{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 6867, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004369674459252786, "grad_norm": 25.289249114823775, "learning_rate": 6.550218340611353e-07, "loss": 3.0338, "step": 10 }, { "epoch": 0.008739348918505571, "grad_norm": 11.656180349008569, "learning_rate": 1.3828238719068415e-06, "loss": 2.81, "step": 20 }, { "epoch": 0.013109023377758357, "grad_norm": 7.437780357675439, "learning_rate": 2.1106259097525473e-06, "loss": 2.485, "step": 30 }, { "epoch": 0.017478697837011142, "grad_norm": 4.1632105620914075, "learning_rate": 2.8384279475982535e-06, "loss": 1.8568, "step": 40 }, { "epoch": 0.021848372296263928, "grad_norm": 2.701149714358883, "learning_rate": 3.5662299854439593e-06, "loss": 1.4461, "step": 50 }, { "epoch": 0.026218046755516713, "grad_norm": 1.5067197468905562, "learning_rate": 4.2940320232896655e-06, "loss": 1.0279, "step": 60 }, { "epoch": 0.0305877212147695, "grad_norm": 3.615460830746041, "learning_rate": 5.021834061135372e-06, "loss": 0.8383, "step": 70 }, { "epoch": 0.034957395674022285, "grad_norm": 2.7841880116071005, "learning_rate": 5.749636098981077e-06, "loss": 0.7115, "step": 80 }, { "epoch": 0.03932707013327507, "grad_norm": 1.4497428109347361, "learning_rate": 6.477438136826783e-06, "loss": 0.6405, "step": 90 }, { "epoch": 0.043696744592527856, "grad_norm": 1.855298707269607, "learning_rate": 7.205240174672489e-06, "loss": 0.6015, "step": 100 }, { "epoch": 0.04806641905178064, "grad_norm": 1.1341246317737723, "learning_rate": 7.933042212518195e-06, "loss": 0.5819, "step": 110 }, { "epoch": 0.05243609351103343, "grad_norm": 0.6453138652786888, "learning_rate": 8.660844250363902e-06, "loss": 0.548, "step": 120 }, { "epoch": 0.05680576797028621, "grad_norm": 1.000591022536797, "learning_rate": 9.388646288209607e-06, "loss": 0.5325, "step": 130 }, { "epoch": 0.061175442429539, "grad_norm": 0.5620960299146275, "learning_rate": 1.0116448326055314e-05, "loss": 0.5191, "step": 140 }, { "epoch": 0.06554511688879179, "grad_norm": 0.6880110939807305, "learning_rate": 1.084425036390102e-05, "loss": 0.4936, "step": 150 }, { "epoch": 0.06991479134804457, "grad_norm": 0.49141234174344767, "learning_rate": 1.1572052401746727e-05, "loss": 0.4848, "step": 160 }, { "epoch": 0.07428446580729736, "grad_norm": 0.5701316809312219, "learning_rate": 1.2299854439592432e-05, "loss": 0.4726, "step": 170 }, { "epoch": 0.07865414026655014, "grad_norm": 0.6460113526829739, "learning_rate": 1.3027656477438139e-05, "loss": 0.4624, "step": 180 }, { "epoch": 0.08302381472580293, "grad_norm": 0.6528876220234264, "learning_rate": 1.3755458515283843e-05, "loss": 0.4529, "step": 190 }, { "epoch": 0.08739348918505571, "grad_norm": 0.6106600189544156, "learning_rate": 1.448326055312955e-05, "loss": 0.4496, "step": 200 }, { "epoch": 0.0917631636443085, "grad_norm": 0.6839889131252415, "learning_rate": 1.5211062590975255e-05, "loss": 0.4416, "step": 210 }, { "epoch": 0.09613283810356128, "grad_norm": 0.7583623922490678, "learning_rate": 1.5938864628820962e-05, "loss": 0.4373, "step": 220 }, { "epoch": 0.10050251256281408, "grad_norm": 0.5358998728274212, "learning_rate": 1.6666666666666667e-05, "loss": 0.4323, "step": 230 }, { "epoch": 0.10487218702206685, "grad_norm": 0.7076410311883243, "learning_rate": 1.7394468704512373e-05, "loss": 0.4298, "step": 240 }, { "epoch": 0.10924186148131965, "grad_norm": 0.56824959821333, "learning_rate": 1.812227074235808e-05, "loss": 0.4338, "step": 250 }, { "epoch": 0.11361153594057243, "grad_norm": 0.5431568870977636, "learning_rate": 1.8850072780203784e-05, "loss": 0.4159, "step": 260 }, { "epoch": 0.11798121039982522, "grad_norm": 0.667974044396845, "learning_rate": 1.9577874818049492e-05, "loss": 0.4225, "step": 270 }, { "epoch": 0.122350884859078, "grad_norm": 0.4963904688817497, "learning_rate": 2.0305676855895198e-05, "loss": 0.4184, "step": 280 }, { "epoch": 0.1267205593183308, "grad_norm": 0.46024205912512767, "learning_rate": 2.1033478893740903e-05, "loss": 0.4138, "step": 290 }, { "epoch": 0.13109023377758358, "grad_norm": 0.4418700047278689, "learning_rate": 2.176128093158661e-05, "loss": 0.4047, "step": 300 }, { "epoch": 0.13545990823683635, "grad_norm": 0.49605777993526634, "learning_rate": 2.2489082969432314e-05, "loss": 0.4104, "step": 310 }, { "epoch": 0.13982958269608914, "grad_norm": 0.6367522014612957, "learning_rate": 2.3216885007278022e-05, "loss": 0.4034, "step": 320 }, { "epoch": 0.14419925715534193, "grad_norm": 0.40893249741965865, "learning_rate": 2.3944687045123728e-05, "loss": 0.3975, "step": 330 }, { "epoch": 0.14856893161459472, "grad_norm": 0.5977284348116177, "learning_rate": 2.4672489082969433e-05, "loss": 0.3999, "step": 340 }, { "epoch": 0.1529386060738475, "grad_norm": 0.6020476803219461, "learning_rate": 2.540029112081514e-05, "loss": 0.3975, "step": 350 }, { "epoch": 0.15730828053310028, "grad_norm": 0.49959249038852227, "learning_rate": 2.6128093158660844e-05, "loss": 0.3926, "step": 360 }, { "epoch": 0.16167795499235307, "grad_norm": 0.5835642198903502, "learning_rate": 2.6855895196506553e-05, "loss": 0.4041, "step": 370 }, { "epoch": 0.16604762945160587, "grad_norm": 0.4867076868793625, "learning_rate": 2.7583697234352258e-05, "loss": 0.3857, "step": 380 }, { "epoch": 0.17041730391085863, "grad_norm": 0.49097610318255575, "learning_rate": 2.831149927219796e-05, "loss": 0.3849, "step": 390 }, { "epoch": 0.17478697837011142, "grad_norm": 0.5557327473737738, "learning_rate": 2.9039301310043672e-05, "loss": 0.3912, "step": 400 }, { "epoch": 0.17915665282936422, "grad_norm": 0.771676099075755, "learning_rate": 2.9767103347889374e-05, "loss": 0.3817, "step": 410 }, { "epoch": 0.183526327288617, "grad_norm": 0.6710435748677224, "learning_rate": 3.0494905385735083e-05, "loss": 0.3919, "step": 420 }, { "epoch": 0.18789600174786977, "grad_norm": 0.7163417130858818, "learning_rate": 3.122270742358079e-05, "loss": 0.382, "step": 430 }, { "epoch": 0.19226567620712257, "grad_norm": 0.5511501311644037, "learning_rate": 3.1950509461426493e-05, "loss": 0.3982, "step": 440 }, { "epoch": 0.19663535066637536, "grad_norm": 0.41376055145170704, "learning_rate": 3.26783114992722e-05, "loss": 0.383, "step": 450 }, { "epoch": 0.20100502512562815, "grad_norm": 0.5361314709561074, "learning_rate": 3.3406113537117904e-05, "loss": 0.3802, "step": 460 }, { "epoch": 0.20537469958488092, "grad_norm": 0.6480092563747167, "learning_rate": 3.413391557496361e-05, "loss": 0.3817, "step": 470 }, { "epoch": 0.2097443740441337, "grad_norm": 0.5445176806555806, "learning_rate": 3.486171761280932e-05, "loss": 0.3708, "step": 480 }, { "epoch": 0.2141140485033865, "grad_norm": 0.6768957968461157, "learning_rate": 3.558951965065502e-05, "loss": 0.3824, "step": 490 }, { "epoch": 0.2184837229626393, "grad_norm": 0.5942773751725828, "learning_rate": 3.6317321688500726e-05, "loss": 0.3715, "step": 500 }, { "epoch": 0.22285339742189206, "grad_norm": 0.5691436120561877, "learning_rate": 3.704512372634644e-05, "loss": 0.3707, "step": 510 }, { "epoch": 0.22722307188114485, "grad_norm": 0.4918785706590604, "learning_rate": 3.777292576419214e-05, "loss": 0.3716, "step": 520 }, { "epoch": 0.23159274634039764, "grad_norm": 0.6178436078018688, "learning_rate": 3.850072780203784e-05, "loss": 0.3782, "step": 530 }, { "epoch": 0.23596242079965044, "grad_norm": 0.4765723691148907, "learning_rate": 3.9228529839883554e-05, "loss": 0.358, "step": 540 }, { "epoch": 0.2403320952589032, "grad_norm": 0.5438674747635697, "learning_rate": 3.995633187772926e-05, "loss": 0.3715, "step": 550 }, { "epoch": 0.244701769718156, "grad_norm": 0.42395909355758743, "learning_rate": 4.068413391557497e-05, "loss": 0.3691, "step": 560 }, { "epoch": 0.24907144417740879, "grad_norm": 0.49325309745663193, "learning_rate": 4.141193595342067e-05, "loss": 0.356, "step": 570 }, { "epoch": 0.2534411186366616, "grad_norm": 0.4259308682116486, "learning_rate": 4.2139737991266375e-05, "loss": 0.3674, "step": 580 }, { "epoch": 0.25781079309591437, "grad_norm": 0.4200523300464016, "learning_rate": 4.286754002911209e-05, "loss": 0.3629, "step": 590 }, { "epoch": 0.26218046755516716, "grad_norm": 0.36702517452638733, "learning_rate": 4.359534206695779e-05, "loss": 0.3627, "step": 600 }, { "epoch": 0.2665501420144199, "grad_norm": 0.38779976127895843, "learning_rate": 4.432314410480349e-05, "loss": 0.3618, "step": 610 }, { "epoch": 0.2709198164736727, "grad_norm": 0.40538273783801576, "learning_rate": 4.50509461426492e-05, "loss": 0.3592, "step": 620 }, { "epoch": 0.2752894909329255, "grad_norm": 0.5051495914506833, "learning_rate": 4.577874818049491e-05, "loss": 0.3544, "step": 630 }, { "epoch": 0.2796591653921783, "grad_norm": 0.43366347927249477, "learning_rate": 4.6506550218340614e-05, "loss": 0.3555, "step": 640 }, { "epoch": 0.28402883985143107, "grad_norm": 0.44468985777461, "learning_rate": 4.723435225618632e-05, "loss": 0.3499, "step": 650 }, { "epoch": 0.28839851431068386, "grad_norm": 0.49145043836490315, "learning_rate": 4.7962154294032025e-05, "loss": 0.353, "step": 660 }, { "epoch": 0.29276818876993665, "grad_norm": 0.4077496645293648, "learning_rate": 4.868995633187774e-05, "loss": 0.3615, "step": 670 }, { "epoch": 0.29713786322918945, "grad_norm": 0.4457778880431833, "learning_rate": 4.9417758369723435e-05, "loss": 0.3528, "step": 680 }, { "epoch": 0.3015075376884422, "grad_norm": 0.41564528785347277, "learning_rate": 4.99999870791001e-05, "loss": 0.3591, "step": 690 }, { "epoch": 0.305877212147695, "grad_norm": 0.3719363783109856, "learning_rate": 4.999953484900578e-05, "loss": 0.3476, "step": 700 }, { "epoch": 0.31024688660694777, "grad_norm": 0.40624762908342826, "learning_rate": 4.999843658727211e-05, "loss": 0.3465, "step": 710 }, { "epoch": 0.31461656106620056, "grad_norm": 0.4085444132524013, "learning_rate": 4.999669232228008e-05, "loss": 0.3531, "step": 720 }, { "epoch": 0.31898623552545335, "grad_norm": 0.43705311239506367, "learning_rate": 4.999430209910455e-05, "loss": 0.3583, "step": 730 }, { "epoch": 0.32335590998470615, "grad_norm": 0.40213824820090827, "learning_rate": 4.9991265979513055e-05, "loss": 0.3618, "step": 740 }, { "epoch": 0.32772558444395894, "grad_norm": 0.3929880609701779, "learning_rate": 4.9987584041964227e-05, "loss": 0.3536, "step": 750 }, { "epoch": 0.33209525890321173, "grad_norm": 0.32003299692943676, "learning_rate": 4.9983256381605764e-05, "loss": 0.3459, "step": 760 }, { "epoch": 0.33646493336246447, "grad_norm": 0.48033731250564854, "learning_rate": 4.997828311027197e-05, "loss": 0.3523, "step": 770 }, { "epoch": 0.34083460782171726, "grad_norm": 0.6598899816976614, "learning_rate": 4.997266435648086e-05, "loss": 0.3553, "step": 780 }, { "epoch": 0.34520428228097005, "grad_norm": 0.34136885817133283, "learning_rate": 4.996640026543084e-05, "loss": 0.3516, "step": 790 }, { "epoch": 0.34957395674022285, "grad_norm": 0.39898829012741177, "learning_rate": 4.9959490998996974e-05, "loss": 0.346, "step": 800 }, { "epoch": 0.35394363119947564, "grad_norm": 0.3974783905204888, "learning_rate": 4.995193673572676e-05, "loss": 0.3404, "step": 810 }, { "epoch": 0.35831330565872843, "grad_norm": 0.3760230310675625, "learning_rate": 4.9943737670835555e-05, "loss": 0.342, "step": 820 }, { "epoch": 0.3626829801179812, "grad_norm": 0.34048339031946, "learning_rate": 4.993489401620153e-05, "loss": 0.3436, "step": 830 }, { "epoch": 0.367052654577234, "grad_norm": 0.3423661342866573, "learning_rate": 4.992540600036014e-05, "loss": 0.3482, "step": 840 }, { "epoch": 0.3714223290364868, "grad_norm": 0.3791528288768608, "learning_rate": 4.9915273868498306e-05, "loss": 0.3399, "step": 850 }, { "epoch": 0.37579200349573955, "grad_norm": 0.33688776489311434, "learning_rate": 4.9904497882448006e-05, "loss": 0.3395, "step": 860 }, { "epoch": 0.38016167795499234, "grad_norm": 0.3364082615107174, "learning_rate": 4.989307832067952e-05, "loss": 0.3385, "step": 870 }, { "epoch": 0.38453135241424513, "grad_norm": 0.387383353709174, "learning_rate": 4.98810154782943e-05, "loss": 0.3409, "step": 880 }, { "epoch": 0.3889010268734979, "grad_norm": 0.35194855655791857, "learning_rate": 4.9868309667017226e-05, "loss": 0.3381, "step": 890 }, { "epoch": 0.3932707013327507, "grad_norm": 0.3413553827551459, "learning_rate": 4.9854961215188676e-05, "loss": 0.3346, "step": 900 }, { "epoch": 0.3976403757920035, "grad_norm": 0.4270395192072039, "learning_rate": 4.984097046775594e-05, "loss": 0.331, "step": 910 }, { "epoch": 0.4020100502512563, "grad_norm": 0.29757358098030284, "learning_rate": 4.9826337786264366e-05, "loss": 0.3456, "step": 920 }, { "epoch": 0.4063797247105091, "grad_norm": 0.3051284353717966, "learning_rate": 4.9811063548848016e-05, "loss": 0.3366, "step": 930 }, { "epoch": 0.41074939916976183, "grad_norm": 0.3323255129002569, "learning_rate": 4.979514815021984e-05, "loss": 0.3432, "step": 940 }, { "epoch": 0.4151190736290146, "grad_norm": 0.352209195838188, "learning_rate": 4.9778592001661536e-05, "loss": 0.3263, "step": 950 }, { "epoch": 0.4194887480882674, "grad_norm": 0.35908324008167625, "learning_rate": 4.9761395531012904e-05, "loss": 0.3309, "step": 960 }, { "epoch": 0.4238584225475202, "grad_norm": 0.36867426742696585, "learning_rate": 4.9743559182660775e-05, "loss": 0.3474, "step": 970 }, { "epoch": 0.428228097006773, "grad_norm": 0.32607477774486465, "learning_rate": 4.972508341752754e-05, "loss": 0.3308, "step": 980 }, { "epoch": 0.4325977714660258, "grad_norm": 0.291319997808853, "learning_rate": 4.9705968713059224e-05, "loss": 0.3475, "step": 990 }, { "epoch": 0.4369674459252786, "grad_norm": 0.3663028617721437, "learning_rate": 4.968621556321319e-05, "loss": 0.3316, "step": 1000 }, { "epoch": 0.4413371203845314, "grad_norm": 0.34488630358290745, "learning_rate": 4.966582447844532e-05, "loss": 0.3311, "step": 1010 }, { "epoch": 0.4457067948437841, "grad_norm": 0.2921917618921613, "learning_rate": 4.964479598569686e-05, "loss": 0.3269, "step": 1020 }, { "epoch": 0.4500764693030369, "grad_norm": 0.2975779757879733, "learning_rate": 4.962313062838077e-05, "loss": 0.3334, "step": 1030 }, { "epoch": 0.4544461437622897, "grad_norm": 0.2855003621138662, "learning_rate": 4.960082896636773e-05, "loss": 0.3266, "step": 1040 }, { "epoch": 0.4588158182215425, "grad_norm": 0.3801047031729062, "learning_rate": 4.957789157597163e-05, "loss": 0.3382, "step": 1050 }, { "epoch": 0.4631854926807953, "grad_norm": 0.33061288510598813, "learning_rate": 4.9554319049934706e-05, "loss": 0.3248, "step": 1060 }, { "epoch": 0.4675551671400481, "grad_norm": 0.27964496325235, "learning_rate": 4.953011199741219e-05, "loss": 0.3356, "step": 1070 }, { "epoch": 0.47192484159930087, "grad_norm": 0.27550609593903164, "learning_rate": 4.950527104395659e-05, "loss": 0.332, "step": 1080 }, { "epoch": 0.47629451605855366, "grad_norm": 0.3183586864368227, "learning_rate": 4.947979683150152e-05, "loss": 0.3359, "step": 1090 }, { "epoch": 0.4806641905178064, "grad_norm": 0.30469809857246144, "learning_rate": 4.9453690018345144e-05, "loss": 0.3303, "step": 1100 }, { "epoch": 0.4850338649770592, "grad_norm": 0.29511788671986083, "learning_rate": 4.942695127913308e-05, "loss": 0.334, "step": 1110 }, { "epoch": 0.489403539436312, "grad_norm": 0.31330241286972, "learning_rate": 4.939958130484106e-05, "loss": 0.3279, "step": 1120 }, { "epoch": 0.4937732138955648, "grad_norm": 0.3121749781938001, "learning_rate": 4.937158080275701e-05, "loss": 0.3232, "step": 1130 }, { "epoch": 0.49814288835481757, "grad_norm": 0.2574872769658399, "learning_rate": 4.934295049646282e-05, "loss": 0.3215, "step": 1140 }, { "epoch": 0.5025125628140703, "grad_norm": 0.3570536771463989, "learning_rate": 4.931369112581557e-05, "loss": 0.3153, "step": 1150 }, { "epoch": 0.5068822372733232, "grad_norm": 0.3356485594524322, "learning_rate": 4.9283803446928524e-05, "loss": 0.3332, "step": 1160 }, { "epoch": 0.5112519117325759, "grad_norm": 0.25570548997297177, "learning_rate": 4.9253288232151496e-05, "loss": 0.3231, "step": 1170 }, { "epoch": 0.5156215861918287, "grad_norm": 0.36361715567442826, "learning_rate": 4.9222146270050926e-05, "loss": 0.3283, "step": 1180 }, { "epoch": 0.5199912606510815, "grad_norm": 0.47664661926355, "learning_rate": 4.919037836538949e-05, "loss": 0.3212, "step": 1190 }, { "epoch": 0.5243609351103343, "grad_norm": 0.30041599138139646, "learning_rate": 4.9157985339105336e-05, "loss": 0.3175, "step": 1200 }, { "epoch": 0.5287306095695871, "grad_norm": 0.3155523920323453, "learning_rate": 4.912496802829083e-05, "loss": 0.3159, "step": 1210 }, { "epoch": 0.5331002840288398, "grad_norm": 0.31302835114380073, "learning_rate": 4.909132728617095e-05, "loss": 0.316, "step": 1220 }, { "epoch": 0.5374699584880926, "grad_norm": 0.32286763621140474, "learning_rate": 4.905706398208121e-05, "loss": 0.3266, "step": 1230 }, { "epoch": 0.5418396329473454, "grad_norm": 0.27978292209445793, "learning_rate": 4.902217900144524e-05, "loss": 0.3161, "step": 1240 }, { "epoch": 0.5462093074065982, "grad_norm": 0.27146167378877767, "learning_rate": 4.898667324575185e-05, "loss": 0.3227, "step": 1250 }, { "epoch": 0.550578981865851, "grad_norm": 0.26552903063873473, "learning_rate": 4.895054763253177e-05, "loss": 0.3252, "step": 1260 }, { "epoch": 0.5549486563251038, "grad_norm": 0.2617825952185592, "learning_rate": 4.8913803095333945e-05, "loss": 0.3187, "step": 1270 }, { "epoch": 0.5593183307843566, "grad_norm": 0.3014488707132253, "learning_rate": 4.887644058370139e-05, "loss": 0.3166, "step": 1280 }, { "epoch": 0.5636880052436094, "grad_norm": 0.3026959575318133, "learning_rate": 4.883846106314664e-05, "loss": 0.3295, "step": 1290 }, { "epoch": 0.5680576797028621, "grad_norm": 0.2825390100309781, "learning_rate": 4.879986551512684e-05, "loss": 0.33, "step": 1300 }, { "epoch": 0.5724273541621149, "grad_norm": 0.28166403202857904, "learning_rate": 4.876065493701837e-05, "loss": 0.3167, "step": 1310 }, { "epoch": 0.5767970286213677, "grad_norm": 0.29566281796643473, "learning_rate": 4.872083034209102e-05, "loss": 0.3206, "step": 1320 }, { "epoch": 0.5811667030806205, "grad_norm": 0.32389000245561744, "learning_rate": 4.868039275948189e-05, "loss": 0.3064, "step": 1330 }, { "epoch": 0.5855363775398733, "grad_norm": 0.4389081716743775, "learning_rate": 4.863934323416871e-05, "loss": 0.3192, "step": 1340 }, { "epoch": 0.589906051999126, "grad_norm": 0.35332108931963935, "learning_rate": 4.859768282694293e-05, "loss": 0.3352, "step": 1350 }, { "epoch": 0.5942757264583789, "grad_norm": 0.3151708364064402, "learning_rate": 4.855541261438222e-05, "loss": 0.3197, "step": 1360 }, { "epoch": 0.5986454009176316, "grad_norm": 0.25857642978839124, "learning_rate": 4.851253368882269e-05, "loss": 0.319, "step": 1370 }, { "epoch": 0.6030150753768844, "grad_norm": 0.3229320492133288, "learning_rate": 4.846904715833065e-05, "loss": 0.3146, "step": 1380 }, { "epoch": 0.6073847498361372, "grad_norm": 0.27348372664224985, "learning_rate": 4.842495414667401e-05, "loss": 0.3246, "step": 1390 }, { "epoch": 0.61175442429539, "grad_norm": 0.2988357859791239, "learning_rate": 4.838025579329319e-05, "loss": 0.3231, "step": 1400 }, { "epoch": 0.6161240987546428, "grad_norm": 0.6545692887255868, "learning_rate": 4.833495325327169e-05, "loss": 0.3183, "step": 1410 }, { "epoch": 0.6204937732138955, "grad_norm": 0.34240755893595265, "learning_rate": 4.828904769730628e-05, "loss": 0.314, "step": 1420 }, { "epoch": 0.6248634476731484, "grad_norm": 0.31136699482382707, "learning_rate": 4.824254031167666e-05, "loss": 0.3166, "step": 1430 }, { "epoch": 0.6292331221324011, "grad_norm": 0.2806552500856881, "learning_rate": 4.819543229821494e-05, "loss": 0.3188, "step": 1440 }, { "epoch": 0.633602796591654, "grad_norm": 0.2886747780478624, "learning_rate": 4.814772487427443e-05, "loss": 0.3209, "step": 1450 }, { "epoch": 0.6379724710509067, "grad_norm": 0.3042721491342838, "learning_rate": 4.809941927269828e-05, "loss": 0.3112, "step": 1460 }, { "epoch": 0.6423421455101594, "grad_norm": 0.34230081928213874, "learning_rate": 4.8050516741787607e-05, "loss": 0.3196, "step": 1470 }, { "epoch": 0.6467118199694123, "grad_norm": 0.25932002739680027, "learning_rate": 4.80010185452692e-05, "loss": 0.3056, "step": 1480 }, { "epoch": 0.651081494428665, "grad_norm": 0.25631785383237754, "learning_rate": 4.7950925962262907e-05, "loss": 0.3181, "step": 1490 }, { "epoch": 0.6554511688879179, "grad_norm": 0.29748380723967977, "learning_rate": 4.7900240287248546e-05, "loss": 0.313, "step": 1500 }, { "epoch": 0.6598208433471706, "grad_norm": 0.28703363194589365, "learning_rate": 4.784896283003249e-05, "loss": 0.3114, "step": 1510 }, { "epoch": 0.6641905178064235, "grad_norm": 0.3131807793908106, "learning_rate": 4.7797094915713774e-05, "loss": 0.3226, "step": 1520 }, { "epoch": 0.6685601922656762, "grad_norm": 0.40851026732676177, "learning_rate": 4.774463788464989e-05, "loss": 0.3112, "step": 1530 }, { "epoch": 0.6729298667249289, "grad_norm": 0.25170773245832345, "learning_rate": 4.7691593092422126e-05, "loss": 0.3082, "step": 1540 }, { "epoch": 0.6772995411841818, "grad_norm": 0.33340910565152576, "learning_rate": 4.7637961909800555e-05, "loss": 0.3092, "step": 1550 }, { "epoch": 0.6816692156434345, "grad_norm": 0.32250097355232266, "learning_rate": 4.758374572270859e-05, "loss": 0.3126, "step": 1560 }, { "epoch": 0.6860388901026874, "grad_norm": 0.3050879782949748, "learning_rate": 4.75289459321872e-05, "loss": 0.3134, "step": 1570 }, { "epoch": 0.6904085645619401, "grad_norm": 0.3084344703441181, "learning_rate": 4.747356395435865e-05, "loss": 0.3134, "step": 1580 }, { "epoch": 0.694778239021193, "grad_norm": 0.2709086004804197, "learning_rate": 4.741760122039e-05, "loss": 0.3085, "step": 1590 }, { "epoch": 0.6991479134804457, "grad_norm": 0.2945664851706884, "learning_rate": 4.7361059176455995e-05, "loss": 0.3163, "step": 1600 }, { "epoch": 0.7035175879396985, "grad_norm": 0.29267861327673594, "learning_rate": 4.730393928370181e-05, "loss": 0.3111, "step": 1610 }, { "epoch": 0.7078872623989513, "grad_norm": 0.3042957229868975, "learning_rate": 4.724624301820524e-05, "loss": 0.3169, "step": 1620 }, { "epoch": 0.712256936858204, "grad_norm": 0.27493617192259484, "learning_rate": 4.718797187093852e-05, "loss": 0.3088, "step": 1630 }, { "epoch": 0.7166266113174569, "grad_norm": 0.2358366517146545, "learning_rate": 4.712912734772987e-05, "loss": 0.314, "step": 1640 }, { "epoch": 0.7209962857767096, "grad_norm": 0.2524376879547846, "learning_rate": 4.7069710969224546e-05, "loss": 0.3073, "step": 1650 }, { "epoch": 0.7253659602359624, "grad_norm": 0.24654147709689114, "learning_rate": 4.700972427084551e-05, "loss": 0.3083, "step": 1660 }, { "epoch": 0.7297356346952152, "grad_norm": 0.31590264112951727, "learning_rate": 4.6949168802753826e-05, "loss": 0.3045, "step": 1670 }, { "epoch": 0.734105309154468, "grad_norm": 0.28200064803403213, "learning_rate": 4.6888046129808546e-05, "loss": 0.3103, "step": 1680 }, { "epoch": 0.7384749836137208, "grad_norm": 0.27151372176819216, "learning_rate": 4.6826357831526266e-05, "loss": 0.3062, "step": 1690 }, { "epoch": 0.7428446580729736, "grad_norm": 0.32000380536906287, "learning_rate": 4.6764105502040353e-05, "loss": 0.3166, "step": 1700 }, { "epoch": 0.7472143325322264, "grad_norm": 0.32141385320999466, "learning_rate": 4.670129075005972e-05, "loss": 0.3113, "step": 1710 }, { "epoch": 0.7515840069914791, "grad_norm": 0.2633398145696732, "learning_rate": 4.6637915198827266e-05, "loss": 0.3085, "step": 1720 }, { "epoch": 0.7559536814507319, "grad_norm": 0.26503287413810733, "learning_rate": 4.6573980486077884e-05, "loss": 0.3044, "step": 1730 }, { "epoch": 0.7603233559099847, "grad_norm": 0.3080909896219821, "learning_rate": 4.6509488263996236e-05, "loss": 0.305, "step": 1740 }, { "epoch": 0.7646930303692375, "grad_norm": 0.29982438488174123, "learning_rate": 4.644444019917396e-05, "loss": 0.3146, "step": 1750 }, { "epoch": 0.7690627048284903, "grad_norm": 0.3742942236206983, "learning_rate": 4.637883797256664e-05, "loss": 0.3015, "step": 1760 }, { "epoch": 0.7734323792877431, "grad_norm": 0.27066015913681, "learning_rate": 4.631268327945038e-05, "loss": 0.3063, "step": 1770 }, { "epoch": 0.7778020537469958, "grad_norm": 0.25760231055948196, "learning_rate": 4.6245977829378e-05, "loss": 0.3078, "step": 1780 }, { "epoch": 0.7821717282062486, "grad_norm": 0.2781431805998506, "learning_rate": 4.61787233461348e-05, "loss": 0.3066, "step": 1790 }, { "epoch": 0.7865414026655014, "grad_norm": 0.29003845209244283, "learning_rate": 4.6110921567694095e-05, "loss": 0.3182, "step": 1800 }, { "epoch": 0.7909110771247542, "grad_norm": 0.2804798339349801, "learning_rate": 4.604257424617225e-05, "loss": 0.3032, "step": 1810 }, { "epoch": 0.795280751584007, "grad_norm": 0.27455148212552744, "learning_rate": 4.59736831477834e-05, "loss": 0.3118, "step": 1820 }, { "epoch": 0.7996504260432598, "grad_norm": 0.2885340877026202, "learning_rate": 4.590425005279385e-05, "loss": 0.3113, "step": 1830 }, { "epoch": 0.8040201005025126, "grad_norm": 0.24756819441365083, "learning_rate": 4.583427675547602e-05, "loss": 0.3003, "step": 1840 }, { "epoch": 0.8083897749617653, "grad_norm": 0.2146912417595887, "learning_rate": 4.5763765064062126e-05, "loss": 0.3032, "step": 1850 }, { "epoch": 0.8127594494210182, "grad_norm": 0.3391087821067352, "learning_rate": 4.5692716800697413e-05, "loss": 0.2937, "step": 1860 }, { "epoch": 0.8171291238802709, "grad_norm": 0.28637221298651927, "learning_rate": 4.562113380139307e-05, "loss": 0.3126, "step": 1870 }, { "epoch": 0.8214987983395237, "grad_norm": 0.26538115487754543, "learning_rate": 4.5549017915978834e-05, "loss": 0.31, "step": 1880 }, { "epoch": 0.8258684727987765, "grad_norm": 0.31160501406483493, "learning_rate": 4.54763710080551e-05, "loss": 0.2992, "step": 1890 }, { "epoch": 0.8302381472580292, "grad_norm": 0.29764530200442213, "learning_rate": 4.540319495494485e-05, "loss": 0.3084, "step": 1900 }, { "epoch": 0.8346078217172821, "grad_norm": 0.24871811816915726, "learning_rate": 4.532949164764511e-05, "loss": 0.309, "step": 1910 }, { "epoch": 0.8389774961765348, "grad_norm": 0.24012947449743874, "learning_rate": 4.525526299077802e-05, "loss": 0.3003, "step": 1920 }, { "epoch": 0.8433471706357877, "grad_norm": 0.3057978022007791, "learning_rate": 4.518051090254172e-05, "loss": 0.3027, "step": 1930 }, { "epoch": 0.8477168450950404, "grad_norm": 0.2460428916521761, "learning_rate": 4.510523731466072e-05, "loss": 0.2987, "step": 1940 }, { "epoch": 0.8520865195542932, "grad_norm": 0.2840413310103727, "learning_rate": 4.5029444172336e-05, "loss": 0.2975, "step": 1950 }, { "epoch": 0.856456194013546, "grad_norm": 0.23565452755897007, "learning_rate": 4.495313343419469e-05, "loss": 0.3054, "step": 1960 }, { "epoch": 0.8608258684727987, "grad_norm": 0.2422537177218105, "learning_rate": 4.4876307072239566e-05, "loss": 0.3044, "step": 1970 }, { "epoch": 0.8651955429320516, "grad_norm": 0.290478590272856, "learning_rate": 4.479896707179796e-05, "loss": 0.3086, "step": 1980 }, { "epoch": 0.8695652173913043, "grad_norm": 0.26516708843391623, "learning_rate": 4.472111543147056e-05, "loss": 0.3042, "step": 1990 }, { "epoch": 0.8739348918505572, "grad_norm": 0.28022265858536916, "learning_rate": 4.464275416307973e-05, "loss": 0.3126, "step": 2000 }, { "epoch": 0.8783045663098099, "grad_norm": 0.3095920868262214, "learning_rate": 4.456388529161746e-05, "loss": 0.3057, "step": 2010 }, { "epoch": 0.8826742407690628, "grad_norm": 0.2746186779327871, "learning_rate": 4.448451085519314e-05, "loss": 0.3078, "step": 2020 }, { "epoch": 0.8870439152283155, "grad_norm": 0.22950821096241686, "learning_rate": 4.4404632904980835e-05, "loss": 0.2948, "step": 2030 }, { "epoch": 0.8914135896875682, "grad_norm": 0.2669312825026787, "learning_rate": 4.432425350516627e-05, "loss": 0.3059, "step": 2040 }, { "epoch": 0.8957832641468211, "grad_norm": 0.2749715744290143, "learning_rate": 4.424337473289351e-05, "loss": 0.3038, "step": 2050 }, { "epoch": 0.9001529386060738, "grad_norm": 0.3115537571810186, "learning_rate": 4.416199867821126e-05, "loss": 0.3044, "step": 2060 }, { "epoch": 0.9045226130653267, "grad_norm": 0.24991144433419615, "learning_rate": 4.408012744401891e-05, "loss": 0.2998, "step": 2070 }, { "epoch": 0.9088922875245794, "grad_norm": 0.27812293392292686, "learning_rate": 4.399776314601212e-05, "loss": 0.3149, "step": 2080 }, { "epoch": 0.9132619619838322, "grad_norm": 0.25486660096547065, "learning_rate": 4.3914907912628196e-05, "loss": 0.3008, "step": 2090 }, { "epoch": 0.917631636443085, "grad_norm": 0.340734308328453, "learning_rate": 4.3831563884991056e-05, "loss": 0.3058, "step": 2100 }, { "epoch": 0.9220013109023377, "grad_norm": 0.2983910352021778, "learning_rate": 4.374773321685594e-05, "loss": 0.3066, "step": 2110 }, { "epoch": 0.9263709853615906, "grad_norm": 0.21099457481460102, "learning_rate": 4.3663418074553694e-05, "loss": 0.2995, "step": 2120 }, { "epoch": 0.9307406598208433, "grad_norm": 0.255241128210839, "learning_rate": 4.357862063693486e-05, "loss": 0.2965, "step": 2130 }, { "epoch": 0.9351103342800962, "grad_norm": 0.3259222994019878, "learning_rate": 4.34933430953133e-05, "loss": 0.2904, "step": 2140 }, { "epoch": 0.9394800087393489, "grad_norm": 0.22687924354850644, "learning_rate": 4.340758765340963e-05, "loss": 0.3105, "step": 2150 }, { "epoch": 0.9438496831986017, "grad_norm": 0.22976444032766097, "learning_rate": 4.3321356527294227e-05, "loss": 0.3109, "step": 2160 }, { "epoch": 0.9482193576578545, "grad_norm": 0.25920931117976254, "learning_rate": 4.3234651945329976e-05, "loss": 0.3005, "step": 2170 }, { "epoch": 0.9525890321171073, "grad_norm": 0.25656535381933304, "learning_rate": 4.314747614811471e-05, "loss": 0.2997, "step": 2180 }, { "epoch": 0.9569587065763601, "grad_norm": 0.2555756515758645, "learning_rate": 4.305983138842325e-05, "loss": 0.2982, "step": 2190 }, { "epoch": 0.9613283810356128, "grad_norm": 0.2568431025737744, "learning_rate": 4.297171993114927e-05, "loss": 0.3088, "step": 2200 }, { "epoch": 0.9656980554948656, "grad_norm": 0.32860647579491037, "learning_rate": 4.288314405324668e-05, "loss": 0.3047, "step": 2210 }, { "epoch": 0.9700677299541184, "grad_norm": 0.2881943240133001, "learning_rate": 4.279410604367088e-05, "loss": 0.3101, "step": 2220 }, { "epoch": 0.9744374044133712, "grad_norm": 0.24463497232043155, "learning_rate": 4.2704608203319505e-05, "loss": 0.3092, "step": 2230 }, { "epoch": 0.978807078872624, "grad_norm": 0.24927261109753046, "learning_rate": 4.261465284497307e-05, "loss": 0.2905, "step": 2240 }, { "epoch": 0.9831767533318768, "grad_norm": 0.25555639280904047, "learning_rate": 4.252424229323512e-05, "loss": 0.3045, "step": 2250 }, { "epoch": 0.9875464277911296, "grad_norm": 0.2603444894123934, "learning_rate": 4.2433378884472196e-05, "loss": 0.292, "step": 2260 }, { "epoch": 0.9919161022503823, "grad_norm": 0.26963858720260425, "learning_rate": 4.234206496675348e-05, "loss": 0.3096, "step": 2270 }, { "epoch": 0.9962857767096351, "grad_norm": 0.3298881111802689, "learning_rate": 4.225030289979006e-05, "loss": 0.2993, "step": 2280 }, { "epoch": 1.0004369674459253, "grad_norm": 0.38079926289714233, "learning_rate": 4.215809505487402e-05, "loss": 0.2904, "step": 2290 }, { "epoch": 1.004806641905178, "grad_norm": 0.2255679963046686, "learning_rate": 4.2065443814817075e-05, "loss": 0.2664, "step": 2300 }, { "epoch": 1.009176316364431, "grad_norm": 0.2601750082000646, "learning_rate": 4.19723515738891e-05, "loss": 0.2749, "step": 2310 }, { "epoch": 1.0135459908236837, "grad_norm": 0.3482203756456138, "learning_rate": 4.1878820737756154e-05, "loss": 0.2747, "step": 2320 }, { "epoch": 1.0179156652829364, "grad_norm": 0.25592594445669803, "learning_rate": 4.1784853723418394e-05, "loss": 0.2634, "step": 2330 }, { "epoch": 1.0222853397421892, "grad_norm": 0.23805283200590813, "learning_rate": 4.1690452959147575e-05, "loss": 0.2726, "step": 2340 }, { "epoch": 1.026655014201442, "grad_norm": 0.2667694721703153, "learning_rate": 4.15956208844243e-05, "loss": 0.2698, "step": 2350 }, { "epoch": 1.0310246886606949, "grad_norm": 0.27229044170106353, "learning_rate": 4.1500359949875e-05, "loss": 0.2638, "step": 2360 }, { "epoch": 1.0353943631199476, "grad_norm": 0.21290804766380314, "learning_rate": 4.1404672617208586e-05, "loss": 0.2647, "step": 2370 }, { "epoch": 1.0397640375792003, "grad_norm": 0.19860864441296505, "learning_rate": 4.130856135915282e-05, "loss": 0.2675, "step": 2380 }, { "epoch": 1.044133712038453, "grad_norm": 0.22365230175166062, "learning_rate": 4.121202865939049e-05, "loss": 0.2583, "step": 2390 }, { "epoch": 1.048503386497706, "grad_norm": 0.29691915429323845, "learning_rate": 4.111507701249513e-05, "loss": 0.2634, "step": 2400 }, { "epoch": 1.0528730609569588, "grad_norm": 0.25558502663468957, "learning_rate": 4.1017708923866595e-05, "loss": 0.2658, "step": 2410 }, { "epoch": 1.0572427354162115, "grad_norm": 0.21526509762412646, "learning_rate": 4.0919926909666364e-05, "loss": 0.2663, "step": 2420 }, { "epoch": 1.0616124098754642, "grad_norm": 0.24947253214008783, "learning_rate": 4.0821733496752426e-05, "loss": 0.2625, "step": 2430 }, { "epoch": 1.065982084334717, "grad_norm": 0.2725574211425989, "learning_rate": 4.072313122261406e-05, "loss": 0.2627, "step": 2440 }, { "epoch": 1.07035175879397, "grad_norm": 0.23216434340822648, "learning_rate": 4.062412263530624e-05, "loss": 0.2607, "step": 2450 }, { "epoch": 1.0747214332532227, "grad_norm": 0.231598735381825, "learning_rate": 4.052471029338375e-05, "loss": 0.2691, "step": 2460 }, { "epoch": 1.0790911077124754, "grad_norm": 0.2241510410201989, "learning_rate": 4.042489676583514e-05, "loss": 0.2683, "step": 2470 }, { "epoch": 1.0834607821717281, "grad_norm": 0.22777647198588957, "learning_rate": 4.032468463201625e-05, "loss": 0.2681, "step": 2480 }, { "epoch": 1.0878304566309809, "grad_norm": 0.2048852902728067, "learning_rate": 4.022407648158365e-05, "loss": 0.261, "step": 2490 }, { "epoch": 1.0922001310902338, "grad_norm": 0.21311944885772335, "learning_rate": 4.0123074914427636e-05, "loss": 0.2599, "step": 2500 }, { "epoch": 1.0965698055494866, "grad_norm": 0.23372317327020886, "learning_rate": 4.0021682540605085e-05, "loss": 0.2611, "step": 2510 }, { "epoch": 1.1009394800087393, "grad_norm": 0.22373708771672665, "learning_rate": 3.991990198027202e-05, "loss": 0.2576, "step": 2520 }, { "epoch": 1.105309154467992, "grad_norm": 0.19798357910753747, "learning_rate": 3.9817735863615866e-05, "loss": 0.2643, "step": 2530 }, { "epoch": 1.109678828927245, "grad_norm": 0.2374312698523212, "learning_rate": 3.9715186830787496e-05, "loss": 0.2705, "step": 2540 }, { "epoch": 1.1140485033864977, "grad_norm": 0.22775645009974038, "learning_rate": 3.961225753183302e-05, "loss": 0.2598, "step": 2550 }, { "epoch": 1.1184181778457505, "grad_norm": 0.2190290494139683, "learning_rate": 3.950895062662524e-05, "loss": 0.2642, "step": 2560 }, { "epoch": 1.1227878523050032, "grad_norm": 0.21441689264750247, "learning_rate": 3.9405268784795046e-05, "loss": 0.2657, "step": 2570 }, { "epoch": 1.1271575267642562, "grad_norm": 0.2437747529044389, "learning_rate": 3.930121468566227e-05, "loss": 0.2683, "step": 2580 }, { "epoch": 1.131527201223509, "grad_norm": 0.2428616188098313, "learning_rate": 3.9196791018166566e-05, "loss": 0.2705, "step": 2590 }, { "epoch": 1.1358968756827617, "grad_norm": 0.22168961342082932, "learning_rate": 3.909200048079786e-05, "loss": 0.2606, "step": 2600 }, { "epoch": 1.1402665501420144, "grad_norm": 0.23303745431696796, "learning_rate": 3.8986845781526645e-05, "loss": 0.2675, "step": 2610 }, { "epoch": 1.1446362246012671, "grad_norm": 0.2040339689571263, "learning_rate": 3.8881329637733995e-05, "loss": 0.2643, "step": 2620 }, { "epoch": 1.14900589906052, "grad_norm": 0.22451105505575797, "learning_rate": 3.877545477614134e-05, "loss": 0.2607, "step": 2630 }, { "epoch": 1.1533755735197728, "grad_norm": 0.29162439467592566, "learning_rate": 3.8669223932739985e-05, "loss": 0.2783, "step": 2640 }, { "epoch": 1.1577452479790256, "grad_norm": 0.2657294242225117, "learning_rate": 3.856263985272046e-05, "loss": 0.2703, "step": 2650 }, { "epoch": 1.1621149224382783, "grad_norm": 0.2373288899156919, "learning_rate": 3.845570529040151e-05, "loss": 0.273, "step": 2660 }, { "epoch": 1.166484596897531, "grad_norm": 0.2056612893995306, "learning_rate": 3.8348423009159004e-05, "loss": 0.2607, "step": 2670 }, { "epoch": 1.170854271356784, "grad_norm": 0.21684923831476527, "learning_rate": 3.824079578135442e-05, "loss": 0.2643, "step": 2680 }, { "epoch": 1.1752239458160367, "grad_norm": 0.21070625788977196, "learning_rate": 3.81328263882633e-05, "loss": 0.2531, "step": 2690 }, { "epoch": 1.1795936202752895, "grad_norm": 0.21664620864901138, "learning_rate": 3.802451762000331e-05, "loss": 0.2653, "step": 2700 }, { "epoch": 1.1839632947345422, "grad_norm": 0.26140109478241513, "learning_rate": 3.791587227546217e-05, "loss": 0.275, "step": 2710 }, { "epoch": 1.188332969193795, "grad_norm": 0.21987139354718674, "learning_rate": 3.780689316222533e-05, "loss": 0.2685, "step": 2720 }, { "epoch": 1.192702643653048, "grad_norm": 0.23992643610247466, "learning_rate": 3.7697583096503375e-05, "loss": 0.2695, "step": 2730 }, { "epoch": 1.1970723181123006, "grad_norm": 0.20915905180397207, "learning_rate": 3.758794490305932e-05, "loss": 0.2693, "step": 2740 }, { "epoch": 1.2014419925715534, "grad_norm": 0.20540976488862947, "learning_rate": 3.747798141513554e-05, "loss": 0.2567, "step": 2750 }, { "epoch": 1.2058116670308061, "grad_norm": 0.20923211990566692, "learning_rate": 3.7367695474380624e-05, "loss": 0.2661, "step": 2760 }, { "epoch": 1.210181341490059, "grad_norm": 0.23642547457204466, "learning_rate": 3.725708993077586e-05, "loss": 0.2814, "step": 2770 }, { "epoch": 1.2145510159493118, "grad_norm": 0.21892121482993648, "learning_rate": 3.714616764256166e-05, "loss": 0.2604, "step": 2780 }, { "epoch": 1.2189206904085645, "grad_norm": 0.2024276201190825, "learning_rate": 3.703493147616367e-05, "loss": 0.2673, "step": 2790 }, { "epoch": 1.2232903648678173, "grad_norm": 0.22976236995873137, "learning_rate": 3.692338430611869e-05, "loss": 0.2703, "step": 2800 }, { "epoch": 1.2276600393270702, "grad_norm": 0.19216988276771937, "learning_rate": 3.68115290150004e-05, "loss": 0.2579, "step": 2810 }, { "epoch": 1.232029713786323, "grad_norm": 0.19440128326379744, "learning_rate": 3.6699368493344856e-05, "loss": 0.2731, "step": 2820 }, { "epoch": 1.2363993882455757, "grad_norm": 0.2481497202537707, "learning_rate": 3.658690563957582e-05, "loss": 0.2734, "step": 2830 }, { "epoch": 1.2407690627048285, "grad_norm": 0.3381697656340553, "learning_rate": 3.647414335992985e-05, "loss": 0.2738, "step": 2840 }, { "epoch": 1.2451387371640812, "grad_norm": 0.20436444573836007, "learning_rate": 3.636108456838115e-05, "loss": 0.2567, "step": 2850 }, { "epoch": 1.2495084116233341, "grad_norm": 0.2351279489083201, "learning_rate": 3.624773218656636e-05, "loss": 0.2626, "step": 2860 }, { "epoch": 1.2538780860825869, "grad_norm": 0.2301071824114486, "learning_rate": 3.613408914370897e-05, "loss": 0.2648, "step": 2870 }, { "epoch": 1.2582477605418396, "grad_norm": 0.19953603043990906, "learning_rate": 3.602015837654369e-05, "loss": 0.2658, "step": 2880 }, { "epoch": 1.2626174350010924, "grad_norm": 0.1892353085889492, "learning_rate": 3.5905942829240505e-05, "loss": 0.2637, "step": 2890 }, { "epoch": 1.266987109460345, "grad_norm": 0.22481202493591268, "learning_rate": 3.57914454533286e-05, "loss": 0.2597, "step": 2900 }, { "epoch": 1.271356783919598, "grad_norm": 0.22336406364798586, "learning_rate": 3.567666920762015e-05, "loss": 0.2644, "step": 2910 }, { "epoch": 1.2757264583788508, "grad_norm": 0.20208174526992362, "learning_rate": 3.556161705813378e-05, "loss": 0.2659, "step": 2920 }, { "epoch": 1.2800961328381035, "grad_norm": 0.21603865257111177, "learning_rate": 3.5446291978017955e-05, "loss": 0.2651, "step": 2930 }, { "epoch": 1.2844658072973563, "grad_norm": 0.23540755364752858, "learning_rate": 3.5330696947474154e-05, "loss": 0.2658, "step": 2940 }, { "epoch": 1.288835481756609, "grad_norm": 0.24821002926831592, "learning_rate": 3.521483495367983e-05, "loss": 0.2728, "step": 2950 }, { "epoch": 1.293205156215862, "grad_norm": 0.21013922636681626, "learning_rate": 3.5098708990711255e-05, "loss": 0.2704, "step": 2960 }, { "epoch": 1.2975748306751147, "grad_norm": 0.21304919224900393, "learning_rate": 3.49823220594661e-05, "loss": 0.2618, "step": 2970 }, { "epoch": 1.3019445051343674, "grad_norm": 0.24975667711364985, "learning_rate": 3.486567716758594e-05, "loss": 0.2626, "step": 2980 }, { "epoch": 1.3063141795936204, "grad_norm": 0.2311154698726339, "learning_rate": 3.474877732937848e-05, "loss": 0.267, "step": 2990 }, { "epoch": 1.3106838540528731, "grad_norm": 0.19422172620264047, "learning_rate": 3.4631625565739686e-05, "loss": 0.2554, "step": 3000 }, { "epoch": 1.3150535285121259, "grad_norm": 0.23668908507343714, "learning_rate": 3.4514224904075725e-05, "loss": 0.262, "step": 3010 }, { "epoch": 1.3194232029713786, "grad_norm": 0.2425545121489129, "learning_rate": 3.439657837822473e-05, "loss": 0.2712, "step": 3020 }, { "epoch": 1.3237928774306313, "grad_norm": 0.20816614733274266, "learning_rate": 3.427868902837838e-05, "loss": 0.2618, "step": 3030 }, { "epoch": 1.3281625518898843, "grad_norm": 0.21426905743676083, "learning_rate": 3.416055990100336e-05, "loss": 0.264, "step": 3040 }, { "epoch": 1.332532226349137, "grad_norm": 0.25308261227305223, "learning_rate": 3.404219404876261e-05, "loss": 0.2718, "step": 3050 }, { "epoch": 1.3369019008083898, "grad_norm": 0.2048445674100599, "learning_rate": 3.392359453043648e-05, "loss": 0.2633, "step": 3060 }, { "epoch": 1.3412715752676425, "grad_norm": 0.18925963744201388, "learning_rate": 3.380476441084363e-05, "loss": 0.2595, "step": 3070 }, { "epoch": 1.3456412497268952, "grad_norm": 0.19086352803124243, "learning_rate": 3.368570676076187e-05, "loss": 0.2603, "step": 3080 }, { "epoch": 1.3500109241861482, "grad_norm": 0.2198440984843584, "learning_rate": 3.356642465684881e-05, "loss": 0.2664, "step": 3090 }, { "epoch": 1.354380598645401, "grad_norm": 0.18680859345628353, "learning_rate": 3.344692118156232e-05, "loss": 0.2636, "step": 3100 }, { "epoch": 1.3587502731046537, "grad_norm": 0.22395752633344654, "learning_rate": 3.332719942308092e-05, "loss": 0.2612, "step": 3110 }, { "epoch": 1.3631199475639064, "grad_norm": 0.22098568814409636, "learning_rate": 3.320726247522391e-05, "loss": 0.2699, "step": 3120 }, { "epoch": 1.3674896220231592, "grad_norm": 0.22968070913382407, "learning_rate": 3.308711343737148e-05, "loss": 0.2661, "step": 3130 }, { "epoch": 1.3718592964824121, "grad_norm": 0.26008237643124177, "learning_rate": 3.2966755414384606e-05, "loss": 0.2651, "step": 3140 }, { "epoch": 1.3762289709416649, "grad_norm": 0.24063805005803016, "learning_rate": 3.284619151652479e-05, "loss": 0.2731, "step": 3150 }, { "epoch": 1.3805986454009176, "grad_norm": 0.2028498732442211, "learning_rate": 3.272542485937369e-05, "loss": 0.2695, "step": 3160 }, { "epoch": 1.3849683198601705, "grad_norm": 0.21864876909957637, "learning_rate": 3.260445856375265e-05, "loss": 0.2578, "step": 3170 }, { "epoch": 1.3893379943194233, "grad_norm": 0.20486326566647226, "learning_rate": 3.248329575564198e-05, "loss": 0.2617, "step": 3180 }, { "epoch": 1.393707668778676, "grad_norm": 0.22685762331857176, "learning_rate": 3.236193956610027e-05, "loss": 0.2702, "step": 3190 }, { "epoch": 1.3980773432379288, "grad_norm": 0.24449078404174052, "learning_rate": 3.224039313118338e-05, "loss": 0.2663, "step": 3200 }, { "epoch": 1.4024470176971815, "grad_norm": 0.20927962147486293, "learning_rate": 3.211865959186345e-05, "loss": 0.2633, "step": 3210 }, { "epoch": 1.4068166921564345, "grad_norm": 0.19147038192773722, "learning_rate": 3.1996742093947724e-05, "loss": 0.2553, "step": 3220 }, { "epoch": 1.4111863666156872, "grad_norm": 0.20878617528947072, "learning_rate": 3.187464378799728e-05, "loss": 0.2693, "step": 3230 }, { "epoch": 1.41555604107494, "grad_norm": 0.20080543211149518, "learning_rate": 3.175236782924556e-05, "loss": 0.2577, "step": 3240 }, { "epoch": 1.4199257155341927, "grad_norm": 0.18004846608679465, "learning_rate": 3.1629917377516886e-05, "loss": 0.2598, "step": 3250 }, { "epoch": 1.4242953899934454, "grad_norm": 0.21396842573353922, "learning_rate": 3.150729559714478e-05, "loss": 0.2767, "step": 3260 }, { "epoch": 1.4286650644526984, "grad_norm": 0.1923072416961986, "learning_rate": 3.138450565689019e-05, "loss": 0.2699, "step": 3270 }, { "epoch": 1.433034738911951, "grad_norm": 0.19507668449483362, "learning_rate": 3.12615507298596e-05, "loss": 0.2578, "step": 3280 }, { "epoch": 1.4374044133712038, "grad_norm": 0.2209013816461278, "learning_rate": 3.1138433993423067e-05, "loss": 0.2583, "step": 3290 }, { "epoch": 1.4417740878304566, "grad_norm": 0.21505689357418759, "learning_rate": 3.101515862913206e-05, "loss": 0.2604, "step": 3300 }, { "epoch": 1.4461437622897093, "grad_norm": 0.2120081735040409, "learning_rate": 3.089172782263728e-05, "loss": 0.2647, "step": 3310 }, { "epoch": 1.4505134367489623, "grad_norm": 0.21281708106812247, "learning_rate": 3.0768144763606336e-05, "loss": 0.2612, "step": 3320 }, { "epoch": 1.454883111208215, "grad_norm": 0.21437704799146978, "learning_rate": 3.064441264564128e-05, "loss": 0.2645, "step": 3330 }, { "epoch": 1.4592527856674677, "grad_norm": 0.19862838211376507, "learning_rate": 3.052053466619613e-05, "loss": 0.2602, "step": 3340 }, { "epoch": 1.4636224601267205, "grad_norm": 0.19494679236440912, "learning_rate": 3.0396514026494223e-05, "loss": 0.2614, "step": 3350 }, { "epoch": 1.4679921345859732, "grad_norm": 0.21500608386986342, "learning_rate": 3.027235393144547e-05, "loss": 0.2535, "step": 3360 }, { "epoch": 1.4723618090452262, "grad_norm": 0.21744607767681662, "learning_rate": 3.014805758956355e-05, "loss": 0.2702, "step": 3370 }, { "epoch": 1.476731483504479, "grad_norm": 0.19546989752420724, "learning_rate": 3.002362821288302e-05, "loss": 0.2626, "step": 3380 }, { "epoch": 1.4811011579637317, "grad_norm": 0.23349607832599673, "learning_rate": 2.989906901687627e-05, "loss": 0.2638, "step": 3390 }, { "epoch": 1.4854708324229846, "grad_norm": 0.2059971151100944, "learning_rate": 2.9774383220370455e-05, "loss": 0.2485, "step": 3400 }, { "epoch": 1.4898405068822373, "grad_norm": 0.23055781158069324, "learning_rate": 2.9649574045464307e-05, "loss": 0.2602, "step": 3410 }, { "epoch": 1.49421018134149, "grad_norm": 0.20627609615535697, "learning_rate": 2.9524644717444867e-05, "loss": 0.2576, "step": 3420 }, { "epoch": 1.4985798558007428, "grad_norm": 0.20913390610964816, "learning_rate": 2.939959846470415e-05, "loss": 0.2599, "step": 3430 }, { "epoch": 1.5029495302599956, "grad_norm": 0.20636051739131434, "learning_rate": 2.92744385186557e-05, "loss": 0.2594, "step": 3440 }, { "epoch": 1.5073192047192485, "grad_norm": 0.19140800764182006, "learning_rate": 2.9149168113651104e-05, "loss": 0.2658, "step": 3450 }, { "epoch": 1.5116888791785013, "grad_norm": 0.2032210431914496, "learning_rate": 2.9023790486896403e-05, "loss": 0.2563, "step": 3460 }, { "epoch": 1.516058553637754, "grad_norm": 0.17353828880925498, "learning_rate": 2.8898308878368426e-05, "loss": 0.2459, "step": 3470 }, { "epoch": 1.5204282280970067, "grad_norm": 0.2936069666923771, "learning_rate": 2.877272653073107e-05, "loss": 0.2638, "step": 3480 }, { "epoch": 1.5247979025562595, "grad_norm": 0.21275909476570462, "learning_rate": 2.8647046689251524e-05, "loss": 0.2561, "step": 3490 }, { "epoch": 1.5291675770155124, "grad_norm": 0.2270172007018007, "learning_rate": 2.8521272601716377e-05, "loss": 0.2663, "step": 3500 }, { "epoch": 1.5335372514747652, "grad_norm": 0.22663439305551406, "learning_rate": 2.8395407518347682e-05, "loss": 0.2615, "step": 3510 }, { "epoch": 1.537906925934018, "grad_norm": 0.202671231223619, "learning_rate": 2.8269454691719026e-05, "loss": 0.2557, "step": 3520 }, { "epoch": 1.5422766003932709, "grad_norm": 0.2562468888209313, "learning_rate": 2.814341737667141e-05, "loss": 0.2588, "step": 3530 }, { "epoch": 1.5466462748525234, "grad_norm": 0.3662647709909802, "learning_rate": 2.801729883022915e-05, "loss": 0.2608, "step": 3540 }, { "epoch": 1.5510159493117763, "grad_norm": 0.19165382212715262, "learning_rate": 2.789110231151575e-05, "loss": 0.264, "step": 3550 }, { "epoch": 1.555385623771029, "grad_norm": 0.20347537896414233, "learning_rate": 2.7764831081669633e-05, "loss": 0.2706, "step": 3560 }, { "epoch": 1.5597552982302818, "grad_norm": 0.21207776671204426, "learning_rate": 2.7638488403759905e-05, "loss": 0.2559, "step": 3570 }, { "epoch": 1.5641249726895348, "grad_norm": 0.19640251514087312, "learning_rate": 2.7512077542702003e-05, "loss": 0.2578, "step": 3580 }, { "epoch": 1.5684946471487873, "grad_norm": 0.2028991496606779, "learning_rate": 2.7385601765173336e-05, "loss": 0.2571, "step": 3590 }, { "epoch": 1.5728643216080402, "grad_norm": 0.1856743694376068, "learning_rate": 2.7259064339528874e-05, "loss": 0.2488, "step": 3600 }, { "epoch": 1.577233996067293, "grad_norm": 0.2457064470133504, "learning_rate": 2.7132468535716678e-05, "loss": 0.2551, "step": 3610 }, { "epoch": 1.5816036705265457, "grad_norm": 0.20779481526227436, "learning_rate": 2.7005817625193397e-05, "loss": 0.2591, "step": 3620 }, { "epoch": 1.5859733449857987, "grad_norm": 0.2057721439702309, "learning_rate": 2.6879114880839757e-05, "loss": 0.2607, "step": 3630 }, { "epoch": 1.5903430194450512, "grad_norm": 0.1998799573650709, "learning_rate": 2.6752363576875933e-05, "loss": 0.2613, "step": 3640 }, { "epoch": 1.5947126939043041, "grad_norm": 0.2082241966847959, "learning_rate": 2.6625566988776974e-05, "loss": 0.2618, "step": 3650 }, { "epoch": 1.5990823683635569, "grad_norm": 0.2688614869913761, "learning_rate": 2.6498728393188155e-05, "loss": 0.2599, "step": 3660 }, { "epoch": 1.6034520428228096, "grad_norm": 0.18461373636795747, "learning_rate": 2.6371851067840302e-05, "loss": 0.2641, "step": 3670 }, { "epoch": 1.6078217172820626, "grad_norm": 0.1929875805388407, "learning_rate": 2.624493829146507e-05, "loss": 0.2669, "step": 3680 }, { "epoch": 1.6121913917413153, "grad_norm": 0.20268394891127728, "learning_rate": 2.6117993343710246e-05, "loss": 0.256, "step": 3690 }, { "epoch": 1.616561066200568, "grad_norm": 0.21393694357661316, "learning_rate": 2.5991019505054963e-05, "loss": 0.2563, "step": 3700 }, { "epoch": 1.6209307406598208, "grad_norm": 0.18887446218025103, "learning_rate": 2.586402005672497e-05, "loss": 0.2522, "step": 3710 }, { "epoch": 1.6253004151190735, "grad_norm": 0.21595530414031858, "learning_rate": 2.5736998280607798e-05, "loss": 0.2613, "step": 3720 }, { "epoch": 1.6296700895783265, "grad_norm": 0.18922395029797706, "learning_rate": 2.5609957459167978e-05, "loss": 0.2575, "step": 3730 }, { "epoch": 1.6340397640375792, "grad_norm": 0.19112241209643366, "learning_rate": 2.548290087536218e-05, "loss": 0.2568, "step": 3740 }, { "epoch": 1.638409438496832, "grad_norm": 0.22587364268663074, "learning_rate": 2.5355831812554437e-05, "loss": 0.2622, "step": 3750 }, { "epoch": 1.642779112956085, "grad_norm": 0.18688321962096968, "learning_rate": 2.522875355443124e-05, "loss": 0.2583, "step": 3760 }, { "epoch": 1.6471487874153374, "grad_norm": 0.18169858700523117, "learning_rate": 2.5101669384916697e-05, "loss": 0.2646, "step": 3770 }, { "epoch": 1.6515184618745904, "grad_norm": 0.22335035223791527, "learning_rate": 2.4974582588087698e-05, "loss": 0.2549, "step": 3780 }, { "epoch": 1.6558881363338431, "grad_norm": 0.19252919635210475, "learning_rate": 2.484749644808902e-05, "loss": 0.2555, "step": 3790 }, { "epoch": 1.6602578107930959, "grad_norm": 0.21321241081815587, "learning_rate": 2.4720414249048442e-05, "loss": 0.2545, "step": 3800 }, { "epoch": 1.6646274852523488, "grad_norm": 0.19828003415897408, "learning_rate": 2.459333927499193e-05, "loss": 0.264, "step": 3810 }, { "epoch": 1.6689971597116013, "grad_norm": 0.19906529314780358, "learning_rate": 2.4466274809758756e-05, "loss": 0.2552, "step": 3820 }, { "epoch": 1.6733668341708543, "grad_norm": 0.194508486417613, "learning_rate": 2.4339224136916572e-05, "loss": 0.2536, "step": 3830 }, { "epoch": 1.677736508630107, "grad_norm": 0.19344412105367936, "learning_rate": 2.4212190539676666e-05, "loss": 0.2509, "step": 3840 }, { "epoch": 1.6821061830893598, "grad_norm": 0.2148479324113871, "learning_rate": 2.4085177300809038e-05, "loss": 0.2572, "step": 3850 }, { "epoch": 1.6864758575486127, "grad_norm": 0.2063112145281865, "learning_rate": 2.395818770255759e-05, "loss": 0.2563, "step": 3860 }, { "epoch": 1.6908455320078652, "grad_norm": 0.21443765436710072, "learning_rate": 2.3831225026555308e-05, "loss": 0.2533, "step": 3870 }, { "epoch": 1.6952152064671182, "grad_norm": 0.22144028468625743, "learning_rate": 2.3704292553739486e-05, "loss": 0.2598, "step": 3880 }, { "epoch": 1.699584880926371, "grad_norm": 0.20759459246891473, "learning_rate": 2.357739356426687e-05, "loss": 0.2524, "step": 3890 }, { "epoch": 1.7039545553856237, "grad_norm": 0.19952693001909003, "learning_rate": 2.3450531337428978e-05, "loss": 0.25, "step": 3900 }, { "epoch": 1.7083242298448766, "grad_norm": 0.202331799126949, "learning_rate": 2.332370915156731e-05, "loss": 0.2565, "step": 3910 }, { "epoch": 1.7126939043041294, "grad_norm": 0.18245339773863017, "learning_rate": 2.3196930283988602e-05, "loss": 0.2601, "step": 3920 }, { "epoch": 1.7170635787633821, "grad_norm": 0.19719505857947925, "learning_rate": 2.307019801088022e-05, "loss": 0.2577, "step": 3930 }, { "epoch": 1.721433253222635, "grad_norm": 0.1825996521424925, "learning_rate": 2.294351560722542e-05, "loss": 0.2533, "step": 3940 }, { "epoch": 1.7258029276818876, "grad_norm": 0.190884343135347, "learning_rate": 2.281688634671874e-05, "loss": 0.2661, "step": 3950 }, { "epoch": 1.7301726021411405, "grad_norm": 0.19435746249585753, "learning_rate": 2.2690313501681427e-05, "loss": 0.2584, "step": 3960 }, { "epoch": 1.7345422766003933, "grad_norm": 0.27047320683777537, "learning_rate": 2.256380034297684e-05, "loss": 0.2601, "step": 3970 }, { "epoch": 1.738911951059646, "grad_norm": 0.18869874921016083, "learning_rate": 2.243735013992593e-05, "loss": 0.2477, "step": 3980 }, { "epoch": 1.743281625518899, "grad_norm": 0.17621543010063825, "learning_rate": 2.2310966160222784e-05, "loss": 0.2534, "step": 3990 }, { "epoch": 1.7476512999781515, "grad_norm": 0.19442141837383498, "learning_rate": 2.2184651669850162e-05, "loss": 0.251, "step": 4000 }, { "epoch": 1.7520209744374045, "grad_norm": 0.19095436328327042, "learning_rate": 2.2058409932995065e-05, "loss": 0.2719, "step": 4010 }, { "epoch": 1.7563906488966572, "grad_norm": 0.20961915473293014, "learning_rate": 2.1932244211964456e-05, "loss": 0.2624, "step": 4020 }, { "epoch": 1.76076032335591, "grad_norm": 0.19712709417400562, "learning_rate": 2.1806157767100905e-05, "loss": 0.2608, "step": 4030 }, { "epoch": 1.7651299978151629, "grad_norm": 0.20535622778489315, "learning_rate": 2.1680153856698326e-05, "loss": 0.2453, "step": 4040 }, { "epoch": 1.7694996722744154, "grad_norm": 0.2035902621859857, "learning_rate": 2.1554235736917838e-05, "loss": 0.2569, "step": 4050 }, { "epoch": 1.7738693467336684, "grad_norm": 0.20980234053003394, "learning_rate": 2.142840666170354e-05, "loss": 0.2626, "step": 4060 }, { "epoch": 1.778239021192921, "grad_norm": 0.19368599267792738, "learning_rate": 2.1302669882698483e-05, "loss": 0.2498, "step": 4070 }, { "epoch": 1.7826086956521738, "grad_norm": 0.18071559574353896, "learning_rate": 2.117702864916063e-05, "loss": 0.2571, "step": 4080 }, { "epoch": 1.7869783701114268, "grad_norm": 0.21458010979547673, "learning_rate": 2.105148620787888e-05, "loss": 0.2526, "step": 4090 }, { "epoch": 1.7913480445706795, "grad_norm": 0.22616881431522837, "learning_rate": 2.0926045803089138e-05, "loss": 0.265, "step": 4100 }, { "epoch": 1.7957177190299323, "grad_norm": 0.19890665856443968, "learning_rate": 2.0800710676390543e-05, "loss": 0.2649, "step": 4110 }, { "epoch": 1.800087393489185, "grad_norm": 0.17682432179771598, "learning_rate": 2.0675484066661664e-05, "loss": 0.2604, "step": 4120 }, { "epoch": 1.8044570679484377, "grad_norm": 0.20320914697663858, "learning_rate": 2.0550369209976775e-05, "loss": 0.2563, "step": 4130 }, { "epoch": 1.8088267424076907, "grad_norm": 0.20590108875725602, "learning_rate": 2.0425369339522275e-05, "loss": 0.2591, "step": 4140 }, { "epoch": 1.8131964168669434, "grad_norm": 0.2319118462107837, "learning_rate": 2.0300487685513127e-05, "loss": 0.2639, "step": 4150 }, { "epoch": 1.8175660913261962, "grad_norm": 0.19092393662348517, "learning_rate": 2.017572747510934e-05, "loss": 0.2596, "step": 4160 }, { "epoch": 1.8219357657854491, "grad_norm": 0.1679728210268398, "learning_rate": 2.0051091932332645e-05, "loss": 0.2548, "step": 4170 }, { "epoch": 1.8263054402447016, "grad_norm": 0.17423669353301824, "learning_rate": 1.992658427798313e-05, "loss": 0.2459, "step": 4180 }, { "epoch": 1.8306751147039546, "grad_norm": 0.21258269158244422, "learning_rate": 1.980220772955602e-05, "loss": 0.2538, "step": 4190 }, { "epoch": 1.8350447891632073, "grad_norm": 0.20979029633680682, "learning_rate": 1.9677965501158534e-05, "loss": 0.2636, "step": 4200 }, { "epoch": 1.83941446362246, "grad_norm": 0.19384700550593356, "learning_rate": 1.955386080342683e-05, "loss": 0.2584, "step": 4210 }, { "epoch": 1.843784138081713, "grad_norm": 0.20279862244923488, "learning_rate": 1.9429896843443023e-05, "loss": 0.2631, "step": 4220 }, { "epoch": 1.8481538125409656, "grad_norm": 0.1824578664689937, "learning_rate": 1.930607682465234e-05, "loss": 0.2507, "step": 4230 }, { "epoch": 1.8525234870002185, "grad_norm": 0.17431928814837297, "learning_rate": 1.9182403946780316e-05, "loss": 0.2509, "step": 4240 }, { "epoch": 1.8568931614594713, "grad_norm": 0.17988192959570887, "learning_rate": 1.9058881405750065e-05, "loss": 0.2572, "step": 4250 }, { "epoch": 1.861262835918724, "grad_norm": 0.18241891945487387, "learning_rate": 1.8935512393599783e-05, "loss": 0.2519, "step": 4260 }, { "epoch": 1.865632510377977, "grad_norm": 0.18134882625693402, "learning_rate": 1.8812300098400193e-05, "loss": 0.2548, "step": 4270 }, { "epoch": 1.8700021848372295, "grad_norm": 0.15950257782600052, "learning_rate": 1.8689247704172187e-05, "loss": 0.2522, "step": 4280 }, { "epoch": 1.8743718592964824, "grad_norm": 0.2121179489033308, "learning_rate": 1.8566358390804513e-05, "loss": 0.2524, "step": 4290 }, { "epoch": 1.8787415337557352, "grad_norm": 0.18725948492764516, "learning_rate": 1.8443635333971643e-05, "loss": 0.2492, "step": 4300 }, { "epoch": 1.883111208214988, "grad_norm": 0.1816822114623371, "learning_rate": 1.8321081705051707e-05, "loss": 0.2581, "step": 4310 }, { "epoch": 1.8874808826742409, "grad_norm": 0.20210446611773902, "learning_rate": 1.8198700671044477e-05, "loss": 0.2568, "step": 4320 }, { "epoch": 1.8918505571334936, "grad_norm": 0.2291538711655988, "learning_rate": 1.8076495394489605e-05, "loss": 0.2517, "step": 4330 }, { "epoch": 1.8962202315927463, "grad_norm": 0.201263140803902, "learning_rate": 1.795446903338487e-05, "loss": 0.2545, "step": 4340 }, { "epoch": 1.9005899060519993, "grad_norm": 0.21692075424280524, "learning_rate": 1.7832624741104526e-05, "loss": 0.2551, "step": 4350 }, { "epoch": 1.9049595805112518, "grad_norm": 0.19176389162742902, "learning_rate": 1.771096566631788e-05, "loss": 0.2594, "step": 4360 }, { "epoch": 1.9093292549705048, "grad_norm": 0.208878353320417, "learning_rate": 1.7589494952907897e-05, "loss": 0.2581, "step": 4370 }, { "epoch": 1.9136989294297575, "grad_norm": 0.22037269811805654, "learning_rate": 1.746821573988994e-05, "loss": 0.2562, "step": 4380 }, { "epoch": 1.9180686038890102, "grad_norm": 0.21401377531726035, "learning_rate": 1.734713116133067e-05, "loss": 0.2588, "step": 4390 }, { "epoch": 1.9224382783482632, "grad_norm": 0.2062036418879069, "learning_rate": 1.7226244346267064e-05, "loss": 0.253, "step": 4400 }, { "epoch": 1.9268079528075157, "grad_norm": 0.17442658987622695, "learning_rate": 1.710555841862553e-05, "loss": 0.255, "step": 4410 }, { "epoch": 1.9311776272667687, "grad_norm": 0.2120783540688887, "learning_rate": 1.698507649714121e-05, "loss": 0.2585, "step": 4420 }, { "epoch": 1.9355473017260214, "grad_norm": 0.18993488021952257, "learning_rate": 1.686480169527737e-05, "loss": 0.2569, "step": 4430 }, { "epoch": 1.9399169761852741, "grad_norm": 0.21107757382996598, "learning_rate": 1.674473712114492e-05, "loss": 0.2541, "step": 4440 }, { "epoch": 1.944286650644527, "grad_norm": 0.17232917352066968, "learning_rate": 1.6624885877422138e-05, "loss": 0.2565, "step": 4450 }, { "epoch": 1.9486563251037796, "grad_norm": 0.20105837715876068, "learning_rate": 1.650525106127449e-05, "loss": 0.2588, "step": 4460 }, { "epoch": 1.9530259995630326, "grad_norm": 0.19233588613180366, "learning_rate": 1.6385835764274522e-05, "loss": 0.2568, "step": 4470 }, { "epoch": 1.9573956740222853, "grad_norm": 0.19522604503125968, "learning_rate": 1.6266643072322067e-05, "loss": 0.2604, "step": 4480 }, { "epoch": 1.961765348481538, "grad_norm": 0.23362836899867961, "learning_rate": 1.614767606556446e-05, "loss": 0.2475, "step": 4490 }, { "epoch": 1.966135022940791, "grad_norm": 0.2022473510573943, "learning_rate": 1.602893781831689e-05, "loss": 0.258, "step": 4500 }, { "epoch": 1.9705046974000437, "grad_norm": 0.19403172327423443, "learning_rate": 1.591043139898304e-05, "loss": 0.2542, "step": 4510 }, { "epoch": 1.9748743718592965, "grad_norm": 0.1721090127598485, "learning_rate": 1.579215986997575e-05, "loss": 0.2506, "step": 4520 }, { "epoch": 1.9792440463185492, "grad_norm": 0.2463041661154928, "learning_rate": 1.5674126287637865e-05, "loss": 0.2603, "step": 4530 }, { "epoch": 1.983613720777802, "grad_norm": 0.19386331867066092, "learning_rate": 1.555633370216329e-05, "loss": 0.2417, "step": 4540 }, { "epoch": 1.987983395237055, "grad_norm": 0.19092835243215459, "learning_rate": 1.5438785157518154e-05, "loss": 0.2636, "step": 4550 }, { "epoch": 1.9923530696963077, "grad_norm": 0.18678755420613113, "learning_rate": 1.532148369136212e-05, "loss": 0.2466, "step": 4560 }, { "epoch": 1.9967227441555604, "grad_norm": 0.19282088494800362, "learning_rate": 1.5204432334969942e-05, "loss": 0.255, "step": 4570 }, { "epoch": 2.0008739348918505, "grad_norm": 0.2204022521779481, "learning_rate": 1.5087634113153079e-05, "loss": 0.2399, "step": 4580 }, { "epoch": 2.0052436093511035, "grad_norm": 0.19203641231313598, "learning_rate": 1.4971092044181553e-05, "loss": 0.2113, "step": 4590 }, { "epoch": 2.009613283810356, "grad_norm": 0.19415273072290806, "learning_rate": 1.485480913970596e-05, "loss": 0.2124, "step": 4600 }, { "epoch": 2.013982958269609, "grad_norm": 0.19328976339568105, "learning_rate": 1.4738788404679636e-05, "loss": 0.2158, "step": 4610 }, { "epoch": 2.018352632728862, "grad_norm": 0.17743627693622602, "learning_rate": 1.462303283728097e-05, "loss": 0.2093, "step": 4620 }, { "epoch": 2.0227223071881144, "grad_norm": 0.18133139847205673, "learning_rate": 1.4507545428835992e-05, "loss": 0.2106, "step": 4630 }, { "epoch": 2.0270919816473674, "grad_norm": 0.18263631533379213, "learning_rate": 1.4392329163741013e-05, "loss": 0.2211, "step": 4640 }, { "epoch": 2.03146165610662, "grad_norm": 0.19891127156804314, "learning_rate": 1.4277387019385544e-05, "loss": 0.2202, "step": 4650 }, { "epoch": 2.035831330565873, "grad_norm": 0.16948509125664413, "learning_rate": 1.4162721966075323e-05, "loss": 0.2189, "step": 4660 }, { "epoch": 2.040201005025126, "grad_norm": 0.16653995158873228, "learning_rate": 1.4048336966955588e-05, "loss": 0.2158, "step": 4670 }, { "epoch": 2.0445706794843783, "grad_norm": 0.19907451828343845, "learning_rate": 1.3934234977934463e-05, "loss": 0.2238, "step": 4680 }, { "epoch": 2.0489403539436313, "grad_norm": 0.18516842868241973, "learning_rate": 1.3820418947606612e-05, "loss": 0.2252, "step": 4690 }, { "epoch": 2.053310028402884, "grad_norm": 0.16551556198595282, "learning_rate": 1.3706891817177036e-05, "loss": 0.2105, "step": 4700 }, { "epoch": 2.0576797028621368, "grad_norm": 0.17239104222906546, "learning_rate": 1.3593656520385035e-05, "loss": 0.2209, "step": 4710 }, { "epoch": 2.0620493773213897, "grad_norm": 0.17949672018554283, "learning_rate": 1.3480715983428432e-05, "loss": 0.2186, "step": 4720 }, { "epoch": 2.0664190517806422, "grad_norm": 0.18425891810972556, "learning_rate": 1.3368073124887958e-05, "loss": 0.2167, "step": 4730 }, { "epoch": 2.070788726239895, "grad_norm": 0.16283627033039932, "learning_rate": 1.3255730855651771e-05, "loss": 0.212, "step": 4740 }, { "epoch": 2.0751584006991477, "grad_norm": 0.1716008439996554, "learning_rate": 1.3143692078840311e-05, "loss": 0.2205, "step": 4750 }, { "epoch": 2.0795280751584007, "grad_norm": 0.17348180182705283, "learning_rate": 1.3031959689731236e-05, "loss": 0.2139, "step": 4760 }, { "epoch": 2.0838977496176536, "grad_norm": 0.18412547976359403, "learning_rate": 1.2920536575684616e-05, "loss": 0.2209, "step": 4770 }, { "epoch": 2.088267424076906, "grad_norm": 0.18288666540251491, "learning_rate": 1.2809425616068287e-05, "loss": 0.2162, "step": 4780 }, { "epoch": 2.092637098536159, "grad_norm": 0.18106613392356596, "learning_rate": 1.26986296821835e-05, "loss": 0.2172, "step": 4790 }, { "epoch": 2.097006772995412, "grad_norm": 0.1675482319347818, "learning_rate": 1.2588151637190687e-05, "loss": 0.217, "step": 4800 }, { "epoch": 2.1013764474546646, "grad_norm": 0.17554932709474122, "learning_rate": 1.2477994336035448e-05, "loss": 0.2085, "step": 4810 }, { "epoch": 2.1057461219139175, "grad_norm": 0.170017739189806, "learning_rate": 1.2368160625374833e-05, "loss": 0.2061, "step": 4820 }, { "epoch": 2.11011579637317, "grad_norm": 0.21585961661871844, "learning_rate": 1.2258653343503747e-05, "loss": 0.2125, "step": 4830 }, { "epoch": 2.114485470832423, "grad_norm": 0.1955796491896077, "learning_rate": 1.2149475320281578e-05, "loss": 0.2252, "step": 4840 }, { "epoch": 2.118855145291676, "grad_norm": 0.17266978222442586, "learning_rate": 1.2040629377059121e-05, "loss": 0.2158, "step": 4850 }, { "epoch": 2.1232248197509285, "grad_norm": 0.16631363410579506, "learning_rate": 1.1932118326605643e-05, "loss": 0.2169, "step": 4860 }, { "epoch": 2.1275944942101814, "grad_norm": 0.1902948343774393, "learning_rate": 1.1823944973036175e-05, "loss": 0.2086, "step": 4870 }, { "epoch": 2.131964168669434, "grad_norm": 0.18739093394887843, "learning_rate": 1.1716112111739095e-05, "loss": 0.2207, "step": 4880 }, { "epoch": 2.136333843128687, "grad_norm": 0.16857112053566867, "learning_rate": 1.1608622529303861e-05, "loss": 0.2137, "step": 4890 }, { "epoch": 2.14070351758794, "grad_norm": 0.16103152072935048, "learning_rate": 1.1501479003448992e-05, "loss": 0.2104, "step": 4900 }, { "epoch": 2.1450731920471924, "grad_norm": 0.1703117596730518, "learning_rate": 1.1394684302950311e-05, "loss": 0.2242, "step": 4910 }, { "epoch": 2.1494428665064453, "grad_norm": 0.15986306603302497, "learning_rate": 1.1288241187569399e-05, "loss": 0.2148, "step": 4920 }, { "epoch": 2.153812540965698, "grad_norm": 0.17368589918363925, "learning_rate": 1.1182152407982225e-05, "loss": 0.2259, "step": 4930 }, { "epoch": 2.158182215424951, "grad_norm": 0.17619963932463548, "learning_rate": 1.1076420705708137e-05, "loss": 0.2099, "step": 4940 }, { "epoch": 2.1625518898842038, "grad_norm": 0.1834089863734305, "learning_rate": 1.0971048813038981e-05, "loss": 0.2172, "step": 4950 }, { "epoch": 2.1669215643434563, "grad_norm": 0.18615170920839888, "learning_rate": 1.0866039452968462e-05, "loss": 0.2154, "step": 4960 }, { "epoch": 2.1712912388027092, "grad_norm": 0.18453919114705294, "learning_rate": 1.0761395339121844e-05, "loss": 0.211, "step": 4970 }, { "epoch": 2.1756609132619618, "grad_norm": 0.18145722544816031, "learning_rate": 1.0657119175685775e-05, "loss": 0.2105, "step": 4980 }, { "epoch": 2.1800305877212147, "grad_norm": 0.17060666896849788, "learning_rate": 1.0553213657338426e-05, "loss": 0.2199, "step": 4990 }, { "epoch": 2.1844002621804677, "grad_norm": 0.15600289701745268, "learning_rate": 1.044968146917986e-05, "loss": 0.2129, "step": 5000 }, { "epoch": 2.18876993663972, "grad_norm": 0.1690986189287265, "learning_rate": 1.0346525286662605e-05, "loss": 0.2112, "step": 5010 }, { "epoch": 2.193139611098973, "grad_norm": 0.17775745464309511, "learning_rate": 1.024374777552258e-05, "loss": 0.2159, "step": 5020 }, { "epoch": 2.197509285558226, "grad_norm": 0.16558227080681498, "learning_rate": 1.0141351591710163e-05, "loss": 0.2201, "step": 5030 }, { "epoch": 2.2018789600174786, "grad_norm": 0.17387309589957087, "learning_rate": 1.0039339381321572e-05, "loss": 0.2159, "step": 5040 }, { "epoch": 2.2062486344767316, "grad_norm": 0.17239160931426226, "learning_rate": 9.937713780530463e-06, "loss": 0.2141, "step": 5050 }, { "epoch": 2.210618308935984, "grad_norm": 0.15649874366855132, "learning_rate": 9.836477415519842e-06, "loss": 0.22, "step": 5060 }, { "epoch": 2.214987983395237, "grad_norm": 0.15640272161867394, "learning_rate": 9.735632902414194e-06, "loss": 0.2149, "step": 5070 }, { "epoch": 2.21935765785449, "grad_norm": 0.17863339401752804, "learning_rate": 9.635182847211827e-06, "loss": 0.2172, "step": 5080 }, { "epoch": 2.2237273323137425, "grad_norm": 0.18301106479467924, "learning_rate": 9.535129845717606e-06, "loss": 0.2199, "step": 5090 }, { "epoch": 2.2280970067729955, "grad_norm": 0.1592869787479507, "learning_rate": 9.435476483475824e-06, "loss": 0.208, "step": 5100 }, { "epoch": 2.232466681232248, "grad_norm": 0.16376505177226974, "learning_rate": 9.336225335703383e-06, "loss": 0.2168, "step": 5110 }, { "epoch": 2.236836355691501, "grad_norm": 0.16813347898302336, "learning_rate": 9.237378967223279e-06, "loss": 0.2178, "step": 5120 }, { "epoch": 2.241206030150754, "grad_norm": 0.16718608214450326, "learning_rate": 9.138939932398301e-06, "loss": 0.211, "step": 5130 }, { "epoch": 2.2455757046100064, "grad_norm": 0.1599670224222752, "learning_rate": 9.040910775065014e-06, "loss": 0.2146, "step": 5140 }, { "epoch": 2.2499453790692594, "grad_norm": 0.17112785799203786, "learning_rate": 8.943294028468044e-06, "loss": 0.2192, "step": 5150 }, { "epoch": 2.2543150535285124, "grad_norm": 0.1745609364310858, "learning_rate": 8.846092215194607e-06, "loss": 0.2156, "step": 5160 }, { "epoch": 2.258684727987765, "grad_norm": 0.16033228990481582, "learning_rate": 8.749307847109292e-06, "loss": 0.2167, "step": 5170 }, { "epoch": 2.263054402447018, "grad_norm": 0.17228898869874218, "learning_rate": 8.652943425289206e-06, "loss": 0.2236, "step": 5180 }, { "epoch": 2.2674240769062703, "grad_norm": 0.15940294702183225, "learning_rate": 8.557001439959307e-06, "loss": 0.2092, "step": 5190 }, { "epoch": 2.2717937513655233, "grad_norm": 0.16358333738978809, "learning_rate": 8.461484370428024e-06, "loss": 0.2174, "step": 5200 }, { "epoch": 2.276163425824776, "grad_norm": 0.1678234097731696, "learning_rate": 8.36639468502326e-06, "loss": 0.2228, "step": 5210 }, { "epoch": 2.280533100284029, "grad_norm": 0.15585182903067626, "learning_rate": 8.271734841028553e-06, "loss": 0.2067, "step": 5220 }, { "epoch": 2.2849027747432817, "grad_norm": 0.1697260189351675, "learning_rate": 8.177507284619567e-06, "loss": 0.2212, "step": 5230 }, { "epoch": 2.2892724492025343, "grad_norm": 0.15819967407271224, "learning_rate": 8.08371445080093e-06, "loss": 0.2049, "step": 5240 }, { "epoch": 2.293642123661787, "grad_norm": 0.1576517005711918, "learning_rate": 7.990358763343272e-06, "loss": 0.2084, "step": 5250 }, { "epoch": 2.29801179812104, "grad_norm": 0.17306944459177356, "learning_rate": 7.897442634720576e-06, "loss": 0.2129, "step": 5260 }, { "epoch": 2.3023814725802927, "grad_norm": 0.1646880072650218, "learning_rate": 7.80496846604788e-06, "loss": 0.2136, "step": 5270 }, { "epoch": 2.3067511470395456, "grad_norm": 0.17723586410215897, "learning_rate": 7.7129386470192e-06, "loss": 0.2126, "step": 5280 }, { "epoch": 2.311120821498798, "grad_norm": 0.17033040933338542, "learning_rate": 7.621355555845788e-06, "loss": 0.2188, "step": 5290 }, { "epoch": 2.315490495958051, "grad_norm": 0.15886758889073826, "learning_rate": 7.530221559194642e-06, "loss": 0.2213, "step": 5300 }, { "epoch": 2.319860170417304, "grad_norm": 0.16468737121869617, "learning_rate": 7.439539012127397e-06, "loss": 0.2145, "step": 5310 }, { "epoch": 2.3242298448765566, "grad_norm": 0.16437936722266838, "learning_rate": 7.349310258039441e-06, "loss": 0.2198, "step": 5320 }, { "epoch": 2.3285995193358096, "grad_norm": 0.15530339481066868, "learning_rate": 7.259537628599356e-06, "loss": 0.2173, "step": 5330 }, { "epoch": 2.332969193795062, "grad_norm": 0.1551121668416178, "learning_rate": 7.170223443688653e-06, "loss": 0.2162, "step": 5340 }, { "epoch": 2.337338868254315, "grad_norm": 0.16687622113049136, "learning_rate": 7.0813700113418534e-06, "loss": 0.2202, "step": 5350 }, { "epoch": 2.341708542713568, "grad_norm": 0.1703042983217557, "learning_rate": 6.9929796276868214e-06, "loss": 0.2114, "step": 5360 }, { "epoch": 2.3460782171728205, "grad_norm": 0.1805361195466261, "learning_rate": 6.905054576885436e-06, "loss": 0.2188, "step": 5370 }, { "epoch": 2.3504478916320735, "grad_norm": 0.1717723903009623, "learning_rate": 6.817597131074566e-06, "loss": 0.2175, "step": 5380 }, { "epoch": 2.3548175660913264, "grad_norm": 0.16616224214054395, "learning_rate": 6.730609550307332e-06, "loss": 0.2022, "step": 5390 }, { "epoch": 2.359187240550579, "grad_norm": 0.18596348012235386, "learning_rate": 6.644094082494745e-06, "loss": 0.2119, "step": 5400 }, { "epoch": 2.363556915009832, "grad_norm": 0.15235707364090476, "learning_rate": 6.5580529633475856e-06, "loss": 0.2164, "step": 5410 }, { "epoch": 2.3679265894690844, "grad_norm": 0.16802244364306884, "learning_rate": 6.47248841631862e-06, "loss": 0.216, "step": 5420 }, { "epoch": 2.3722962639283374, "grad_norm": 0.1637680124738261, "learning_rate": 6.3874026525451825e-06, "loss": 0.2102, "step": 5430 }, { "epoch": 2.37666593838759, "grad_norm": 0.17586865595817228, "learning_rate": 6.302797870792007e-06, "loss": 0.2124, "step": 5440 }, { "epoch": 2.381035612846843, "grad_norm": 0.18140383469789903, "learning_rate": 6.21867625739439e-06, "loss": 0.2025, "step": 5450 }, { "epoch": 2.385405287306096, "grad_norm": 0.17499929296202876, "learning_rate": 6.1350399862017435e-06, "loss": 0.2098, "step": 5460 }, { "epoch": 2.3897749617653483, "grad_norm": 0.1791296518139122, "learning_rate": 6.051891218521377e-06, "loss": 0.216, "step": 5470 }, { "epoch": 2.3941446362246013, "grad_norm": 0.15189044349380637, "learning_rate": 5.969232103062647e-06, "loss": 0.2111, "step": 5480 }, { "epoch": 2.3985143106838542, "grad_norm": 0.15438326021674537, "learning_rate": 5.8870647758814595e-06, "loss": 0.2154, "step": 5490 }, { "epoch": 2.4028839851431067, "grad_norm": 0.16744269488259383, "learning_rate": 5.805391360325049e-06, "loss": 0.2164, "step": 5500 }, { "epoch": 2.4072536596023597, "grad_norm": 0.16224327611508244, "learning_rate": 5.724213966977093e-06, "loss": 0.2156, "step": 5510 }, { "epoch": 2.4116233340616122, "grad_norm": 0.15833500293564692, "learning_rate": 5.643534693603214e-06, "loss": 0.2105, "step": 5520 }, { "epoch": 2.415993008520865, "grad_norm": 0.1714858172159274, "learning_rate": 5.563355625096733e-06, "loss": 0.223, "step": 5530 }, { "epoch": 2.420362682980118, "grad_norm": 0.15925562136983235, "learning_rate": 5.483678833424797e-06, "loss": 0.2227, "step": 5540 }, { "epoch": 2.4247323574393707, "grad_norm": 0.15987450234850412, "learning_rate": 5.404506377574853e-06, "loss": 0.2108, "step": 5550 }, { "epoch": 2.4291020318986236, "grad_norm": 0.17169417547908242, "learning_rate": 5.325840303501431e-06, "loss": 0.2173, "step": 5560 }, { "epoch": 2.433471706357876, "grad_norm": 0.1528840485878302, "learning_rate": 5.247682644073257e-06, "loss": 0.2153, "step": 5570 }, { "epoch": 2.437841380817129, "grad_norm": 0.15059281611348097, "learning_rate": 5.170035419020749e-06, "loss": 0.2151, "step": 5580 }, { "epoch": 2.442211055276382, "grad_norm": 0.151962454547107, "learning_rate": 5.09290063488381e-06, "loss": 0.2089, "step": 5590 }, { "epoch": 2.4465807297356346, "grad_norm": 0.1550399769234159, "learning_rate": 5.016280284959957e-06, "loss": 0.2086, "step": 5600 }, { "epoch": 2.4509504041948875, "grad_norm": 0.1671431037785645, "learning_rate": 4.9401763492528455e-06, "loss": 0.2236, "step": 5610 }, { "epoch": 2.4553200786541405, "grad_norm": 0.15796464427217982, "learning_rate": 4.8645907944210925e-06, "loss": 0.216, "step": 5620 }, { "epoch": 2.459689753113393, "grad_norm": 0.14876003608506433, "learning_rate": 4.789525573727418e-06, "loss": 0.2117, "step": 5630 }, { "epoch": 2.464059427572646, "grad_norm": 0.1630939492277196, "learning_rate": 4.714982626988229e-06, "loss": 0.2175, "step": 5640 }, { "epoch": 2.4684291020318985, "grad_norm": 0.14186087192994887, "learning_rate": 4.640963880523446e-06, "loss": 0.206, "step": 5650 }, { "epoch": 2.4727987764911514, "grad_norm": 0.17254468396934777, "learning_rate": 4.56747124710675e-06, "loss": 0.2107, "step": 5660 }, { "epoch": 2.4771684509504044, "grad_norm": 0.1529419585360116, "learning_rate": 4.494506625916126e-06, "loss": 0.2144, "step": 5670 }, { "epoch": 2.481538125409657, "grad_norm": 0.15607398100185668, "learning_rate": 4.4220719024848115e-06, "loss": 0.2142, "step": 5680 }, { "epoch": 2.48590779986891, "grad_norm": 0.15809466504991854, "learning_rate": 4.350168948652561e-06, "loss": 0.2174, "step": 5690 }, { "epoch": 2.4902774743281624, "grad_norm": 0.1492303870885846, "learning_rate": 4.278799622517274e-06, "loss": 0.2074, "step": 5700 }, { "epoch": 2.4946471487874153, "grad_norm": 0.16513731313738675, "learning_rate": 4.2079657683869786e-06, "loss": 0.2143, "step": 5710 }, { "epoch": 2.4990168232466683, "grad_norm": 0.15080321770571686, "learning_rate": 4.137669216732162e-06, "loss": 0.2113, "step": 5720 }, { "epoch": 2.503386497705921, "grad_norm": 0.16345890216718448, "learning_rate": 4.067911784138484e-06, "loss": 0.2106, "step": 5730 }, { "epoch": 2.5077561721651738, "grad_norm": 0.17340636441290833, "learning_rate": 3.998695273259834e-06, "loss": 0.2175, "step": 5740 }, { "epoch": 2.5121258466244267, "grad_norm": 0.16847468221465162, "learning_rate": 3.930021472771722e-06, "loss": 0.2115, "step": 5750 }, { "epoch": 2.5164955210836792, "grad_norm": 0.15526232478991495, "learning_rate": 3.861892157325089e-06, "loss": 0.213, "step": 5760 }, { "epoch": 2.520865195542932, "grad_norm": 0.25090581544684, "learning_rate": 3.7943090875004307e-06, "loss": 0.2183, "step": 5770 }, { "epoch": 2.5252348700021847, "grad_norm": 0.16224420787021895, "learning_rate": 3.727274009762288e-06, "loss": 0.2159, "step": 5780 }, { "epoch": 2.5296045444614377, "grad_norm": 0.16731499429058524, "learning_rate": 3.6607886564141437e-06, "loss": 0.2182, "step": 5790 }, { "epoch": 2.53397421892069, "grad_norm": 0.14710180105464096, "learning_rate": 3.5948547455536357e-06, "loss": 0.2069, "step": 5800 }, { "epoch": 2.538343893379943, "grad_norm": 0.14618076750205436, "learning_rate": 3.5294739810281725e-06, "loss": 0.2139, "step": 5810 }, { "epoch": 2.542713567839196, "grad_norm": 0.16047915844682772, "learning_rate": 3.4646480523908812e-06, "loss": 0.2143, "step": 5820 }, { "epoch": 2.5470832422984486, "grad_norm": 0.1625409155668901, "learning_rate": 3.400378634856974e-06, "loss": 0.2125, "step": 5830 }, { "epoch": 2.5514529167577016, "grad_norm": 0.15053743645135229, "learning_rate": 3.3366673892604448e-06, "loss": 0.206, "step": 5840 }, { "epoch": 2.5558225912169545, "grad_norm": 0.15823053094178416, "learning_rate": 3.2735159620111285e-06, "loss": 0.2164, "step": 5850 }, { "epoch": 2.560192265676207, "grad_norm": 0.15401338006783896, "learning_rate": 3.2109259850522045e-06, "loss": 0.2221, "step": 5860 }, { "epoch": 2.56456194013546, "grad_norm": 0.14487505839909515, "learning_rate": 3.148899075817985e-06, "loss": 0.2197, "step": 5870 }, { "epoch": 2.5689316145947125, "grad_norm": 0.14831509300063359, "learning_rate": 3.0874368371921176e-06, "loss": 0.2111, "step": 5880 }, { "epoch": 2.5733012890539655, "grad_norm": 0.1545521810148666, "learning_rate": 3.0265408574661926e-06, "loss": 0.2114, "step": 5890 }, { "epoch": 2.577670963513218, "grad_norm": 0.1447781651428092, "learning_rate": 2.9662127102986742e-06, "loss": 0.2102, "step": 5900 }, { "epoch": 2.582040637972471, "grad_norm": 0.152773353283902, "learning_rate": 2.9064539546742375e-06, "loss": 0.2136, "step": 5910 }, { "epoch": 2.586410312431724, "grad_norm": 0.146829743337425, "learning_rate": 2.8472661348634883e-06, "loss": 0.2153, "step": 5920 }, { "epoch": 2.5907799868909764, "grad_norm": 0.14888204953981307, "learning_rate": 2.7886507803830675e-06, "loss": 0.2179, "step": 5930 }, { "epoch": 2.5951496613502294, "grad_norm": 0.15843710893798904, "learning_rate": 2.7306094059560832e-06, "loss": 0.2108, "step": 5940 }, { "epoch": 2.5995193358094824, "grad_norm": 0.15587202489987187, "learning_rate": 2.6731435114730254e-06, "loss": 0.2097, "step": 5950 }, { "epoch": 2.603889010268735, "grad_norm": 0.16081101909646806, "learning_rate": 2.616254581952962e-06, "loss": 0.2157, "step": 5960 }, { "epoch": 2.608258684727988, "grad_norm": 0.15806090701968617, "learning_rate": 2.5599440875051783e-06, "loss": 0.2103, "step": 5970 }, { "epoch": 2.612628359187241, "grad_norm": 0.16111470687957785, "learning_rate": 2.504213483291193e-06, "loss": 0.2135, "step": 5980 }, { "epoch": 2.6169980336464933, "grad_norm": 0.1699471590218342, "learning_rate": 2.449064209487148e-06, "loss": 0.2, "step": 5990 }, { "epoch": 2.6213677081057463, "grad_norm": 0.16497405965950862, "learning_rate": 2.3944976912465915e-06, "loss": 0.2047, "step": 6000 }, { "epoch": 2.625737382564999, "grad_norm": 0.15089249806595506, "learning_rate": 2.340515338663646e-06, "loss": 0.2122, "step": 6010 }, { "epoch": 2.6301070570242517, "grad_norm": 0.15487442969708834, "learning_rate": 2.287118546736572e-06, "loss": 0.2151, "step": 6020 }, { "epoch": 2.6344767314835043, "grad_norm": 0.15284632753949368, "learning_rate": 2.2343086953317326e-06, "loss": 0.2135, "step": 6030 }, { "epoch": 2.638846405942757, "grad_norm": 0.15605926035866713, "learning_rate": 2.18208714914791e-06, "loss": 0.2219, "step": 6040 }, { "epoch": 2.64321608040201, "grad_norm": 0.15577630551734745, "learning_rate": 2.130455257681063e-06, "loss": 0.22, "step": 6050 }, { "epoch": 2.6475857548612627, "grad_norm": 0.14652493649973405, "learning_rate": 2.0794143551894267e-06, "loss": 0.2056, "step": 6060 }, { "epoch": 2.6519554293205156, "grad_norm": 0.15500949330246658, "learning_rate": 2.0289657606590656e-06, "loss": 0.2262, "step": 6070 }, { "epoch": 2.6563251037797686, "grad_norm": 0.14474074629136158, "learning_rate": 1.979110777769763e-06, "loss": 0.2072, "step": 6080 }, { "epoch": 2.660694778239021, "grad_norm": 0.15767067661666023, "learning_rate": 1.92985069486134e-06, "loss": 0.2181, "step": 6090 }, { "epoch": 2.665064452698274, "grad_norm": 0.1486160435579505, "learning_rate": 1.8811867849003683e-06, "loss": 0.2079, "step": 6100 }, { "epoch": 2.669434127157527, "grad_norm": 0.16264415321256365, "learning_rate": 1.833120305447275e-06, "loss": 0.2149, "step": 6110 }, { "epoch": 2.6738038016167796, "grad_norm": 0.1517206341115815, "learning_rate": 1.785652498623816e-06, "loss": 0.2124, "step": 6120 }, { "epoch": 2.678173476076032, "grad_norm": 0.1576531182111802, "learning_rate": 1.7387845910810275e-06, "loss": 0.2158, "step": 6130 }, { "epoch": 2.682543150535285, "grad_norm": 0.14631491706090213, "learning_rate": 1.6925177939674935e-06, "loss": 0.2039, "step": 6140 }, { "epoch": 2.686912824994538, "grad_norm": 0.14550164873727042, "learning_rate": 1.6468533028980382e-06, "loss": 0.2135, "step": 6150 }, { "epoch": 2.6912824994537905, "grad_norm": 0.1488011611255984, "learning_rate": 1.601792297922866e-06, "loss": 0.2073, "step": 6160 }, { "epoch": 2.6956521739130435, "grad_norm": 0.1445169239138001, "learning_rate": 1.5573359434970459e-06, "loss": 0.2084, "step": 6170 }, { "epoch": 2.7000218483722964, "grad_norm": 0.16105995620000982, "learning_rate": 1.513485388450403e-06, "loss": 0.2226, "step": 6180 }, { "epoch": 2.704391522831549, "grad_norm": 0.15980200476353595, "learning_rate": 1.4702417659578637e-06, "loss": 0.2102, "step": 6190 }, { "epoch": 2.708761197290802, "grad_norm": 0.1512193825053304, "learning_rate": 1.4276061935101586e-06, "loss": 0.2162, "step": 6200 }, { "epoch": 2.713130871750055, "grad_norm": 0.15427596474826816, "learning_rate": 1.385579772884929e-06, "loss": 0.2076, "step": 6210 }, { "epoch": 2.7175005462093074, "grad_norm": 0.14654105315448618, "learning_rate": 1.3441635901182804e-06, "loss": 0.2087, "step": 6220 }, { "epoch": 2.7218702206685603, "grad_norm": 0.1421434600281987, "learning_rate": 1.303358715476713e-06, "loss": 0.2123, "step": 6230 }, { "epoch": 2.726239895127813, "grad_norm": 0.15821815935479802, "learning_rate": 1.2631662034294389e-06, "loss": 0.2135, "step": 6240 }, { "epoch": 2.730609569587066, "grad_norm": 0.15325308542480637, "learning_rate": 1.2235870926211619e-06, "loss": 0.2116, "step": 6250 }, { "epoch": 2.7349792440463183, "grad_norm": 0.15504478719541256, "learning_rate": 1.1846224058452316e-06, "loss": 0.2173, "step": 6260 }, { "epoch": 2.7393489185055713, "grad_norm": 0.14735665917665872, "learning_rate": 1.1462731500171964e-06, "loss": 0.2088, "step": 6270 }, { "epoch": 2.7437185929648242, "grad_norm": 0.1575294237133432, "learning_rate": 1.1085403161488012e-06, "loss": 0.2111, "step": 6280 }, { "epoch": 2.7480882674240767, "grad_norm": 0.15834490229089185, "learning_rate": 1.0714248793223718e-06, "loss": 0.2099, "step": 6290 }, { "epoch": 2.7524579418833297, "grad_norm": 0.15434184488840358, "learning_rate": 1.034927798665608e-06, "loss": 0.2157, "step": 6300 }, { "epoch": 2.7568276163425827, "grad_norm": 0.1417103700203199, "learning_rate": 9.990500173268164e-07, "loss": 0.1969, "step": 6310 }, { "epoch": 2.761197290801835, "grad_norm": 0.1477883455121692, "learning_rate": 9.63792462450519e-07, "loss": 0.2113, "step": 6320 }, { "epoch": 2.765566965261088, "grad_norm": 0.1436289751022934, "learning_rate": 9.291560451535175e-07, "loss": 0.216, "step": 6330 }, { "epoch": 2.769936639720341, "grad_norm": 0.13939995964947724, "learning_rate": 8.951416605013113e-07, "loss": 0.2086, "step": 6340 }, { "epoch": 2.7743063141795936, "grad_norm": 0.1552874482797402, "learning_rate": 8.617501874850053e-07, "loss": 0.2024, "step": 6350 }, { "epoch": 2.7786759886388466, "grad_norm": 0.14716505432534613, "learning_rate": 8.289824889985809e-07, "loss": 0.2132, "step": 6360 }, { "epoch": 2.783045663098099, "grad_norm": 0.15513883471726814, "learning_rate": 7.968394118165884e-07, "loss": 0.2108, "step": 6370 }, { "epoch": 2.787415337557352, "grad_norm": 0.1516681629243742, "learning_rate": 7.653217865722817e-07, "loss": 0.2148, "step": 6380 }, { "epoch": 2.7917850120166046, "grad_norm": 0.13746311918081686, "learning_rate": 7.344304277361325e-07, "loss": 0.21, "step": 6390 }, { "epoch": 2.7961546864758575, "grad_norm": 0.14446336792204623, "learning_rate": 7.041661335948025e-07, "loss": 0.2067, "step": 6400 }, { "epoch": 2.8005243609351105, "grad_norm": 0.14132680276862664, "learning_rate": 6.745296862305073e-07, "loss": 0.2123, "step": 6410 }, { "epoch": 2.804894035394363, "grad_norm": 0.1451979511933469, "learning_rate": 6.455218515008049e-07, "loss": 0.2124, "step": 6420 }, { "epoch": 2.809263709853616, "grad_norm": 0.14820242866536618, "learning_rate": 6.171433790188002e-07, "loss": 0.2077, "step": 6430 }, { "epoch": 2.813633384312869, "grad_norm": 0.14216621753713982, "learning_rate": 5.893950021337829e-07, "loss": 0.2107, "step": 6440 }, { "epoch": 2.8180030587721214, "grad_norm": 0.15976207650503466, "learning_rate": 5.62277437912273e-07, "loss": 0.21, "step": 6450 }, { "epoch": 2.8223727332313744, "grad_norm": 0.153968196976785, "learning_rate": 5.357913871194859e-07, "loss": 0.2117, "step": 6460 }, { "epoch": 2.826742407690627, "grad_norm": 0.14911110207889341, "learning_rate": 5.099375342012297e-07, "loss": 0.2098, "step": 6470 }, { "epoch": 2.83111208214988, "grad_norm": 0.1474434611710356, "learning_rate": 4.847165472662146e-07, "loss": 0.21, "step": 6480 }, { "epoch": 2.8354817566091324, "grad_norm": 0.1496356931440323, "learning_rate": 4.6012907806878836e-07, "loss": 0.2103, "step": 6490 }, { "epoch": 2.8398514310683853, "grad_norm": 0.14150544300570939, "learning_rate": 4.361757619920942e-07, "loss": 0.2027, "step": 6500 }, { "epoch": 2.8442211055276383, "grad_norm": 0.1426708708944942, "learning_rate": 4.128572180316509e-07, "loss": 0.209, "step": 6510 }, { "epoch": 2.848590779986891, "grad_norm": 0.1413810502175028, "learning_rate": 3.901740487793598e-07, "loss": 0.2041, "step": 6520 }, { "epoch": 2.8529604544461438, "grad_norm": 0.14979561972176883, "learning_rate": 3.68126840407923e-07, "loss": 0.2182, "step": 6530 }, { "epoch": 2.8573301289053967, "grad_norm": 0.1473216052536028, "learning_rate": 3.467161626557164e-07, "loss": 0.2139, "step": 6540 }, { "epoch": 2.8616998033646492, "grad_norm": 0.14985145553034246, "learning_rate": 3.259425688120349e-07, "loss": 0.2082, "step": 6550 }, { "epoch": 2.866069477823902, "grad_norm": 0.15545775571977008, "learning_rate": 3.0580659570282886e-07, "loss": 0.2157, "step": 6560 }, { "epoch": 2.870439152283155, "grad_norm": 0.15918210555105547, "learning_rate": 2.8630876367680414e-07, "loss": 0.2237, "step": 6570 }, { "epoch": 2.8748088267424077, "grad_norm": 0.15194693986634103, "learning_rate": 2.674495765919938e-07, "loss": 0.2123, "step": 6580 }, { "epoch": 2.8791785012016606, "grad_norm": 0.1430322734744411, "learning_rate": 2.4922952180272685e-07, "loss": 0.2098, "step": 6590 }, { "epoch": 2.883548175660913, "grad_norm": 0.1520249769246339, "learning_rate": 2.316490701470414e-07, "loss": 0.2083, "step": 6600 }, { "epoch": 2.887917850120166, "grad_norm": 0.1414670723338776, "learning_rate": 2.147086759345024e-07, "loss": 0.2174, "step": 6610 }, { "epoch": 2.8922875245794186, "grad_norm": 0.15432609391643473, "learning_rate": 1.9840877693448888e-07, "loss": 0.2188, "step": 6620 }, { "epoch": 2.8966571990386716, "grad_norm": 0.20516987809668177, "learning_rate": 1.8274979436485318e-07, "loss": 0.2098, "step": 6630 }, { "epoch": 2.9010268734979245, "grad_norm": 0.15061209809714965, "learning_rate": 1.677321328810516e-07, "loss": 0.2195, "step": 6640 }, { "epoch": 2.905396547957177, "grad_norm": 0.14858475969482018, "learning_rate": 1.5335618056568357e-07, "loss": 0.2008, "step": 6650 }, { "epoch": 2.90976622241643, "grad_norm": 0.14732131616505734, "learning_rate": 1.3962230891846618e-07, "loss": 0.204, "step": 6660 }, { "epoch": 2.914135896875683, "grad_norm": 0.1471839238116558, "learning_rate": 1.2653087284662245e-07, "loss": 0.2053, "step": 6670 }, { "epoch": 2.9185055713349355, "grad_norm": 0.14230258450646716, "learning_rate": 1.1408221065572489e-07, "loss": 0.2092, "step": 6680 }, { "epoch": 2.9228752457941884, "grad_norm": 0.14615525467051937, "learning_rate": 1.0227664404094673e-07, "loss": 0.2136, "step": 6690 }, { "epoch": 2.927244920253441, "grad_norm": 0.14188191357523552, "learning_rate": 9.111447807874373e-08, "loss": 0.2117, "step": 6700 }, { "epoch": 2.931614594712694, "grad_norm": 0.15024449813442625, "learning_rate": 8.059600121897993e-08, "loss": 0.2009, "step": 6710 }, { "epoch": 2.9359842691719464, "grad_norm": 0.1455540408659374, "learning_rate": 7.072148527746403e-08, "loss": 0.2119, "step": 6720 }, { "epoch": 2.9403539436311994, "grad_norm": 0.14596395136001977, "learning_rate": 6.149118542893572e-08, "loss": 0.207, "step": 6730 }, { "epoch": 2.9447236180904524, "grad_norm": 0.14287836532940817, "learning_rate": 5.290534020046256e-08, "loss": 0.2094, "step": 6740 }, { "epoch": 2.949093292549705, "grad_norm": 0.15794469085833213, "learning_rate": 4.4964171465278205e-08, "loss": 0.2151, "step": 6750 }, { "epoch": 2.953462967008958, "grad_norm": 0.14587559340793407, "learning_rate": 3.7667884437050935e-08, "loss": 0.2097, "step": 6760 }, { "epoch": 2.957832641468211, "grad_norm": 0.13474674551351107, "learning_rate": 3.101666766457956e-08, "loss": 0.2053, "step": 6770 }, { "epoch": 2.9622023159274633, "grad_norm": 0.15225735642450763, "learning_rate": 2.501069302692227e-08, "loss": 0.2217, "step": 6780 }, { "epoch": 2.9665719903867163, "grad_norm": 0.1439021184857997, "learning_rate": 1.965011572895026e-08, "loss": 0.2097, "step": 6790 }, { "epoch": 2.970941664845969, "grad_norm": 0.14547625590719865, "learning_rate": 1.493507429734531e-08, "loss": 0.2193, "step": 6800 }, { "epoch": 2.9753113393052217, "grad_norm": 0.1436338621780316, "learning_rate": 1.0865690577011035e-08, "loss": 0.2088, "step": 6810 }, { "epoch": 2.9796810137644747, "grad_norm": 0.1423627565327296, "learning_rate": 7.442069727930933e-09, "loss": 0.2111, "step": 6820 }, { "epoch": 2.984050688223727, "grad_norm": 0.14127833584305693, "learning_rate": 4.6643002224427966e-09, "loss": 0.2054, "step": 6830 }, { "epoch": 2.98842036268298, "grad_norm": 0.1483387905589872, "learning_rate": 2.532453842965521e-09, "loss": 0.2068, "step": 6840 }, { "epoch": 2.9927900371422327, "grad_norm": 0.1427593851265455, "learning_rate": 1.0465856801311625e-09, "loss": 0.2112, "step": 6850 }, { "epoch": 2.9971597116014856, "grad_norm": 0.14044301870209314, "learning_rate": 2.06734131366626e-10, "loss": 0.2079, "step": 6860 }, { "epoch": 3.0, "step": 6867, "total_flos": 1.345474644612114e+19, "train_loss": 0.2895994404266022, "train_runtime": 478363.3479, "train_samples_per_second": 0.459, "train_steps_per_second": 0.014 } ], "logging_steps": 10, "max_steps": 6867, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.345474644612114e+19, "train_batch_size": 2, "trial_name": null, "trial_params": null }