{ "best_global_step": 2028, "best_metric": 0.12891627848148346, "best_model_checkpoint": "/mnt/d/ai/training/openbiollm-phi/outputs/final-adapter/checkpoint-2028", "epoch": 2.0, "eval_steps": 500, "global_step": 2028, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "entropy": 0.8023399017751217, "epoch": 0.009865581452706869, "grad_norm": 4.375, "learning_rate": 1.8e-05, "loss": 1.1496441841125489, "mean_token_accuracy": 0.785130063444376, "num_tokens": 28284.0, "step": 10 }, { "entropy": 0.5005267957225442, "epoch": 0.019731162905413737, "grad_norm": 0.7890625, "learning_rate": 3.8e-05, "loss": 0.4015222072601318, "mean_token_accuracy": 0.931036177277565, "num_tokens": 57055.0, "step": 20 }, { "entropy": 0.20316241607069968, "epoch": 0.029596744358120607, "grad_norm": 1.046875, "learning_rate": 5.8e-05, "loss": 0.19807834625244142, "mean_token_accuracy": 0.968285684287548, "num_tokens": 85550.0, "step": 30 }, { "entropy": 0.17202282957732679, "epoch": 0.039462325810827474, "grad_norm": 0.51953125, "learning_rate": 7.800000000000001e-05, "loss": 0.15795693397521973, "mean_token_accuracy": 0.9717290192842484, "num_tokens": 113950.0, "step": 40 }, { "entropy": 0.158708918094635, "epoch": 0.049327907263534344, "grad_norm": 0.19921875, "learning_rate": 9.8e-05, "loss": 0.14324311017990113, "mean_token_accuracy": 0.9745404750108719, "num_tokens": 142366.0, "step": 50 }, { "entropy": 0.14976419135928154, "epoch": 0.059193488716241215, "grad_norm": 0.0986328125, "learning_rate": 0.000118, "loss": 0.1319120168685913, "mean_token_accuracy": 0.9765947729349136, "num_tokens": 170629.0, "step": 60 }, { "entropy": 0.14688248625025152, "epoch": 0.06905907016894808, "grad_norm": 0.306640625, "learning_rate": 0.000138, "loss": 0.13508867025375365, "mean_token_accuracy": 0.9753234691917896, "num_tokens": 198826.0, "step": 70 }, { "entropy": 0.13914804700762035, "epoch": 0.07892465162165495, "grad_norm": 0.126953125, "learning_rate": 0.00015800000000000002, "loss": 0.1278115391731262, "mean_token_accuracy": 0.9766607381403446, "num_tokens": 227301.0, "step": 80 }, { "entropy": 0.14371367944404484, "epoch": 0.08879023307436182, "grad_norm": 0.35546875, "learning_rate": 0.00017800000000000002, "loss": 0.1317190170288086, "mean_token_accuracy": 0.9761511102318764, "num_tokens": 255497.0, "step": 90 }, { "entropy": 0.13979870928451418, "epoch": 0.09865581452706869, "grad_norm": 0.22265625, "learning_rate": 0.00019800000000000002, "loss": 0.13048830032348632, "mean_token_accuracy": 0.9765857473015785, "num_tokens": 283740.0, "step": 100 }, { "entropy": 0.14078679895028473, "epoch": 0.10852139597977556, "grad_norm": 0.044677734375, "learning_rate": 0.00019998924691878362, "loss": 0.12818093299865724, "mean_token_accuracy": 0.9770116224884987, "num_tokens": 311969.0, "step": 110 }, { "entropy": 0.13612987883388997, "epoch": 0.11838697743248243, "grad_norm": 0.146484375, "learning_rate": 0.0001999520787428256, "loss": 0.12426202297210694, "mean_token_accuracy": 0.9769940137863159, "num_tokens": 340527.0, "step": 120 }, { "entropy": 0.13964395532384516, "epoch": 0.1282525588851893, "grad_norm": 0.1220703125, "learning_rate": 0.0001998883725841002, "loss": 0.1288180708885193, "mean_token_accuracy": 0.9765414282679558, "num_tokens": 368794.0, "step": 130 }, { "entropy": 0.13610617620870472, "epoch": 0.13811814033789616, "grad_norm": 0.11083984375, "learning_rate": 0.00019979814535704243, "loss": 0.12604674100875854, "mean_token_accuracy": 0.976799800992012, "num_tokens": 397372.0, "step": 140 }, { "entropy": 0.1376406972296536, "epoch": 0.14798372179060304, "grad_norm": 0.0830078125, "learning_rate": 0.00019968142101761826, "loss": 0.1287875294685364, "mean_token_accuracy": 0.9766498073935509, "num_tokens": 425822.0, "step": 150 }, { "entropy": 0.13571005184203386, "epoch": 0.1578493032433099, "grad_norm": 0.125, "learning_rate": 0.00019953823055696412, "loss": 0.12613801956176757, "mean_token_accuracy": 0.9775339424610138, "num_tokens": 454265.0, "step": 160 }, { "entropy": 0.1362694211304188, "epoch": 0.16771488469601678, "grad_norm": 0.06298828125, "learning_rate": 0.0001993686119931585, "loss": 0.12453360557556152, "mean_token_accuracy": 0.9771473035216331, "num_tokens": 482868.0, "step": 170 }, { "entropy": 0.14204563926905395, "epoch": 0.17758046614872364, "grad_norm": 0.07080078125, "learning_rate": 0.000199172610361128, "loss": 0.12911393642425537, "mean_token_accuracy": 0.97633573487401, "num_tokens": 511076.0, "step": 180 }, { "entropy": 0.1398276717402041, "epoch": 0.18744604760143052, "grad_norm": 0.1220703125, "learning_rate": 0.00019895027770069, "loss": 0.12636626958847047, "mean_token_accuracy": 0.9771449662744999, "num_tokens": 539269.0, "step": 190 }, { "entropy": 0.1380731706507504, "epoch": 0.19731162905413738, "grad_norm": 0.0380859375, "learning_rate": 0.00019870167304273607, "loss": 0.1280285596847534, "mean_token_accuracy": 0.977227246016264, "num_tokens": 567464.0, "step": 200 }, { "entropy": 0.13736872393637894, "epoch": 0.20717721050684423, "grad_norm": 0.003021240234375, "learning_rate": 0.00019842686239355842, "loss": 0.12631702423095703, "mean_token_accuracy": 0.9771020725369454, "num_tokens": 595781.0, "step": 210 }, { "entropy": 0.12889715228229762, "epoch": 0.21704279195955112, "grad_norm": 0.1162109375, "learning_rate": 0.00019812591871732517, "loss": 0.1197730541229248, "mean_token_accuracy": 0.9784359365701676, "num_tokens": 624587.0, "step": 220 }, { "entropy": 0.13859685016795992, "epoch": 0.22690837341225797, "grad_norm": 0.1328125, "learning_rate": 0.00019779892191670767, "loss": 0.12636926174163818, "mean_token_accuracy": 0.9770418994128705, "num_tokens": 653000.0, "step": 230 }, { "entropy": 0.13383242171257734, "epoch": 0.23677395486496486, "grad_norm": 0.041015625, "learning_rate": 0.00019744595881166562, "loss": 0.12312976121902466, "mean_token_accuracy": 0.9780343927443027, "num_tokens": 681615.0, "step": 240 }, { "entropy": 0.14151692641898989, "epoch": 0.24663953631767171, "grad_norm": 0.08203125, "learning_rate": 0.00019706712311639616, "loss": 0.12832958698272706, "mean_token_accuracy": 0.9766872271895408, "num_tokens": 709713.0, "step": 250 }, { "entropy": 0.14054928198456765, "epoch": 0.2565051177703786, "grad_norm": 0.1279296875, "learning_rate": 0.00019666251541445163, "loss": 0.12945976257324218, "mean_token_accuracy": 0.9767507568001748, "num_tokens": 737780.0, "step": 260 }, { "entropy": 0.1373858324252069, "epoch": 0.2663706992230855, "grad_norm": 0.0272216796875, "learning_rate": 0.00019623224313203432, "loss": 0.12566013336181642, "mean_token_accuracy": 0.9775523208081722, "num_tokens": 766198.0, "step": 270 }, { "entropy": 0.13666383242234587, "epoch": 0.2762362806757923, "grad_norm": 0.1455078125, "learning_rate": 0.0001957764205094737, "loss": 0.12408857345581055, "mean_token_accuracy": 0.9773378528654575, "num_tokens": 794690.0, "step": 280 }, { "entropy": 0.13422702737152575, "epoch": 0.2861018621284992, "grad_norm": 0.0291748046875, "learning_rate": 0.00019529516857089512, "loss": 0.12166121006011962, "mean_token_accuracy": 0.9774681590497494, "num_tokens": 823273.0, "step": 290 }, { "entropy": 0.1338389404118061, "epoch": 0.2959674435812061, "grad_norm": 0.2001953125, "learning_rate": 0.00019478861509208688, "loss": 0.1265432357788086, "mean_token_accuracy": 0.9771662212908268, "num_tokens": 851790.0, "step": 300 }, { "entropy": 0.13281645821407437, "epoch": 0.3058330250339129, "grad_norm": 0.0281982421875, "learning_rate": 0.0001942568945665749, "loss": 0.12200206518173218, "mean_token_accuracy": 0.9776300400495529, "num_tokens": 880577.0, "step": 310 }, { "entropy": 0.13494051322340966, "epoch": 0.3156986064866198, "grad_norm": 0.146484375, "learning_rate": 0.0001937001481699136, "loss": 0.12529757022857665, "mean_token_accuracy": 0.9775164686143398, "num_tokens": 909005.0, "step": 320 }, { "entropy": 0.14295281181111932, "epoch": 0.3255641879393267, "grad_norm": 0.017822265625, "learning_rate": 0.0001931185237222032, "loss": 0.12978450059890748, "mean_token_accuracy": 0.9768691435456276, "num_tokens": 937077.0, "step": 330 }, { "entropy": 0.13115282114595175, "epoch": 0.33542976939203356, "grad_norm": 0.0185546875, "learning_rate": 0.00019251217564884187, "loss": 0.11987617015838622, "mean_token_accuracy": 0.9781784944236278, "num_tokens": 965862.0, "step": 340 }, { "entropy": 0.13873312631621956, "epoch": 0.3452953508447404, "grad_norm": 0.005279541015625, "learning_rate": 0.00019188126493952518, "loss": 0.12614299058914186, "mean_token_accuracy": 0.9771831274032593, "num_tokens": 994161.0, "step": 350 }, { "entropy": 0.13355796597898006, "epoch": 0.3551609322974473, "grad_norm": 0.37890625, "learning_rate": 0.00019122595910550191, "loss": 0.12263578176498413, "mean_token_accuracy": 0.9777961499989033, "num_tokens": 1022807.0, "step": 360 }, { "entropy": 0.13650657385587692, "epoch": 0.36502651375015416, "grad_norm": 0.0247802734375, "learning_rate": 0.0001905464321350988, "loss": 0.12675031423568725, "mean_token_accuracy": 0.9774383820593358, "num_tokens": 1051131.0, "step": 370 }, { "entropy": 0.13800616860389708, "epoch": 0.37489209520286104, "grad_norm": 0.0196533203125, "learning_rate": 0.00018984286444752524, "loss": 0.12497754096984863, "mean_token_accuracy": 0.9774119921028614, "num_tokens": 1079501.0, "step": 380 }, { "entropy": 0.13551138881593944, "epoch": 0.38475767665556787, "grad_norm": 0.051513671875, "learning_rate": 0.00018911544284497093, "loss": 0.12369418144226074, "mean_token_accuracy": 0.9777314260601997, "num_tokens": 1107999.0, "step": 390 }, { "entropy": 0.1347731866873801, "epoch": 0.39462325810827475, "grad_norm": 0.01226806640625, "learning_rate": 0.0001883643604630084, "loss": 0.12635958194732666, "mean_token_accuracy": 0.9772857747972011, "num_tokens": 1136346.0, "step": 400 }, { "entropy": 0.1420691579580307, "epoch": 0.40448883956098164, "grad_norm": 0.007598876953125, "learning_rate": 0.00018758981671931416, "loss": 0.1292852520942688, "mean_token_accuracy": 0.9766104318201542, "num_tokens": 1164402.0, "step": 410 }, { "entropy": 0.1375271413475275, "epoch": 0.41435442101368847, "grad_norm": 0.010009765625, "learning_rate": 0.00018679201726072204, "loss": 0.12546764612197875, "mean_token_accuracy": 0.9771398045122623, "num_tokens": 1192710.0, "step": 420 }, { "entropy": 0.1342703066766262, "epoch": 0.42422000246639535, "grad_norm": 0.0191650390625, "learning_rate": 0.00018597117390862228, "loss": 0.12357202768325806, "mean_token_accuracy": 0.9779355488717556, "num_tokens": 1221103.0, "step": 430 }, { "entropy": 0.13978215726092458, "epoch": 0.43408558391910224, "grad_norm": 0.10498046875, "learning_rate": 0.0001851275046027217, "loss": 0.13000580072402954, "mean_token_accuracy": 0.9766978964209556, "num_tokens": 1249258.0, "step": 440 }, { "entropy": 0.13839555103331805, "epoch": 0.4439511653718091, "grad_norm": 0.029296875, "learning_rate": 0.00018426123334317905, "loss": 0.12596620321273805, "mean_token_accuracy": 0.9773071080446243, "num_tokens": 1277491.0, "step": 450 }, { "entropy": 0.13596669165417552, "epoch": 0.45381674682451595, "grad_norm": 0.0150146484375, "learning_rate": 0.00018337259013113125, "loss": 0.12328847646713256, "mean_token_accuracy": 0.9780720941722393, "num_tokens": 1305867.0, "step": 460 }, { "entropy": 0.13245578771457076, "epoch": 0.46368232827722283, "grad_norm": 0.091796875, "learning_rate": 0.00018246181090762677, "loss": 0.12181999683380126, "mean_token_accuracy": 0.9777809880673886, "num_tokens": 1334412.0, "step": 470 }, { "entropy": 0.13981428062543272, "epoch": 0.4735479097299297, "grad_norm": 0.0262451171875, "learning_rate": 0.00018152913749098132, "loss": 0.1279812455177307, "mean_token_accuracy": 0.9768534265458584, "num_tokens": 1362565.0, "step": 480 }, { "entropy": 0.13838341003283858, "epoch": 0.4834134911826366, "grad_norm": 0.10546875, "learning_rate": 0.00018057481751257356, "loss": 0.12548565864562988, "mean_token_accuracy": 0.9772889137268066, "num_tokens": 1390996.0, "step": 490 }, { "entropy": 0.13627033829689025, "epoch": 0.49327907263534343, "grad_norm": 0.05908203125, "learning_rate": 0.00017959910435109717, "loss": 0.12579237222671508, "mean_token_accuracy": 0.9775010101497174, "num_tokens": 1419383.0, "step": 500 }, { "entropy": 0.12956608394160868, "epoch": 0.5031446540880503, "grad_norm": 0.052734375, "learning_rate": 0.00017860225706528703, "loss": 0.11810436248779296, "mean_token_accuracy": 0.9790474943816662, "num_tokens": 1448074.0, "step": 510 }, { "entropy": 0.134104797616601, "epoch": 0.5130102355407572, "grad_norm": 0.00762939453125, "learning_rate": 0.00017758454032513723, "loss": 0.12165459394454955, "mean_token_accuracy": 0.9784664116799832, "num_tokens": 1476462.0, "step": 520 }, { "entropy": 0.13819981999695302, "epoch": 0.5228758169934641, "grad_norm": 0.05859375, "learning_rate": 0.0001765462243416294, "loss": 0.12643526792526244, "mean_token_accuracy": 0.9772073484957218, "num_tokens": 1504622.0, "step": 530 }, { "entropy": 0.131108792219311, "epoch": 0.532741398446171, "grad_norm": 0.04736328125, "learning_rate": 0.00017548758479498972, "loss": 0.12083779573440552, "mean_token_accuracy": 0.9781857557594776, "num_tokens": 1533231.0, "step": 540 }, { "entropy": 0.13788466211408376, "epoch": 0.5426069798988777, "grad_norm": 0.0072021484375, "learning_rate": 0.00017440890276149395, "loss": 0.1245079755783081, "mean_token_accuracy": 0.9772788502275944, "num_tokens": 1561554.0, "step": 550 }, { "entropy": 0.13741337759420275, "epoch": 0.5524725613515846, "grad_norm": 0.00628662109375, "learning_rate": 0.00017331046463883967, "loss": 0.1247126579284668, "mean_token_accuracy": 0.9774496532976628, "num_tokens": 1589793.0, "step": 560 }, { "entropy": 0.13634313521906732, "epoch": 0.5623381428042915, "grad_norm": 0.150390625, "learning_rate": 0.0001721925620701055, "loss": 0.12345428466796875, "mean_token_accuracy": 0.9776806205511093, "num_tokens": 1618206.0, "step": 570 }, { "entropy": 0.14105848912149668, "epoch": 0.5722037242569984, "grad_norm": 0.07421875, "learning_rate": 0.00017105549186631822, "loss": 0.12956223487854004, "mean_token_accuracy": 0.9768215775489807, "num_tokens": 1646278.0, "step": 580 }, { "entropy": 0.13872844530269504, "epoch": 0.5820693057097053, "grad_norm": 0.1240234375, "learning_rate": 0.0001698995559276468, "loss": 0.12983914613723754, "mean_token_accuracy": 0.9765517644584178, "num_tokens": 1674561.0, "step": 590 }, { "entropy": 0.14561426928266882, "epoch": 0.5919348871624122, "grad_norm": 0.0074462890625, "learning_rate": 0.00016872506116324617, "loss": 0.13211548328399658, "mean_token_accuracy": 0.9762530513107777, "num_tokens": 1702452.0, "step": 600 }, { "entropy": 0.13719567554071546, "epoch": 0.601800468615119, "grad_norm": 0.02197265625, "learning_rate": 0.0001675323194097706, "loss": 0.12331978082656861, "mean_token_accuracy": 0.9778655581176281, "num_tokens": 1730903.0, "step": 610 }, { "entropy": 0.13551129242405296, "epoch": 0.6116660500678258, "grad_norm": 0.06103515625, "learning_rate": 0.00016632164734857883, "loss": 0.12384191751480103, "mean_token_accuracy": 0.9774068400263787, "num_tokens": 1759380.0, "step": 620 }, { "entropy": 0.1343284510076046, "epoch": 0.6215316315205327, "grad_norm": 0.02978515625, "learning_rate": 0.00016509336642165273, "loss": 0.12308658361434936, "mean_token_accuracy": 0.9781653232872486, "num_tokens": 1787756.0, "step": 630 }, { "entropy": 0.1336012206040323, "epoch": 0.6313972129732396, "grad_norm": 0.00323486328125, "learning_rate": 0.00016384780274625264, "loss": 0.12076139450073242, "mean_token_accuracy": 0.9782443180680275, "num_tokens": 1816167.0, "step": 640 }, { "entropy": 0.13287872709333898, "epoch": 0.6412627944259465, "grad_norm": 0.0228271484375, "learning_rate": 0.00016258528702833047, "loss": 0.12263975143432618, "mean_token_accuracy": 0.9782126508653164, "num_tokens": 1844651.0, "step": 650 }, { "entropy": 0.13714895490556955, "epoch": 0.6511283758786534, "grad_norm": 0.04443359375, "learning_rate": 0.00016130615447472522, "loss": 0.12450298070907592, "mean_token_accuracy": 0.9775685340166091, "num_tokens": 1872913.0, "step": 660 }, { "entropy": 0.13559102937579154, "epoch": 0.6609939573313602, "grad_norm": 0.0869140625, "learning_rate": 0.00016001074470416314, "loss": 0.12432575225830078, "mean_token_accuracy": 0.9774284318089486, "num_tokens": 1901325.0, "step": 670 }, { "entropy": 0.14654233315959572, "epoch": 0.6708595387840671, "grad_norm": 0.05908203125, "learning_rate": 0.0001586994016570865, "loss": 0.133248770236969, "mean_token_accuracy": 0.9761250920593738, "num_tokens": 1929156.0, "step": 680 }, { "entropy": 0.136425125785172, "epoch": 0.680725120236774, "grad_norm": 0.0035552978515625, "learning_rate": 0.0001573724735043349, "loss": 0.12281620502471924, "mean_token_accuracy": 0.9778434507548809, "num_tokens": 1957527.0, "step": 690 }, { "entropy": 0.13412187108770013, "epoch": 0.6905907016894808, "grad_norm": 0.009033203125, "learning_rate": 0.00015603031255470375, "loss": 0.12007204294204712, "mean_token_accuracy": 0.9783773384988308, "num_tokens": 1986049.0, "step": 700 }, { "entropy": 0.13671441841870546, "epoch": 0.7004562831421877, "grad_norm": 0.0057373046875, "learning_rate": 0.00015467327516140346, "loss": 0.12455042600631713, "mean_token_accuracy": 0.9777524806559086, "num_tokens": 2014375.0, "step": 710 }, { "entropy": 0.13356605730950832, "epoch": 0.7103218645948945, "grad_norm": 0.01092529296875, "learning_rate": 0.0001533017216274454, "loss": 0.12249385118484497, "mean_token_accuracy": 0.9779535219073295, "num_tokens": 2042809.0, "step": 720 }, { "entropy": 0.13847268400713802, "epoch": 0.7201874460476014, "grad_norm": 0.103515625, "learning_rate": 0.0001519160161099787, "loss": 0.1281079888343811, "mean_token_accuracy": 0.9772497572004795, "num_tokens": 2070893.0, "step": 730 }, { "entropy": 0.1359175430610776, "epoch": 0.7300530275003083, "grad_norm": 1.15625, "learning_rate": 0.0001505165265236041, "loss": 0.12455817461013793, "mean_token_accuracy": 0.977594693005085, "num_tokens": 2099191.0, "step": 740 }, { "entropy": 0.13774135578423738, "epoch": 0.7399186089530152, "grad_norm": 0.01409912109375, "learning_rate": 0.00014910362444268998, "loss": 0.12687115669250487, "mean_token_accuracy": 0.9774483822286129, "num_tokens": 2127334.0, "step": 750 }, { "entropy": 0.13371777413412927, "epoch": 0.7497841904057221, "grad_norm": 0.20703125, "learning_rate": 0.00014767768500271672, "loss": 0.12280063629150391, "mean_token_accuracy": 0.9777301244437695, "num_tokens": 2155825.0, "step": 760 }, { "entropy": 0.14284388041123747, "epoch": 0.7596497718584289, "grad_norm": 0.039794921875, "learning_rate": 0.0001462390868006756, "loss": 0.13144956827163695, "mean_token_accuracy": 0.9766757681965827, "num_tokens": 2183894.0, "step": 770 }, { "entropy": 0.14144302653148771, "epoch": 0.7695153533111357, "grad_norm": 0.017822265625, "learning_rate": 0.00014478821179454872, "loss": 0.12898389101028443, "mean_token_accuracy": 0.9770457074046135, "num_tokens": 2211972.0, "step": 780 }, { "entropy": 0.14034299813210965, "epoch": 0.7793809347638426, "grad_norm": 0.017578125, "learning_rate": 0.00014332544520189616, "loss": 0.12705788612365723, "mean_token_accuracy": 0.9771609790623188, "num_tokens": 2240230.0, "step": 790 }, { "entropy": 0.13051901962608098, "epoch": 0.7892465162165495, "grad_norm": 0.01116943359375, "learning_rate": 0.0001418511753975784, "loss": 0.11829437017440796, "mean_token_accuracy": 0.9785726971924305, "num_tokens": 2268986.0, "step": 800 }, { "entropy": 0.13454419681802393, "epoch": 0.7991120976692564, "grad_norm": 0.01904296875, "learning_rate": 0.00014036579381063971, "loss": 0.1226949691772461, "mean_token_accuracy": 0.9776648931205273, "num_tokens": 2297484.0, "step": 810 }, { "entropy": 0.1361277119256556, "epoch": 0.8089776791219633, "grad_norm": 0.0294189453125, "learning_rate": 0.00013886969482038126, "loss": 0.1240857720375061, "mean_token_accuracy": 0.9777307607233524, "num_tokens": 2325691.0, "step": 820 }, { "entropy": 0.1346275858581066, "epoch": 0.8188432605746702, "grad_norm": 0.01177978515625, "learning_rate": 0.00013736327565165037, "loss": 0.12228076457977295, "mean_token_accuracy": 0.9780977793037892, "num_tokens": 2354153.0, "step": 830 }, { "entropy": 0.13894635848701, "epoch": 0.8287088420273769, "grad_norm": 0.007476806640625, "learning_rate": 0.00013584693626937465, "loss": 0.12682583332061767, "mean_token_accuracy": 0.9772831581532955, "num_tokens": 2382275.0, "step": 840 }, { "entropy": 0.13055060487240552, "epoch": 0.8385744234800838, "grad_norm": 0.0072021484375, "learning_rate": 0.0001343210792723685, "loss": 0.11919403076171875, "mean_token_accuracy": 0.9786668792366982, "num_tokens": 2410859.0, "step": 850 }, { "entropy": 0.1351913319900632, "epoch": 0.8484400049327907, "grad_norm": 0.05615234375, "learning_rate": 0.00013278610978644025, "loss": 0.12403948307037353, "mean_token_accuracy": 0.9774477735161782, "num_tokens": 2439275.0, "step": 860 }, { "entropy": 0.135931574460119, "epoch": 0.8583055863854976, "grad_norm": 0.00885009765625, "learning_rate": 0.00013124243535682825, "loss": 0.1243247628211975, "mean_token_accuracy": 0.9779405005276203, "num_tokens": 2467530.0, "step": 870 }, { "entropy": 0.13914883006364107, "epoch": 0.8681711678382045, "grad_norm": 0.05712890625, "learning_rate": 0.0001296904658399951, "loss": 0.1256848931312561, "mean_token_accuracy": 0.9774674661457539, "num_tokens": 2495718.0, "step": 880 }, { "entropy": 0.1410721031948924, "epoch": 0.8780367492909114, "grad_norm": 0.00433349609375, "learning_rate": 0.00012813061329480747, "loss": 0.12887948751449585, "mean_token_accuracy": 0.9768504619598388, "num_tokens": 2523765.0, "step": 890 }, { "entropy": 0.1413712895475328, "epoch": 0.8879023307436182, "grad_norm": 0.005523681640625, "learning_rate": 0.0001265632918731318, "loss": 0.12763890027999877, "mean_token_accuracy": 0.9769708946347236, "num_tokens": 2551816.0, "step": 900 }, { "entropy": 0.13617176273837686, "epoch": 0.8977679121963251, "grad_norm": 0.07373046875, "learning_rate": 0.00012498891770987394, "loss": 0.12494395971298218, "mean_token_accuracy": 0.9774423398077488, "num_tokens": 2580036.0, "step": 910 }, { "entropy": 0.13287583040073514, "epoch": 0.9076334936490319, "grad_norm": 0.03564453125, "learning_rate": 0.0001234079088124925, "loss": 0.12190787792205811, "mean_token_accuracy": 0.9780240021646023, "num_tokens": 2608493.0, "step": 920 }, { "entropy": 0.1400419868528843, "epoch": 0.9174990751017388, "grad_norm": 0.01123046875, "learning_rate": 0.00012182068495001475, "loss": 0.12783472537994384, "mean_token_accuracy": 0.9770628847181797, "num_tokens": 2636593.0, "step": 930 }, { "entropy": 0.13885226137936116, "epoch": 0.9273646565544457, "grad_norm": 0.045654296875, "learning_rate": 0.00012022766754158521, "loss": 0.12473338842391968, "mean_token_accuracy": 0.9772628821432591, "num_tokens": 2664857.0, "step": 940 }, { "entropy": 0.1308811165392399, "epoch": 0.9372302380071525, "grad_norm": 0.0037689208984375, "learning_rate": 0.00011862927954457566, "loss": 0.12016142606735229, "mean_token_accuracy": 0.9784671284258366, "num_tokens": 2693546.0, "step": 950 }, { "entropy": 0.13529244558885695, "epoch": 0.9470958194598594, "grad_norm": 0.05908203125, "learning_rate": 0.0001170259453422871, "loss": 0.12330687046051025, "mean_token_accuracy": 0.977809751778841, "num_tokens": 2721832.0, "step": 960 }, { "entropy": 0.12781465742737055, "epoch": 0.9569614009125663, "grad_norm": 0.12890625, "learning_rate": 0.00011541809063127307, "loss": 0.11694331169128418, "mean_token_accuracy": 0.9792409002780914, "num_tokens": 2750552.0, "step": 970 }, { "entropy": 0.1382864972576499, "epoch": 0.9668269823652732, "grad_norm": 0.003021240234375, "learning_rate": 0.0001138061423083139, "loss": 0.12697868347167968, "mean_token_accuracy": 0.9772090330719948, "num_tokens": 2778782.0, "step": 980 }, { "entropy": 0.14215879468247294, "epoch": 0.97669256381798, "grad_norm": 0.12060546875, "learning_rate": 0.00011219052835707293, "loss": 0.12857697010040284, "mean_token_accuracy": 0.9767957553267479, "num_tokens": 2806804.0, "step": 990 }, { "entropy": 0.12928050421178341, "epoch": 0.9865581452706869, "grad_norm": 0.0035552978515625, "learning_rate": 0.00011057167773446373, "loss": 0.11841636896133423, "mean_token_accuracy": 0.9786058440804482, "num_tokens": 2835477.0, "step": 1000 }, { "entropy": 0.13619811739772558, "epoch": 0.9964237267233937, "grad_norm": 0.0084228515625, "learning_rate": 0.0001089500202567589, "loss": 0.12360646724700927, "mean_token_accuracy": 0.9778580211102963, "num_tokens": 2863853.0, "step": 1010 }, { "epoch": 1.0, "eval_entropy": 0.13515008005076612, "eval_loss": 0.12946189939975739, "eval_mean_token_accuracy": 0.9778799934264941, "eval_num_tokens": 2874093.0, "eval_runtime": 101.3679, "eval_samples_per_second": 19.996, "eval_steps_per_second": 10.003, "step": 1014 }, { "entropy": 0.13946586963418242, "epoch": 1.005919348871624, "grad_norm": 0.006988525390625, "learning_rate": 0.00010732598648547083, "loss": 0.12724997997283935, "mean_token_accuracy": 0.9772353125857068, "num_tokens": 2890914.0, "step": 1020 }, { "entropy": 0.1288278905674815, "epoch": 1.015784930324331, "grad_norm": 0.005279541015625, "learning_rate": 0.00010570000761303482, "loss": 0.11608068943023682, "mean_token_accuracy": 0.9791852064430714, "num_tokens": 2919674.0, "step": 1030 }, { "entropy": 0.1328219858929515, "epoch": 1.0256505117770378, "grad_norm": 0.0186767578125, "learning_rate": 0.00010407251534832416, "loss": 0.11962487697601318, "mean_token_accuracy": 0.9785868972539902, "num_tokens": 2948257.0, "step": 1040 }, { "entropy": 0.12928575854748486, "epoch": 1.0355160932297447, "grad_norm": 0.004974365234375, "learning_rate": 0.00010244394180202837, "loss": 0.11730986833572388, "mean_token_accuracy": 0.9789474971592427, "num_tokens": 2977019.0, "step": 1050 }, { "entropy": 0.13721467312425376, "epoch": 1.0453816746824516, "grad_norm": 0.0037841796875, "learning_rate": 0.00010081471937192505, "loss": 0.12385734319686889, "mean_token_accuracy": 0.9777599819004535, "num_tokens": 3005276.0, "step": 1060 }, { "entropy": 0.1354451086372137, "epoch": 1.0552472561351585, "grad_norm": 0.00494384765625, "learning_rate": 9.918528062807496e-05, "loss": 0.12247976064682006, "mean_token_accuracy": 0.9780112229287624, "num_tokens": 3033671.0, "step": 1070 }, { "entropy": 0.1386636977083981, "epoch": 1.0651128375878653, "grad_norm": 0.0184326171875, "learning_rate": 9.755605819797164e-05, "loss": 0.1263806104660034, "mean_token_accuracy": 0.9775540590286255, "num_tokens": 3061794.0, "step": 1080 }, { "entropy": 0.13344219541177155, "epoch": 1.0749784190405722, "grad_norm": 0.031494140625, "learning_rate": 9.592748465167588e-05, "loss": 0.1197545051574707, "mean_token_accuracy": 0.9783686861395836, "num_tokens": 3090296.0, "step": 1090 }, { "entropy": 0.12862191488966346, "epoch": 1.0848440004932791, "grad_norm": 0.00689697265625, "learning_rate": 9.429999238696517e-05, "loss": 0.11654337644577026, "mean_token_accuracy": 0.9792077913880348, "num_tokens": 3118992.0, "step": 1100 }, { "entropy": 0.12970471139997244, "epoch": 1.094709581945986, "grad_norm": 0.73828125, "learning_rate": 9.267401351452915e-05, "loss": 0.1205396056175232, "mean_token_accuracy": 0.9789305351674557, "num_tokens": 3147469.0, "step": 1110 }, { "entropy": 0.13936769934371113, "epoch": 1.1045751633986929, "grad_norm": 0.0032501220703125, "learning_rate": 9.104997974324113e-05, "loss": 0.12835370302200316, "mean_token_accuracy": 0.9773115016520023, "num_tokens": 3175526.0, "step": 1120 }, { "entropy": 0.1322865524329245, "epoch": 1.1144407448513998, "grad_norm": 0.001922607421875, "learning_rate": 8.942832226553628e-05, "loss": 0.11992882490158081, "mean_token_accuracy": 0.9786051079630852, "num_tokens": 3204044.0, "step": 1130 }, { "entropy": 0.1381893863901496, "epoch": 1.1243063263041067, "grad_norm": 0.039306640625, "learning_rate": 8.780947164292708e-05, "loss": 0.12567477226257323, "mean_token_accuracy": 0.9773389711976052, "num_tokens": 3232314.0, "step": 1140 }, { "entropy": 0.13791207429021596, "epoch": 1.1341719077568135, "grad_norm": 0.0033416748046875, "learning_rate": 8.619385769168614e-05, "loss": 0.12569307088851928, "mean_token_accuracy": 0.9776455730199813, "num_tokens": 3260497.0, "step": 1150 }, { "entropy": 0.1384853098541498, "epoch": 1.1440374892095202, "grad_norm": 0.00787353515625, "learning_rate": 8.458190936872696e-05, "loss": 0.12491084337234497, "mean_token_accuracy": 0.9776448138058186, "num_tokens": 3288751.0, "step": 1160 }, { "entropy": 0.13668620772659779, "epoch": 1.153903070662227, "grad_norm": 0.0036773681640625, "learning_rate": 8.29740546577129e-05, "loss": 0.12487212419509888, "mean_token_accuracy": 0.9778970316052437, "num_tokens": 3317022.0, "step": 1170 }, { "entropy": 0.13441719906404614, "epoch": 1.163768652114934, "grad_norm": 0.0047607421875, "learning_rate": 8.137072045542437e-05, "loss": 0.1219482660293579, "mean_token_accuracy": 0.9780469901859761, "num_tokens": 3345411.0, "step": 1180 }, { "entropy": 0.1383935773745179, "epoch": 1.1736342335676408, "grad_norm": 0.004791259765625, "learning_rate": 7.977233245841482e-05, "loss": 0.12580922842025757, "mean_token_accuracy": 0.9776074849069118, "num_tokens": 3373582.0, "step": 1190 }, { "entropy": 0.1389146512374282, "epoch": 1.1834998150203477, "grad_norm": 0.0927734375, "learning_rate": 7.817931504998528e-05, "loss": 0.1258576035499573, "mean_token_accuracy": 0.9774652354419231, "num_tokens": 3401794.0, "step": 1200 }, { "entropy": 0.13396938629448413, "epoch": 1.1933653964730546, "grad_norm": 0.004547119140625, "learning_rate": 7.659209118750754e-05, "loss": 0.12248883247375489, "mean_token_accuracy": 0.9782519124448299, "num_tokens": 3430151.0, "step": 1210 }, { "entropy": 0.13288255603983998, "epoch": 1.2032309779257615, "grad_norm": 0.0026702880859375, "learning_rate": 7.501108229012607e-05, "loss": 0.12038558721542358, "mean_token_accuracy": 0.9784954778850079, "num_tokens": 3458645.0, "step": 1220 }, { "entropy": 0.1470490033738315, "epoch": 1.2130965593784684, "grad_norm": 0.109375, "learning_rate": 7.343670812686823e-05, "loss": 0.13351836204528808, "mean_token_accuracy": 0.9757956475019455, "num_tokens": 3486461.0, "step": 1230 }, { "entropy": 0.14223379427567123, "epoch": 1.2229621408311753, "grad_norm": 0.0022430419921875, "learning_rate": 7.186938670519254e-05, "loss": 0.1276126980781555, "mean_token_accuracy": 0.9769740261137485, "num_tokens": 3514581.0, "step": 1240 }, { "entropy": 0.13168479865416885, "epoch": 1.2328277222838822, "grad_norm": 0.004669189453125, "learning_rate": 7.030953416000492e-05, "loss": 0.12016587257385254, "mean_token_accuracy": 0.9782513029873371, "num_tokens": 3543143.0, "step": 1250 }, { "entropy": 0.1365566598251462, "epoch": 1.242693303736589, "grad_norm": 0.004425048828125, "learning_rate": 6.875756464317178e-05, "loss": 0.12377066612243652, "mean_token_accuracy": 0.9776980735361576, "num_tokens": 3571467.0, "step": 1260 }, { "entropy": 0.1351331824436784, "epoch": 1.252558885189296, "grad_norm": 0.003173828125, "learning_rate": 6.721389021355982e-05, "loss": 0.12217887639999389, "mean_token_accuracy": 0.9781767517328263, "num_tokens": 3599835.0, "step": 1270 }, { "entropy": 0.13429148057475687, "epoch": 1.2624244666420026, "grad_norm": 0.0546875, "learning_rate": 6.567892072763151e-05, "loss": 0.12223434448242188, "mean_token_accuracy": 0.9783378601074219, "num_tokens": 3628185.0, "step": 1280 }, { "entropy": 0.142876058537513, "epoch": 1.2722900480947095, "grad_norm": 0.06982421875, "learning_rate": 6.415306373062536e-05, "loss": 0.12888083457946778, "mean_token_accuracy": 0.9767387375235558, "num_tokens": 3656119.0, "step": 1290 }, { "entropy": 0.13675054740160703, "epoch": 1.2821556295474164, "grad_norm": 0.0047607421875, "learning_rate": 6.263672434834965e-05, "loss": 0.1242174506187439, "mean_token_accuracy": 0.9776271350681782, "num_tokens": 3684428.0, "step": 1300 }, { "entropy": 0.13780908789485694, "epoch": 1.2920212110001232, "grad_norm": 0.010498046875, "learning_rate": 6.113030517961875e-05, "loss": 0.12436658143997192, "mean_token_accuracy": 0.9775737129151821, "num_tokens": 3712674.0, "step": 1310 }, { "entropy": 0.1413337049074471, "epoch": 1.3018867924528301, "grad_norm": 0.00433349609375, "learning_rate": 5.963420618936031e-05, "loss": 0.1274978280067444, "mean_token_accuracy": 0.9770147010684014, "num_tokens": 3740764.0, "step": 1320 }, { "entropy": 0.13298388719558715, "epoch": 1.311752373905537, "grad_norm": 0.0247802734375, "learning_rate": 5.814882460242167e-05, "loss": 0.12207503318786621, "mean_token_accuracy": 0.9782201044261456, "num_tokens": 3769154.0, "step": 1330 }, { "entropy": 0.1320271048694849, "epoch": 1.321617955358244, "grad_norm": 0.005401611328125, "learning_rate": 5.667455479810386e-05, "loss": 0.12129522562026977, "mean_token_accuracy": 0.9786344654858112, "num_tokens": 3797564.0, "step": 1340 }, { "entropy": 0.13690719436854124, "epoch": 1.3314835368109508, "grad_norm": 0.1328125, "learning_rate": 5.521178820545132e-05, "loss": 0.12443944215774536, "mean_token_accuracy": 0.9775426663458348, "num_tokens": 3825875.0, "step": 1350 }, { "entropy": 0.13706938559189438, "epoch": 1.3413491182636577, "grad_norm": 0.00148773193359375, "learning_rate": 5.37609131993244e-05, "loss": 0.12503232955932617, "mean_token_accuracy": 0.9777161441743374, "num_tokens": 3854095.0, "step": 1360 }, { "entropy": 0.12949093943461776, "epoch": 1.3512146997163645, "grad_norm": 0.0050048828125, "learning_rate": 5.232231499728331e-05, "loss": 0.11770267486572265, "mean_token_accuracy": 0.9790447227656841, "num_tokens": 3882689.0, "step": 1370 }, { "entropy": 0.13218411868438124, "epoch": 1.3610802811690714, "grad_norm": 0.00469970703125, "learning_rate": 5.089637555731006e-05, "loss": 0.1202161192893982, "mean_token_accuracy": 0.9784768708050251, "num_tokens": 3911186.0, "step": 1380 }, { "entropy": 0.13080717977136375, "epoch": 1.3709458626217783, "grad_norm": 0.0194091796875, "learning_rate": 4.9483473476395935e-05, "loss": 0.11989362239837646, "mean_token_accuracy": 0.9788383729755878, "num_tokens": 3939716.0, "step": 1390 }, { "entropy": 0.13561268495395778, "epoch": 1.3808114440744852, "grad_norm": 0.001983642578125, "learning_rate": 4.80839838900213e-05, "loss": 0.12290310859680176, "mean_token_accuracy": 0.9780489586293697, "num_tokens": 3968064.0, "step": 1400 }, { "entropy": 0.1323835564777255, "epoch": 1.390677025527192, "grad_norm": 0.15234375, "learning_rate": 4.669827837255461e-05, "loss": 0.12190349102020263, "mean_token_accuracy": 0.9783749088644982, "num_tokens": 3996514.0, "step": 1410 }, { "entropy": 0.13479993510991334, "epoch": 1.400542606979899, "grad_norm": 0.00433349609375, "learning_rate": 4.532672483859653e-05, "loss": 0.12221817970275879, "mean_token_accuracy": 0.9781881980597973, "num_tokens": 4024924.0, "step": 1420 }, { "entropy": 0.1331013419665396, "epoch": 1.4104081884326058, "grad_norm": 0.008056640625, "learning_rate": 4.396968744529626e-05, "loss": 0.1203757643699646, "mean_token_accuracy": 0.9784512914717197, "num_tokens": 4053373.0, "step": 1430 }, { "entropy": 0.13283769460394979, "epoch": 1.4202737698853127, "grad_norm": 0.00433349609375, "learning_rate": 4.2627526495665116e-05, "loss": 0.11865881681442261, "mean_token_accuracy": 0.9784345299005508, "num_tokens": 4081976.0, "step": 1440 }, { "entropy": 0.13691943241283297, "epoch": 1.4301393513380196, "grad_norm": 0.005279541015625, "learning_rate": 4.130059834291354e-05, "loss": 0.1245947003364563, "mean_token_accuracy": 0.9778485208749771, "num_tokens": 4110243.0, "step": 1450 }, { "entropy": 0.13957461677491664, "epoch": 1.4400049327907263, "grad_norm": 0.005157470703125, "learning_rate": 3.998925529583686e-05, "loss": 0.1272833228111267, "mean_token_accuracy": 0.977188840508461, "num_tokens": 4138341.0, "step": 1460 }, { "entropy": 0.1353292756713927, "epoch": 1.4498705142434332, "grad_norm": 0.00885009765625, "learning_rate": 3.869384552527477e-05, "loss": 0.12350196838378906, "mean_token_accuracy": 0.977997999638319, "num_tokens": 4166695.0, "step": 1470 }, { "entropy": 0.13581356341019274, "epoch": 1.45973609569614, "grad_norm": 0.0035552978515625, "learning_rate": 3.7414712971669555e-05, "loss": 0.12334957122802734, "mean_token_accuracy": 0.9777064241468907, "num_tokens": 4195030.0, "step": 1480 }, { "entropy": 0.13494674442335963, "epoch": 1.469601677148847, "grad_norm": 0.0006103515625, "learning_rate": 3.615219725374739e-05, "loss": 0.12299468517303466, "mean_token_accuracy": 0.9781503833830356, "num_tokens": 4223406.0, "step": 1490 }, { "entropy": 0.13651033407077193, "epoch": 1.4794672586015538, "grad_norm": 0.04345703125, "learning_rate": 3.490663357834728e-05, "loss": 0.12479565143585206, "mean_token_accuracy": 0.9776611216366291, "num_tokens": 4251632.0, "step": 1500 }, { "entropy": 0.13344744117930532, "epoch": 1.4893328400542607, "grad_norm": 0.017333984375, "learning_rate": 3.367835265142122e-05, "loss": 0.12157227993011474, "mean_token_accuracy": 0.9784086912870407, "num_tokens": 4280034.0, "step": 1510 }, { "entropy": 0.133386751729995, "epoch": 1.4991984215069676, "grad_norm": 0.005126953125, "learning_rate": 3.2467680590229376e-05, "loss": 0.12114341259002685, "mean_token_accuracy": 0.9782067142426968, "num_tokens": 4308424.0, "step": 1520 }, { "entropy": 0.14430891815572977, "epoch": 1.5090640029596745, "grad_norm": 0.0072021484375, "learning_rate": 3.127493883675383e-05, "loss": 0.13027212619781495, "mean_token_accuracy": 0.976651220023632, "num_tokens": 4336449.0, "step": 1530 }, { "entropy": 0.14111009733751417, "epoch": 1.5189295844123814, "grad_norm": 0.0037994384765625, "learning_rate": 3.010044407235323e-05, "loss": 0.12620443105697632, "mean_token_accuracy": 0.9771548487246037, "num_tokens": 4364618.0, "step": 1540 }, { "entropy": 0.14147729156538844, "epoch": 1.528795165865088, "grad_norm": 0.003326416015625, "learning_rate": 2.89445081336818e-05, "loss": 0.1275140404701233, "mean_token_accuracy": 0.9768898740410805, "num_tokens": 4392696.0, "step": 1550 }, { "entropy": 0.13267950126901268, "epoch": 1.538660747317795, "grad_norm": 0.07177734375, "learning_rate": 2.7807437929894486e-05, "loss": 0.12011861801147461, "mean_token_accuracy": 0.9784411750733852, "num_tokens": 4421270.0, "step": 1560 }, { "entropy": 0.1457324031740427, "epoch": 1.5485263287705018, "grad_norm": 0.0223388671875, "learning_rate": 2.6689535361160356e-05, "loss": 0.1316989541053772, "mean_token_accuracy": 0.9763911448419094, "num_tokens": 4449066.0, "step": 1570 }, { "entropy": 0.13126388657838106, "epoch": 1.5583919102232087, "grad_norm": 0.0019683837890625, "learning_rate": 2.5591097238506045e-05, "loss": 0.11896783113479614, "mean_token_accuracy": 0.9786751069128513, "num_tokens": 4477621.0, "step": 1580 }, { "entropy": 0.13537672525271774, "epoch": 1.5682574916759155, "grad_norm": 0.028076171875, "learning_rate": 2.45124152050103e-05, "loss": 0.12315057516098023, "mean_token_accuracy": 0.9778838872909545, "num_tokens": 4505949.0, "step": 1590 }, { "entropy": 0.13635533889755608, "epoch": 1.5781230731286224, "grad_norm": 0.00140380859375, "learning_rate": 2.3453775658370626e-05, "loss": 0.12194361686706542, "mean_token_accuracy": 0.9778023555874824, "num_tokens": 4534320.0, "step": 1600 }, { "entropy": 0.1336421871557832, "epoch": 1.5879886545813293, "grad_norm": 0.0050048828125, "learning_rate": 2.2415459674862795e-05, "loss": 0.12157649993896484, "mean_token_accuracy": 0.9782549634575843, "num_tokens": 4562694.0, "step": 1610 }, { "entropy": 0.13699853476136922, "epoch": 1.5978542360340362, "grad_norm": 0.0028228759765625, "learning_rate": 2.1397742934712982e-05, "loss": 0.12470865249633789, "mean_token_accuracy": 0.9777181230485439, "num_tokens": 4590893.0, "step": 1620 }, { "entropy": 0.13491226183250546, "epoch": 1.607719817486743, "grad_norm": 0.00049591064453125, "learning_rate": 2.040089564890284e-05, "loss": 0.12282907962799072, "mean_token_accuracy": 0.9781803645193576, "num_tokens": 4619224.0, "step": 1630 }, { "entropy": 0.13295670710504054, "epoch": 1.61758539893945, "grad_norm": 0.017578125, "learning_rate": 1.942518248742644e-05, "loss": 0.12057149410247803, "mean_token_accuracy": 0.978556378185749, "num_tokens": 4647757.0, "step": 1640 }, { "entropy": 0.13519622730091213, "epoch": 1.6274509803921569, "grad_norm": 0.000568389892578125, "learning_rate": 1.8470862509018706e-05, "loss": 0.1230826735496521, "mean_token_accuracy": 0.9779294766485691, "num_tokens": 4676100.0, "step": 1650 }, { "entropy": 0.14108815789222717, "epoch": 1.6373165618448637, "grad_norm": 0.0022430419921875, "learning_rate": 1.7538189092373247e-05, "loss": 0.1281256318092346, "mean_token_accuracy": 0.9770424336194992, "num_tokens": 4704175.0, "step": 1660 }, { "entropy": 0.1354208650998771, "epoch": 1.6471821432975706, "grad_norm": 0.0008087158203125, "learning_rate": 1.662740986886875e-05, "loss": 0.12380268573760986, "mean_token_accuracy": 0.9778368383646011, "num_tokens": 4732510.0, "step": 1670 }, { "entropy": 0.13397875968366862, "epoch": 1.6570477247502775, "grad_norm": 0.0023345947265625, "learning_rate": 1.5738766656820968e-05, "loss": 0.1210429310798645, "mean_token_accuracy": 0.9783028393983841, "num_tokens": 4760963.0, "step": 1680 }, { "entropy": 0.13994298363104463, "epoch": 1.6669133062029844, "grad_norm": 0.001556396484375, "learning_rate": 1.4872495397278285e-05, "loss": 0.12799561023712158, "mean_token_accuracy": 0.9769411139190197, "num_tokens": 4789053.0, "step": 1690 }, { "entropy": 0.1352415460161865, "epoch": 1.6767788876556913, "grad_norm": 0.008544921875, "learning_rate": 1.4028826091377723e-05, "loss": 0.12407561540603637, "mean_token_accuracy": 0.9780896320939064, "num_tokens": 4817292.0, "step": 1700 }, { "entropy": 0.1349070666357875, "epoch": 1.6866444691083982, "grad_norm": 0.006317138671875, "learning_rate": 1.320798273927798e-05, "loss": 0.12214746475219726, "mean_token_accuracy": 0.9781498357653617, "num_tokens": 4845683.0, "step": 1710 }, { "entropy": 0.13030224703252316, "epoch": 1.696510050561105, "grad_norm": 0.0020904541015625, "learning_rate": 1.241018328068585e-05, "loss": 0.11913927793502807, "mean_token_accuracy": 0.9788830585777759, "num_tokens": 4874202.0, "step": 1720 }, { "entropy": 0.13764284029603005, "epoch": 1.706375632013812, "grad_norm": 0.00177764892578125, "learning_rate": 1.1635639536991617e-05, "loss": 0.12407644987106323, "mean_token_accuracy": 0.9776008367538452, "num_tokens": 4902478.0, "step": 1730 }, { "entropy": 0.13921196358278393, "epoch": 1.7162412134665188, "grad_norm": 0.00244140625, "learning_rate": 1.0884557155029074e-05, "loss": 0.12623913288116456, "mean_token_accuracy": 0.9771126128733159, "num_tokens": 4930635.0, "step": 1740 }, { "entropy": 0.1401450714096427, "epoch": 1.7261067949192257, "grad_norm": 0.0120849609375, "learning_rate": 1.0157135552474772e-05, "loss": 0.12569124698638917, "mean_token_accuracy": 0.9772847659885884, "num_tokens": 4958778.0, "step": 1750 }, { "entropy": 0.13814257495105267, "epoch": 1.7359723763719326, "grad_norm": 0.01483154296875, "learning_rate": 9.453567864901214e-06, "loss": 0.12515150308609008, "mean_token_accuracy": 0.9776085413992405, "num_tokens": 4986920.0, "step": 1760 }, { "entropy": 0.1348606870509684, "epoch": 1.7458379578246392, "grad_norm": 0.0010223388671875, "learning_rate": 8.77404089449808e-06, "loss": 0.1228646993637085, "mean_token_accuracy": 0.9780511312186718, "num_tokens": 5015229.0, "step": 1770 }, { "entropy": 0.13221357660368085, "epoch": 1.7557035392773461, "grad_norm": 0.018798828125, "learning_rate": 8.118735060474825e-06, "loss": 0.12112637758255004, "mean_token_accuracy": 0.9786153517663478, "num_tokens": 5043676.0, "step": 1780 }, { "entropy": 0.14117911662906407, "epoch": 1.765569120730053, "grad_norm": 0.2890625, "learning_rate": 7.487824351158135e-06, "loss": 0.12821227312088013, "mean_token_accuracy": 0.976915530115366, "num_tokens": 5071700.0, "step": 1790 }, { "entropy": 0.1357238058000803, "epoch": 1.77543470218276, "grad_norm": 0.0206298828125, "learning_rate": 6.881476277796817e-06, "loss": 0.12314530611038207, "mean_token_accuracy": 0.9780463971197605, "num_tokens": 5099970.0, "step": 1800 }, { "entropy": 0.13418944757431744, "epoch": 1.7853002836354668, "grad_norm": 0.0078125, "learning_rate": 6.299851830086389e-06, "loss": 0.12375905513763427, "mean_token_accuracy": 0.9779085904359818, "num_tokens": 5128223.0, "step": 1810 }, { "entropy": 0.13244303213432432, "epoch": 1.7951658650881737, "grad_norm": 0.0096435546875, "learning_rate": 5.743105433425144e-06, "loss": 0.12159106731414795, "mean_token_accuracy": 0.9783015891909599, "num_tokens": 5156694.0, "step": 1820 }, { "entropy": 0.13863758454099298, "epoch": 1.8050314465408805, "grad_norm": 0.00628662109375, "learning_rate": 5.2113849079131305e-06, "loss": 0.12690427303314208, "mean_token_accuracy": 0.9773243874311447, "num_tokens": 5184809.0, "step": 1830 }, { "entropy": 0.13508734861388802, "epoch": 1.8148970279935872, "grad_norm": 0.0023651123046875, "learning_rate": 4.704831429104884e-06, "loss": 0.12359409332275391, "mean_token_accuracy": 0.9781290367245674, "num_tokens": 5213122.0, "step": 1840 }, { "entropy": 0.1349564053118229, "epoch": 1.824762609446294, "grad_norm": 0.0006103515625, "learning_rate": 4.22357949052632e-06, "loss": 0.12259750366210938, "mean_token_accuracy": 0.9780478946864605, "num_tokens": 5241629.0, "step": 1850 }, { "entropy": 0.12540355436503886, "epoch": 1.834628190899001, "grad_norm": 0.059814453125, "learning_rate": 3.7677568679656994e-06, "loss": 0.11663299798965454, "mean_token_accuracy": 0.9793126709759236, "num_tokens": 5270456.0, "step": 1860 }, { "entropy": 0.13347775135189294, "epoch": 1.8444937723517079, "grad_norm": 0.001556396484375, "learning_rate": 3.337484585548356e-06, "loss": 0.12143052816390991, "mean_token_accuracy": 0.9782666966319085, "num_tokens": 5298948.0, "step": 1870 }, { "entropy": 0.1408340737223625, "epoch": 1.8543593538044147, "grad_norm": 0.048828125, "learning_rate": 2.9328768836038413e-06, "loss": 0.12787405252456666, "mean_token_accuracy": 0.9770998515188694, "num_tokens": 5327021.0, "step": 1880 }, { "entropy": 0.1395120346918702, "epoch": 1.8642249352571216, "grad_norm": 0.002166748046875, "learning_rate": 2.554041188334366e-06, "loss": 0.12588475942611693, "mean_token_accuracy": 0.9772776275873184, "num_tokens": 5355124.0, "step": 1890 }, { "entropy": 0.1324211777187884, "epoch": 1.8740905167098285, "grad_norm": 0.001129150390625, "learning_rate": 2.201078083292363e-06, "loss": 0.12047438621520996, "mean_token_accuracy": 0.9783697292208672, "num_tokens": 5383672.0, "step": 1900 }, { "entropy": 0.13247792813926934, "epoch": 1.8839560981625354, "grad_norm": 0.007171630859375, "learning_rate": 1.8740812826748333e-06, "loss": 0.11918172836303711, "mean_token_accuracy": 0.9785584852099418, "num_tokens": 5412243.0, "step": 1910 }, { "entropy": 0.13208765974268316, "epoch": 1.8938216796152423, "grad_norm": 0.009033203125, "learning_rate": 1.5731376064415992e-06, "loss": 0.1188730239868164, "mean_token_accuracy": 0.9786086544394493, "num_tokens": 5440849.0, "step": 1920 }, { "entropy": 0.1309529710561037, "epoch": 1.9036872610679492, "grad_norm": 0.00335693359375, "learning_rate": 1.2983269572639623e-06, "loss": 0.11905847787857056, "mean_token_accuracy": 0.9788044765591621, "num_tokens": 5469424.0, "step": 1930 }, { "entropy": 0.13058362044394017, "epoch": 1.913552842520656, "grad_norm": 0.003082275390625, "learning_rate": 1.0497222993099852e-06, "loss": 0.11901161670684815, "mean_token_accuracy": 0.9787779085338115, "num_tokens": 5498042.0, "step": 1940 }, { "entropy": 0.12557834926992656, "epoch": 1.923418423973363, "grad_norm": 0.0020904541015625, "learning_rate": 8.273896388720115e-07, "loss": 0.11433438062667847, "mean_token_accuracy": 0.9796731486916542, "num_tokens": 5526928.0, "step": 1950 }, { "entropy": 0.13270391179248692, "epoch": 1.9332840054260698, "grad_norm": 0.007080078125, "learning_rate": 6.313880068414935e-07, "loss": 0.1210409164428711, "mean_token_accuracy": 0.9784272678196431, "num_tokens": 5555384.0, "step": 1960 }, { "entropy": 0.12834761645644904, "epoch": 1.9431495868787767, "grad_norm": 0.006317138671875, "learning_rate": 4.6176944303589766e-07, "loss": 0.11687668561935424, "mean_token_accuracy": 0.9792519778013229, "num_tokens": 5584133.0, "step": 1970 }, { "entropy": 0.13026652419939638, "epoch": 1.9530151683314836, "grad_norm": 0.00433349609375, "learning_rate": 3.185789823817564e-07, "loss": 0.11823811531066894, "mean_token_accuracy": 0.9788157038390637, "num_tokens": 5612762.0, "step": 1980 }, { "entropy": 0.13472106643021106, "epoch": 1.9628807497841905, "grad_norm": 0.000713348388671875, "learning_rate": 2.0185464295758848e-07, "loss": 0.12169078588485718, "mean_token_accuracy": 0.9781753793358803, "num_tokens": 5641199.0, "step": 1990 }, { "entropy": 0.12945695519447326, "epoch": 1.9727463312368974, "grad_norm": 0.0020294189453125, "learning_rate": 1.1162741589980652e-07, "loss": 0.11810406446456909, "mean_token_accuracy": 0.9790156573057175, "num_tokens": 5669762.0, "step": 2000 }, { "entropy": 0.134550541639328, "epoch": 1.9826119126896042, "grad_norm": 0.0023040771484375, "learning_rate": 4.7921257174399215e-08, "loss": 0.12251890897750854, "mean_token_accuracy": 0.977896050363779, "num_tokens": 5698114.0, "step": 2010 }, { "entropy": 0.13257609438151122, "epoch": 1.9924774941423111, "grad_norm": 0.004669189453125, "learning_rate": 1.0753081216385852e-08, "loss": 0.12118911743164062, "mean_token_accuracy": 0.9785582162439823, "num_tokens": 5726514.0, "step": 2020 }, { "epoch": 2.0, "eval_entropy": 0.13473600795018603, "eval_loss": 0.12891627848148346, "eval_mean_token_accuracy": 0.9780772614055837, "eval_num_tokens": 5748186.0, "eval_runtime": 105.8696, "eval_samples_per_second": 19.146, "eval_steps_per_second": 9.578, "step": 2028 } ], "logging_steps": 10, "max_steps": 2028, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.7308452509514138e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }