{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9069992330521192, "eval_steps": 500, "global_step": 6800, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.001333822401547234, "grad_norm": 5.80256772259428, "learning_rate": 4e-06, "loss": 1.0498, "step": 10 }, { "epoch": 0.002667644803094468, "grad_norm": 33.895696082107904, "learning_rate": 8e-06, "loss": 1.0653, "step": 20 }, { "epoch": 0.004001467204641702, "grad_norm": 5.523348234283539, "learning_rate": 1.2e-05, "loss": 1.0341, "step": 30 }, { "epoch": 0.005335289606188936, "grad_norm": 11.1556403156453, "learning_rate": 1.6e-05, "loss": 0.9692, "step": 40 }, { "epoch": 0.00666911200773617, "grad_norm": 3.7375231126561825, "learning_rate": 1.9999999999999998e-05, "loss": 0.9554, "step": 50 }, { "epoch": 0.008002934409283404, "grad_norm": 8.43538339698909, "learning_rate": 2.4e-05, "loss": 0.8965, "step": 60 }, { "epoch": 0.009336756810830639, "grad_norm": 13.403454896011478, "learning_rate": 2.8e-05, "loss": 0.8273, "step": 70 }, { "epoch": 0.010670579212377872, "grad_norm": 3.95522050766088, "learning_rate": 2.9999966406213696e-05, "loss": 0.7837, "step": 80 }, { "epoch": 0.012004401613925107, "grad_norm": 36.799552052300854, "learning_rate": 2.9999697656826056e-05, "loss": 0.8288, "step": 90 }, { "epoch": 0.01333822401547234, "grad_norm": 1.6305479563258536, "learning_rate": 2.9999160162865885e-05, "loss": 0.7778, "step": 100 }, { "epoch": 0.014672046417019574, "grad_norm": 2.159536648784889, "learning_rate": 2.9998353933963273e-05, "loss": 0.7616, "step": 110 }, { "epoch": 0.016005868818566808, "grad_norm": 3.397321425707004, "learning_rate": 2.999727898456315e-05, "loss": 0.7594, "step": 120 }, { "epoch": 0.017339691220114042, "grad_norm": 4.772220837365037, "learning_rate": 2.999593533392503e-05, "loss": 0.756, "step": 130 }, { "epoch": 0.018673513621661277, "grad_norm": 2.4845945633126885, "learning_rate": 2.9994323006122654e-05, "loss": 0.7601, "step": 140 }, { "epoch": 0.02000733602320851, "grad_norm": 3.591682569169127, "learning_rate": 2.9992442030043557e-05, "loss": 0.7894, "step": 150 }, { "epoch": 0.021341158424755743, "grad_norm": 2.5679458807474416, "learning_rate": 2.9990292439388565e-05, "loss": 0.7093, "step": 160 }, { "epoch": 0.022674980826302978, "grad_norm": 1.9412569107551652, "learning_rate": 2.9987874272671168e-05, "loss": 0.706, "step": 170 }, { "epoch": 0.024008803227850213, "grad_norm": 3.2667097270489, "learning_rate": 2.9985187573216855e-05, "loss": 0.7586, "step": 180 }, { "epoch": 0.025342625629397444, "grad_norm": 4.4208737375400675, "learning_rate": 2.998223238916232e-05, "loss": 0.6985, "step": 190 }, { "epoch": 0.02667644803094468, "grad_norm": 5.515966302183704, "learning_rate": 2.9979008773454618e-05, "loss": 0.7323, "step": 200 }, { "epoch": 0.028010270432491914, "grad_norm": 2.964165450396077, "learning_rate": 2.997551678385019e-05, "loss": 0.7603, "step": 210 }, { "epoch": 0.02934409283403915, "grad_norm": 3.0952916783456197, "learning_rate": 2.997175648291384e-05, "loss": 0.7421, "step": 220 }, { "epoch": 0.03067791523558638, "grad_norm": 4.213588693904103, "learning_rate": 2.996772793801763e-05, "loss": 0.7322, "step": 230 }, { "epoch": 0.032011737637133615, "grad_norm": 1.8568586103139084, "learning_rate": 2.996343122133965e-05, "loss": 0.6922, "step": 240 }, { "epoch": 0.033345560038680847, "grad_norm": 4.494146778909846, "learning_rate": 2.9958866409862745e-05, "loss": 0.7244, "step": 250 }, { "epoch": 0.034679382440228085, "grad_norm": 7.438170074282725, "learning_rate": 2.9954033585373108e-05, "loss": 0.7093, "step": 260 }, { "epoch": 0.036013204841775316, "grad_norm": 2.3744787346857015, "learning_rate": 2.994893283445885e-05, "loss": 0.6983, "step": 270 }, { "epoch": 0.037347027243322554, "grad_norm": 1.4722011682616383, "learning_rate": 2.9943564248508415e-05, "loss": 0.6781, "step": 280 }, { "epoch": 0.038680849644869786, "grad_norm": 3.3397620832486075, "learning_rate": 2.9937927923708966e-05, "loss": 0.7399, "step": 290 }, { "epoch": 0.04001467204641702, "grad_norm": 5.05063397044549, "learning_rate": 2.993202396104465e-05, "loss": 0.7671, "step": 300 }, { "epoch": 0.041348494447964255, "grad_norm": 3.0128431385936767, "learning_rate": 2.9925852466294795e-05, "loss": 0.7015, "step": 310 }, { "epoch": 0.04268231684951149, "grad_norm": 2.0161342716764237, "learning_rate": 2.9919413550032014e-05, "loss": 0.7009, "step": 320 }, { "epoch": 0.04401613925105872, "grad_norm": 1.3114004070324985, "learning_rate": 2.991270732762022e-05, "loss": 0.7153, "step": 330 }, { "epoch": 0.045349961652605957, "grad_norm": 18.493625676806268, "learning_rate": 2.990573391921255e-05, "loss": 0.7518, "step": 340 }, { "epoch": 0.04668378405415319, "grad_norm": 2.9526764059703567, "learning_rate": 2.989849344974924e-05, "loss": 0.7133, "step": 350 }, { "epoch": 0.048017606455700426, "grad_norm": 5.26274958582726, "learning_rate": 2.9890986048955368e-05, "loss": 0.7139, "step": 360 }, { "epoch": 0.04935142885724766, "grad_norm": 3.5319788357887933, "learning_rate": 2.9883211851338516e-05, "loss": 0.7084, "step": 370 }, { "epoch": 0.05068525125879489, "grad_norm": 7.607269935902469, "learning_rate": 2.9875170996186392e-05, "loss": 0.7309, "step": 380 }, { "epoch": 0.05201907366034213, "grad_norm": 2.3456663308287253, "learning_rate": 2.986686362756431e-05, "loss": 0.6827, "step": 390 }, { "epoch": 0.05335289606188936, "grad_norm": 2.176182050789012, "learning_rate": 2.9858289894312617e-05, "loss": 0.6995, "step": 400 }, { "epoch": 0.0546867184634366, "grad_norm": 11.171630173781537, "learning_rate": 2.9849449950044036e-05, "loss": 0.7335, "step": 410 }, { "epoch": 0.05602054086498383, "grad_norm": 6.63441431767892, "learning_rate": 2.984034395314088e-05, "loss": 0.7031, "step": 420 }, { "epoch": 0.05735436326653106, "grad_norm": 2.861620412225736, "learning_rate": 2.983097206675227e-05, "loss": 0.6559, "step": 430 }, { "epoch": 0.0586881856680783, "grad_norm": 5.523165036486206, "learning_rate": 2.9821334458791156e-05, "loss": 0.726, "step": 440 }, { "epoch": 0.06002200806962553, "grad_norm": 3.5602243751368197, "learning_rate": 2.9811431301931344e-05, "loss": 0.7202, "step": 450 }, { "epoch": 0.06135583047117276, "grad_norm": 11.333380381168622, "learning_rate": 2.9801262773604377e-05, "loss": 0.7189, "step": 460 }, { "epoch": 0.06268965287271999, "grad_norm": 14.159758615106613, "learning_rate": 2.9790829055996398e-05, "loss": 0.7267, "step": 470 }, { "epoch": 0.06402347527426723, "grad_norm": 9.009079485918289, "learning_rate": 2.978013033604483e-05, "loss": 0.748, "step": 480 }, { "epoch": 0.06535729767581447, "grad_norm": 1.9682648681675994, "learning_rate": 2.976916680543506e-05, "loss": 0.7369, "step": 490 }, { "epoch": 0.06669112007736169, "grad_norm": 2.9278164598232777, "learning_rate": 2.975793866059701e-05, "loss": 0.7037, "step": 500 }, { "epoch": 0.06802494247890893, "grad_norm": 5.5563562303649885, "learning_rate": 2.9746446102701606e-05, "loss": 0.6986, "step": 510 }, { "epoch": 0.06935876488045617, "grad_norm": 4.036767303783137, "learning_rate": 2.9734689337657157e-05, "loss": 0.7119, "step": 520 }, { "epoch": 0.07069258728200341, "grad_norm": 1.9856990692088847, "learning_rate": 2.9722668576105703e-05, "loss": 0.7205, "step": 530 }, { "epoch": 0.07202640968355063, "grad_norm": 5.200308739226583, "learning_rate": 2.971038403341921e-05, "loss": 0.6918, "step": 540 }, { "epoch": 0.07336023208509787, "grad_norm": 2.237349124701919, "learning_rate": 2.9697835929695727e-05, "loss": 0.7339, "step": 550 }, { "epoch": 0.07469405448664511, "grad_norm": 1.6388680632753365, "learning_rate": 2.968502448975544e-05, "loss": 0.7086, "step": 560 }, { "epoch": 0.07602787688819233, "grad_norm": 2.8545575025135244, "learning_rate": 2.967194994313663e-05, "loss": 0.678, "step": 570 }, { "epoch": 0.07736169928973957, "grad_norm": 2.674647983669599, "learning_rate": 2.9658612524091594e-05, "loss": 0.7119, "step": 580 }, { "epoch": 0.07869552169128681, "grad_norm": 2.489047760330112, "learning_rate": 2.9645012471582406e-05, "loss": 0.7382, "step": 590 }, { "epoch": 0.08002934409283403, "grad_norm": 5.509352102248308, "learning_rate": 2.9631150029276662e-05, "loss": 0.738, "step": 600 }, { "epoch": 0.08136316649438127, "grad_norm": 3.6489235270404015, "learning_rate": 2.9617025445543114e-05, "loss": 0.7018, "step": 610 }, { "epoch": 0.08269698889592851, "grad_norm": 2.7813651243235697, "learning_rate": 2.9602638973447218e-05, "loss": 0.7381, "step": 620 }, { "epoch": 0.08403081129747574, "grad_norm": 8.271390523006518, "learning_rate": 2.9587990870746574e-05, "loss": 0.7168, "step": 630 }, { "epoch": 0.08536463369902297, "grad_norm": 1.2460611751687307, "learning_rate": 2.9573081399886356e-05, "loss": 0.7004, "step": 640 }, { "epoch": 0.08669845610057021, "grad_norm": 1.704626418994062, "learning_rate": 2.9557910827994568e-05, "loss": 0.738, "step": 650 }, { "epoch": 0.08803227850211744, "grad_norm": 3.275051693107957, "learning_rate": 2.9542479426877283e-05, "loss": 0.7017, "step": 660 }, { "epoch": 0.08936610090366467, "grad_norm": 11.389990685570503, "learning_rate": 2.9526787473013753e-05, "loss": 0.7107, "step": 670 }, { "epoch": 0.09069992330521191, "grad_norm": 5.591277359184055, "learning_rate": 2.9510835247551485e-05, "loss": 0.7141, "step": 680 }, { "epoch": 0.09203374570675915, "grad_norm": 3.180111568581053, "learning_rate": 2.949462303630116e-05, "loss": 0.6987, "step": 690 }, { "epoch": 0.09336756810830638, "grad_norm": 3.8428068166831753, "learning_rate": 2.9478151129731567e-05, "loss": 0.7373, "step": 700 }, { "epoch": 0.09470139050985361, "grad_norm": 2.231397231771392, "learning_rate": 2.9461419822964348e-05, "loss": 0.6962, "step": 710 }, { "epoch": 0.09603521291140085, "grad_norm": 18.287201889017563, "learning_rate": 2.9444429415768726e-05, "loss": 0.6723, "step": 720 }, { "epoch": 0.09736903531294808, "grad_norm": 4.340932687135137, "learning_rate": 2.942718021255617e-05, "loss": 0.7151, "step": 730 }, { "epoch": 0.09870285771449532, "grad_norm": 2.7813821825484446, "learning_rate": 2.940967252237488e-05, "loss": 0.7332, "step": 740 }, { "epoch": 0.10003668011604255, "grad_norm": 2.3251782912937475, "learning_rate": 2.9391906658904296e-05, "loss": 0.6751, "step": 750 }, { "epoch": 0.10137050251758978, "grad_norm": 8.123799866292751, "learning_rate": 2.937388294044946e-05, "loss": 0.6886, "step": 760 }, { "epoch": 0.10270432491913702, "grad_norm": 1.528579329214318, "learning_rate": 2.9355601689935315e-05, "loss": 0.7146, "step": 770 }, { "epoch": 0.10403814732068425, "grad_norm": 2.0278953433974825, "learning_rate": 2.933706323490092e-05, "loss": 0.7453, "step": 780 }, { "epoch": 0.10537196972223148, "grad_norm": 1.4306270659678864, "learning_rate": 2.9318267907493583e-05, "loss": 0.6702, "step": 790 }, { "epoch": 0.10670579212377872, "grad_norm": 1.5178081087799355, "learning_rate": 2.9299216044462903e-05, "loss": 0.7346, "step": 800 }, { "epoch": 0.10803961452532596, "grad_norm": 9.506616797760028, "learning_rate": 2.927990798715475e-05, "loss": 0.6558, "step": 810 }, { "epoch": 0.1093734369268732, "grad_norm": 2.4597311302505767, "learning_rate": 2.926034408150513e-05, "loss": 0.726, "step": 820 }, { "epoch": 0.11070725932842042, "grad_norm": 12.372180964422007, "learning_rate": 2.9240524678034016e-05, "loss": 0.7308, "step": 830 }, { "epoch": 0.11204108172996766, "grad_norm": 1.4488469801164658, "learning_rate": 2.9220450131839037e-05, "loss": 0.7072, "step": 840 }, { "epoch": 0.1133749041315149, "grad_norm": 8.602946960846197, "learning_rate": 2.920012080258912e-05, "loss": 0.7234, "step": 850 }, { "epoch": 0.11470872653306212, "grad_norm": 1.441195423452674, "learning_rate": 2.9179537054518085e-05, "loss": 0.6934, "step": 860 }, { "epoch": 0.11604254893460936, "grad_norm": 4.318952956999577, "learning_rate": 2.9158699256418056e-05, "loss": 0.6534, "step": 870 }, { "epoch": 0.1173763713361566, "grad_norm": 9.733179695623866, "learning_rate": 2.9137607781632913e-05, "loss": 0.71, "step": 880 }, { "epoch": 0.11871019373770382, "grad_norm": 7.397049093836735, "learning_rate": 2.911626300805155e-05, "loss": 0.7386, "step": 890 }, { "epoch": 0.12004401613925106, "grad_norm": 2.920812240139869, "learning_rate": 2.9094665318101155e-05, "loss": 0.6789, "step": 900 }, { "epoch": 0.1213778385407983, "grad_norm": 1.7031296196271206, "learning_rate": 2.9072815098740326e-05, "loss": 0.715, "step": 910 }, { "epoch": 0.12271166094234552, "grad_norm": 1.5630656172291801, "learning_rate": 2.9050712741452136e-05, "loss": 0.7136, "step": 920 }, { "epoch": 0.12404548334389276, "grad_norm": 7.870543414771234, "learning_rate": 2.902835864223715e-05, "loss": 0.6669, "step": 930 }, { "epoch": 0.12537930574543998, "grad_norm": 4.843671834991794, "learning_rate": 2.9005753201606287e-05, "loss": 0.7281, "step": 940 }, { "epoch": 0.12671312814698724, "grad_norm": 3.010503818258016, "learning_rate": 2.8982896824573678e-05, "loss": 0.7018, "step": 950 }, { "epoch": 0.12804695054853446, "grad_norm": 2.5552186559589654, "learning_rate": 2.8959789920649394e-05, "loss": 0.7338, "step": 960 }, { "epoch": 0.12938077295008168, "grad_norm": 12.306055851495117, "learning_rate": 2.893643290383212e-05, "loss": 0.6732, "step": 970 }, { "epoch": 0.13071459535162894, "grad_norm": 2.16185926525944, "learning_rate": 2.891282619260172e-05, "loss": 0.7108, "step": 980 }, { "epoch": 0.13204841775317616, "grad_norm": 5.992378798792086, "learning_rate": 2.8888970209911754e-05, "loss": 0.6525, "step": 990 }, { "epoch": 0.13338224015472339, "grad_norm": 2.986272238787896, "learning_rate": 2.8864865383181893e-05, "loss": 0.6655, "step": 1000 }, { "epoch": 0.13471606255627064, "grad_norm": 12.855377354582437, "learning_rate": 2.8840512144290273e-05, "loss": 0.6826, "step": 1010 }, { "epoch": 0.13604988495781786, "grad_norm": 2.045979893776702, "learning_rate": 2.8815910929565734e-05, "loss": 0.6616, "step": 1020 }, { "epoch": 0.1373837073593651, "grad_norm": 6.623264301300591, "learning_rate": 2.879106217978002e-05, "loss": 0.6935, "step": 1030 }, { "epoch": 0.13871752976091234, "grad_norm": 2.67990218211766, "learning_rate": 2.8765966340139892e-05, "loss": 0.6671, "step": 1040 }, { "epoch": 0.14005135216245956, "grad_norm": 2.699521523924172, "learning_rate": 2.8740623860279116e-05, "loss": 0.6763, "step": 1050 }, { "epoch": 0.14138517456400682, "grad_norm": 4.1129898011507535, "learning_rate": 2.871503519425044e-05, "loss": 0.7159, "step": 1060 }, { "epoch": 0.14271899696555404, "grad_norm": 2.4592021333659146, "learning_rate": 2.8689200800517448e-05, "loss": 0.6551, "step": 1070 }, { "epoch": 0.14405281936710126, "grad_norm": 5.138500389099849, "learning_rate": 2.866312114194634e-05, "loss": 0.7214, "step": 1080 }, { "epoch": 0.14538664176864852, "grad_norm": 2.822433730666048, "learning_rate": 2.8636796685797657e-05, "loss": 0.6862, "step": 1090 }, { "epoch": 0.14672046417019574, "grad_norm": 3.086468537427806, "learning_rate": 2.8610227903717876e-05, "loss": 0.6784, "step": 1100 }, { "epoch": 0.14805428657174297, "grad_norm": 2.079766793749202, "learning_rate": 2.8583415271730994e-05, "loss": 0.7065, "step": 1110 }, { "epoch": 0.14938810897329022, "grad_norm": 1.659870509072264, "learning_rate": 2.855635927022998e-05, "loss": 0.7197, "step": 1120 }, { "epoch": 0.15072193137483744, "grad_norm": 7.870626779339635, "learning_rate": 2.8529060383968175e-05, "loss": 0.7305, "step": 1130 }, { "epoch": 0.15205575377638467, "grad_norm": 3.0600340899893537, "learning_rate": 2.850151910205061e-05, "loss": 0.6922, "step": 1140 }, { "epoch": 0.15338957617793192, "grad_norm": 3.6147451373702806, "learning_rate": 2.847373591792523e-05, "loss": 0.7044, "step": 1150 }, { "epoch": 0.15472339857947914, "grad_norm": 4.740777951553679, "learning_rate": 2.844571132937407e-05, "loss": 0.6794, "step": 1160 }, { "epoch": 0.15605722098102637, "grad_norm": 3.377522973717319, "learning_rate": 2.841744583850431e-05, "loss": 0.673, "step": 1170 }, { "epoch": 0.15739104338257362, "grad_norm": 4.250656077289992, "learning_rate": 2.838893995173932e-05, "loss": 0.6975, "step": 1180 }, { "epoch": 0.15872486578412084, "grad_norm": 11.73693900915769, "learning_rate": 2.836019417980955e-05, "loss": 0.6572, "step": 1190 }, { "epoch": 0.16005868818566807, "grad_norm": 2.729291714043308, "learning_rate": 2.8331209037743387e-05, "loss": 0.7247, "step": 1200 }, { "epoch": 0.16139251058721532, "grad_norm": 2.347985877636318, "learning_rate": 2.8301985044857947e-05, "loss": 0.7199, "step": 1210 }, { "epoch": 0.16272633298876255, "grad_norm": 2.2534314586033113, "learning_rate": 2.8272522724749743e-05, "loss": 0.6835, "step": 1220 }, { "epoch": 0.16406015539030977, "grad_norm": 3.159583116387406, "learning_rate": 2.8242822605285323e-05, "loss": 0.7122, "step": 1230 }, { "epoch": 0.16539397779185702, "grad_norm": 2.086588782887239, "learning_rate": 2.8212885218591812e-05, "loss": 0.6949, "step": 1240 }, { "epoch": 0.16672780019340425, "grad_norm": 7.284236966547317, "learning_rate": 2.8182711101047362e-05, "loss": 0.6641, "step": 1250 }, { "epoch": 0.16806162259495147, "grad_norm": 3.0369619450249594, "learning_rate": 2.815230079327156e-05, "loss": 0.6731, "step": 1260 }, { "epoch": 0.16939544499649872, "grad_norm": 1.4144726574636068, "learning_rate": 2.8121654840115734e-05, "loss": 0.6898, "step": 1270 }, { "epoch": 0.17072926739804595, "grad_norm": 3.66202356670303, "learning_rate": 2.809077379065319e-05, "loss": 0.7174, "step": 1280 }, { "epoch": 0.17206308979959317, "grad_norm": 4.778073521019285, "learning_rate": 2.805965819816937e-05, "loss": 0.6186, "step": 1290 }, { "epoch": 0.17339691220114042, "grad_norm": 3.9620427201734576, "learning_rate": 2.802830862015196e-05, "loss": 0.684, "step": 1300 }, { "epoch": 0.17473073460268765, "grad_norm": 4.170199740083487, "learning_rate": 2.799672561828087e-05, "loss": 0.7102, "step": 1310 }, { "epoch": 0.17606455700423487, "grad_norm": 2.2612205048804714, "learning_rate": 2.79649097584182e-05, "loss": 0.7451, "step": 1320 }, { "epoch": 0.17739837940578213, "grad_norm": 1.7156828128822517, "learning_rate": 2.7932861610598077e-05, "loss": 0.6641, "step": 1330 }, { "epoch": 0.17873220180732935, "grad_norm": 7.960733847217257, "learning_rate": 2.7900581749016466e-05, "loss": 0.7365, "step": 1340 }, { "epoch": 0.1800660242088766, "grad_norm": 2.5364939682563756, "learning_rate": 2.7868070752020865e-05, "loss": 0.7078, "step": 1350 }, { "epoch": 0.18139984661042383, "grad_norm": 2.7446281678776137, "learning_rate": 2.7835329202099944e-05, "loss": 0.7214, "step": 1360 }, { "epoch": 0.18273366901197105, "grad_norm": 3.2416602016145886, "learning_rate": 2.7802357685873117e-05, "loss": 0.6757, "step": 1370 }, { "epoch": 0.1840674914135183, "grad_norm": 5.225459736579946, "learning_rate": 2.7769156794080033e-05, "loss": 0.7381, "step": 1380 }, { "epoch": 0.18540131381506553, "grad_norm": 5.176692689501482, "learning_rate": 2.7735727121569967e-05, "loss": 0.7354, "step": 1390 }, { "epoch": 0.18673513621661275, "grad_norm": 2.7441883232342574, "learning_rate": 2.770206926729121e-05, "loss": 0.6937, "step": 1400 }, { "epoch": 0.18806895861816, "grad_norm": 2.9792116246243525, "learning_rate": 2.7668183834280284e-05, "loss": 0.6641, "step": 1410 }, { "epoch": 0.18940278101970723, "grad_norm": 2.4645298487410723, "learning_rate": 2.763407142965117e-05, "loss": 0.6274, "step": 1420 }, { "epoch": 0.19073660342125445, "grad_norm": 7.245032878035033, "learning_rate": 2.759973266458444e-05, "loss": 0.6962, "step": 1430 }, { "epoch": 0.1920704258228017, "grad_norm": 5.642209662597534, "learning_rate": 2.756516815431627e-05, "loss": 0.7016, "step": 1440 }, { "epoch": 0.19340424822434893, "grad_norm": 2.9804981875184526, "learning_rate": 2.7530378518127445e-05, "loss": 0.7331, "step": 1450 }, { "epoch": 0.19473807062589615, "grad_norm": 7.496561660992361, "learning_rate": 2.7495364379332256e-05, "loss": 0.7234, "step": 1460 }, { "epoch": 0.1960718930274434, "grad_norm": 1.6139389803246291, "learning_rate": 2.7460126365267335e-05, "loss": 0.7013, "step": 1470 }, { "epoch": 0.19740571542899063, "grad_norm": 4.618678334755141, "learning_rate": 2.7424665107280402e-05, "loss": 0.6892, "step": 1480 }, { "epoch": 0.19873953783053785, "grad_norm": 15.494190234738744, "learning_rate": 2.738898124071898e-05, "loss": 0.6785, "step": 1490 }, { "epoch": 0.2000733602320851, "grad_norm": 3.1680363319798954, "learning_rate": 2.735307540491898e-05, "loss": 0.669, "step": 1500 }, { "epoch": 0.20140718263363233, "grad_norm": 2.5397562341036224, "learning_rate": 2.7316948243193273e-05, "loss": 0.6726, "step": 1510 }, { "epoch": 0.20274100503517956, "grad_norm": 4.139021422606072, "learning_rate": 2.7280600402820146e-05, "loss": 0.6706, "step": 1520 }, { "epoch": 0.2040748274367268, "grad_norm": 2.7422468825646065, "learning_rate": 2.724403253503171e-05, "loss": 0.7078, "step": 1530 }, { "epoch": 0.20540864983827403, "grad_norm": 2.744225768808104, "learning_rate": 2.7207245295002242e-05, "loss": 0.6821, "step": 1540 }, { "epoch": 0.20674247223982126, "grad_norm": 2.234040668790152, "learning_rate": 2.7170239341836436e-05, "loss": 0.7451, "step": 1550 }, { "epoch": 0.2080762946413685, "grad_norm": 2.531733996425376, "learning_rate": 2.7133015338557585e-05, "loss": 0.7205, "step": 1560 }, { "epoch": 0.20941011704291573, "grad_norm": 2.9772483856455616, "learning_rate": 2.7095573952095727e-05, "loss": 0.7274, "step": 1570 }, { "epoch": 0.21074393944446296, "grad_norm": 3.317235333047955, "learning_rate": 2.705791585327568e-05, "loss": 0.7309, "step": 1580 }, { "epoch": 0.2120777618460102, "grad_norm": 1.9652386793628944, "learning_rate": 2.7020041716805014e-05, "loss": 0.7157, "step": 1590 }, { "epoch": 0.21341158424755743, "grad_norm": 2.93724058913164, "learning_rate": 2.6981952221261986e-05, "loss": 0.7123, "step": 1600 }, { "epoch": 0.21474540664910466, "grad_norm": 6.395577225750395, "learning_rate": 2.6943648049083366e-05, "loss": 0.6991, "step": 1610 }, { "epoch": 0.2160792290506519, "grad_norm": 2.4292347967714973, "learning_rate": 2.6905129886552208e-05, "loss": 0.7004, "step": 1620 }, { "epoch": 0.21741305145219914, "grad_norm": 1.8304810950546353, "learning_rate": 2.6866398423785568e-05, "loss": 0.6941, "step": 1630 }, { "epoch": 0.2187468738537464, "grad_norm": 2.762870839632077, "learning_rate": 2.682745435472212e-05, "loss": 0.6928, "step": 1640 }, { "epoch": 0.2200806962552936, "grad_norm": 3.4172019229090917, "learning_rate": 2.6788298377109748e-05, "loss": 0.7344, "step": 1650 }, { "epoch": 0.22141451865684084, "grad_norm": 2.7483538989548175, "learning_rate": 2.6748931192493017e-05, "loss": 0.7367, "step": 1660 }, { "epoch": 0.2227483410583881, "grad_norm": 7.314729269236597, "learning_rate": 2.670935350620063e-05, "loss": 0.6849, "step": 1670 }, { "epoch": 0.2240821634599353, "grad_norm": 3.8688065039432527, "learning_rate": 2.6669566027332767e-05, "loss": 0.6812, "step": 1680 }, { "epoch": 0.22541598586148254, "grad_norm": 7.10517346658295, "learning_rate": 2.6629569468748404e-05, "loss": 0.6089, "step": 1690 }, { "epoch": 0.2267498082630298, "grad_norm": 2.4198822683275147, "learning_rate": 2.658936454705251e-05, "loss": 0.6666, "step": 1700 }, { "epoch": 0.22808363066457701, "grad_norm": 2.4915285584652054, "learning_rate": 2.6548951982583246e-05, "loss": 0.7088, "step": 1710 }, { "epoch": 0.22941745306612424, "grad_norm": 2.2849831540010537, "learning_rate": 2.650833249939903e-05, "loss": 0.7149, "step": 1720 }, { "epoch": 0.2307512754676715, "grad_norm": 1.5098088938051029, "learning_rate": 2.6467506825265573e-05, "loss": 0.7254, "step": 1730 }, { "epoch": 0.23208509786921871, "grad_norm": 3.4800248296443814, "learning_rate": 2.642647569164284e-05, "loss": 0.6916, "step": 1740 }, { "epoch": 0.23341892027076594, "grad_norm": 7.281500947090542, "learning_rate": 2.638523983367194e-05, "loss": 0.6831, "step": 1750 }, { "epoch": 0.2347527426723132, "grad_norm": 3.0161864395495446, "learning_rate": 2.634379999016198e-05, "loss": 0.6999, "step": 1760 }, { "epoch": 0.23608656507386042, "grad_norm": 2.0917745352156762, "learning_rate": 2.6302156903576784e-05, "loss": 0.7112, "step": 1770 }, { "epoch": 0.23742038747540764, "grad_norm": 1.918811185774526, "learning_rate": 2.6260311320021628e-05, "loss": 0.6725, "step": 1780 }, { "epoch": 0.2387542098769549, "grad_norm": 3.0697413876733695, "learning_rate": 2.6218263989229855e-05, "loss": 0.7133, "step": 1790 }, { "epoch": 0.24008803227850212, "grad_norm": 6.14274393655379, "learning_rate": 2.617601566454944e-05, "loss": 0.6678, "step": 1800 }, { "epoch": 0.24142185468004934, "grad_norm": 4.259979200715344, "learning_rate": 2.613356710292951e-05, "loss": 0.7013, "step": 1810 }, { "epoch": 0.2427556770815966, "grad_norm": 3.1011058557692808, "learning_rate": 2.6090919064906766e-05, "loss": 0.7027, "step": 1820 }, { "epoch": 0.24408949948314382, "grad_norm": 3.677900978078831, "learning_rate": 2.6048072314591854e-05, "loss": 0.711, "step": 1830 }, { "epoch": 0.24542332188469104, "grad_norm": 2.368576699713982, "learning_rate": 2.600502761965569e-05, "loss": 0.6917, "step": 1840 }, { "epoch": 0.2467571442862383, "grad_norm": 3.0346306894457, "learning_rate": 2.59617857513157e-05, "loss": 0.69, "step": 1850 }, { "epoch": 0.24809096668778552, "grad_norm": 3.1228131080916204, "learning_rate": 2.591834748432198e-05, "loss": 0.695, "step": 1860 }, { "epoch": 0.24942478908933274, "grad_norm": 2.6886660685401034, "learning_rate": 2.5874713596943465e-05, "loss": 0.6681, "step": 1870 }, { "epoch": 0.25075861149087997, "grad_norm": 1.7244460999561722, "learning_rate": 2.5830884870953933e-05, "loss": 0.6737, "step": 1880 }, { "epoch": 0.25209243389242725, "grad_norm": 2.4283725332509842, "learning_rate": 2.578686209161803e-05, "loss": 0.6598, "step": 1890 }, { "epoch": 0.2534262562939745, "grad_norm": 5.496556851547161, "learning_rate": 2.5742646047677186e-05, "loss": 0.6931, "step": 1900 }, { "epoch": 0.2547600786955217, "grad_norm": 1.2751270156124934, "learning_rate": 2.5698237531335493e-05, "loss": 0.7043, "step": 1910 }, { "epoch": 0.2560939010970689, "grad_norm": 8.807017683974516, "learning_rate": 2.56536373382455e-05, "loss": 0.6234, "step": 1920 }, { "epoch": 0.25742772349861615, "grad_norm": 3.6331868296726277, "learning_rate": 2.5608846267493974e-05, "loss": 0.6763, "step": 1930 }, { "epoch": 0.25876154590016337, "grad_norm": 5.094905230807839, "learning_rate": 2.5563865121587563e-05, "loss": 0.6692, "step": 1940 }, { "epoch": 0.26009536830171065, "grad_norm": 2.0520732769663237, "learning_rate": 2.5518694706438445e-05, "loss": 0.7008, "step": 1950 }, { "epoch": 0.2614291907032579, "grad_norm": 2.1265138955486336, "learning_rate": 2.5473335831349842e-05, "loss": 0.6623, "step": 1960 }, { "epoch": 0.2627630131048051, "grad_norm": 4.532469697105077, "learning_rate": 2.5427789309001577e-05, "loss": 0.7099, "step": 1970 }, { "epoch": 0.2640968355063523, "grad_norm": 1.8912900905557881, "learning_rate": 2.538205595543548e-05, "loss": 0.712, "step": 1980 }, { "epoch": 0.26543065790789955, "grad_norm": 9.714825687307293, "learning_rate": 2.5336136590040767e-05, "loss": 0.6418, "step": 1990 }, { "epoch": 0.26676448030944677, "grad_norm": 4.375615975749738, "learning_rate": 2.529003203553937e-05, "loss": 0.6933, "step": 2000 }, { "epoch": 0.26809830271099405, "grad_norm": 5.945657366701919, "learning_rate": 2.5243743117971186e-05, "loss": 0.6748, "step": 2010 }, { "epoch": 0.2694321251125413, "grad_norm": 7.453951551881255, "learning_rate": 2.5197270666679295e-05, "loss": 0.7004, "step": 2020 }, { "epoch": 0.2707659475140885, "grad_norm": 2.3916662603858665, "learning_rate": 2.515061551429509e-05, "loss": 0.6961, "step": 2030 }, { "epoch": 0.2720997699156357, "grad_norm": 3.5972047868369104, "learning_rate": 2.5103778496723334e-05, "loss": 0.7058, "step": 2040 }, { "epoch": 0.27343359231718295, "grad_norm": 4.525268184238612, "learning_rate": 2.5056760453127242e-05, "loss": 0.6704, "step": 2050 }, { "epoch": 0.2747674147187302, "grad_norm": 5.9581146555788465, "learning_rate": 2.5009562225913385e-05, "loss": 0.6722, "step": 2060 }, { "epoch": 0.27610123712027745, "grad_norm": 4.163590223716233, "learning_rate": 2.4962184660716645e-05, "loss": 0.6933, "step": 2070 }, { "epoch": 0.2774350595218247, "grad_norm": 2.0180801697563258, "learning_rate": 2.4914628606385022e-05, "loss": 0.6982, "step": 2080 }, { "epoch": 0.2787688819233719, "grad_norm": 2.3996169579330373, "learning_rate": 2.4866894914964462e-05, "loss": 0.6832, "step": 2090 }, { "epoch": 0.2801027043249191, "grad_norm": 20.07054133895426, "learning_rate": 2.481898444168357e-05, "loss": 0.6871, "step": 2100 }, { "epoch": 0.28143652672646635, "grad_norm": 3.563765719247629, "learning_rate": 2.4770898044938284e-05, "loss": 0.703, "step": 2110 }, { "epoch": 0.28277034912801363, "grad_norm": 1.9816905810381245, "learning_rate": 2.4722636586276522e-05, "loss": 0.7132, "step": 2120 }, { "epoch": 0.28410417152956086, "grad_norm": 4.0053115388283205, "learning_rate": 2.4674200930382712e-05, "loss": 0.6991, "step": 2130 }, { "epoch": 0.2854379939311081, "grad_norm": 1.9643538302216321, "learning_rate": 2.4625591945062326e-05, "loss": 0.7182, "step": 2140 }, { "epoch": 0.2867718163326553, "grad_norm": 1.7027289253737494, "learning_rate": 2.4576810501226318e-05, "loss": 0.6856, "step": 2150 }, { "epoch": 0.28810563873420253, "grad_norm": 3.394597130806682, "learning_rate": 2.4527857472875515e-05, "loss": 0.7013, "step": 2160 }, { "epoch": 0.28943946113574975, "grad_norm": 2.766786923916393, "learning_rate": 2.447873373708498e-05, "loss": 0.6913, "step": 2170 }, { "epoch": 0.29077328353729703, "grad_norm": 6.781532105937228, "learning_rate": 2.4429440173988275e-05, "loss": 0.7401, "step": 2180 }, { "epoch": 0.29210710593884426, "grad_norm": 2.6220209383444946, "learning_rate": 2.43799776667617e-05, "loss": 0.7287, "step": 2190 }, { "epoch": 0.2934409283403915, "grad_norm": 4.597566226152422, "learning_rate": 2.4330347101608492e-05, "loss": 0.6664, "step": 2200 }, { "epoch": 0.2947747507419387, "grad_norm": 3.15622915128866, "learning_rate": 2.428054936774289e-05, "loss": 0.6757, "step": 2210 }, { "epoch": 0.29610857314348593, "grad_norm": 3.5777836932521065, "learning_rate": 2.423058535737427e-05, "loss": 0.7396, "step": 2220 }, { "epoch": 0.29744239554503316, "grad_norm": 2.505384749600403, "learning_rate": 2.418045596569111e-05, "loss": 0.7156, "step": 2230 }, { "epoch": 0.29877621794658044, "grad_norm": 15.640998645324629, "learning_rate": 2.4130162090844976e-05, "loss": 0.7016, "step": 2240 }, { "epoch": 0.30011004034812766, "grad_norm": 6.1147200283733865, "learning_rate": 2.4079704633934427e-05, "loss": 0.6835, "step": 2250 }, { "epoch": 0.3014438627496749, "grad_norm": 2.4704828096249907, "learning_rate": 2.4029084498988864e-05, "loss": 0.717, "step": 2260 }, { "epoch": 0.3027776851512221, "grad_norm": 3.624817679194012, "learning_rate": 2.3978302592952332e-05, "loss": 0.6863, "step": 2270 }, { "epoch": 0.30411150755276933, "grad_norm": 7.1778372122735155, "learning_rate": 2.392735982566728e-05, "loss": 0.7057, "step": 2280 }, { "epoch": 0.30544532995431656, "grad_norm": 1.541203747230883, "learning_rate": 2.387625710985826e-05, "loss": 0.6755, "step": 2290 }, { "epoch": 0.30677915235586384, "grad_norm": 5.290753363343769, "learning_rate": 2.3824995361115552e-05, "loss": 0.7214, "step": 2300 }, { "epoch": 0.30811297475741106, "grad_norm": 11.18524078914846, "learning_rate": 2.3773575497878784e-05, "loss": 0.687, "step": 2310 }, { "epoch": 0.3094467971589583, "grad_norm": 2.8473409260968854, "learning_rate": 2.372199844142048e-05, "loss": 0.6588, "step": 2320 }, { "epoch": 0.3107806195605055, "grad_norm": 3.6509202763742894, "learning_rate": 2.3670265115829523e-05, "loss": 0.7146, "step": 2330 }, { "epoch": 0.31211444196205274, "grad_norm": 2.86323212169014, "learning_rate": 2.3618376447994633e-05, "loss": 0.6965, "step": 2340 }, { "epoch": 0.31344826436359996, "grad_norm": 1.6724444694024563, "learning_rate": 2.3566333367587737e-05, "loss": 0.6827, "step": 2350 }, { "epoch": 0.31478208676514724, "grad_norm": 3.7438462947121876, "learning_rate": 2.3514136807047318e-05, "loss": 0.677, "step": 2360 }, { "epoch": 0.31611590916669446, "grad_norm": 3.150319939971515, "learning_rate": 2.3461787701561724e-05, "loss": 0.6926, "step": 2370 }, { "epoch": 0.3174497315682417, "grad_norm": 1.9724696911512674, "learning_rate": 2.340928698905239e-05, "loss": 0.7269, "step": 2380 }, { "epoch": 0.3187835539697889, "grad_norm": 2.6615995505256604, "learning_rate": 2.335663561015704e-05, "loss": 0.719, "step": 2390 }, { "epoch": 0.32011737637133614, "grad_norm": 3.648818329043563, "learning_rate": 2.3303834508212845e-05, "loss": 0.6593, "step": 2400 }, { "epoch": 0.3214511987728834, "grad_norm": 5.032935766388129, "learning_rate": 2.325088462923951e-05, "loss": 0.7018, "step": 2410 }, { "epoch": 0.32278502117443064, "grad_norm": 5.116190153583237, "learning_rate": 2.319778692192233e-05, "loss": 0.6138, "step": 2420 }, { "epoch": 0.32411884357597787, "grad_norm": 8.77553429349065, "learning_rate": 2.3144542337595196e-05, "loss": 0.6995, "step": 2430 }, { "epoch": 0.3254526659775251, "grad_norm": 4.020402137418298, "learning_rate": 2.3091151830223537e-05, "loss": 0.6935, "step": 2440 }, { "epoch": 0.3267864883790723, "grad_norm": 2.326990350307363, "learning_rate": 2.3037616356387237e-05, "loss": 0.6657, "step": 2450 }, { "epoch": 0.32812031078061954, "grad_norm": 1.9450305290081706, "learning_rate": 2.2983936875263495e-05, "loss": 0.6884, "step": 2460 }, { "epoch": 0.3294541331821668, "grad_norm": 2.4083218262957407, "learning_rate": 2.2930114348609655e-05, "loss": 0.6324, "step": 2470 }, { "epoch": 0.33078795558371404, "grad_norm": 4.469293094525185, "learning_rate": 2.2876149740745935e-05, "loss": 0.7054, "step": 2480 }, { "epoch": 0.33212177798526127, "grad_norm": 3.0408327884382613, "learning_rate": 2.28220440185382e-05, "loss": 0.6996, "step": 2490 }, { "epoch": 0.3334556003868085, "grad_norm": 2.5340984000691273, "learning_rate": 2.2767798151380597e-05, "loss": 0.6908, "step": 2500 }, { "epoch": 0.3347894227883557, "grad_norm": 2.4867165525033, "learning_rate": 2.27134131111782e-05, "loss": 0.6838, "step": 2510 }, { "epoch": 0.33612324518990294, "grad_norm": 14.755496795057269, "learning_rate": 2.2658889872329628e-05, "loss": 0.7072, "step": 2520 }, { "epoch": 0.3374570675914502, "grad_norm": 11.498768616138861, "learning_rate": 2.2604229411709518e-05, "loss": 0.6837, "step": 2530 }, { "epoch": 0.33879088999299745, "grad_norm": 1.6627733851927542, "learning_rate": 2.25494327086511e-05, "loss": 0.6948, "step": 2540 }, { "epoch": 0.34012471239454467, "grad_norm": 4.465322393758394, "learning_rate": 2.2494500744928583e-05, "loss": 0.706, "step": 2550 }, { "epoch": 0.3414585347960919, "grad_norm": 2.5329140738676714, "learning_rate": 2.243943450473963e-05, "loss": 0.6652, "step": 2560 }, { "epoch": 0.3427923571976391, "grad_norm": 2.6213955428320963, "learning_rate": 2.2384234974687658e-05, "loss": 0.7123, "step": 2570 }, { "epoch": 0.34412617959918634, "grad_norm": 2.8450668136715827, "learning_rate": 2.2328903143764216e-05, "loss": 0.6748, "step": 2580 }, { "epoch": 0.3454600020007336, "grad_norm": 9.246863580911334, "learning_rate": 2.2273440003331237e-05, "loss": 0.6774, "step": 2590 }, { "epoch": 0.34679382440228085, "grad_norm": 2.610989556515575, "learning_rate": 2.2217846547103275e-05, "loss": 0.7042, "step": 2600 }, { "epoch": 0.3481276468038281, "grad_norm": 7.325969061692186, "learning_rate": 2.216212377112972e-05, "loss": 0.6834, "step": 2610 }, { "epoch": 0.3494614692053753, "grad_norm": 3.001379331751721, "learning_rate": 2.2106272673776934e-05, "loss": 0.7033, "step": 2620 }, { "epoch": 0.3507952916069225, "grad_norm": 3.463073346975308, "learning_rate": 2.2050294255710375e-05, "loss": 0.6839, "step": 2630 }, { "epoch": 0.35212911400846975, "grad_norm": 3.524564101951424, "learning_rate": 2.1994189519876663e-05, "loss": 0.6948, "step": 2640 }, { "epoch": 0.353462936410017, "grad_norm": 3.152341329769827, "learning_rate": 2.19379594714856e-05, "loss": 0.6767, "step": 2650 }, { "epoch": 0.35479675881156425, "grad_norm": 4.2343916663936305, "learning_rate": 2.188160511799219e-05, "loss": 0.6755, "step": 2660 }, { "epoch": 0.3561305812131115, "grad_norm": 2.7909676165285813, "learning_rate": 2.1825127469078555e-05, "loss": 0.6694, "step": 2670 }, { "epoch": 0.3574644036146587, "grad_norm": 1.8765416483232782, "learning_rate": 2.1768527536635868e-05, "loss": 0.7031, "step": 2680 }, { "epoch": 0.3587982260162059, "grad_norm": 13.262978009985517, "learning_rate": 2.171180633474621e-05, "loss": 0.7371, "step": 2690 }, { "epoch": 0.3601320484177532, "grad_norm": 3.886717400478723, "learning_rate": 2.1654964879664407e-05, "loss": 0.7109, "step": 2700 }, { "epoch": 0.3614658708193004, "grad_norm": 2.040560351248799, "learning_rate": 2.1598004189799826e-05, "loss": 0.7274, "step": 2710 }, { "epoch": 0.36279969322084765, "grad_norm": 24.610089275348535, "learning_rate": 2.1540925285698122e-05, "loss": 0.6886, "step": 2720 }, { "epoch": 0.3641335156223949, "grad_norm": 3.6439264742220216, "learning_rate": 2.148372919002295e-05, "loss": 0.681, "step": 2730 }, { "epoch": 0.3654673380239421, "grad_norm": 5.83580774778366, "learning_rate": 2.142641692753765e-05, "loss": 0.6502, "step": 2740 }, { "epoch": 0.3668011604254893, "grad_norm": 1.8530940550203352, "learning_rate": 2.1368989525086893e-05, "loss": 0.6854, "step": 2750 }, { "epoch": 0.3681349828270366, "grad_norm": 5.003536499561226, "learning_rate": 2.1311448011578255e-05, "loss": 0.6699, "step": 2760 }, { "epoch": 0.36946880522858383, "grad_norm": 2.6889933495770912, "learning_rate": 2.125379341796382e-05, "loss": 0.741, "step": 2770 }, { "epoch": 0.37080262763013105, "grad_norm": 2.0672372686575575, "learning_rate": 2.1196026777221684e-05, "loss": 0.693, "step": 2780 }, { "epoch": 0.3721364500316783, "grad_norm": 3.023122371840424, "learning_rate": 2.1138149124337448e-05, "loss": 0.7227, "step": 2790 }, { "epoch": 0.3734702724332255, "grad_norm": 5.98908480573641, "learning_rate": 2.108016149628569e-05, "loss": 0.6875, "step": 2800 }, { "epoch": 0.3748040948347727, "grad_norm": 13.324804502845906, "learning_rate": 2.102206493201137e-05, "loss": 0.6693, "step": 2810 }, { "epoch": 0.37613791723632, "grad_norm": 2.877158805709884, "learning_rate": 2.096386047241123e-05, "loss": 0.6752, "step": 2820 }, { "epoch": 0.37747173963786723, "grad_norm": 3.417018003930411, "learning_rate": 2.0905549160315116e-05, "loss": 0.6874, "step": 2830 }, { "epoch": 0.37880556203941446, "grad_norm": 6.197947611584602, "learning_rate": 2.084713204046734e-05, "loss": 0.6995, "step": 2840 }, { "epoch": 0.3801393844409617, "grad_norm": 2.4400537269180327, "learning_rate": 2.078861015950793e-05, "loss": 0.718, "step": 2850 }, { "epoch": 0.3814732068425089, "grad_norm": 3.4313321352162878, "learning_rate": 2.072998456595387e-05, "loss": 0.6928, "step": 2860 }, { "epoch": 0.38280702924405613, "grad_norm": 3.323108743280233, "learning_rate": 2.0671256310180334e-05, "loss": 0.7141, "step": 2870 }, { "epoch": 0.3841408516456034, "grad_norm": 2.270407423855968, "learning_rate": 2.0612426444401874e-05, "loss": 0.6677, "step": 2880 }, { "epoch": 0.38547467404715063, "grad_norm": 4.473087793045971, "learning_rate": 2.0553496022653535e-05, "loss": 0.706, "step": 2890 }, { "epoch": 0.38680849644869786, "grad_norm": 4.498504602131192, "learning_rate": 2.0494466100772006e-05, "loss": 0.6783, "step": 2900 }, { "epoch": 0.3881423188502451, "grad_norm": 1.8721168603816298, "learning_rate": 2.0435337736376677e-05, "loss": 0.7327, "step": 2910 }, { "epoch": 0.3894761412517923, "grad_norm": 2.1819398242824093, "learning_rate": 2.03761119888507e-05, "loss": 0.6798, "step": 2920 }, { "epoch": 0.39080996365333953, "grad_norm": 29.747303047069977, "learning_rate": 2.031678991932201e-05, "loss": 0.7045, "step": 2930 }, { "epoch": 0.3921437860548868, "grad_norm": 4.708328967247123, "learning_rate": 2.0257372590644314e-05, "loss": 0.6896, "step": 2940 }, { "epoch": 0.39347760845643404, "grad_norm": 2.873510721340991, "learning_rate": 2.0197861067378044e-05, "loss": 0.6802, "step": 2950 }, { "epoch": 0.39481143085798126, "grad_norm": 4.540574995423212, "learning_rate": 2.0138256415771275e-05, "loss": 0.6219, "step": 2960 }, { "epoch": 0.3961452532595285, "grad_norm": 11.817372765224325, "learning_rate": 2.0078559703740654e-05, "loss": 0.65, "step": 2970 }, { "epoch": 0.3974790756610757, "grad_norm": 11.004144754692504, "learning_rate": 2.0018772000852216e-05, "loss": 0.7056, "step": 2980 }, { "epoch": 0.398812898062623, "grad_norm": 1.7365475356133573, "learning_rate": 1.9958894378302265e-05, "loss": 0.6827, "step": 2990 }, { "epoch": 0.4001467204641702, "grad_norm": 4.31426545646336, "learning_rate": 1.989892790889817e-05, "loss": 0.6796, "step": 3000 }, { "epoch": 0.40148054286571744, "grad_norm": 2.534413468413497, "learning_rate": 1.9838873667039134e-05, "loss": 0.6825, "step": 3010 }, { "epoch": 0.40281436526726466, "grad_norm": 2.5821079814088, "learning_rate": 1.9778732728696937e-05, "loss": 0.6522, "step": 3020 }, { "epoch": 0.4041481876688119, "grad_norm": 10.45675108188373, "learning_rate": 1.9718506171396694e-05, "loss": 0.6752, "step": 3030 }, { "epoch": 0.4054820100703591, "grad_norm": 10.969680268488736, "learning_rate": 1.965819507419751e-05, "loss": 0.7195, "step": 3040 }, { "epoch": 0.4068158324719064, "grad_norm": 9.540053007670354, "learning_rate": 1.9597800517673165e-05, "loss": 0.6762, "step": 3050 }, { "epoch": 0.4081496548734536, "grad_norm": 8.551702443669248, "learning_rate": 1.9537323583892753e-05, "loss": 0.7292, "step": 3060 }, { "epoch": 0.40948347727500084, "grad_norm": 3.0994689178852903, "learning_rate": 1.9476765356401304e-05, "loss": 0.6764, "step": 3070 }, { "epoch": 0.41081729967654806, "grad_norm": 3.1013298812228163, "learning_rate": 1.9416126920200344e-05, "loss": 0.6484, "step": 3080 }, { "epoch": 0.4121511220780953, "grad_norm": 2.00628497131861, "learning_rate": 1.9355409361728482e-05, "loss": 0.7094, "step": 3090 }, { "epoch": 0.4134849444796425, "grad_norm": 5.224082004633703, "learning_rate": 1.9294613768841932e-05, "loss": 0.7279, "step": 3100 }, { "epoch": 0.4148187668811898, "grad_norm": 18.62631978728915, "learning_rate": 1.9233741230795022e-05, "loss": 0.662, "step": 3110 }, { "epoch": 0.416152589282737, "grad_norm": 3.6495526914982968, "learning_rate": 1.9172792838220686e-05, "loss": 0.6836, "step": 3120 }, { "epoch": 0.41748641168428424, "grad_norm": 2.304337917905853, "learning_rate": 1.9111769683110914e-05, "loss": 0.6901, "step": 3130 }, { "epoch": 0.41882023408583147, "grad_norm": 8.427846401703292, "learning_rate": 1.905067285879719e-05, "loss": 0.6606, "step": 3140 }, { "epoch": 0.4201540564873787, "grad_norm": 2.2306668115119104, "learning_rate": 1.8989503459930908e-05, "loss": 0.7434, "step": 3150 }, { "epoch": 0.4214878788889259, "grad_norm": 2.231586663842237, "learning_rate": 1.892826258246376e-05, "loss": 0.7184, "step": 3160 }, { "epoch": 0.4228217012904732, "grad_norm": 5.804571835994344, "learning_rate": 1.886695132362808e-05, "loss": 0.7073, "step": 3170 }, { "epoch": 0.4241555236920204, "grad_norm": 4.7472512172058785, "learning_rate": 1.8805570781917228e-05, "loss": 0.7102, "step": 3180 }, { "epoch": 0.42548934609356764, "grad_norm": 1.723627694530291, "learning_rate": 1.8744122057065856e-05, "loss": 0.6828, "step": 3190 }, { "epoch": 0.42682316849511487, "grad_norm": 1.9952068710149184, "learning_rate": 1.868260625003024e-05, "loss": 0.6545, "step": 3200 }, { "epoch": 0.4281569908966621, "grad_norm": 4.588444559005735, "learning_rate": 1.8621024462968553e-05, "loss": 0.67, "step": 3210 }, { "epoch": 0.4294908132982093, "grad_norm": 2.155634253115107, "learning_rate": 1.85593777992211e-05, "loss": 0.7173, "step": 3220 }, { "epoch": 0.4308246356997566, "grad_norm": 3.3412948579128194, "learning_rate": 1.849766736329056e-05, "loss": 0.6364, "step": 3230 }, { "epoch": 0.4321584581013038, "grad_norm": 2.1344417176214607, "learning_rate": 1.8435894260822208e-05, "loss": 0.6919, "step": 3240 }, { "epoch": 0.43349228050285105, "grad_norm": 3.8410669902748764, "learning_rate": 1.8374059598584084e-05, "loss": 0.6524, "step": 3250 }, { "epoch": 0.43482610290439827, "grad_norm": 2.609728029777106, "learning_rate": 1.831216448444717e-05, "loss": 0.688, "step": 3260 }, { "epoch": 0.4361599253059455, "grad_norm": 2.182084710285402, "learning_rate": 1.8250210027365562e-05, "loss": 0.7327, "step": 3270 }, { "epoch": 0.4374937477074928, "grad_norm": 1.0672619638672702, "learning_rate": 1.818819733735657e-05, "loss": 0.7137, "step": 3280 }, { "epoch": 0.43882757010904, "grad_norm": 1.7248236414002174, "learning_rate": 1.812612752548084e-05, "loss": 0.6848, "step": 3290 }, { "epoch": 0.4401613925105872, "grad_norm": 2.717100059326369, "learning_rate": 1.806400170382246e-05, "loss": 0.6582, "step": 3300 }, { "epoch": 0.44149521491213445, "grad_norm": 2.7420980324781348, "learning_rate": 1.8001820985469026e-05, "loss": 0.6976, "step": 3310 }, { "epoch": 0.4428290373136817, "grad_norm": 3.9917362204420357, "learning_rate": 1.7939586484491704e-05, "loss": 0.7259, "step": 3320 }, { "epoch": 0.4441628597152289, "grad_norm": 3.2371945093430514, "learning_rate": 1.787729931592525e-05, "loss": 0.6883, "step": 3330 }, { "epoch": 0.4454966821167762, "grad_norm": 2.439245137250377, "learning_rate": 1.781496059574807e-05, "loss": 0.6876, "step": 3340 }, { "epoch": 0.4468305045183234, "grad_norm": 4.525984025887397, "learning_rate": 1.7752571440862178e-05, "loss": 0.6724, "step": 3350 }, { "epoch": 0.4481643269198706, "grad_norm": 2.3388903272276518, "learning_rate": 1.7690132969073223e-05, "loss": 0.7065, "step": 3360 }, { "epoch": 0.44949814932141785, "grad_norm": 6.946538587379132, "learning_rate": 1.7627646299070457e-05, "loss": 0.6444, "step": 3370 }, { "epoch": 0.4508319717229651, "grad_norm": 1.5334789635428385, "learning_rate": 1.7565112550406663e-05, "loss": 0.6597, "step": 3380 }, { "epoch": 0.4521657941245123, "grad_norm": 1.7438745925855814, "learning_rate": 1.7502532843478134e-05, "loss": 0.736, "step": 3390 }, { "epoch": 0.4534996165260596, "grad_norm": 2.352884928297456, "learning_rate": 1.743990829950458e-05, "loss": 0.7209, "step": 3400 }, { "epoch": 0.4548334389276068, "grad_norm": 2.589791551987411, "learning_rate": 1.737724004050903e-05, "loss": 0.6873, "step": 3410 }, { "epoch": 0.45616726132915403, "grad_norm": 1.5018800238986845, "learning_rate": 1.731452918929774e-05, "loss": 0.6993, "step": 3420 }, { "epoch": 0.45750108373070125, "grad_norm": 1.618737845945941, "learning_rate": 1.7251776869440097e-05, "loss": 0.719, "step": 3430 }, { "epoch": 0.4588349061322485, "grad_norm": 4.764891120811521, "learning_rate": 1.718898420524845e-05, "loss": 0.7066, "step": 3440 }, { "epoch": 0.4601687285337957, "grad_norm": 30.008073864717016, "learning_rate": 1.7126152321757985e-05, "loss": 0.7234, "step": 3450 }, { "epoch": 0.461502550935343, "grad_norm": 4.718402571866902, "learning_rate": 1.7063282344706577e-05, "loss": 0.671, "step": 3460 }, { "epoch": 0.4628363733368902, "grad_norm": 3.279168331496427, "learning_rate": 1.7000375400514602e-05, "loss": 0.6748, "step": 3470 }, { "epoch": 0.46417019573843743, "grad_norm": 4.202866783860852, "learning_rate": 1.693743261626476e-05, "loss": 0.7135, "step": 3480 }, { "epoch": 0.46550401813998465, "grad_norm": 2.959211747400748, "learning_rate": 1.68744551196819e-05, "loss": 0.6684, "step": 3490 }, { "epoch": 0.4668378405415319, "grad_norm": 3.7208053935256085, "learning_rate": 1.6811444039112787e-05, "loss": 0.6842, "step": 3500 }, { "epoch": 0.4681716629430791, "grad_norm": 1.8411337183473255, "learning_rate": 1.6748400503505905e-05, "loss": 0.6796, "step": 3510 }, { "epoch": 0.4695054853446264, "grad_norm": 1.5569024338481647, "learning_rate": 1.6685325642391223e-05, "loss": 0.7357, "step": 3520 }, { "epoch": 0.4708393077461736, "grad_norm": 2.30459532472586, "learning_rate": 1.662222058585996e-05, "loss": 0.6825, "step": 3530 }, { "epoch": 0.47217313014772083, "grad_norm": 1.6593076444414934, "learning_rate": 1.6559086464544334e-05, "loss": 0.7067, "step": 3540 }, { "epoch": 0.47350695254926806, "grad_norm": 2.6738168898709356, "learning_rate": 1.6495924409597305e-05, "loss": 0.665, "step": 3550 }, { "epoch": 0.4748407749508153, "grad_norm": 10.974918207024547, "learning_rate": 1.6432735552672317e-05, "loss": 0.705, "step": 3560 }, { "epoch": 0.4761745973523625, "grad_norm": 4.279092732465272, "learning_rate": 1.636952102590301e-05, "loss": 0.6858, "step": 3570 }, { "epoch": 0.4775084197539098, "grad_norm": 8.958608602390235, "learning_rate": 1.630628196188295e-05, "loss": 0.7022, "step": 3580 }, { "epoch": 0.478842242155457, "grad_norm": 1.2316277268276075, "learning_rate": 1.6243019493645315e-05, "loss": 0.7091, "step": 3590 }, { "epoch": 0.48017606455700423, "grad_norm": 1.6977852924595596, "learning_rate": 1.617973475464262e-05, "loss": 0.6725, "step": 3600 }, { "epoch": 0.48150988695855146, "grad_norm": 9.102696583046576, "learning_rate": 1.6116428878726396e-05, "loss": 0.706, "step": 3610 }, { "epoch": 0.4828437093600987, "grad_norm": 2.983654314671525, "learning_rate": 1.6053103000126874e-05, "loss": 0.6663, "step": 3620 }, { "epoch": 0.48417753176164596, "grad_norm": 2.9273555172026304, "learning_rate": 1.598975825343267e-05, "loss": 0.6986, "step": 3630 }, { "epoch": 0.4855113541631932, "grad_norm": 2.4687475856334613, "learning_rate": 1.5926395773570447e-05, "loss": 0.7192, "step": 3640 }, { "epoch": 0.4868451765647404, "grad_norm": 4.171039626246759, "learning_rate": 1.5863016695784604e-05, "loss": 0.6702, "step": 3650 }, { "epoch": 0.48817899896628764, "grad_norm": 3.8655482044779337, "learning_rate": 1.5799622155616887e-05, "loss": 0.6568, "step": 3660 }, { "epoch": 0.48951282136783486, "grad_norm": 2.8245022157946362, "learning_rate": 1.5736213288886112e-05, "loss": 0.7075, "step": 3670 }, { "epoch": 0.4908466437693821, "grad_norm": 2.1969432272158556, "learning_rate": 1.567279123166776e-05, "loss": 0.7043, "step": 3680 }, { "epoch": 0.49218046617092936, "grad_norm": 3.7154807458182835, "learning_rate": 1.560935712027364e-05, "loss": 0.6467, "step": 3690 }, { "epoch": 0.4935142885724766, "grad_norm": 4.060155573527941, "learning_rate": 1.5545912091231543e-05, "loss": 0.6957, "step": 3700 }, { "epoch": 0.4948481109740238, "grad_norm": 2.057087008440973, "learning_rate": 1.548245728126486e-05, "loss": 0.6656, "step": 3710 }, { "epoch": 0.49618193337557104, "grad_norm": 1.975534767472513, "learning_rate": 1.5418993827272224e-05, "loss": 0.6867, "step": 3720 }, { "epoch": 0.49751575577711826, "grad_norm": 11.237169875747464, "learning_rate": 1.5355522866307144e-05, "loss": 0.693, "step": 3730 }, { "epoch": 0.4988495781786655, "grad_norm": 2.7505125088389066, "learning_rate": 1.529204553555762e-05, "loss": 0.6715, "step": 3740 }, { "epoch": 0.5001834005802127, "grad_norm": 14.47964311360144, "learning_rate": 1.522856297232579e-05, "loss": 0.6638, "step": 3750 }, { "epoch": 0.5015172229817599, "grad_norm": 1.4576903787797197, "learning_rate": 1.5165076314007529e-05, "loss": 0.6461, "step": 3760 }, { "epoch": 0.5028510453833072, "grad_norm": 4.190097060433623, "learning_rate": 1.5101586698072095e-05, "loss": 0.6997, "step": 3770 }, { "epoch": 0.5041848677848545, "grad_norm": 2.6358802196743887, "learning_rate": 1.5038095262041725e-05, "loss": 0.6805, "step": 3780 }, { "epoch": 0.5055186901864017, "grad_norm": 2.9885793100944484, "learning_rate": 1.4974603143471268e-05, "loss": 0.663, "step": 3790 }, { "epoch": 0.506852512587949, "grad_norm": 3.364287860442736, "learning_rate": 1.4911111479927804e-05, "loss": 0.6851, "step": 3800 }, { "epoch": 0.5081863349894962, "grad_norm": 6.415730527817265, "learning_rate": 1.4847621408970266e-05, "loss": 0.6544, "step": 3810 }, { "epoch": 0.5095201573910434, "grad_norm": 1.6327349630681778, "learning_rate": 1.4784134068129043e-05, "loss": 0.6629, "step": 3820 }, { "epoch": 0.5108539797925906, "grad_norm": 3.0622996050606783, "learning_rate": 1.4720650594885614e-05, "loss": 0.6651, "step": 3830 }, { "epoch": 0.5121878021941378, "grad_norm": 5.445942430441996, "learning_rate": 1.4657172126652167e-05, "loss": 0.664, "step": 3840 }, { "epoch": 0.5135216245956851, "grad_norm": 4.518334654823446, "learning_rate": 1.459369980075121e-05, "loss": 0.6959, "step": 3850 }, { "epoch": 0.5148554469972323, "grad_norm": 1.8471627413065406, "learning_rate": 1.4530234754395207e-05, "loss": 0.6774, "step": 3860 }, { "epoch": 0.5161892693987795, "grad_norm": 3.6484122755334525, "learning_rate": 1.4466778124666192e-05, "loss": 0.6825, "step": 3870 }, { "epoch": 0.5175230918003267, "grad_norm": 2.087118207544068, "learning_rate": 1.4403331048495404e-05, "loss": 0.6985, "step": 3880 }, { "epoch": 0.5188569142018741, "grad_norm": 11.878313425481934, "learning_rate": 1.4339894662642914e-05, "loss": 0.6764, "step": 3890 }, { "epoch": 0.5201907366034213, "grad_norm": 2.5453717997032115, "learning_rate": 1.4276470103677257e-05, "loss": 0.7091, "step": 3900 }, { "epoch": 0.5215245590049685, "grad_norm": 4.791248513372535, "learning_rate": 1.4213058507955072e-05, "loss": 0.644, "step": 3910 }, { "epoch": 0.5228583814065157, "grad_norm": 2.1955258954683545, "learning_rate": 1.4149661011600734e-05, "loss": 0.6954, "step": 3920 }, { "epoch": 0.524192203808063, "grad_norm": 3.5143987933185676, "learning_rate": 1.4086278750486017e-05, "loss": 0.6848, "step": 3930 }, { "epoch": 0.5255260262096102, "grad_norm": 3.168504700204386, "learning_rate": 1.4022912860209709e-05, "loss": 0.6752, "step": 3940 }, { "epoch": 0.5268598486111574, "grad_norm": 1.9655682723891459, "learning_rate": 1.3959564476077308e-05, "loss": 0.6904, "step": 3950 }, { "epoch": 0.5281936710127046, "grad_norm": 1.6897897373972772, "learning_rate": 1.389623473308065e-05, "loss": 0.6929, "step": 3960 }, { "epoch": 0.5295274934142519, "grad_norm": 4.400154605229998, "learning_rate": 1.3832924765877587e-05, "loss": 0.726, "step": 3970 }, { "epoch": 0.5308613158157991, "grad_norm": 2.790842978581456, "learning_rate": 1.3769635708771654e-05, "loss": 0.6724, "step": 3980 }, { "epoch": 0.5321951382173463, "grad_norm": 1.5712798066752716, "learning_rate": 1.3706368695691745e-05, "loss": 0.6703, "step": 3990 }, { "epoch": 0.5335289606188935, "grad_norm": 5.340886291219129, "learning_rate": 1.3643124860171801e-05, "loss": 0.6595, "step": 4000 }, { "epoch": 0.5348627830204409, "grad_norm": 1.985940330857511, "learning_rate": 1.35799053353305e-05, "loss": 0.6892, "step": 4010 }, { "epoch": 0.5361966054219881, "grad_norm": 3.917331449757074, "learning_rate": 1.3516711253850949e-05, "loss": 0.6417, "step": 4020 }, { "epoch": 0.5375304278235353, "grad_norm": 1.66962823795828, "learning_rate": 1.3453543747960393e-05, "loss": 0.6784, "step": 4030 }, { "epoch": 0.5388642502250826, "grad_norm": 4.181035760200595, "learning_rate": 1.3390403949409943e-05, "loss": 0.7115, "step": 4040 }, { "epoch": 0.5401980726266298, "grad_norm": 2.4193575665243214, "learning_rate": 1.3327292989454273e-05, "loss": 0.7104, "step": 4050 }, { "epoch": 0.541531895028177, "grad_norm": 2.0442192962046275, "learning_rate": 1.3264211998831374e-05, "loss": 0.7008, "step": 4060 }, { "epoch": 0.5428657174297242, "grad_norm": 3.0689852808863183, "learning_rate": 1.3201162107742285e-05, "loss": 0.677, "step": 4070 }, { "epoch": 0.5441995398312715, "grad_norm": 2.22632841251654, "learning_rate": 1.3138144445830841e-05, "loss": 0.6223, "step": 4080 }, { "epoch": 0.5455333622328187, "grad_norm": 8.813265719863766, "learning_rate": 1.3075160142163442e-05, "loss": 0.6791, "step": 4090 }, { "epoch": 0.5468671846343659, "grad_norm": 2.461550778463616, "learning_rate": 1.3012210325208818e-05, "loss": 0.7165, "step": 4100 }, { "epoch": 0.5482010070359131, "grad_norm": 2.1304508310591896, "learning_rate": 1.2949296122817813e-05, "loss": 0.6905, "step": 4110 }, { "epoch": 0.5495348294374603, "grad_norm": 2.1733622775851535, "learning_rate": 1.2886418662203174e-05, "loss": 0.6963, "step": 4120 }, { "epoch": 0.5508686518390077, "grad_norm": 2.654530675610581, "learning_rate": 1.282357906991936e-05, "loss": 0.6796, "step": 4130 }, { "epoch": 0.5522024742405549, "grad_norm": 2.6976858995246085, "learning_rate": 1.276077847184236e-05, "loss": 0.6922, "step": 4140 }, { "epoch": 0.5535362966421021, "grad_norm": 2.5591371381474857, "learning_rate": 1.2698017993149504e-05, "loss": 0.7047, "step": 4150 }, { "epoch": 0.5548701190436494, "grad_norm": 6.439964637422321, "learning_rate": 1.2635298758299336e-05, "loss": 0.6722, "step": 4160 }, { "epoch": 0.5562039414451966, "grad_norm": 1.6222259612163727, "learning_rate": 1.2572621891011426e-05, "loss": 0.6646, "step": 4170 }, { "epoch": 0.5575377638467438, "grad_norm": 3.410425968580818, "learning_rate": 1.2509988514246272e-05, "loss": 0.6894, "step": 4180 }, { "epoch": 0.558871586248291, "grad_norm": 2.7111542804682327, "learning_rate": 1.2447399750185166e-05, "loss": 0.7196, "step": 4190 }, { "epoch": 0.5602054086498383, "grad_norm": 3.3657872237953868, "learning_rate": 1.2384856720210086e-05, "loss": 0.7052, "step": 4200 }, { "epoch": 0.5615392310513855, "grad_norm": 3.4383001609998143, "learning_rate": 1.2322360544883608e-05, "loss": 0.664, "step": 4210 }, { "epoch": 0.5628730534529327, "grad_norm": 4.31412552867304, "learning_rate": 1.2259912343928831e-05, "loss": 0.6923, "step": 4220 }, { "epoch": 0.5642068758544799, "grad_norm": 2.9738159323747655, "learning_rate": 1.2197513236209312e-05, "loss": 0.6787, "step": 4230 }, { "epoch": 0.5655406982560273, "grad_norm": 14.42279175461777, "learning_rate": 1.213516433970902e-05, "loss": 0.7313, "step": 4240 }, { "epoch": 0.5668745206575745, "grad_norm": 2.6156276324588195, "learning_rate": 1.2072866771512306e-05, "loss": 0.6856, "step": 4250 }, { "epoch": 0.5682083430591217, "grad_norm": 2.692794641012978, "learning_rate": 1.201062164778389e-05, "loss": 0.6587, "step": 4260 }, { "epoch": 0.5695421654606689, "grad_norm": 3.01896569407463, "learning_rate": 1.1948430083748864e-05, "loss": 0.7225, "step": 4270 }, { "epoch": 0.5708759878622162, "grad_norm": 2.266424840293995, "learning_rate": 1.1886293193672707e-05, "loss": 0.6847, "step": 4280 }, { "epoch": 0.5722098102637634, "grad_norm": 2.2789387948762987, "learning_rate": 1.1824212090841321e-05, "loss": 0.7011, "step": 4290 }, { "epoch": 0.5735436326653106, "grad_norm": 2.826447974943076, "learning_rate": 1.1762187887541088e-05, "loss": 0.689, "step": 4300 }, { "epoch": 0.5748774550668578, "grad_norm": 2.565293440960005, "learning_rate": 1.1700221695038944e-05, "loss": 0.7077, "step": 4310 }, { "epoch": 0.5762112774684051, "grad_norm": 4.459154190124916, "learning_rate": 1.1638314623562459e-05, "loss": 0.6885, "step": 4320 }, { "epoch": 0.5775450998699523, "grad_norm": 1.8187338733285852, "learning_rate": 1.1576467782279953e-05, "loss": 0.7103, "step": 4330 }, { "epoch": 0.5788789222714995, "grad_norm": 4.078050868504266, "learning_rate": 1.1514682279280621e-05, "loss": 0.6742, "step": 4340 }, { "epoch": 0.5802127446730467, "grad_norm": 2.4612673583806233, "learning_rate": 1.1452959221554684e-05, "loss": 0.6941, "step": 4350 }, { "epoch": 0.5815465670745941, "grad_norm": 8.05059787591381, "learning_rate": 1.1391299714973553e-05, "loss": 0.7072, "step": 4360 }, { "epoch": 0.5828803894761413, "grad_norm": 5.041675641180621, "learning_rate": 1.1329704864270005e-05, "loss": 0.6914, "step": 4370 }, { "epoch": 0.5842142118776885, "grad_norm": 3.8176735967050672, "learning_rate": 1.1268175773018409e-05, "loss": 0.6489, "step": 4380 }, { "epoch": 0.5855480342792357, "grad_norm": 2.068471874891413, "learning_rate": 1.1206713543614942e-05, "loss": 0.7182, "step": 4390 }, { "epoch": 0.586881856680783, "grad_norm": 4.7154770167485065, "learning_rate": 1.1145319277257834e-05, "loss": 0.6961, "step": 4400 }, { "epoch": 0.5882156790823302, "grad_norm": 3.3453200032391917, "learning_rate": 1.108399407392765e-05, "loss": 0.701, "step": 4410 }, { "epoch": 0.5895495014838774, "grad_norm": 3.462978751346215, "learning_rate": 1.1022739032367572e-05, "loss": 0.6504, "step": 4420 }, { "epoch": 0.5908833238854246, "grad_norm": 3.9283885591229075, "learning_rate": 1.0961555250063718e-05, "loss": 0.7025, "step": 4430 }, { "epoch": 0.5922171462869719, "grad_norm": 2.2363832425317463, "learning_rate": 1.090044382322548e-05, "loss": 0.7106, "step": 4440 }, { "epoch": 0.5935509686885191, "grad_norm": 2.4683539157329544, "learning_rate": 1.083940584676588e-05, "loss": 0.6919, "step": 4450 }, { "epoch": 0.5948847910900663, "grad_norm": 1.6027050129978238, "learning_rate": 1.077844241428195e-05, "loss": 0.6579, "step": 4460 }, { "epoch": 0.5962186134916136, "grad_norm": 4.272201666240297, "learning_rate": 1.071755461803515e-05, "loss": 0.6992, "step": 4470 }, { "epoch": 0.5975524358931609, "grad_norm": 4.847908056514074, "learning_rate": 1.0656743548931784e-05, "loss": 0.6858, "step": 4480 }, { "epoch": 0.5988862582947081, "grad_norm": 1.899776347699883, "learning_rate": 1.0596010296503469e-05, "loss": 0.7175, "step": 4490 }, { "epoch": 0.6002200806962553, "grad_norm": 3.6851504324405533, "learning_rate": 1.0535355948887598e-05, "loss": 0.6731, "step": 4500 }, { "epoch": 0.6015539030978025, "grad_norm": 20.935216614062877, "learning_rate": 1.0474781592807854e-05, "loss": 0.6548, "step": 4510 }, { "epoch": 0.6028877254993498, "grad_norm": 5.577424675925709, "learning_rate": 1.0414288313554746e-05, "loss": 0.7263, "step": 4520 }, { "epoch": 0.604221547900897, "grad_norm": 2.9726973141053334, "learning_rate": 1.0353877194966152e-05, "loss": 0.7446, "step": 4530 }, { "epoch": 0.6055553703024442, "grad_norm": 2.021480129071628, "learning_rate": 1.0293549319407901e-05, "loss": 0.7137, "step": 4540 }, { "epoch": 0.6068891927039914, "grad_norm": 1.9390208520343517, "learning_rate": 1.0233305767754391e-05, "loss": 0.6998, "step": 4550 }, { "epoch": 0.6082230151055387, "grad_norm": 2.2439008274229337, "learning_rate": 1.0173147619369212e-05, "loss": 0.6977, "step": 4560 }, { "epoch": 0.6095568375070859, "grad_norm": 3.002628922946286, "learning_rate": 1.0113075952085815e-05, "loss": 0.7119, "step": 4570 }, { "epoch": 0.6108906599086331, "grad_norm": 1.8784698804400835, "learning_rate": 1.0053091842188196e-05, "loss": 0.6813, "step": 4580 }, { "epoch": 0.6122244823101805, "grad_norm": 3.6775461109208702, "learning_rate": 9.993196364391614e-06, "loss": 0.6963, "step": 4590 }, { "epoch": 0.6135583047117277, "grad_norm": 3.0082378136289636, "learning_rate": 9.93339059182334e-06, "loss": 0.6761, "step": 4600 }, { "epoch": 0.6148921271132749, "grad_norm": 2.0259105048263297, "learning_rate": 9.873675596003424e-06, "loss": 0.6645, "step": 4610 }, { "epoch": 0.6162259495148221, "grad_norm": 7.087002002369676, "learning_rate": 9.8140524468255e-06, "loss": 0.6836, "step": 4620 }, { "epoch": 0.6175597719163693, "grad_norm": 6.82917662319771, "learning_rate": 9.754522212537614e-06, "loss": 0.6546, "step": 4630 }, { "epoch": 0.6188935943179166, "grad_norm": 2.7798504683532546, "learning_rate": 9.695085959723088e-06, "loss": 0.6879, "step": 4640 }, { "epoch": 0.6202274167194638, "grad_norm": 2.9169362806410124, "learning_rate": 9.63574475328141e-06, "loss": 0.7287, "step": 4650 }, { "epoch": 0.621561239121011, "grad_norm": 1.9790125803612642, "learning_rate": 9.576499656409158e-06, "loss": 0.6933, "step": 4660 }, { "epoch": 0.6228950615225582, "grad_norm": 3.533798783312709, "learning_rate": 9.517351730580939e-06, "loss": 0.6763, "step": 4670 }, { "epoch": 0.6242288839241055, "grad_norm": 4.906070778847422, "learning_rate": 9.458302035530384e-06, "loss": 0.7089, "step": 4680 }, { "epoch": 0.6255627063256527, "grad_norm": 3.448200148869349, "learning_rate": 9.399351629231154e-06, "loss": 0.6911, "step": 4690 }, { "epoch": 0.6268965287271999, "grad_norm": 7.159835250493477, "learning_rate": 9.340501567877989e-06, "loss": 0.6387, "step": 4700 }, { "epoch": 0.6282303511287473, "grad_norm": 7.692987979738203, "learning_rate": 9.281752905867778e-06, "loss": 0.6239, "step": 4710 }, { "epoch": 0.6295641735302945, "grad_norm": 3.0193719069272187, "learning_rate": 9.223106695780677e-06, "loss": 0.6755, "step": 4720 }, { "epoch": 0.6308979959318417, "grad_norm": 2.31533598338176, "learning_rate": 9.164563988361242e-06, "loss": 0.6772, "step": 4730 }, { "epoch": 0.6322318183333889, "grad_norm": 4.426498530040442, "learning_rate": 9.106125832499604e-06, "loss": 0.689, "step": 4740 }, { "epoch": 0.6335656407349362, "grad_norm": 3.055489437274157, "learning_rate": 9.047793275212686e-06, "loss": 0.687, "step": 4750 }, { "epoch": 0.6348994631364834, "grad_norm": 7.12383434732346, "learning_rate": 8.989567361625427e-06, "loss": 0.6604, "step": 4760 }, { "epoch": 0.6362332855380306, "grad_norm": 2.6553339556706788, "learning_rate": 8.931449134952075e-06, "loss": 0.6866, "step": 4770 }, { "epoch": 0.6375671079395778, "grad_norm": 3.1179277884273806, "learning_rate": 8.873439636477484e-06, "loss": 0.6599, "step": 4780 }, { "epoch": 0.638900930341125, "grad_norm": 2.4632186731676993, "learning_rate": 8.815539905538459e-06, "loss": 0.6957, "step": 4790 }, { "epoch": 0.6402347527426723, "grad_norm": 2.243620092515075, "learning_rate": 8.757750979505137e-06, "loss": 0.678, "step": 4800 }, { "epoch": 0.6415685751442195, "grad_norm": 1.559849971379389, "learning_rate": 8.700073893762408e-06, "loss": 0.675, "step": 4810 }, { "epoch": 0.6429023975457668, "grad_norm": 2.174037487987736, "learning_rate": 8.642509681691347e-06, "loss": 0.6654, "step": 4820 }, { "epoch": 0.6442362199473141, "grad_norm": 4.920990395145698, "learning_rate": 8.585059374650717e-06, "loss": 0.6839, "step": 4830 }, { "epoch": 0.6455700423488613, "grad_norm": 2.4413434562237115, "learning_rate": 8.527724001958476e-06, "loss": 0.7275, "step": 4840 }, { "epoch": 0.6469038647504085, "grad_norm": 1.8344146906183378, "learning_rate": 8.470504590873346e-06, "loss": 0.6961, "step": 4850 }, { "epoch": 0.6482376871519557, "grad_norm": 1.9274949192717368, "learning_rate": 8.413402166576397e-06, "loss": 0.6802, "step": 4860 }, { "epoch": 0.649571509553503, "grad_norm": 2.823435230180327, "learning_rate": 8.3564177521527e-06, "loss": 0.6545, "step": 4870 }, { "epoch": 0.6509053319550502, "grad_norm": 12.01208015208349, "learning_rate": 8.29955236857297e-06, "loss": 0.6608, "step": 4880 }, { "epoch": 0.6522391543565974, "grad_norm": 2.7110104949534146, "learning_rate": 8.242807034675289e-06, "loss": 0.6825, "step": 4890 }, { "epoch": 0.6535729767581446, "grad_norm": 3.5400540139770467, "learning_rate": 8.186182767146848e-06, "loss": 0.7173, "step": 4900 }, { "epoch": 0.6549067991596919, "grad_norm": 2.22571512795563, "learning_rate": 8.12968058050574e-06, "loss": 0.6979, "step": 4910 }, { "epoch": 0.6562406215612391, "grad_norm": 9.112812839574897, "learning_rate": 8.073301487082768e-06, "loss": 0.7059, "step": 4920 }, { "epoch": 0.6575744439627863, "grad_norm": 3.3720066783743867, "learning_rate": 8.017046497003308e-06, "loss": 0.6896, "step": 4930 }, { "epoch": 0.6589082663643336, "grad_norm": 6.1591898293726945, "learning_rate": 7.960916618169233e-06, "loss": 0.697, "step": 4940 }, { "epoch": 0.6602420887658809, "grad_norm": 2.7295855620488716, "learning_rate": 7.904912856240833e-06, "loss": 0.6892, "step": 4950 }, { "epoch": 0.6615759111674281, "grad_norm": 2.345166503103055, "learning_rate": 7.849036214618802e-06, "loss": 0.6748, "step": 4960 }, { "epoch": 0.6629097335689753, "grad_norm": 7.498544170227416, "learning_rate": 7.793287694426263e-06, "loss": 0.7016, "step": 4970 }, { "epoch": 0.6642435559705225, "grad_norm": 2.0475845854151986, "learning_rate": 7.737668294490834e-06, "loss": 0.7179, "step": 4980 }, { "epoch": 0.6655773783720698, "grad_norm": 3.114340952412098, "learning_rate": 7.68217901132672e-06, "loss": 0.6793, "step": 4990 }, { "epoch": 0.666911200773617, "grad_norm": 3.83883079292243, "learning_rate": 7.626820839116876e-06, "loss": 0.6876, "step": 5000 }, { "epoch": 0.6682450231751642, "grad_norm": 8.432894431693887, "learning_rate": 7.571594769695181e-06, "loss": 0.6874, "step": 5010 }, { "epoch": 0.6695788455767114, "grad_norm": 2.674070025495073, "learning_rate": 7.51650179252867e-06, "loss": 0.7172, "step": 5020 }, { "epoch": 0.6709126679782587, "grad_norm": 5.973778009949667, "learning_rate": 7.461542894699818e-06, "loss": 0.7133, "step": 5030 }, { "epoch": 0.6722464903798059, "grad_norm": 3.2677383520376795, "learning_rate": 7.406719060888837e-06, "loss": 0.7314, "step": 5040 }, { "epoch": 0.6735803127813531, "grad_norm": 4.172418337526236, "learning_rate": 7.352031273356045e-06, "loss": 0.6805, "step": 5050 }, { "epoch": 0.6749141351829004, "grad_norm": 2.1121612232039477, "learning_rate": 7.297480511924263e-06, "loss": 0.664, "step": 5060 }, { "epoch": 0.6762479575844477, "grad_norm": 4.737396987962653, "learning_rate": 7.243067753961267e-06, "loss": 0.7202, "step": 5070 }, { "epoch": 0.6775817799859949, "grad_norm": 2.2892995457734053, "learning_rate": 7.188793974362254e-06, "loss": 0.6675, "step": 5080 }, { "epoch": 0.6789156023875421, "grad_norm": 2.5303725617403647, "learning_rate": 7.13466014553241e-06, "loss": 0.7053, "step": 5090 }, { "epoch": 0.6802494247890893, "grad_norm": 2.1911360902224315, "learning_rate": 7.080667237369468e-06, "loss": 0.6363, "step": 5100 }, { "epoch": 0.6815832471906366, "grad_norm": 1.7722416009046082, "learning_rate": 7.0268162172463215e-06, "loss": 0.6647, "step": 5110 }, { "epoch": 0.6829170695921838, "grad_norm": 2.3107884233221396, "learning_rate": 6.973108049993714e-06, "loss": 0.6566, "step": 5120 }, { "epoch": 0.684250891993731, "grad_norm": 2.374293674041196, "learning_rate": 6.919543697882938e-06, "loss": 0.6772, "step": 5130 }, { "epoch": 0.6855847143952782, "grad_norm": 1.323614967454432, "learning_rate": 6.866124120608596e-06, "loss": 0.7142, "step": 5140 }, { "epoch": 0.6869185367968255, "grad_norm": 2.5080725989534463, "learning_rate": 6.812850275271412e-06, "loss": 0.6672, "step": 5150 }, { "epoch": 0.6882523591983727, "grad_norm": 2.4821500591867527, "learning_rate": 6.759723116361077e-06, "loss": 0.6752, "step": 5160 }, { "epoch": 0.68958618159992, "grad_norm": 1.5591146033827414, "learning_rate": 6.706743595739151e-06, "loss": 0.6816, "step": 5170 }, { "epoch": 0.6909200040014672, "grad_norm": 3.7818920756836967, "learning_rate": 6.653912662622009e-06, "loss": 0.6865, "step": 5180 }, { "epoch": 0.6922538264030145, "grad_norm": 1.4437006855258527, "learning_rate": 6.601231263563832e-06, "loss": 0.6606, "step": 5190 }, { "epoch": 0.6935876488045617, "grad_norm": 16.033675993643524, "learning_rate": 6.548700342439648e-06, "loss": 0.6438, "step": 5200 }, { "epoch": 0.6949214712061089, "grad_norm": 4.180984666556721, "learning_rate": 6.496320840428426e-06, "loss": 0.6929, "step": 5210 }, { "epoch": 0.6962552936076561, "grad_norm": 2.58067260639148, "learning_rate": 6.444093695996205e-06, "loss": 0.6869, "step": 5220 }, { "epoch": 0.6975891160092034, "grad_norm": 5.857865039816727, "learning_rate": 6.392019844879289e-06, "loss": 0.691, "step": 5230 }, { "epoch": 0.6989229384107506, "grad_norm": 1.7969091660941678, "learning_rate": 6.340100220067473e-06, "loss": 0.6887, "step": 5240 }, { "epoch": 0.7002567608122978, "grad_norm": 4.64525683921728, "learning_rate": 6.28833575178733e-06, "loss": 0.7081, "step": 5250 }, { "epoch": 0.701590583213845, "grad_norm": 30.247747163980556, "learning_rate": 6.23672736748555e-06, "loss": 0.6977, "step": 5260 }, { "epoch": 0.7029244056153923, "grad_norm": 4.753750976434006, "learning_rate": 6.1852759918123145e-06, "loss": 0.6724, "step": 5270 }, { "epoch": 0.7042582280169395, "grad_norm": 1.9245050933564345, "learning_rate": 6.133982546604735e-06, "loss": 0.6498, "step": 5280 }, { "epoch": 0.7055920504184868, "grad_norm": 8.580113915174046, "learning_rate": 6.082847950870334e-06, "loss": 0.6798, "step": 5290 }, { "epoch": 0.706925872820034, "grad_norm": 5.254260741119434, "learning_rate": 6.031873120770585e-06, "loss": 0.7094, "step": 5300 }, { "epoch": 0.7082596952215813, "grad_norm": 4.267395769845368, "learning_rate": 5.9810589696044935e-06, "loss": 0.6447, "step": 5310 }, { "epoch": 0.7095935176231285, "grad_norm": 2.1531210048068203, "learning_rate": 5.9304064077922274e-06, "loss": 0.6686, "step": 5320 }, { "epoch": 0.7109273400246757, "grad_norm": 3.0627830530266467, "learning_rate": 5.879916342858821e-06, "loss": 0.6987, "step": 5330 }, { "epoch": 0.712261162426223, "grad_norm": 3.5786494907624546, "learning_rate": 5.829589679417901e-06, "loss": 0.7058, "step": 5340 }, { "epoch": 0.7135949848277702, "grad_norm": 4.212959667448277, "learning_rate": 5.779427319155485e-06, "loss": 0.6698, "step": 5350 }, { "epoch": 0.7149288072293174, "grad_norm": 5.081476320526283, "learning_rate": 5.7294301608138274e-06, "loss": 0.6978, "step": 5360 }, { "epoch": 0.7162626296308646, "grad_norm": 2.2390204735144006, "learning_rate": 5.679599100175312e-06, "loss": 0.6923, "step": 5370 }, { "epoch": 0.7175964520324118, "grad_norm": 3.722611802307121, "learning_rate": 5.629935030046409e-06, "loss": 0.7111, "step": 5380 }, { "epoch": 0.7189302744339591, "grad_norm": 8.85305178336297, "learning_rate": 5.580438840241671e-06, "loss": 0.7014, "step": 5390 }, { "epoch": 0.7202640968355064, "grad_norm": 2.7228108037923406, "learning_rate": 5.531111417567799e-06, "loss": 0.6918, "step": 5400 }, { "epoch": 0.7215979192370536, "grad_norm": 2.774305095096996, "learning_rate": 5.48195364580775e-06, "loss": 0.6986, "step": 5410 }, { "epoch": 0.7229317416386009, "grad_norm": 3.3079542242294053, "learning_rate": 5.432966405704895e-06, "loss": 0.702, "step": 5420 }, { "epoch": 0.7242655640401481, "grad_norm": 4.313463849702513, "learning_rate": 5.384150574947258e-06, "loss": 0.7034, "step": 5430 }, { "epoch": 0.7255993864416953, "grad_norm": 10.27608427466511, "learning_rate": 5.335507028151768e-06, "loss": 0.6847, "step": 5440 }, { "epoch": 0.7269332088432425, "grad_norm": 1.3643574661197668, "learning_rate": 5.2870366368486074e-06, "loss": 0.6652, "step": 5450 }, { "epoch": 0.7282670312447898, "grad_norm": 2.0571595612294313, "learning_rate": 5.238740269465584e-06, "loss": 0.7057, "step": 5460 }, { "epoch": 0.729600853646337, "grad_norm": 2.177709522875291, "learning_rate": 5.190618791312581e-06, "loss": 0.7147, "step": 5470 }, { "epoch": 0.7309346760478842, "grad_norm": 3.4529351229700134, "learning_rate": 5.142673064566048e-06, "loss": 0.6676, "step": 5480 }, { "epoch": 0.7322684984494314, "grad_norm": 2.0423403296749534, "learning_rate": 5.094903948253557e-06, "loss": 0.6519, "step": 5490 }, { "epoch": 0.7336023208509787, "grad_norm": 2.6755745244049045, "learning_rate": 5.047312298238407e-06, "loss": 0.6553, "step": 5500 }, { "epoch": 0.7349361432525259, "grad_norm": 2.0735822085238365, "learning_rate": 4.999898967204293e-06, "loss": 0.6853, "step": 5510 }, { "epoch": 0.7362699656540732, "grad_norm": 1.4637677526008275, "learning_rate": 4.952664804640032e-06, "loss": 0.6382, "step": 5520 }, { "epoch": 0.7376037880556204, "grad_norm": 10.3457254327091, "learning_rate": 4.905610656824338e-06, "loss": 0.6963, "step": 5530 }, { "epoch": 0.7389376104571677, "grad_norm": 3.7186152329588307, "learning_rate": 4.858737366810661e-06, "loss": 0.6892, "step": 5540 }, { "epoch": 0.7402714328587149, "grad_norm": 3.866784770301919, "learning_rate": 4.812045774412074e-06, "loss": 0.6976, "step": 5550 }, { "epoch": 0.7416052552602621, "grad_norm": 1.2902307476254877, "learning_rate": 4.765536716186247e-06, "loss": 0.651, "step": 5560 }, { "epoch": 0.7429390776618093, "grad_norm": 5.660668096573102, "learning_rate": 4.719211025420436e-06, "loss": 0.6715, "step": 5570 }, { "epoch": 0.7442729000633566, "grad_norm": 1.7865867976347491, "learning_rate": 4.673069532116575e-06, "loss": 0.683, "step": 5580 }, { "epoch": 0.7456067224649038, "grad_norm": 8.841085187491435, "learning_rate": 4.627113062976379e-06, "loss": 0.7104, "step": 5590 }, { "epoch": 0.746940544866451, "grad_norm": 1.8568471185779514, "learning_rate": 4.581342441386563e-06, "loss": 0.6738, "step": 5600 }, { "epoch": 0.7482743672679982, "grad_norm": 8.613029244471035, "learning_rate": 4.53575848740406e-06, "loss": 0.6907, "step": 5610 }, { "epoch": 0.7496081896695455, "grad_norm": 5.379674431405216, "learning_rate": 4.490362017741346e-06, "loss": 0.6547, "step": 5620 }, { "epoch": 0.7509420120710927, "grad_norm": 40.42031518782124, "learning_rate": 4.445153845751808e-06, "loss": 0.6615, "step": 5630 }, { "epoch": 0.75227583447264, "grad_norm": 2.262382554857583, "learning_rate": 4.4001347814151625e-06, "loss": 0.7268, "step": 5640 }, { "epoch": 0.7536096568741872, "grad_norm": 4.773999070319672, "learning_rate": 4.355305631322943e-06, "loss": 0.6671, "step": 5650 }, { "epoch": 0.7549434792757345, "grad_norm": 2.4606314332551893, "learning_rate": 4.31066719866406e-06, "loss": 0.6989, "step": 5660 }, { "epoch": 0.7562773016772817, "grad_norm": 3.715537855764248, "learning_rate": 4.266220283210403e-06, "loss": 0.7022, "step": 5670 }, { "epoch": 0.7576111240788289, "grad_norm": 2.812731415446895, "learning_rate": 4.221965681302506e-06, "loss": 0.6742, "step": 5680 }, { "epoch": 0.7589449464803761, "grad_norm": 2.21543044943041, "learning_rate": 4.177904185835289e-06, "loss": 0.6574, "step": 5690 }, { "epoch": 0.7602787688819234, "grad_norm": 2.5996433086745623, "learning_rate": 4.134036586243852e-06, "loss": 0.6634, "step": 5700 }, { "epoch": 0.7616125912834706, "grad_norm": 1.4311723346543919, "learning_rate": 4.0903636684893205e-06, "loss": 0.6986, "step": 5710 }, { "epoch": 0.7629464136850178, "grad_norm": 1.5544477881800915, "learning_rate": 4.046886215044773e-06, "loss": 0.6982, "step": 5720 }, { "epoch": 0.764280236086565, "grad_norm": 50.87643723015965, "learning_rate": 4.003605004881224e-06, "loss": 0.6944, "step": 5730 }, { "epoch": 0.7656140584881123, "grad_norm": 3.3621842813994958, "learning_rate": 3.960520813453654e-06, "loss": 0.6388, "step": 5740 }, { "epoch": 0.7669478808896596, "grad_norm": 1.6373066358476196, "learning_rate": 3.917634412687132e-06, "loss": 0.6381, "step": 5750 }, { "epoch": 0.7682817032912068, "grad_norm": 4.021243191453306, "learning_rate": 3.874946570962977e-06, "loss": 0.679, "step": 5760 }, { "epoch": 0.769615525692754, "grad_norm": 2.1771219666869404, "learning_rate": 3.832458053104985e-06, "loss": 0.65, "step": 5770 }, { "epoch": 0.7709493480943013, "grad_norm": 4.117022111093457, "learning_rate": 3.790169620365742e-06, "loss": 0.6968, "step": 5780 }, { "epoch": 0.7722831704958485, "grad_norm": 1.7554846375395396, "learning_rate": 3.748082030412971e-06, "loss": 0.6943, "step": 5790 }, { "epoch": 0.7736169928973957, "grad_norm": 1.0196062550048701, "learning_rate": 3.7061960373159603e-06, "loss": 0.6867, "step": 5800 }, { "epoch": 0.7749508152989429, "grad_norm": 1.7419968573435696, "learning_rate": 3.66451239153206e-06, "loss": 0.6265, "step": 5810 }, { "epoch": 0.7762846377004902, "grad_norm": 2.3677756205266065, "learning_rate": 3.623031839893226e-06, "loss": 0.7055, "step": 5820 }, { "epoch": 0.7776184601020374, "grad_norm": 3.9343316788594915, "learning_rate": 3.5817551255926473e-06, "loss": 0.6789, "step": 5830 }, { "epoch": 0.7789522825035846, "grad_norm": 2.48678634987911, "learning_rate": 3.5406829881714254e-06, "loss": 0.6917, "step": 5840 }, { "epoch": 0.7802861049051318, "grad_norm": 3.0165585668876678, "learning_rate": 3.4998161635053274e-06, "loss": 0.6703, "step": 5850 }, { "epoch": 0.7816199273066791, "grad_norm": 1.6901003264945378, "learning_rate": 3.459155383791601e-06, "loss": 0.6551, "step": 5860 }, { "epoch": 0.7829537497082264, "grad_norm": 3.1065648861476727, "learning_rate": 3.4187013775358515e-06, "loss": 0.6593, "step": 5870 }, { "epoch": 0.7842875721097736, "grad_norm": 3.5522700882458516, "learning_rate": 3.3784548695389993e-06, "loss": 0.6649, "step": 5880 }, { "epoch": 0.7856213945113208, "grad_norm": 3.2401137004728997, "learning_rate": 3.338416580884284e-06, "loss": 0.7241, "step": 5890 }, { "epoch": 0.7869552169128681, "grad_norm": 4.874888492189262, "learning_rate": 3.2985872289243466e-06, "loss": 0.6394, "step": 5900 }, { "epoch": 0.7882890393144153, "grad_norm": 2.987557037958223, "learning_rate": 3.2589675272683855e-06, "loss": 0.6662, "step": 5910 }, { "epoch": 0.7896228617159625, "grad_norm": 5.926328730580206, "learning_rate": 3.2195581857693595e-06, "loss": 0.6915, "step": 5920 }, { "epoch": 0.7909566841175097, "grad_norm": 2.907571799653725, "learning_rate": 3.180359910511275e-06, "loss": 0.6684, "step": 5930 }, { "epoch": 0.792290506519057, "grad_norm": 8.236684978263545, "learning_rate": 3.1413734037965386e-06, "loss": 0.6998, "step": 5940 }, { "epoch": 0.7936243289206042, "grad_norm": 3.2966043804161584, "learning_rate": 3.102599364133366e-06, "loss": 0.7251, "step": 5950 }, { "epoch": 0.7949581513221514, "grad_norm": 2.1556446497268507, "learning_rate": 3.0640384862232756e-06, "loss": 0.6811, "step": 5960 }, { "epoch": 0.7962919737236986, "grad_norm": 3.636606511761579, "learning_rate": 3.0256914609486367e-06, "loss": 0.6736, "step": 5970 }, { "epoch": 0.797625796125246, "grad_norm": 1.5667682334090924, "learning_rate": 2.9875589753602926e-06, "loss": 0.7156, "step": 5980 }, { "epoch": 0.7989596185267932, "grad_norm": 2.0397471013105246, "learning_rate": 2.9496417126652476e-06, "loss": 0.6691, "step": 5990 }, { "epoch": 0.8002934409283404, "grad_norm": 16.740171367173748, "learning_rate": 2.911940352214437e-06, "loss": 0.6677, "step": 6000 }, { "epoch": 0.8016272633298877, "grad_norm": 131.63779763284177, "learning_rate": 2.874455569490535e-06, "loss": 0.7027, "step": 6010 }, { "epoch": 0.8029610857314349, "grad_norm": 2.5173921242626873, "learning_rate": 2.8371880360958764e-06, "loss": 0.6399, "step": 6020 }, { "epoch": 0.8042949081329821, "grad_norm": 10.518961542895665, "learning_rate": 2.800138419740408e-06, "loss": 0.6518, "step": 6030 }, { "epoch": 0.8056287305345293, "grad_norm": 3.3475848872480505, "learning_rate": 2.76330738422973e-06, "loss": 0.6879, "step": 6040 }, { "epoch": 0.8069625529360765, "grad_norm": 2.1730977739719166, "learning_rate": 2.7266955894532046e-06, "loss": 0.6781, "step": 6050 }, { "epoch": 0.8082963753376238, "grad_norm": 2.072678851059398, "learning_rate": 2.6903036913721285e-06, "loss": 0.7262, "step": 6060 }, { "epoch": 0.809630197739171, "grad_norm": 1.6941591901687694, "learning_rate": 2.6541323420079832e-06, "loss": 0.6686, "step": 6070 }, { "epoch": 0.8109640201407182, "grad_norm": 4.28150571630244, "learning_rate": 2.6181821894307534e-06, "loss": 0.7151, "step": 6080 }, { "epoch": 0.8122978425422654, "grad_norm": 5.224988669059186, "learning_rate": 2.582453877747313e-06, "loss": 0.6516, "step": 6090 }, { "epoch": 0.8136316649438128, "grad_norm": 2.736915062547764, "learning_rate": 2.546948047089889e-06, "loss": 0.692, "step": 6100 }, { "epoch": 0.81496548734536, "grad_norm": 3.209445281777465, "learning_rate": 2.5116653336045905e-06, "loss": 0.6929, "step": 6110 }, { "epoch": 0.8162993097469072, "grad_norm": 18.35451398413344, "learning_rate": 2.4766063694400064e-06, "loss": 0.6409, "step": 6120 }, { "epoch": 0.8176331321484545, "grad_norm": 2.7769421422561162, "learning_rate": 2.4417717827358895e-06, "loss": 0.6802, "step": 6130 }, { "epoch": 0.8189669545500017, "grad_norm": 5.202682155516338, "learning_rate": 2.4071621976118928e-06, "loss": 0.7161, "step": 6140 }, { "epoch": 0.8203007769515489, "grad_norm": 1.8785422878709268, "learning_rate": 2.3727782341563915e-06, "loss": 0.6637, "step": 6150 }, { "epoch": 0.8216345993530961, "grad_norm": 4.160045067102897, "learning_rate": 2.3386205084153754e-06, "loss": 0.6853, "step": 6160 }, { "epoch": 0.8229684217546434, "grad_norm": 11.6299591971655, "learning_rate": 2.304689632381407e-06, "loss": 0.6898, "step": 6170 }, { "epoch": 0.8243022441561906, "grad_norm": 2.19668666417576, "learning_rate": 2.2709862139826554e-06, "loss": 0.681, "step": 6180 }, { "epoch": 0.8256360665577378, "grad_norm": 9.10806081724238, "learning_rate": 2.237510857072013e-06, "loss": 0.6929, "step": 6190 }, { "epoch": 0.826969888959285, "grad_norm": 2.0779769959484438, "learning_rate": 2.204264161416265e-06, "loss": 0.6801, "step": 6200 }, { "epoch": 0.8283037113608323, "grad_norm": 5.811721585561884, "learning_rate": 2.171246722685354e-06, "loss": 0.6646, "step": 6210 }, { "epoch": 0.8296375337623796, "grad_norm": 2.5237647632697993, "learning_rate": 2.1384591324417e-06, "loss": 0.66, "step": 6220 }, { "epoch": 0.8309713561639268, "grad_norm": 2.5679386388987986, "learning_rate": 2.1059019781296073e-06, "loss": 0.6674, "step": 6230 }, { "epoch": 0.832305178565474, "grad_norm": 1.9554398513591067, "learning_rate": 2.0735758430647316e-06, "loss": 0.6719, "step": 6240 }, { "epoch": 0.8336390009670213, "grad_norm": 5.185838548176553, "learning_rate": 2.041481306423638e-06, "loss": 0.708, "step": 6250 }, { "epoch": 0.8349728233685685, "grad_norm": 3.1308842001803523, "learning_rate": 2.0096189432334194e-06, "loss": 0.6997, "step": 6260 }, { "epoch": 0.8363066457701157, "grad_norm": 2.3929519797200474, "learning_rate": 1.977989324361394e-06, "loss": 0.706, "step": 6270 }, { "epoch": 0.8376404681716629, "grad_norm": 1.4916686908701087, "learning_rate": 1.946593016504877e-06, "loss": 0.6761, "step": 6280 }, { "epoch": 0.8389742905732102, "grad_norm": 4.139454571827269, "learning_rate": 1.915430582181031e-06, "loss": 0.6654, "step": 6290 }, { "epoch": 0.8403081129747574, "grad_norm": 4.3252701691493, "learning_rate": 1.8845025797167792e-06, "loss": 0.6656, "step": 6300 }, { "epoch": 0.8416419353763046, "grad_norm": 2.512715902588831, "learning_rate": 1.8538095632388135e-06, "loss": 0.7088, "step": 6310 }, { "epoch": 0.8429757577778518, "grad_norm": 2.55125098116761, "learning_rate": 1.8233520826636563e-06, "loss": 0.6996, "step": 6320 }, { "epoch": 0.8443095801793992, "grad_norm": 9.147529276811154, "learning_rate": 1.7931306836878154e-06, "loss": 0.6894, "step": 6330 }, { "epoch": 0.8456434025809464, "grad_norm": 1.448576833860535, "learning_rate": 1.763145907777997e-06, "loss": 0.687, "step": 6340 }, { "epoch": 0.8469772249824936, "grad_norm": 2.025078656588097, "learning_rate": 1.7333982921614194e-06, "loss": 0.7409, "step": 6350 }, { "epoch": 0.8483110473840408, "grad_norm": 2.0074920992381062, "learning_rate": 1.703888369816174e-06, "loss": 0.6691, "step": 6360 }, { "epoch": 0.8496448697855881, "grad_norm": 2.1486420135259428, "learning_rate": 1.6746166694616821e-06, "loss": 0.729, "step": 6370 }, { "epoch": 0.8509786921871353, "grad_norm": 2.787722139958403, "learning_rate": 1.6455837155492198e-06, "loss": 0.6569, "step": 6380 }, { "epoch": 0.8523125145886825, "grad_norm": 5.294352693264194, "learning_rate": 1.616790028252526e-06, "loss": 0.6792, "step": 6390 }, { "epoch": 0.8536463369902297, "grad_norm": 3.2433926835727918, "learning_rate": 1.5882361234584758e-06, "loss": 0.6569, "step": 6400 }, { "epoch": 0.854980159391777, "grad_norm": 3.4990155334127864, "learning_rate": 1.559922512757847e-06, "loss": 0.699, "step": 6410 }, { "epoch": 0.8563139817933242, "grad_norm": 3.4556853179717772, "learning_rate": 1.5318497034361435e-06, "loss": 0.6851, "step": 6420 }, { "epoch": 0.8576478041948714, "grad_norm": 5.153635922949773, "learning_rate": 1.5040181984645112e-06, "loss": 0.6503, "step": 6430 }, { "epoch": 0.8589816265964186, "grad_norm": 2.0203574757537304, "learning_rate": 1.4764284964907287e-06, "loss": 0.6641, "step": 6440 }, { "epoch": 0.860315448997966, "grad_norm": 2.3763023620326607, "learning_rate": 1.449081091830271e-06, "loss": 0.7082, "step": 6450 }, { "epoch": 0.8616492713995132, "grad_norm": 3.3414602550452943, "learning_rate": 1.421976474457456e-06, "loss": 0.6918, "step": 6460 }, { "epoch": 0.8629830938010604, "grad_norm": 3.3400330606982944, "learning_rate": 1.3951151299966526e-06, "loss": 0.7078, "step": 6470 }, { "epoch": 0.8643169162026076, "grad_norm": 3.0397063605389225, "learning_rate": 1.3684975397135996e-06, "loss": 0.6943, "step": 6480 }, { "epoch": 0.8656507386041549, "grad_norm": 2.355039946587831, "learning_rate": 1.3421241805067714e-06, "loss": 0.6257, "step": 6490 }, { "epoch": 0.8669845610057021, "grad_norm": 4.221674830663121, "learning_rate": 1.3159955248988354e-06, "loss": 0.7075, "step": 6500 }, { "epoch": 0.8683183834072493, "grad_norm": 4.367585575202146, "learning_rate": 1.2901120410281864e-06, "loss": 0.7036, "step": 6510 }, { "epoch": 0.8696522058087965, "grad_norm": 3.214721825699072, "learning_rate": 1.2644741926405595e-06, "loss": 0.6788, "step": 6520 }, { "epoch": 0.8709860282103438, "grad_norm": 2.1651228637720012, "learning_rate": 1.2390824390807204e-06, "loss": 0.6985, "step": 6530 }, { "epoch": 0.872319850611891, "grad_norm": 3.9985415660915318, "learning_rate": 1.2139372352842331e-06, "loss": 0.6495, "step": 6540 }, { "epoch": 0.8736536730134382, "grad_norm": 1.3249366637153097, "learning_rate": 1.189039031769319e-06, "loss": 0.6893, "step": 6550 }, { "epoch": 0.8749874954149855, "grad_norm": 2.6059489343870137, "learning_rate": 1.1643882746287738e-06, "loss": 0.6938, "step": 6560 }, { "epoch": 0.8763213178165328, "grad_norm": 2.834432912148686, "learning_rate": 1.139985405521975e-06, "loss": 0.6619, "step": 6570 }, { "epoch": 0.87765514021808, "grad_norm": 2.4479368709514997, "learning_rate": 1.1158308616669777e-06, "loss": 0.6799, "step": 6580 }, { "epoch": 0.8789889626196272, "grad_norm": 2.375391877620538, "learning_rate": 1.0919250758326777e-06, "loss": 0.6649, "step": 6590 }, { "epoch": 0.8803227850211744, "grad_norm": 2.011168410197422, "learning_rate": 1.0682684763310541e-06, "loss": 0.7059, "step": 6600 }, { "epoch": 0.8816566074227217, "grad_norm": 6.9489058477193915, "learning_rate": 1.0448614870094975e-06, "loss": 0.7003, "step": 6610 }, { "epoch": 0.8829904298242689, "grad_norm": 2.89629768289524, "learning_rate": 1.0217045272432174e-06, "loss": 0.6807, "step": 6620 }, { "epoch": 0.8843242522258161, "grad_norm": 3.1365127590849733, "learning_rate": 9.98798011927725e-07, "loss": 0.6501, "step": 6630 }, { "epoch": 0.8856580746273633, "grad_norm": 2.478817175158367, "learning_rate": 9.76142351471408e-07, "loss": 0.6748, "step": 6640 }, { "epoch": 0.8869918970289106, "grad_norm": 4.623604022524473, "learning_rate": 9.537379517881633e-07, "loss": 0.7036, "step": 6650 }, { "epoch": 0.8883257194304578, "grad_norm": 2.322496920810461, "learning_rate": 9.315852142901366e-07, "loss": 0.7309, "step": 6660 }, { "epoch": 0.889659541832005, "grad_norm": 3.4181763329821613, "learning_rate": 9.096845358805278e-07, "loss": 0.6462, "step": 6670 }, { "epoch": 0.8909933642335524, "grad_norm": 2.815040204182841, "learning_rate": 8.880363089464749e-07, "loss": 0.6749, "step": 6680 }, { "epoch": 0.8923271866350996, "grad_norm": 2.205505383761902, "learning_rate": 8.666409213520305e-07, "loss": 0.666, "step": 6690 }, { "epoch": 0.8936610090366468, "grad_norm": 5.542301323288899, "learning_rate": 8.454987564312039e-07, "loss": 0.6528, "step": 6700 }, { "epoch": 0.894994831438194, "grad_norm": 5.18279779413142, "learning_rate": 8.246101929811056e-07, "loss": 0.6471, "step": 6710 }, { "epoch": 0.8963286538397413, "grad_norm": 2.5900897201256914, "learning_rate": 8.039756052551473e-07, "loss": 0.6738, "step": 6720 }, { "epoch": 0.8976624762412885, "grad_norm": 4.210976680085005, "learning_rate": 7.835953629563453e-07, "loss": 0.7009, "step": 6730 }, { "epoch": 0.8989962986428357, "grad_norm": 1.7338703555444437, "learning_rate": 7.634698312306915e-07, "loss": 0.6956, "step": 6740 }, { "epoch": 0.9003301210443829, "grad_norm": 1.5255307415290653, "learning_rate": 7.435993706606197e-07, "loss": 0.6423, "step": 6750 }, { "epoch": 0.9016639434459301, "grad_norm": 11.474970905208894, "learning_rate": 7.239843372585297e-07, "loss": 0.699, "step": 6760 }, { "epoch": 0.9029977658474774, "grad_norm": 2.8700032141797083, "learning_rate": 7.04625082460425e-07, "loss": 0.6863, "step": 6770 }, { "epoch": 0.9043315882490246, "grad_norm": 2.145342230538711, "learning_rate": 6.855219531196083e-07, "loss": 0.6456, "step": 6780 }, { "epoch": 0.9056654106505718, "grad_norm": 7.88073666842204, "learning_rate": 6.666752915004648e-07, "loss": 0.6738, "step": 6790 }, { "epoch": 0.9069992330521192, "grad_norm": 2.41261114546358, "learning_rate": 6.48085435272337e-07, "loss": 0.6573, "step": 6800 } ], "logging_steps": 10, "max_steps": 7497, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 400, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.8588280013665075e+19, "train_batch_size": 4, "trial_name": null, "trial_params": null }