669712431e-05,
"loss": 1.7851,
"step": 14167
},
{
"epoch": 0.36127915927148735,
"grad_norm": 1.4461426734924316,
"learning_rate": 2.990797641805408e-05,
"loss": 1.7759,
"step": 14198
},
{
"epoch": 0.36206797839653426,
"grad_norm": 1.452197790145874,
"learning_rate": 2.982560075313704e-05,
"loss": 1.7877,
"step": 14229
},
{
"epoch": 0.3628567975215812,
"grad_norm": 1.4651702642440796,
"learning_rate": 2.9743170631971368e-05,
"loss": 1.7609,
"step": 14260
},
{
"epoch": 0.3636456166466281,
"grad_norm": 1.4937199354171753,
"learning_rate": 2.9660686984769792e-05,
"loss": 1.7794,
"step": 14291
},
{
"epoch": 0.36443443577167506,
"grad_norm": 1.4670535326004028,
"learning_rate": 2.9578150742349047e-05,
"loss": 1.7891,
"step": 14322
},
{
"epoch": 0.36522325489672197,
"grad_norm": 1.5605027675628662,
"learning_rate": 2.949556283611942e-05,
"loss": 1.7941,
"step": 14353
},
{
"epoch": 0.3660120740217689,
"grad_norm": 1.4724050760269165,
"learning_rate": 2.9412924198074206e-05,
"loss": 1.7944,
"step": 14384
},
{
"epoch": 0.3668008931468158,
"grad_norm": 1.4741649627685547,
"learning_rate": 2.9330235760779208e-05,
"loss": 1.7881,
"step": 14415
},
{
"epoch": 0.3675897122718627,
"grad_norm": 1.9677305221557617,
"learning_rate": 2.9247498457362188e-05,
"loss": 1.7911,
"step": 14446
},
{
"epoch": 0.3683785313969096,
"grad_norm": 1.5587711334228516,
"learning_rate": 2.9164713221502373e-05,
"loss": 1.7955,
"step": 14477
},
{
"epoch": 0.36916735052195654,
"grad_norm": 1.4210267066955566,
"learning_rate": 2.9081880987419912e-05,
"loss": 1.7656,
"step": 14508
},
{
"epoch": 0.36995616964700345,
"grad_norm": 1.5188663005828857,
"learning_rate": 2.8999002689865296e-05,
"loss": 1.7779,
"step": 14539
},
{
"epoch": 0.37074498877205037,
"grad_norm": 1.460797905921936,
"learning_rate": 2.8916079264108852e-05,
"loss": 1.7929,
"step": 14570
},
{
"epoch": 0.3715338078970973,
"grad_norm": 1.464895486831665,
"learning_rate": 2.883311164593017e-05,
"loss": 1.777,
"step": 14601
},
{
"epoch": 0.3723226270221442,
"grad_norm": 1.5290796756744385,
"learning_rate": 2.875010077160754e-05,
"loss": 1.7837,
"step": 14632
},
{
"epoch": 0.3731114461471911,
"grad_norm": 1.4018083810806274,
"learning_rate": 2.866704757790741e-05,
"loss": 1.7685,
"step": 14663
},
{
"epoch": 0.373900265272238,
"grad_norm": 1.3743735551834106,
"learning_rate": 2.858395300207376e-05,
"loss": 1.7752,
"step": 14694
},
{
"epoch": 0.37468908439728493,
"grad_norm": 1.4071861505508423,
"learning_rate": 2.8500817981817607e-05,
"loss": 1.7894,
"step": 14725
},
{
"epoch": 0.37547790352233185,
"grad_norm": 1.533968448638916,
"learning_rate": 2.8417643455306336e-05,
"loss": 1.7815,
"step": 14756
},
{
"epoch": 0.37626672264737876,
"grad_norm": 1.4698586463928223,
"learning_rate": 2.8334430361153185e-05,
"loss": 1.7693,
"step": 14787
},
{
"epoch": 0.3770555417724257,
"grad_norm": 1.4075795412063599,
"learning_rate": 2.8251179638406612e-05,
"loss": 1.7701,
"step": 14818
},
{
"epoch": 0.3778443608974726,
"grad_norm": 1.4033679962158203,
"learning_rate": 2.8167892226539704e-05,
"loss": 1.7728,
"step": 14849
},
{
"epoch": 0.3786331800225195,
"grad_norm": 1.4572257995605469,
"learning_rate": 2.8084569065439588e-05,
"loss": 1.8019,
"step": 14880
},
{
"epoch": 0.3794219991475664,
"grad_norm": 1.5333317518234253,
"learning_rate": 2.8001211095396807e-05,
"loss": 1.7979,
"step": 14911
},
{
"epoch": 0.3802108182726133,
"grad_norm": 1.4421522617340088,
"learning_rate": 2.791781925709473e-05,
"loss": 1.7768,
"step": 14942
},
{
"epoch": 0.38099963739766024,
"grad_norm": 1.5021952390670776,
"learning_rate": 2.7834394491598908e-05,
"loss": 1.7758,
"step": 14973
},
{
"epoch": 0.38178845652270715,
"grad_norm": 1.462990641593933,
"learning_rate": 2.7750937740346485e-05,
"loss": 1.757,
"step": 15004
},
{
"epoch": 0.38257727564775407,
"grad_norm": 1.4034866094589233,
"learning_rate": 2.7667449945135564e-05,
"loss": 1.7658,
"step": 15035
},
{
"epoch": 0.383366094772801,
"grad_norm": 1.5529896020889282,
"learning_rate": 2.7583932048114557e-05,
"loss": 1.787,
"step": 15066
},
{
"epoch": 0.38415491389784795,
"grad_norm": 1.3766757249832153,
"learning_rate": 2.7500384991771587e-05,
"loss": 1.7857,
"step": 15097
},
{
"epoch": 0.38494373302289486,
"grad_norm": 1.3775665760040283,
"learning_rate": 2.7416809718923825e-05,
"loss": 1.7961,
"step": 15128
},
{
"epoch": 0.3857325521479418,
"grad_norm": 1.4085272550582886,
"learning_rate": 2.7333207172706864e-05,
"loss": 1.7818,
"step": 15159
},
{
"epoch": 0.3865213712729887,
"grad_norm": 1.441758394241333,
"learning_rate": 2.7249578296564088e-05,
"loss": 1.7746,
"step": 15190
},
{
"epoch": 0.3873101903980356,
"grad_norm": 1.4011828899383545,
"learning_rate": 2.7165924034235973e-05,
"loss": 1.7704,
"step": 15221
},
{
"epoch": 0.3880990095230825,
"grad_norm": 1.4673304557800293,
"learning_rate": 2.708224532974953e-05,
"loss": 1.7863,
"step": 15252
},
{
"epoch": 0.38888782864812943,
"grad_norm": 1.4282735586166382,
"learning_rate": 2.6998543127407538e-05,
"loss": 1.76,
"step": 15283
},
{
"epoch": 0.38967664777317634,
"grad_norm": 1.3983831405639648,
"learning_rate": 2.6914818371777988e-05,
"loss": 1.7803,
"step": 15314
},
{
"epoch": 0.39046546689822326,
"grad_norm": 1.5473729372024536,
"learning_rate": 2.6831072007683373e-05,
"loss": 1.7787,
"step": 15345
},
{
"epoch": 0.39125428602327017,
"grad_norm": 1.565489649772644,
"learning_rate": 2.6747304980190018e-05,
"loss": 1.755,
"step": 15376
},
{
"epoch": 0.3920431051483171,
"grad_norm": 1.4918326139450073,
"learning_rate": 2.6663518234597453e-05,
"loss": 1.8007,
"step": 15407
},
{
"epoch": 0.392831924273364,
"grad_norm": 1.5468804836273193,
"learning_rate": 2.6579712716427696e-05,
"loss": 1.7574,
"step": 15438
},
{
"epoch": 0.3936207433984109,
"grad_norm": 1.4871866703033447,
"learning_rate": 2.6495889371414652e-05,
"loss": 1.7757,
"step": 15469
},
{
"epoch": 0.3944095625234578,
"grad_norm": 1.5485950708389282,
"learning_rate": 2.6412049145493367e-05,
"loss": 1.79,
"step": 15500
},
{
"epoch": 0.39519838164850474,
"grad_norm": 1.5302681922912598,
"learning_rate": 2.632819298478939e-05,
"loss": 1.779,
"step": 15531
},
{
"epoch": 0.39598720077355165,
"grad_norm": 1.5713484287261963,
"learning_rate": 2.6244321835608105e-05,
"loss": 1.7526,
"step": 15562
},
{
"epoch": 0.39677601989859856,
"grad_norm": 1.4450056552886963,
"learning_rate": 2.6160436644424024e-05,
"loss": 1.7896,
"step": 15593
},
{
"epoch": 0.3975648390236455,
"grad_norm": 1.5404566526412964,
"learning_rate": 2.6076538357870133e-05,
"loss": 1.7612,
"step": 15624
},
{
"epoch": 0.3983536581486924,
"grad_norm": 1.5850070714950562,
"learning_rate": 2.5992627922727196e-05,
"loss": 1.7588,
"step": 15655
},
{
"epoch": 0.3991424772737393,
"grad_norm": 1.4891109466552734,
"learning_rate": 2.5908706285913066e-05,
"loss": 1.768,
"step": 15686
},
{
"epoch": 0.3999312963987862,
"grad_norm": 1.4907901287078857,
"learning_rate": 2.5824774394472008e-05,
"loss": 1.7672,
"step": 15717
},
{
"epoch": 0.40072011552383313,
"grad_norm": 1.418935775756836,
"learning_rate": 2.5740833195563996e-05,
"loss": 1.7812,
"step": 15748
},
{
"epoch": 0.40150893464888004,
"grad_norm": 1.4996947050094604,
"learning_rate": 2.5656883636454067e-05,
"loss": 1.7833,
"step": 15779
},
{
"epoch": 0.40229775377392696,
"grad_norm": 1.5073673725128174,
"learning_rate": 2.557292666450159e-05,
"loss": 1.7768,
"step": 15810
},
{
"epoch": 0.40308657289897387,
"grad_norm": 1.4004729986190796,
"learning_rate": 2.5488963227149566e-05,
"loss": 1.7688,
"step": 15841
},
{
"epoch": 0.40387539202402084,
"grad_norm": 1.4226566553115845,
"learning_rate": 2.5404994271913983e-05,
"loss": 1.7758,
"step": 15872
},
{
"epoch": 0.40466421114906775,
"grad_norm": 1.3709113597869873,
"learning_rate": 2.5321020746373085e-05,
"loss": 1.7664,
"step": 15903
},
{
"epoch": 0.40545303027411467,
"grad_norm": 1.3796721696853638,
"learning_rate": 2.52370435981567e-05,
"loss": 1.7584,
"step": 15934
},
{
"epoch": 0.4062418493991616,
"grad_norm": 1.455452561378479,
"learning_rate": 2.5153063774935533e-05,
"loss": 1.7745,
"step": 15965
},
{
"epoch": 0.4070306685242085,
"grad_norm": 1.509347677230835,
"learning_rate": 2.506908222441045e-05,
"loss": 1.7763,
"step": 15996
},
{
"epoch": 0.4078194876492554,
"grad_norm": 1.3093947172164917,
"learning_rate": 2.498509989430187e-05,
"loss": 1.7565,
"step": 16027
},
{
"epoch": 0.4086083067743023,
"grad_norm": 1.4819965362548828,
"learning_rate": 2.4901117732338958e-05,
"loss": 1.7678,
"step": 16058
},
{
"epoch": 0.40939712589934923,
"grad_norm": 1.4977960586547852,
"learning_rate": 2.481713668624899e-05,
"loss": 1.7673,
"step": 16089
},
{
"epoch": 0.41018594502439615,
"grad_norm": 1.4152425527572632,
"learning_rate": 2.4733157703746663e-05,
"loss": 1.759,
"step": 16120
},
{
"epoch": 0.41097476414944306,
"grad_norm": 1.3504704236984253,
"learning_rate": 2.4649181732523392e-05,
"loss": 1.773,
"step": 16151
},
{
"epoch": 0.41176358327449,
"grad_norm": 1.3932607173919678,
"learning_rate": 2.4565209720236582e-05,
"loss": 1.7724,
"step": 16182
},
{
"epoch": 0.4125524023995369,
"grad_norm": 1.423255443572998,
"learning_rate": 2.4481242614498975e-05,
"loss": 1.7504,
"step": 16213
},
{
"epoch": 0.4133412215245838,
"grad_norm": 1.5146458148956299,
"learning_rate": 2.439728136286796e-05,
"loss": 1.7572,
"step": 16244
},
{
"epoch": 0.4141300406496307,
"grad_norm": 1.4159959554672241,
"learning_rate": 2.4313326912834852e-05,
"loss": 1.7495,
"step": 16275
},
{
"epoch": 0.41491885977467763,
"grad_norm": 1.4505484104156494,
"learning_rate": 2.4229380211814206e-05,
"loss": 1.7748,
"step": 16306
},
{
"epoch": 0.41570767889972454,
"grad_norm": 1.4519730806350708,
"learning_rate": 2.4145442207133124e-05,
"loss": 1.7635,
"step": 16337
},
{
"epoch": 0.41649649802477146,
"grad_norm": 1.452431082725525,
"learning_rate": 2.406151384602059e-05,
"loss": 1.7624,
"step": 16368
},
{
"epoch": 0.41728531714981837,
"grad_norm": 1.4315119981765747,
"learning_rate": 2.3977596075596747e-05,
"loss": 1.7765,
"step": 16399
},
{
"epoch": 0.4180741362748653,
"grad_norm": 1.4047067165374756,
"learning_rate": 2.3893689842862223e-05,
"loss": 1.755,
"step": 16430
},
{
"epoch": 0.4188629553999122,
"grad_norm": 1.426621913909912,
"learning_rate": 2.3809796094687475e-05,
"loss": 1.7598,
"step": 16461
},
{
"epoch": 0.4196517745249591,
"grad_norm": 1.4108635187149048,
"learning_rate": 2.372591577780202e-05,
"loss": 1.7652,
"step": 16492
},
{
"epoch": 0.420440593650006,
"grad_norm": 1.4988287687301636,
"learning_rate": 2.3642049838783838e-05,
"loss": 1.7763,
"step": 16523
},
{
"epoch": 0.42122941277505294,
"grad_norm": 1.4525630474090576,
"learning_rate": 2.3558199224048666e-05,
"loss": 1.7607,
"step": 16554
},
{
"epoch": 0.42201823190009985,
"grad_norm": 1.512402892112732,
"learning_rate": 2.347436487983929e-05,
"loss": 1.7625,
"step": 16585
},
{
"epoch": 0.42280705102514676,
"grad_norm": 1.4328192472457886,
"learning_rate": 2.3390547752214888e-05,
"loss": 1.7598,
"step": 16616
},
{
"epoch": 0.42359587015019373,
"grad_norm": 1.4110822677612305,
"learning_rate": 2.330674878704035e-05,
"loss": 1.757,
"step": 16647
},
{
"epoch": 0.42438468927524065,
"grad_norm": 1.4538228511810303,
"learning_rate": 2.322296892997561e-05,
"loss": 1.7503,
"step": 16678
},
{
"epoch": 0.42517350840028756,
"grad_norm": 1.4495991468429565,
"learning_rate": 2.313920912646497e-05,
"loss": 1.7593,
"step": 16709
},
{
"epoch": 0.42596232752533447,
"grad_norm": 1.5201659202575684,
"learning_rate": 2.305547032172643e-05,
"loss": 1.7512,
"step": 16740
},
{
"epoch": 0.4267511466503814,
"grad_norm": 1.4683400392532349,
"learning_rate": 2.2971753460741014e-05,
"loss": 1.7792,
"step": 16771
},
{
"epoch": 0.4275399657754283,
"grad_norm": 1.4335435628890991,
"learning_rate": 2.288805948824212e-05,
"loss": 1.7495,
"step": 16802
},
{
"epoch": 0.4283287849004752,
"grad_norm": 1.494997501373291,
"learning_rate": 2.2804389348704858e-05,
"loss": 1.7806,
"step": 16833
},
{
"epoch": 0.4291176040255221,
"grad_norm": 1.5237140655517578,
"learning_rate": 2.2720743986335374e-05,
"loss": 1.7348,
"step": 16864
},
{
"epoch": 0.42990642315056904,
"grad_norm": 1.4462862014770508,
"learning_rate": 2.2637124345060233e-05,
"loss": 1.7663,
"step": 16895
},
{
"epoch": 0.43069524227561595,
"grad_norm": 1.4371618032455444,
"learning_rate": 2.2553531368515695e-05,
"loss": 1.7699,
"step": 16926
},
{
"epoch": 0.43148406140066287,
"grad_norm": 1.4182896614074707,
"learning_rate": 2.2469966000037144e-05,
"loss": 1.7651,
"step": 16957
},
{
"epoch": 0.4322728805257098,
"grad_norm": 1.3784195184707642,
"learning_rate": 2.2386429182648417e-05,
"loss": 1.7596,
"step": 16988
},
{
"epoch": 0.4330616996507567,
"grad_norm": 1.398327350616455,
"learning_rate": 2.230292185905114e-05,
"loss": 1.7588,
"step": 17019
},
{
"epoch": 0.4338505187758036,
"grad_norm": 1.4239211082458496,
"learning_rate": 2.2219444971614116e-05,
"loss": 1.7656,
"step": 17050
},
{
"epoch": 0.4346393379008505,
"grad_norm": 1.4027754068374634,
"learning_rate": 2.2135999462362655e-05,
"loss": 1.7485,
"step": 17081
},
{
"epoch": 0.43542815702589743,
"grad_norm": 1.442612648010254,
"learning_rate": 2.2052586272968003e-05,
"loss": 1.7682,
"step": 17112
},
{
"epoch": 0.43621697615094435,
"grad_norm": 1.3537038564682007,
"learning_rate": 2.196920634473666e-05,
"loss": 1.7511,
"step": 17143
},
{
"epoch": 0.43700579527599126,
"grad_norm": 1.3696125745773315,
"learning_rate": 2.1885860618599787e-05,
"loss": 1.767,
"step": 17174
},
{
"epoch": 0.4377946144010382,
"grad_norm": 1.5365840196609497,
"learning_rate": 2.1802550035102577e-05,
"loss": 1.7527,
"step": 17205
},
{
"epoch": 0.4385834335260851,
"grad_norm": 1.4375520944595337,
"learning_rate": 2.171927553439363e-05,
"loss": 1.7577,
"step": 17236
},
{
"epoch": 0.439372252651132,
"grad_norm": 1.4054752588272095,
"learning_rate": 2.1636038056214376e-05,
"loss": 1.7479,
"step": 17267
},
{
"epoch": 0.4401610717761789,
"grad_norm": 1.4836634397506714,
"learning_rate": 2.155283853988844e-05,
"loss": 1.7463,
"step": 17298
},
{
"epoch": 0.4409498909012258,
"grad_norm": 1.4966789484024048,
"learning_rate": 2.146967792431106e-05,
"loss": 1.7539,
"step": 17329
},
{
"epoch": 0.44173871002627274,
"grad_norm": 1.3743985891342163,
"learning_rate": 2.138655714793849e-05,
"loss": 1.7501,
"step": 17360
},
{
"epoch": 0.44252752915131965,
"grad_norm": 1.4786440134048462,
"learning_rate": 2.1303477148777367e-05,
"loss": 1.7438,
"step": 17391
},
{
"epoch": 0.4433163482763666,
"grad_norm": 1.3931723833084106,
"learning_rate": 2.122043886437421e-05,
"loss": 1.7569,
"step": 17422
},
{
"epoch": 0.44410516740141354,
"grad_norm": 1.457221508026123,
"learning_rate": 2.1137443231804765e-05,
"loss": 1.7459,
"step": 17453
},
{
"epoch": 0.44489398652646045,
"grad_norm": 1.4700186252593994,
"learning_rate": 2.105449118766347e-05,
"loss": 1.749,
"step": 17484
},
{
"epoch": 0.44568280565150736,
"grad_norm": 1.4787609577178955,
"learning_rate": 2.097158366805287e-05,
"loss": 1.7433,
"step": 17515
},
{
"epoch": 0.4464716247765543,
"grad_norm": 1.435116171836853,
"learning_rate": 2.0888721608573047e-05,
"loss": 1.7492,
"step": 17546
},
{
"epoch": 0.4472604439016012,
"grad_norm": 1.3931212425231934,
"learning_rate": 2.0805905944311087e-05,
"loss": 1.7698,
"step": 17577
},
{
"epoch": 0.4480492630266481,
"grad_norm": 1.4713780879974365,
"learning_rate": 2.0723137609830497e-05,
"loss": 1.7599,
"step": 17608
},
{
"epoch": 0.448838082151695,
"grad_norm": 1.3709975481033325,
"learning_rate": 2.0640417539160686e-05,
"loss": 1.7615,
"step": 17639
},
{
"epoch": 0.44962690127674193,
"grad_norm": 1.4991896152496338,
"learning_rate": 2.0557746665786427e-05,
"loss": 1.7541,
"step": 17670
},
{
"epoch": 0.45041572040178884,
"grad_norm": 1.4068297147750854,
"learning_rate": 2.0475125922637256e-05,
"loss": 1.7383,
"step": 17701
},
{
"epoch": 0.45120453952683576,
"grad_norm": 1.4071435928344727,
"learning_rate": 2.0392556242077047e-05,
"loss": 1.754,
"step": 17732
},
{
"epoch": 0.45199335865188267,
"grad_norm": 1.5196751356124878,
"learning_rate": 2.031003855589343e-05,
"loss": 1.7406,
"step": 17763
},
{
"epoch": 0.4527821777769296,
"grad_norm": 1.4674859046936035,
"learning_rate": 2.022757379528727e-05,
"loss": 1.7496,
"step": 17794
},
{
"epoch": 0.4535709969019765,
"grad_norm": 1.377008318901062,
"learning_rate": 2.0145162890862184e-05,
"loss": 1.7573,
"step": 17825
},
{
"epoch": 0.4543598160270234,
"grad_norm": 1.3753769397735596,
"learning_rate": 2.0062806772614022e-05,
"loss": 1.7312,
"step": 17856
},
{
"epoch": 0.4551486351520703,
"grad_norm": 1.3770841360092163,
"learning_rate": 1.9980506369920392e-05,
"loss": 1.7378,
"step": 17887
},
{
"epoch": 0.45593745427711724,
"grad_norm": 1.644900918006897,
"learning_rate": 1.989826261153015e-05,
"loss": 1.7423,
"step": 17918
},
{
"epoch": 0.45672627340216415,
"grad_norm": 1.423464059829712,
"learning_rate": 1.9816076425552923e-05,
"loss": 1.745,
"step": 17949
},
{
"epoch": 0.45751509252721106,
"grad_norm": 1.4013458490371704,
"learning_rate": 1.9733948739448676e-05,
"loss": 1.7444,
"step": 17980
},
{
"epoch": 0.458303911652258,
"grad_norm": 1.4134126901626587,
"learning_rate": 1.9651880480017155e-05,
"loss": 1.7318,
"step": 18011
},
{
"epoch": 0.4590927307773049,
"grad_norm": 1.389404535293579,
"learning_rate": 1.9569872573387516e-05,
"loss": 1.7531,
"step": 18042
},
{
"epoch": 0.4598815499023518,
"grad_norm": 1.4825111627578735,
"learning_rate": 1.9487925945007854e-05,
"loss": 1.7178,
"step": 18073
},
{
"epoch": 0.4606703690273987,
"grad_norm": 1.335856556892395,
"learning_rate": 1.9406041519634726e-05,
"loss": 1.7569,
"step": 18104
},
{
"epoch": 0.46145918815244563,
"grad_norm": 1.5451414585113525,
"learning_rate": 1.932422022132275e-05,
"loss": 1.7608,
"step": 18135
},
{
"epoch": 0.4622480072774926,
"grad_norm": 1.458856225013733,
"learning_rate": 1.924246297341414e-05,
"loss": 1.7381,
"step": 18166
},
{
"epoch": 0.4630368264025395,
"grad_norm": 1.5138990879058838,
"learning_rate": 1.9160770698528338e-05,
"loss": 1.7505,
"step": 18197
},
{
"epoch": 0.4638256455275864,
"grad_norm": 1.433817744255066,
"learning_rate": 1.907914431855156e-05,
"loss": 1.7406,
"step": 18228
},
{
"epoch": 0.46461446465263334,
"grad_norm": 1.3636925220489502,
"learning_rate": 1.8997584754626412e-05,
"loss": 1.7533,
"step": 18259
},
{
"epoch": 0.46540328377768025,
"grad_norm": 1.4643160104751587,
"learning_rate": 1.8916092927141486e-05,
"loss": 1.7329,
"step": 18290
},
{
"epoch": 0.46619210290272717,
"grad_norm": 1.3883280754089355,
"learning_rate": 1.883466975572098e-05,
"loss": 1.7386,
"step": 18321
},
{
"epoch": 0.4669809220277741,
"grad_norm": 1.4294878244400024,
"learning_rate": 1.8753316159214312e-05,
"loss": 1.7553,
"step": 18352
},
{
"epoch": 0.467769741152821,
"grad_norm": 1.4422011375427246,
"learning_rate": 1.8672033055685766e-05,
"loss": 1.7333,
"step": 18383
},
{
"epoch": 0.4685585602778679,
"grad_norm": 1.4707412719726562,
"learning_rate": 1.8590821362404116e-05,
"loss": 1.7463,
"step": 18414
},
{
"epoch": 0.4693473794029148,
"grad_norm": 1.4446028470993042,
"learning_rate": 1.8509681995832294e-05,
"loss": 1.7262,
"step": 18445
},
{
"epoch": 0.47013619852796174,
"grad_norm": 1.3790693283081055,
"learning_rate": 1.8428615871617004e-05,
"loss": 1.7442,
"step": 18476
},
{
"epoch": 0.47092501765300865,
"grad_norm": 1.4067668914794922,
"learning_rate": 1.8347623904578448e-05,
"loss": 1.731,
"step": 18507
},
{
"epoch": 0.47171383677805556,
"grad_norm": 1.496756672859192,
"learning_rate": 1.8266707008699975e-05,
"loss": 1.7546,
"step": 18538
},
{
"epoch": 0.4725026559031025,
"grad_norm": 1.4508312940597534,
"learning_rate": 1.818586609711774e-05,
"loss": 1.748,
"step": 18569
},
{
"epoch": 0.4732914750281494,
"grad_norm": 1.4680043458938599,
"learning_rate": 1.8105102082110462e-05,
"loss": 1.7334,
"step": 18600
},
{
"epoch": 0.4740802941531963,
"grad_norm": 1.512750267982483,
"learning_rate": 1.8024415875089058e-05,
"loss": 1.7437,
"step": 18631
},
{
"epoch": 0.4748691132782432,
"grad_norm": 1.4424457550048828,
"learning_rate": 1.7943808386586407e-05,
"loss": 1.7454,
"step": 18662
},
{
"epoch": 0.47565793240329013,
"grad_norm": 1.47055983543396,
"learning_rate": 1.7863280526247073e-05,
"loss": 1.7132,
"step": 18693
},
{
"epoch": 0.47644675152833704,
"grad_norm": 1.3706848621368408,
"learning_rate": 1.7782833202817003e-05,
"loss": 1.7329,
"step": 18724
},
{
"epoch": 0.47723557065338396,
"grad_norm": 1.3911017179489136,
"learning_rate": 1.7702467324133327e-05,
"loss": 1.7322,
"step": 18755
},
{
"epoch": 0.47802438977843087,
"grad_norm": 1.3935508728027344,
"learning_rate": 1.7622183797114042e-05,
"loss": 1.7463,
"step": 18786
},
{
"epoch": 0.4788132089034778,
"grad_norm": 1.3628978729248047,
"learning_rate": 1.7541983527747838e-05,
"loss": 1.7529,
"step": 18817
},
{
"epoch": 0.4796020280285247,
"grad_norm": 1.3993347883224487,
"learning_rate": 1.746186742108387e-05,
"loss": 1.7517,
"step": 18848
},
{
"epoch": 0.4803908471535716,
"grad_norm": 1.4427608251571655,
"learning_rate": 1.73818363812215e-05,
"loss": 1.7499,
"step": 18879
},
{
"epoch": 0.4811796662786185,
"grad_norm": 1.4312776327133179,
"learning_rate": 1.7301891311300153e-05,
"loss": 1.7471,
"step": 18910
},
{
"epoch": 0.4819684854036655,
"grad_norm": 1.373559832572937,
"learning_rate": 1.7222033113489055e-05,
"loss": 1.7555,
"step": 18941
},
{
"epoch": 0.4827573045287124,
"grad_norm": 1.383086085319519,
"learning_rate": 1.7142262688977127e-05,
"loss": 1.7446,
"step": 18972
},
{
"epoch": 0.4835461236537593,
"grad_norm": 1.459486722946167,
"learning_rate": 1.7062580937962764e-05,
"loss": 1.7523,
"step": 19003
},
{
"epoch": 0.48433494277880623,
"grad_norm": 1.5249037742614746,
"learning_rate": 1.698298875964369e-05,
"loss": 1.7249,
"step": 19034
},
{
"epoch": 0.48512376190385315,
"grad_norm": 1.431281566619873,
"learning_rate": 1.690348705220684e-05,
"loss": 1.7133,
"step": 19065
},
{
"epoch": 0.48591258102890006,
"grad_norm": 1.4861342906951904,
"learning_rate": 1.6824076712818156e-05,
"loss": 1.7377,
"step": 19096
},
{
"epoch": 0.486701400153947,
"grad_norm": 1.3854913711547852,
"learning_rate": 1.6744758637612533e-05,
"loss": 1.7292,
"step": 19127
},
{
"epoch": 0.4874902192789939,
"grad_norm": 1.482332468032837,
"learning_rate": 1.6665533721683664e-05,
"loss": 1.7505,
"step": 19158
},
{
"epoch": 0.4882790384040408,
"grad_norm": 1.3565430641174316,
"learning_rate": 1.6586402859073974e-05,
"loss": 1.742,
"step": 19189
},
{
"epoch": 0.4890678575290877,
"grad_norm": 1.445395588874817,
"learning_rate": 1.6507366942764463e-05,
"loss": 1.7387,
"step": 19220
},
{
"epoch": 0.4898566766541346,
"grad_norm": 1.4928544759750366,
"learning_rate": 1.6428426864664732e-05,
"loss": 1.762,
"step": 19251
},
{
"epoch": 0.49064549577918154,
"grad_norm": 1.38858962059021,
"learning_rate": 1.6349583515602816e-05,
"loss": 1.7486,
"step": 19282
},
{
"epoch": 0.49143431490422845,
"grad_norm": 1.3937194347381592,
"learning_rate": 1.6270837785315208e-05,
"loss": 1.7542,
"step": 19313
},
{
"epoch": 0.49222313402927537,
"grad_norm": 1.501042127609253,
"learning_rate": 1.619219056243676e-05,
"loss": 1.7274,
"step": 19344
},
{
"epoch": 0.4930119531543223,
"grad_norm": 1.3143610954284668,
"learning_rate": 1.6113642734490698e-05,
"loss": 1.7137,
"step": 19375
},
{
"epoch": 0.4938007722793692,
"grad_norm": 1.4225116968154907,
"learning_rate": 1.6035195187878577e-05,
"loss": 1.7414,
"step": 19406
},
{
"epoch": 0.4945895914044161,
"grad_norm": 1.4575517177581787,
"learning_rate": 1.5956848807870305e-05,
"loss": 1.7044,
"step": 19437
},
{
"epoch": 0.495378410529463,
"grad_norm": 1.4163532257080078,
"learning_rate": 1.587860447859413e-05,
"loss": 1.7365,
"step": 19468
},
{
"epoch": 0.49616722965450993,
"grad_norm": 1.504955530166626,
"learning_rate": 1.5800463083026686e-05,
"loss": 1.759,
"step": 19499
},
{
"epoch": 0.49695604877955685,
"grad_norm": 1.4385664463043213,
"learning_rate": 1.572242550298298e-05,
"loss": 1.7193,
"step": 19530
},
{
"epoch": 0.49774486790460376,
"grad_norm": 1.384371280670166,
"learning_rate": 1.56444926191065e-05,
"loss": 1.7224,
"step": 19561
},
{
"epoch": 0.4985336870296507,
"grad_norm": 1.392520785331726,
"learning_rate": 1.5566665310859257e-05,
"loss": 1.7246,
"step": 19592
},
{
"epoch": 0.4993225061546976,
"grad_norm": 1.421629786491394,
"learning_rate": 1.5488944456511846e-05,
"loss": 1.7314,
"step": 19623
},
{
"epoch": 0.5001113252797446,
"grad_norm": 1.405013918876648,
"learning_rate": 1.5411330933133546e-05,
"loss": 1.7329,
"step": 19654
},
{
"epoch": 0.5009001444047915,
"grad_norm": 1.372490406036377,
"learning_rate": 1.533382561658241e-05,
"loss": 1.7253,
"step": 19685
},
{
"epoch": 0.5016889635298384,
"grad_norm": 1.4861042499542236,
"learning_rate": 1.525642938149541e-05,
"loss": 1.7317,
"step": 19716
},
{
"epoch": 0.5024777826548853,
"grad_norm": 1.4108079671859741,
"learning_rate": 1.5179143101278536e-05,
"loss": 1.7391,
"step": 19747
},
{
"epoch": 0.5032666017799322,
"grad_norm": 1.4616518020629883,
"learning_rate": 1.5101967648096955e-05,
"loss": 1.7129,
"step": 19778
},
{
"epoch": 0.5040554209049791,
"grad_norm": 1.3660775423049927,
"learning_rate": 1.5024903892865172e-05,
"loss": 1.7149,
"step": 19809
},
{
"epoch": 0.504844240030026,
"grad_norm": 1.4286696910858154,
"learning_rate": 1.4947952705237184e-05,
"loss": 1.6982,
"step": 19840
},
{
"epoch": 0.505633059155073,
"grad_norm": 1.4002240896224976,
"learning_rate": 1.4871114953596682e-05,
"loss": 1.7319,
"step": 19871
},
{
"epoch": 0.5064218782801199,
"grad_norm": 1.4784040451049805,
"learning_rate": 1.4794391505047256e-05,
"loss": 1.7193,
"step": 19902
},
{
"epoch": 0.5072106974051668,
"grad_norm": 1.4451029300689697,
"learning_rate": 1.4717783225402596e-05,
"loss": 1.7225,
"step": 19933
},
{
"epoch": 0.5079995165302137,
"grad_norm": 1.438502550125122,
"learning_rate": 1.4641290979176735e-05,
"loss": 1.7186,
"step": 19964
},
{
"epoch": 0.5087883356552606,
"grad_norm": 1.4443246126174927,
"learning_rate": 1.4564915629574246e-05,
"loss": 1.7242,
"step": 19995
},
{
"epoch": 0.5095771547803075,
"grad_norm": 1.5206542015075684,
"learning_rate": 1.4488658038480601e-05,
"loss": 1.7406,
"step": 20026
},
{
"epoch": 0.5103659739053544,
"grad_norm": 1.4452012777328491,
"learning_rate": 1.4412519066452323e-05,
"loss": 1.7218,
"step": 20057
},
{
"epoch": 0.5111547930304013,
"grad_norm": 1.4169068336486816,
"learning_rate": 1.4336499572707373e-05,
"loss": 1.7365,
"step": 20088
},
{
"epoch": 0.5119436121554483,
"grad_norm": 1.475844383239746,
"learning_rate": 1.4260600415115433e-05,
"loss": 1.7264,
"step": 20119
},
{
"epoch": 0.5127324312804952,
"grad_norm": 1.4148321151733398,
"learning_rate": 1.4184822450188137e-05,
"loss": 1.7348,
"step": 20150
},
{
"epoch": 0.5135212504055421,
"grad_norm": 1.4532842636108398,
"learning_rate": 1.410916653306954e-05,
"loss": 1.7021,
"step": 20181
},
{
"epoch": 0.514310069530589,
"grad_norm": 1.4390312433242798,
"learning_rate": 1.403363351752639e-05,
"loss": 1.7205,
"step": 20212
},
{
"epoch": 0.5150988886556359,
"grad_norm": 1.4490697383880615,
"learning_rate": 1.3958224255938485e-05,
"loss": 1.7235,
"step": 20243
},
{
"epoch": 0.5158877077806828,
"grad_norm": 1.4487396478652954,
"learning_rate": 1.388293959928911e-05,
"loss": 1.7325,
"step": 20274
},
{
"epoch": 0.5166765269057297,
"grad_norm": 1.3987274169921875,
"learning_rate": 1.3807780397155379e-05,
"loss": 1.707,
"step": 20305
},
{
"epoch": 0.5174653460307767,
"grad_norm": 1.4041749238967896,
"learning_rate": 1.3732747497698655e-05,
"loss": 1.7179,
"step": 20336
},
{
"epoch": 0.5182541651558236,
"grad_norm": 1.442674994468689,
"learning_rate": 1.3657841747655038e-05,
"loss": 1.7382,
"step": 20367
},
{
"epoch": 0.5190429842808705,
"grad_norm": 1.4303447008132935,
"learning_rate": 1.3583063992325706e-05,
"loss": 1.7375,
"step": 20398
},
{
"epoch": 0.5198318034059174,
"grad_norm": 1.4175372123718262,
"learning_rate": 1.3508415075567496e-05,
"loss": 1.7126,
"step": 20429
},
{
"epoch": 0.5206206225309643,
"grad_norm": 1.4764102697372437,
"learning_rate": 1.343389583978327e-05,
"loss": 1.7372,
"step": 20460
},
{
"epoch": 0.5214094416560112,
"grad_norm": 1.4597842693328857,
"learning_rate": 1.3359507125912468e-05,
"loss": 1.7247,
"step": 20491
},
{
"epoch": 0.5221982607810581,
"grad_norm": 1.4972703456878662,
"learning_rate": 1.3285249773421627e-05,
"loss": 1.7158,
"step": 20522
},
{
"epoch": 0.522987079906105,
"grad_norm": 1.4102123975753784,
"learning_rate": 1.3211124620294884e-05,
"loss": 1.74,
"step": 20553
},
{
"epoch": 0.523775899031152,
"grad_norm": 1.3836309909820557,
"learning_rate": 1.313713250302451e-05,
"loss": 1.733,
"step": 20584
},
{
"epoch": 0.5245647181561989,
"grad_norm": 1.4065951108932495,
"learning_rate": 1.3063274256601479e-05,
"loss": 1.7291,
"step": 20615
},
{
"epoch": 0.5253535372812458,
"grad_norm": 1.4294134378433228,
"learning_rate": 1.2989550714506086e-05,
"loss": 1.7196,
"step": 20646
},
{
"epoch": 0.5261423564062927,
"grad_norm": 1.438848853111267,
"learning_rate": 1.291596270869846e-05,
"loss": 1.7294,
"step": 20677
},
{
"epoch": 0.5269311755313396,
"grad_norm": 1.3648425340652466,
"learning_rate": 1.284251106960927e-05,
"loss": 1.725,
"step": 20708
},
{
"epoch": 0.5277199946563865,
"grad_norm": 1.4666975736618042,
"learning_rate": 1.2769196626130263e-05,
"loss": 1.7272,
"step": 20739
},
{
"epoch": 0.5285088137814334,
"grad_norm": 1.4472864866256714,
"learning_rate": 1.2696020205604969e-05,
"loss": 1.7216,
"step": 20770
},
{
"epoch": 0.5292976329064804,
"grad_norm": 1.4326000213623047,
"learning_rate": 1.2622982633819359e-05,
"loss": 1.7263,
"step": 20801
},
{
"epoch": 0.5300864520315273,
"grad_norm": 1.468807578086853,
"learning_rate": 1.2550084734992484e-05,
"loss": 1.7366,
"step": 20832
},
{
"epoch": 0.5308752711565742,
"grad_norm": 1.3874242305755615,
"learning_rate": 1.247732733176724e-05,
"loss": 1.7235,
"step": 20863
},
{
"epoch": 0.5316640902816212,
"grad_norm": 1.4644588232040405,
"learning_rate": 1.2404711245201044e-05,
"loss": 1.7363,
"step": 20894
},
{
"epoch": 0.5324529094066681,
"grad_norm": 1.4458835124969482,
"learning_rate": 1.2332237294756535e-05,
"loss": 1.7062,
"step": 20925
},
{
"epoch": 0.533241728531715,
"grad_norm": 1.4956963062286377,
"learning_rate": 1.225990629829241e-05,
"loss": 1.7244,
"step": 20956
},
{
"epoch": 0.534030547656762,
"grad_norm": 1.4594619274139404,
"learning_rate": 1.2187719072054136e-05,
"loss": 1.7074,
"step": 20987
},
{
"epoch": 0.5348193667818089,
"grad_norm": 1.4499660730361938,
"learning_rate": 1.2115676430664735e-05,
"loss": 1.7154,
"step": 21018
},
{
"epoch": 0.5356081859068558,
"grad_norm": 1.5303255319595337,
"learning_rate": 1.2043779187115647e-05,
"loss": 1.7284,
"step": 21049
},
{
"epoch": 0.5363970050319027,
"grad_norm": 1.3913129568099976,
"learning_rate": 1.1972028152757476e-05,
"loss": 1.7328,
"step": 21080
},
{
"epoch": 0.5371858241569496,
"grad_norm": 1.4718728065490723,
"learning_rate": 1.1900424137290889e-05,
"loss": 1.7208,
"step": 21111
},
{
"epoch": 0.5379746432819965,
"grad_norm": 1.3919767141342163,
"learning_rate": 1.1828967948757482e-05,
"loss": 1.7143,
"step": 21142
},
{
"epoch": 0.5387634624070434,
"grad_norm": 1.4659541845321655,
"learning_rate": 1.175766039353062e-05,
"loss": 1.7111,
"step": 21173
},
{
"epoch": 0.5395522815320903,
"grad_norm": 1.4828646183013916,
"learning_rate": 1.1686502276306382e-05,
"loss": 1.7113,
"step": 21204
},
{
"epoch": 0.5403411006571373,
"grad_norm": 1.458970308303833,
"learning_rate": 1.1615494400094445e-05,
"loss": 1.7199,
"step": 21235
},
{
"epoch": 0.5411299197821842,
"grad_norm": 1.4522119760513306,
"learning_rate": 1.1544637566209029e-05,
"loss": 1.7052,
"step": 21266
},
{
"epoch": 0.5419187389072311,
"grad_norm": 1.4456357955932617,
"learning_rate": 1.1473932574259886e-05,
"loss": 1.7201,
"step": 21297
},
{
"epoch": 0.542707558032278,
"grad_norm": 1.4089595079421997,
"learning_rate": 1.1403380222143247e-05,
"loss": 1.705,
"step": 21328
},
{
"epoch": 0.5434963771573249,
"grad_norm": 1.4137688875198364,
"learning_rate": 1.1332981306032808e-05,
"loss": 1.7135,
"step": 21359
},
{
"epoch": 0.5442851962823718,
"grad_norm": 1.4155645370483398,
"learning_rate": 1.1262736620370762e-05,
"loss": 1.7159,
"step": 21390
},
{
"epoch": 0.5450740154074187,
"grad_norm": 1.5558688640594482,
"learning_rate": 1.1192646957858854e-05,
"loss": 1.7283,
"step": 21421
},
{
"epoch": 0.5458628345324656,
"grad_norm": 1.5027565956115723,
"learning_rate": 1.1122713109449381e-05,
"loss": 1.7135,
"step": 21452
},
{
"epoch": 0.5466516536575126,
"grad_norm": 1.499029517173767,
"learning_rate": 1.105293586433634e-05,
"loss": 1.7208,
"step": 21483
},
{
"epoch": 0.5474404727825595,
"grad_norm": 1.4107885360717773,
"learning_rate": 1.0983316009946446e-05,
"loss": 1.7199,
"step": 21514
},
{
"epoch": 0.5482292919076064,
"grad_norm": 1.3750280141830444,
"learning_rate": 1.0913854331930282e-05,
"loss": 1.6973,
"step": 21545
},
{
"epoch": 0.5490181110326533,
"grad_norm": 1.4192049503326416,
"learning_rate": 1.0844551614153456e-05,
"loss": 1.7034,
"step": 21576
},
{
"epoch": 0.5498069301577002,
"grad_norm": 1.422545075416565,
"learning_rate": 1.0775408638687725e-05,
"loss": 1.7168,
"step": 21607
},
{
"epoch": 0.5505957492827471,
"grad_norm": 1.4749089479446411,
"learning_rate": 1.0706426185802165e-05,
"loss": 1.7169,
"step": 21638
},
{
"epoch": 0.551384568407794,
"grad_norm": 1.4102238416671753,
"learning_rate": 1.0637605033954371e-05,
"loss": 1.7195,
"step": 21669
},
{
"epoch": 0.552173387532841,
"grad_norm": 1.4288344383239746,
"learning_rate": 1.05689459597817e-05,
"loss": 1.704,
"step": 21700
},
{
"epoch": 0.5529622066578879,
"grad_norm": 1.4801214933395386,
"learning_rate": 1.050044973809246e-05,
"loss": 1.7011,
"step": 21731
},
{
"epoch": 0.5537510257829348,
"grad_norm": 1.4600056409835815,
"learning_rate": 1.043211714185722e-05,
"loss": 1.714,
"step": 21762
},
{
"epoch": 0.5545398449079817,
"grad_norm": 1.405286192893982,
"learning_rate": 1.036394894220003e-05,
"loss": 1.7098,
"step": 21793
},
{
"epoch": 0.5553286640330286,
"grad_norm": 1.4454749822616577,
"learning_rate": 1.0295945908389751e-05,
"loss": 1.7152,
"step": 21824
},
{
"epoch": 0.5561174831580755,
"grad_norm": 1.5381968021392822,
"learning_rate": 1.0228108807831393e-05,
"loss": 1.7199,
"step": 21855
},
{
"epoch": 0.5569063022831224,
"grad_norm": 1.426140546798706,
"learning_rate": 1.01604384060574e-05,
"loss": 1.7317,
"step": 21886
},
{
"epoch": 0.5576951214081693,
"grad_norm": 1.6093019247055054,
"learning_rate": 1.009293546671907e-05,
"loss": 1.6993,
"step": 21917
},
{
"epoch": 0.5584839405332163,
"grad_norm": 1.344679355621338,
"learning_rate": 1.002560075157791e-05,
"loss": 1.7258,
"step": 21948
},
{
"epoch": 0.5592727596582632,
"grad_norm": 1.3664970397949219,
"learning_rate": 9.958435020496995e-06,
"loss": 1.71,
"step": 21979
},
{
"epoch": 0.5600615787833101,
"grad_norm": 1.457160234451294,
"learning_rate": 9.89143903143249e-06,
"loss": 1.7173,
"step": 22010
},
{
"epoch": 0.560850397908357,
"grad_norm": 1.3795216083526611,
"learning_rate": 9.824613540425038e-06,
"loss": 1.6924,
"step": 22041
},
{
"epoch": 0.5616392170334039,
"grad_norm": 1.3805229663848877,
"learning_rate": 9.757959301591197e-06,
"loss": 1.7149,
"step": 22072
},
{
"epoch": 0.5624280361584508,
"grad_norm": 1.4958610534667969,
"learning_rate": 9.691477067115017e-06,
"loss": 1.7161,
"step": 22103
},
{
"epoch": 0.5632168552834977,
"grad_norm": 1.4804275035858154,
"learning_rate": 9.625167587239467e-06,
"loss": 1.7023,
"step": 22134
},
{
"epoch": 0.5640056744085447,
"grad_norm": 1.3880009651184082,
"learning_rate": 9.559031610258007e-06,
"loss": 1.7312,
"step": 22165
},
{
"epoch": 0.5647944935335916,
"grad_norm": 1.429051160812378,
"learning_rate": 9.493069882506164e-06,
"loss": 1.7102,
"step": 22196
},
{
"epoch": 0.5655833126586385,
"grad_norm": 1.4571672677993774,
"learning_rate": 9.427283148353056e-06,
"loss": 1.7234,
"step": 22227
},
{
"epoch": 0.5663721317836854,
"grad_norm": 1.4143497943878174,
"learning_rate": 9.361672150193052e-06,
"loss": 1.7129,
"step": 22258
},
{
"epoch": 0.5671609509087323,
"grad_norm": 1.4296061992645264,
"learning_rate": 9.29623762843734e-06,
"loss": 1.7291,
"step": 22289
},
{
"epoch": 0.5679497700337792,
"grad_norm": 1.4027940034866333,
"learning_rate": 9.230980321505594e-06,
"loss": 1.7199,
"step": 22320
},
{
"epoch": 0.5687385891588261,
"grad_norm": 1.4574463367462158,
"learning_rate": 9.165900965817668e-06,
"loss": 1.713,
"step": 22351
},
{
"epoch": 0.569527408283873,
"grad_norm": 1.4593865871429443,
"learning_rate": 9.101000295785245e-06,
"loss": 1.7153,
"step": 22382
},
{
"epoch": 0.57031622740892,
"grad_norm": 1.4154292345046997,
"learning_rate": 9.036279043803565e-06,
"loss": 1.7046,
"step": 22413
},
{
"epoch": 0.571105046533967,
"grad_norm": 1.4303706884384155,
"learning_rate": 8.971737940243147e-06,
"loss": 1.6939,
"step": 22444
},
{
"epoch": 0.5718938656590139,
"grad_norm": 1.4045100212097168,
"learning_rate": 8.907377713441592e-06,
"loss": 1.7022,
"step": 22475
},
{
"epoch": 0.5726826847840608,
"grad_norm": 1.4179104566574097,
"learning_rate": 8.843199089695293e-06,
"loss": 1.6987,
"step": 22506
},
{
"epoch": 0.5734715039091077,
"grad_norm": 1.4246447086334229,
"learning_rate": 8.779202793251311e-06,
"loss": 1.7187,
"step": 22537
},
{
"epoch": 0.5742603230341546,
"grad_norm": 1.3932377099990845,
"learning_rate": 8.715389546299149e-06,
"loss": 1.7074,
"step": 22568
},
{
"epoch": 0.5750491421592016,
"grad_norm": 1.3958845138549805,
"learning_rate": 8.651760068962617e-06,
"loss": 1.6959,
"step": 22599
},
{
"epoch": 0.5758379612842485,
"grad_norm": 1.4608432054519653,
"learning_rate": 8.588315079291733e-06,
"loss": 1.6956,
"step": 22630
},
{
"epoch": 0.5766267804092954,
"grad_norm": 1.4236079454421997,
"learning_rate": 8.52505529325457e-06,
"loss": 1.6964,
"step": 22661
},
{
"epoch": 0.5774155995343423,
"grad_norm": 1.4737471342086792,
"learning_rate": 8.461981424729216e-06,
"loss": 1.7316,
"step": 22692
},
{
"epoch": 0.5782044186593892,
"grad_norm": 1.4301661252975464,
"learning_rate": 8.399094185495725e-06,
"loss": 1.6969,
"step": 22723
},
{
"epoch": 0.5789932377844361,
"grad_norm": 1.5357433557510376,
"learning_rate": 8.336394285228017e-06,
"loss": 1.7211,
"step": 22754
},
{
"epoch": 0.579782056909483,
"grad_norm": 1.4420846700668335,
"learning_rate": 8.273882431485952e-06,
"loss": 1.7177,
"step": 22785
},
{
"epoch": 0.58057087603453,
"grad_norm": 1.402849793434143,
"learning_rate": 8.211559329707316e-06,
"loss": 1.7008,
"step": 22816
},
{
"epoch": 0.5813596951595769,
"grad_norm": 1.4084275960922241,
"learning_rate": 8.149425683199823e-06,
"loss": 1.7052,
"step": 22847
},
{
"epoch": 0.5821485142846238,
"grad_norm": 1.406717300415039,
"learning_rate": 8.08748219313325e-06,
"loss": 1.7201,
"step": 22878
},
{
"epoch": 0.5829373334096707,
"grad_norm": 1.9726225137710571,
"learning_rate": 8.025729558531453e-06,
"loss": 1.7197,
"step": 22909
},
{
"epoch": 0.5837261525347176,
"grad_norm": 1.4474542140960693,
"learning_rate": 7.964168476264508e-06,
"loss": 1.7314,
"step": 22940
},
{
"epoch": 0.5845149716597645,
"grad_norm": 1.5046030282974243,
"learning_rate": 7.902799641040884e-06,
"loss": 1.7128,
"step": 22971
},
{
"epoch": 0.5853037907848114,
"grad_norm": 1.4233800172805786,
"learning_rate": 7.841623745399523e-06,
"loss": 1.7026,
"step": 23002
},
{
"epoch": 0.5860926099098583,
"grad_norm": 1.4411020278930664,
"learning_rate": 7.780641479702114e-06,
"loss": 1.7039,
"step": 23033
},
{
"epoch": 0.5868814290349053,
"grad_norm": 1.3648072481155396,
"learning_rate": 7.719853532125227e-06,
"loss": 1.6997,
"step": 23064
},
{
"epoch": 0.5876702481599522,
"grad_norm": 1.3941482305526733,
"learning_rate": 7.65926058865258e-06,
"loss": 1.6947,
"step": 23095
},
{
"epoch": 0.5884590672849991,
"grad_norm": 1.4287201166152954,
"learning_rate": 7.598863333067313e-06,
"loss": 1.7081,
"step": 23126
},
{
"epoch": 0.589247886410046,
"grad_norm": 1.4891555309295654,
"learning_rate": 7.538662446944253e-06,
"loss": 1.6999,
"step": 23157
},
{
"epoch": 0.5900367055350929,
"grad_norm": 1.4390950202941895,
"learning_rate": 7.478658609642211e-06,
"loss": 1.71,
"step": 23188
},
{
"epoch": 0.5908255246601398,
"grad_norm": 1.4771630764007568,
"learning_rate": 7.418852498296327e-06,
"loss": 1.6975,
"step": 23219
},
{
"epoch": 0.5916143437851867,
"grad_norm": 1.4118000268936157,
"learning_rate": 7.359244787810457e-06,
"loss": 1.7028,
"step": 23250
},
{
"epoch": 0.5924031629102336,
"grad_norm": 1.4303267002105713,
"learning_rate": 7.299836150849493e-06,
"loss": 1.7052,
"step": 23281
},
{
"epoch": 0.5931919820352806,
"grad_norm": 1.3951334953308105,
"learning_rate": 7.240627257831847e-06,
"loss": 1.711,
"step": 23312
},
{
"epoch": 0.5939808011603275,
"grad_norm": 1.4434106349945068,
"learning_rate": 7.1816187769218195e-06,
"loss": 1.71,
"step": 23343
},
{
"epoch": 0.5947696202853744,
"grad_norm": 1.4348808526992798,
"learning_rate": 7.1228113740220895e-06,
"loss": 1.7104,
"step": 23374
},
{
"epoch": 0.5955584394104213,
"grad_norm": 1.4280933141708374,
"learning_rate": 7.064205712766226e-06,
"loss": 1.6948,
"step": 23405
},
{
"epoch": 0.5963472585354682,
"grad_norm": 1.4204617738723755,
"learning_rate": 7.005802454511129e-06,
"loss": 1.7016,
"step": 23436
},
{
"epoch": 0.5971360776605151,
"grad_norm": 1.3821487426757812,
"learning_rate": 6.947602258329639e-06,
"loss": 1.6919,
"step": 23467
},
{
"epoch": 0.597924896785562,
"grad_norm": 1.4799888134002686,
"learning_rate": 6.889605781003078e-06,
"loss": 1.7245,
"step": 23498
},
{
"epoch": 0.598713715910609,
"grad_norm": 1.4447741508483887,
"learning_rate": 6.831813677013776e-06,
"loss": 1.7352,
"step": 23529
},
{
"epoch": 0.5995025350356559,
"grad_norm": 1.5367285013198853,
"learning_rate": 6.774226598537792e-06,
"loss": 1.7047,
"step": 23560
},
{
"epoch": 0.6002913541607028,
"grad_norm": 1.4005663394927979,
"learning_rate": 6.716845195437482e-06,
"loss": 1.7021,
"step": 23591
},
{
"epoch": 0.6010801732857497,
"grad_norm": 1.4289170503616333,
"learning_rate": 6.659670115254168e-06,
"loss": 1.7093,
"step": 23622
},
{
"epoch": 0.6018689924107966,
"grad_norm": 1.5853567123413086,
"learning_rate": 6.602702003200872e-06,
"loss": 1.7075,
"step": 23653
},
{
"epoch": 0.6026578115358435,
"grad_norm": 1.580708622932434,
"learning_rate": 6.545941502154992e-06,
"loss": 1.7041,
"step": 23684
},
{
"epoch": 0.6034466306608904,
"grad_norm": 1.477163553237915,
"learning_rate": 6.489389252651057e-06,
"loss": 1.7145,
"step": 23715
},
{
"epoch": 0.6042354497859374,
"grad_norm": 1.428688883781433,
"learning_rate": 6.4330458928735325e-06,
"loss": 1.6906,
"step": 23746
},
{
"epoch": 0.6050242689109843,
"grad_norm": 1.4114421606063843,
"learning_rate": 6.376912058649559e-06,
"loss": 1.7116,
"step": 23777
},
{
"epoch": 0.6058130880360312,
"grad_norm": 1.472838044166565,
"learning_rate": 6.320988383441845e-06,
"loss": 1.6997,
"step": 23808
},
{
"epoch": 0.6066019071610781,
"grad_norm": 1.4391415119171143,
"learning_rate": 6.265275498341452e-06,
"loss": 1.7061,
"step": 23839
},
{
"epoch": 0.607390726286125,
"grad_norm": 1.4201316833496094,
"learning_rate": 6.209774032060714e-06,
"loss": 1.6964,
"step": 23870
},
{
"epoch": 0.6081795454111719,
"grad_norm": 1.3875731229782104,
"learning_rate": 6.1544846109261365e-06,
"loss": 1.7004,
"step": 23901
},
{
"epoch": 0.6089683645362188,
"grad_norm": 1.4405038356781006,
"learning_rate": 6.099407858871342e-06,
"loss": 1.7059,
"step": 23932
},
{
"epoch": 0.6097571836612657,
"grad_norm": 1.4719831943511963,
"learning_rate": 6.044544397429958e-06,
"loss": 1.7063,
"step": 23963
},
{
"epoch": 0.6105460027863128,
"grad_norm": 1.4867260456085205,
"learning_rate": 5.989894845728708e-06,
"loss": 1.7054,
"step": 23994
},
{
"epoch": 0.6113348219113597,
"grad_norm": 1.4767833948135376,
"learning_rate": 5.9354598204803605e-06,
"loss": 1.6951,
"step": 24025
},
{
"epoch": 0.6121236410364066,
"grad_norm": 1.4202378988265991,
"learning_rate": 5.881239935976762e-06,
"loss": 1.7042,
"step": 24056
},
{
"epoch": 0.6129124601614535,
"grad_norm": 1.4658666849136353,
"learning_rate": 5.827235804081954e-06,
"loss": 1.7011,
"step": 24087
},
{
"epoch": 0.6137012792865004,
"grad_norm": 1.437771201133728,
"learning_rate": 5.773448034225221e-06,
"loss": 1.7033,
"step": 24118
},
{
"epoch": 0.6144900984115473,
"grad_norm": 1.4407992362976074,
"learning_rate": 5.719877233394228e-06,
"loss": 1.6841,
"step": 24149
},
{
"epoch": 0.6152789175365942,
"grad_norm": 1.434173822402954,
"learning_rate": 5.666524006128191e-06,
"loss": 1.6893,
"step": 24180
},
{
"epoch": 0.6160677366616412,
"grad_norm": 1.5241893529891968,
"learning_rate": 5.613388954511015e-06,
"loss": 1.707,
"step": 24211
},
{
"epoch": 0.6168565557866881,
"grad_norm": 1.4565976858139038,
"learning_rate": 5.560472678164552e-06,
"loss": 1.695,
"step": 24242
},
{
"epoch": 0.617645374911735,
"grad_norm": 1.458123803138733,
"learning_rate": 5.507775774241775e-06,
"loss": 1.6988,
"step": 24273
},
{
"epoch": 0.6184341940367819,
"grad_norm": 1.4085556268692017,
"learning_rate": 5.4552988374200945e-06,
"loss": 1.6986,
"step": 24304
},
{
"epoch": 0.6192230131618288,
"grad_norm": 1.444799542427063,
"learning_rate": 5.403042459894597e-06,
"loss": 1.7036,
"step": 24335
},
{
"epoch": 0.6200118322868757,
"grad_norm": 1.417597770690918,
"learning_rate": 5.3510072313714135e-06,
"loss": 1.7016,
"step": 24366
},
{
"epoch": 0.6208006514119226,
"grad_norm": 1.4727599620819092,
"learning_rate": 5.2991937390610205e-06,
"loss": 1.7198,
"step": 24397
},
{
"epoch": 0.6215894705369696,
"grad_norm": 1.407718300819397,
"learning_rate": 5.247602567671625e-06,
"loss": 1.6932,
"step": 24428
},
{
"epoch": 0.6223782896620165,
"grad_norm": 1.424126148223877,
"learning_rate": 5.196234299402603e-06,
"loss": 1.6927,
"step": 24459
},
{
"epoch": 0.6231671087870634,
"grad_norm": 1.5233465433120728,
"learning_rate": 5.145089513937865e-06,
"loss": 1.7072,
"step": 24490
},
{
"epoch": 0.6239559279121103,
"grad_norm": 1.427517056465149,
"learning_rate": 5.094168788439369e-06,
"loss": 1.6981,
"step": 24521
},
{
"epoch": 0.6247447470371572,
"grad_norm": 1.4485392570495605,
"learning_rate": 5.043472697540594e-06,
"loss": 1.6855,
"step": 24552
},
{
"epoch": 0.6255335661622041,
"grad_norm": 1.4278972148895264,
"learning_rate": 4.993001813340012e-06,
"loss": 1.6945,
"step": 24583
},
{
"epoch": 0.626322385287251,
"grad_norm": 1.392105221748352,
"learning_rate": 4.942756705394702e-06,
"loss": 1.6946,
"step": 24614
},
{
"epoch": 0.627111204412298,
"grad_norm": 1.4224188327789307,
"learning_rate": 4.892737940713884e-06,
"loss": 1.7071,
"step": 24645
},
{
"epoch": 0.6279000235373449,
"grad_norm": 1.4652680158615112,
"learning_rate": 4.842946083752511e-06,
"loss": 1.6967,
"step": 24676
},
{
"epoch": 0.6286888426623918,
"grad_norm": 1.490435004234314,
"learning_rate": 4.79338169640493e-06,
"loss": 1.6873,
"step": 24707
},
{
"epoch": 0.6294776617874387,
"grad_norm": 1.54020357131958,
"learning_rate": 4.74404533799851e-06,
"loss": 1.7026,
"step": 24738
},
{
"epoch": 0.6302664809124856,
"grad_norm": 1.3947267532348633,
"learning_rate": 4.694937565287344e-06,
"loss": 1.6959,
"step": 24769
},
{
"epoch": 0.6310553000375325,
"grad_norm": 1.4161572456359863,
"learning_rate": 4.646058932445985e-06,
"loss": 1.6909,
"step": 24800
},
{
"epoch": 0.6318441191625794,
"grad_norm": 1.4541959762573242,
"learning_rate": 4.597409991063148e-06,
"loss": 1.6961,
"step": 24831
},
{
"epoch": 0.6326329382876263,
"grad_norm": 1.410683035850525,
"learning_rate": 4.5489912901355375e-06,
"loss": 1.6846,
"step": 24862
},
{
"epoch": 0.6334217574126733,
"grad_norm": 1.4031442403793335,
"learning_rate": 4.500803376061608e-06,
"loss": 1.6855,
"step": 24893
},
{
"epoch": 0.6342105765377202,
"grad_norm": 1.3770359754562378,
"learning_rate": 4.45284679263541e-06,
"loss": 1.6989,
"step": 24924
},
{
"epoch": 0.6349993956627671,
"grad_norm": 1.4767192602157593,
"learning_rate": 4.4051220810404775e-06,
"loss": 1.6911,
"step": 24955
},
{
"epoch": 0.635788214787814,
"grad_norm": 1.4399274587631226,
"learning_rate": 4.3576297798437025e-06,
"loss": 1.7003,
"step": 24986
},
{
"epoch": 0.6365770339128609,
"grad_norm": 1.3938783407211304,
"learning_rate": 4.3103704249892436e-06,
"loss": 1.7098,
"step": 25017
},
{
"epoch": 0.6373658530379078,
"grad_norm": 1.4374542236328125,
"learning_rate": 4.263344549792487e-06,
"loss": 1.6949,
"step": 25048
},
{
"epoch": 0.6381546721629547,
"grad_norm": 1.443415641784668,
"learning_rate": 4.216552684934056e-06,
"loss": 1.7002,
"step": 25079
},
{
"epoch": 0.6389434912880017,
"grad_norm": 1.455540418624878,
"learning_rate": 4.169995358453777e-06,
"loss": 1.7018,
"step": 25110
},
{
"epoch": 0.6397323104130486,
"grad_norm": 1.4947654008865356,
"learning_rate": 4.123673095744757e-06,
"loss": 1.681,
"step": 25141
},
{
"epoch": 0.6405211295380955,
"grad_norm": 1.4933280944824219,
"learning_rate": 4.077586419547435e-06,
"loss": 1.703,
"step": 25172
},
{
"epoch": 0.6413099486631424,
"grad_norm": 1.4724138975143433,
"learning_rate": 4.03173584994368e-06,
"loss": 1.6987,
"step": 25203
},
{
"epoch": 0.6420987677881893,
"grad_norm": 1.370006799697876,
"learning_rate": 3.986121904350948e-06,
"loss": 1.6881,
"step": 25234
},
{
"epoch": 0.6428875869132362,
"grad_norm": 1.5258022546768188,
"learning_rate": 3.940745097516407e-06,
"loss": 1.6856,
"step": 25265
},
{
"epoch": 0.6436764060382831,
"grad_norm": 1.3982164859771729,
"learning_rate": 3.89560594151116e-06,
"loss": 1.6956,
"step": 25296
},
{
"epoch": 0.64446522516333,
"grad_norm": 1.457051396369934,
"learning_rate": 3.850704945724456e-06,
"loss": 1.7038,
"step": 25327
},
{
"epoch": 0.645254044288377,
"grad_norm": 1.4047811031341553,
"learning_rate": 3.8060426168579077e-06,
"loss": 1.6984,
"step": 25358
},
{
"epoch": 0.6460428634134239,
"grad_norm": 1.3755521774291992,
"learning_rate": 3.7616194589198407e-06,
"loss": 1.7016,
"step": 25389
},
{
"epoch": 0.6468316825384708,
"grad_norm": 1.4575284719467163,
"learning_rate": 3.7174359732195574e-06,
"loss": 1.6907,
"step": 25420
},
{
"epoch": 0.6476205016635177,
"grad_norm": 1.563887357711792,
"learning_rate": 3.673492658361677e-06,
"loss": 1.7141,
"step": 25451
},
{
"epoch": 0.6484093207885646,
"grad_norm": 1.4307068586349487,
"learning_rate": 3.6297900102405467e-06,
"loss": 1.7018,
"step": 25482
},
{
"epoch": 0.6491981399136116,
"grad_norm": 1.4639099836349487,
"learning_rate": 3.586328522034607e-06,
"loss": 1.7162,
"step": 25513
},
{
"epoch": 0.6499869590386586,
"grad_norm": 1.4759560823440552,
"learning_rate": 3.543108684200838e-06,
"loss": 1.6893,
"step": 25544
},
{
"epoch": 0.6507757781637055,
"grad_norm": 1.4981391429901123,
"learning_rate": 3.5001309844692464e-06,
"loss": 1.7037,
"step": 25575
},
{
"epoch": 0.6515645972887524,
"grad_norm": 1.4637056589126587,
"learning_rate": 3.4573959078373215e-06,
"loss": 1.683,
"step": 25606
},
{
"epoch": 0.6523534164137993,
"grad_norm": 1.5560393333435059,
"learning_rate": 3.4149039365646063e-06,
"loss": 1.6843,
"step": 25637
},
{
"epoch": 0.6531422355388462,
"grad_norm": 1.4658019542694092,
"learning_rate": 3.3726555501672143e-06,
"loss": 1.6883,
"step": 25668
},
{
"epoch": 0.6539310546638931,
"grad_norm": 1.397363543510437,
"learning_rate": 3.33065122541244e-06,
"loss": 1.7005,
"step": 25699
},
{
"epoch": 0.65471987378894,
"grad_norm": 1.439571738243103,
"learning_rate": 3.288891436313385e-06,
"loss": 1.7144,
"step": 25730
},
{
"epoch": 0.655508692913987,
"grad_norm": 1.4690093994140625,
"learning_rate": 3.2473766541235963e-06,
"loss": 1.6918,
"step": 25761
},
{
"epoch": 0.6562975120390339,
"grad_norm": 1.4217287302017212,
"learning_rate": 3.2061073473317466e-06,
"loss": 1.6902,
"step": 25792
},
{
"epoch": 0.6570863311640808,
"grad_norm": 1.3401572704315186,
"learning_rate": 3.1650839816563444e-06,
"loss": 1.6949,
"step": 25823
},
{
"epoch": 0.6578751502891277,
"grad_norm": 1.4836351871490479,
"learning_rate": 3.1243070200405093e-06,
"loss": 1.6989,
"step": 25854
},
{
"epoch": 0.6586639694141746,
"grad_norm": 1.3935353755950928,
"learning_rate": 3.0837769226467e-06,
"loss": 1.6926,
"step": 25885
},
{
"epoch": 0.6594527885392215,
"grad_norm": 1.4989404678344727,
"learning_rate": 3.0434941468515666e-06,
"loss": 1.6857,
"step": 25916
},
{
"epoch": 0.6602416076642684,
"grad_norm": 1.4674372673034668,
"learning_rate": 3.003459147240753e-06,
"loss": 1.6912,
"step": 25947
},
{
"epoch": 0.6610304267893153,
"grad_norm": 1.5865478515625,
"learning_rate": 2.9636723756037875e-06,
"loss": 1.6904,
"step": 25978
},
{
"epoch": 0.6618192459143623,
"grad_norm": 1.440338134765625,
"learning_rate": 2.9241342809289833e-06,
"loss": 1.6981,
"step": 26009
},
{
"epoch": 0.6626080650394092,
"grad_norm": 1.3906199932098389,
"learning_rate": 2.8848453093983594e-06,
"loss": 1.6854,
"step": 26040
},
{
"epoch": 0.6633968841644561,
"grad_norm": 1.475035309791565,
"learning_rate": 2.8458059043826257e-06,
"loss": 1.704,
"step": 26071
},
{
"epoch": 0.664185703289503,
"grad_norm": 1.4185906648635864,
"learning_rate": 2.807016506436172e-06,
"loss": 1.6873,
"step": 26102
},
{
"epoch": 0.6649745224145499,
"grad_norm": 1.5231366157531738,
"learning_rate": 2.7684775532920566e-06,
"loss": 1.7009,
"step": 26133
},
{
"epoch": 0.6657633415395968,
"grad_norm": 1.427589774131775,
"learning_rate": 2.7301894798571425e-06,
"loss": 1.7065,
"step": 26164
},
{
"epoch": 0.6665521606646437,
"grad_norm": 1.4298368692398071,
"learning_rate": 2.6921527182071386e-06,
"loss": 1.6944,
"step": 26195
},
{
"epoch": 0.6673409797896906,
"grad_norm": 1.498779058456421,
"learning_rate": 2.654367697581725e-06,
"loss": 1.6845,
"step": 26226
},
{
"epoch": 0.6681297989147376,
"grad_norm": 1.4032225608825684,
"learning_rate": 2.6168348443797175e-06,
"loss": 1.6936,
"step": 26257
},
{
"epoch": 0.6689186180397845,
"grad_norm": 1.4571456909179688,
"learning_rate": 2.5795545821542757e-06,
"loss": 1.7055,
"step": 26288
},
{
"epoch": 0.6697074371648314,
"grad_norm": 1.4156779050827026,
"learning_rate": 2.54252733160808e-06,
"loss": 1.6861,
"step": 26319
},
{
"epoch": 0.6704962562898783,
"grad_norm": 1.5022954940795898,
"learning_rate": 2.5057535105886294e-06,
"loss": 1.6834,
"step": 26350
},
{
"epoch": 0.6712850754149252,
"grad_norm": 1.4164525270462036,
"learning_rate": 2.4692335340834953e-06,
"loss": 1.699,
"step": 26381
},
{
"epoch": 0.6720738945399721,
"grad_norm": 1.459639072418213,
"learning_rate": 2.432967814215639e-06,
"loss": 1.6987,
"step": 26412
},
{
"epoch": 0.672862713665019,
"grad_norm": 1.4693511724472046,
"learning_rate": 2.396956760238794e-06,
"loss": 1.701,
"step": 26443
},
{
"epoch": 0.673651532790066,
"grad_norm": 1.3714548349380493,
"learning_rate": 2.361200778532796e-06,
"loss": 1.6754,
"step": 26474
},
{
"epoch": 0.6744403519151129,
"grad_norm": 1.4285922050476074,
"learning_rate": 2.325700272599049e-06,
"loss": 1.6907,
"step": 26505
},
{
"epoch": 0.6752291710401598,
"grad_norm": 1.5615297555923462,
"learning_rate": 2.2904556430559415e-06,
"loss": 1.6932,
"step": 26536
},
{
"epoch": 0.6760179901652067,
"grad_norm": 1.3956187963485718,
"learning_rate": 2.2554672876343106e-06,
"loss": 1.7064,
"step": 26567
},
{
"epoch": 0.6768068092902536,
"grad_norm": 1.4564794301986694,
"learning_rate": 2.220735601173002e-06,
"loss": 1.6922,
"step": 26598
},
{
"epoch": 0.6775956284153005,
"grad_norm": 1.4553749561309814,
"learning_rate": 2.186260975614382e-06,
"loss": 1.7055,
"step": 26629
},
{
"epoch": 0.6783844475403474,
"grad_norm": 1.4266986846923828,
"learning_rate": 2.1520437999999034e-06,
"loss": 1.7145,
"step": 26660
},
{
"epoch": 0.6791732666653943,
"grad_norm": 1.4530359506607056,
"learning_rate": 2.1180844604657526e-06,
"loss": 1.6916,
"step": 26691
},
{
"epoch": 0.6799620857904413,
"grad_norm": 1.4178498983383179,
"learning_rate": 2.084383340238455e-06,
"loss": 1.6766,
"step": 26722
},
{
"epoch": 0.6807509049154882,
"grad_norm": 1.393988847732544,
"learning_rate": 2.0509408196305704e-06,
"loss": 1.6873,
"step": 26753
},
{
"epoch": 0.6815397240405351,
"grad_norm": 1.3752752542495728,
"learning_rate": 2.017757276036403e-06,
"loss": 1.6984,
"step": 26784
},
{
"epoch": 0.682328543165582,
"grad_norm": 1.394559383392334,
"learning_rate": 1.984833083927726e-06,
"loss": 1.7032,
"step": 26815
},
{
"epoch": 0.6831173622906289,
"grad_norm": 1.4148964881896973,
"learning_rate": 1.952168614849581e-06,
"loss": 1.6844,
"step": 26846
},
{
"epoch": 0.6839061814156758,
"grad_norm": 1.4353492259979248,
"learning_rate": 1.919764237416058e-06,
"loss": 1.7102,
"step": 26877
},
{
"epoch": 0.6846950005407227,
"grad_norm": 1.3867477178573608,
"learning_rate": 1.8876203173061463e-06,
"loss": 1.6931,
"step": 26908
},
{
"epoch": 0.6854838196657697,
"grad_norm": 1.3678532838821411,
"learning_rate": 1.8557372172596206e-06,
"loss": 1.7009,
"step": 26939
},
{
"epoch": 0.6862726387908166,
"grad_norm": 1.485137939453125,
"learning_rate": 1.8241152970729341e-06,
"loss": 1.7016,
"step": 26970
},
{
"epoch": 0.6870614579158635,
"grad_norm": 1.4369994401931763,
"learning_rate": 1.7927549135951572e-06,
"loss": 1.6963,
"step": 27001
},
{
"epoch": 0.6878502770409104,
"grad_norm": 1.4508947134017944,
"learning_rate": 1.7616564207239477e-06,
"loss": 1.6831,
"step": 27032
},
{
"epoch": 0.6886390961659574,
"grad_norm": 1.3917666673660278,
"learning_rate": 1.730820169401584e-06,
"loss": 1.6708,
"step": 27063
},
{
"epoch": 0.6894279152910043,
"grad_norm": 1.4687188863754272,
"learning_rate": 1.7002465076109558e-06,
"loss": 1.6764,
"step": 27094
},
{
"epoch": 0.6902167344160512,
"grad_norm": 1.4053486585617065,
"learning_rate": 1.6699357803716898e-06,
"loss": 1.6931,
"step": 27125
},
{
"epoch": 0.6910055535410982,
"grad_norm": 1.4645016193389893,
"learning_rate": 1.6398883297362305e-06,
"loss": 1.7035,
"step": 27156
},
{
"epoch": 0.6917943726661451,
"grad_norm": 1.434180498123169,
"learning_rate": 1.6101044947859606e-06,
"loss": 1.6693,
"step": 27187
},
{
"epoch": 0.692583191791192,
"grad_norm": 1.451497197151184,
"learning_rate": 1.5805846116274114e-06,
"loss": 1.6776,
"step": 27218
},
{
"epoch": 0.6933720109162389,
"grad_norm": 1.4085174798965454,
"learning_rate": 1.5513290133884611e-06,
"loss": 1.684,
"step": 27249
},
{
"epoch": 0.6941608300412858,
"grad_norm": 1.3959869146347046,
"learning_rate": 1.5223380302145512e-06,
"loss": 1.6798,
"step": 27280
},
{
"epoch": 0.6949496491663327,
"grad_norm": 1.377614974975586,
"learning_rate": 1.4936119892649925e-06,
"loss": 1.6889,
"step": 27311
},
{
"epoch": 0.6957384682913796,
"grad_norm": 1.4654227495193481,
"learning_rate": 1.4651512147092482e-06,
"loss": 1.7027,
"step": 27342
},
{
"epoch": 0.6965272874164266,
"grad_norm": 1.336857795715332,
"learning_rate": 1.4369560277232908e-06,
"loss": 1.6756,
"step": 27373
},
{
"epoch": 0.6973161065414735,
"grad_norm": 1.4347259998321533,
"learning_rate": 1.409026746485978e-06,
"loss": 1.6831,
"step": 27404
},
{
"epoch": 0.6981049256665204,
"grad_norm": 1.5176235437393188,
"learning_rate": 1.3813636861754464e-06,
"loss": 1.6864,
"step": 27435
},
{
"epoch": 0.6988937447915673,
"grad_norm": 1.4501276016235352,
"learning_rate": 1.3539671589655773e-06,
"loss": 1.6941,
"step": 27466
}
],
"logging_steps": 31,
"max_steps": 30517,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3052,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.037550548620044e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}