{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 27975, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0017873100983020554, "grad_norm": 0.8679524064064026, "learning_rate": 4.99106344950849e-05, "loss": 0.5902, "step": 50 }, { "epoch": 0.0035746201966041107, "grad_norm": 4.405018329620361, "learning_rate": 4.9821268990169795e-05, "loss": 0.2634, "step": 100 }, { "epoch": 0.005361930294906166, "grad_norm": 0.18726997077465057, "learning_rate": 4.973190348525469e-05, "loss": 0.1303, "step": 150 }, { "epoch": 0.0071492403932082215, "grad_norm": 0.007119085639715195, "learning_rate": 4.9642537980339595e-05, "loss": 0.1407, "step": 200 }, { "epoch": 0.008936550491510277, "grad_norm": 0.07268306612968445, "learning_rate": 4.9553172475424484e-05, "loss": 0.1344, "step": 250 }, { "epoch": 0.010723860589812333, "grad_norm": 0.007479673717170954, "learning_rate": 4.946380697050939e-05, "loss": 0.0245, "step": 300 }, { "epoch": 0.012511170688114389, "grad_norm": 0.0016641393303871155, "learning_rate": 4.9374441465594284e-05, "loss": 0.1117, "step": 350 }, { "epoch": 0.014298480786416443, "grad_norm": 0.01625875197350979, "learning_rate": 4.928507596067918e-05, "loss": 0.03, "step": 400 }, { "epoch": 0.0160857908847185, "grad_norm": 0.004859230015426874, "learning_rate": 4.9195710455764076e-05, "loss": 0.0676, "step": 450 }, { "epoch": 0.017873100983020553, "grad_norm": 12.948219299316406, "learning_rate": 4.910634495084897e-05, "loss": 0.1749, "step": 500 }, { "epoch": 0.01966041108132261, "grad_norm": 2.3208250999450684, "learning_rate": 4.901697944593387e-05, "loss": 0.1308, "step": 550 }, { "epoch": 0.021447721179624665, "grad_norm": 0.0003704081173054874, "learning_rate": 4.8927613941018765e-05, "loss": 0.0956, "step": 600 }, { "epoch": 0.02323503127792672, "grad_norm": 0.010676690377295017, "learning_rate": 4.883824843610367e-05, "loss": 0.0391, "step": 650 }, { "epoch": 0.025022341376228777, "grad_norm": 0.021889682859182358, "learning_rate": 4.874888293118856e-05, "loss": 0.0802, "step": 700 }, { "epoch": 0.02680965147453083, "grad_norm": 0.0009336157818324864, "learning_rate": 4.865951742627346e-05, "loss": 0.1714, "step": 750 }, { "epoch": 0.028596961572832886, "grad_norm": 0.9198712706565857, "learning_rate": 4.857015192135836e-05, "loss": 0.0459, "step": 800 }, { "epoch": 0.03038427167113494, "grad_norm": 0.21598006784915924, "learning_rate": 4.848078641644326e-05, "loss": 0.0471, "step": 850 }, { "epoch": 0.032171581769437, "grad_norm": 11.548770904541016, "learning_rate": 4.839142091152815e-05, "loss": 0.0597, "step": 900 }, { "epoch": 0.03395889186773905, "grad_norm": 0.00023208574566524476, "learning_rate": 4.8302055406613053e-05, "loss": 0.0156, "step": 950 }, { "epoch": 0.035746201966041107, "grad_norm": 0.0003857246192637831, "learning_rate": 4.821268990169795e-05, "loss": 0.152, "step": 1000 }, { "epoch": 0.03753351206434316, "grad_norm": 0.0007178910309448838, "learning_rate": 4.8123324396782846e-05, "loss": 0.0664, "step": 1050 }, { "epoch": 0.03932082216264522, "grad_norm": 0.00014497812662739307, "learning_rate": 4.803395889186774e-05, "loss": 0.0932, "step": 1100 }, { "epoch": 0.041108132260947276, "grad_norm": 1.2417834997177124, "learning_rate": 4.794459338695264e-05, "loss": 0.0247, "step": 1150 }, { "epoch": 0.04289544235924933, "grad_norm": 0.008595147170126438, "learning_rate": 4.7855227882037535e-05, "loss": 0.0341, "step": 1200 }, { "epoch": 0.044682752457551385, "grad_norm": 0.0008264633943326771, "learning_rate": 4.776586237712243e-05, "loss": 0.0409, "step": 1250 }, { "epoch": 0.04647006255585344, "grad_norm": 0.0013328195782378316, "learning_rate": 4.7676496872207335e-05, "loss": 0.1012, "step": 1300 }, { "epoch": 0.04825737265415549, "grad_norm": 0.010135401971638203, "learning_rate": 4.7587131367292224e-05, "loss": 0.0313, "step": 1350 }, { "epoch": 0.050044682752457555, "grad_norm": 0.00528075871989131, "learning_rate": 4.749776586237713e-05, "loss": 0.1651, "step": 1400 }, { "epoch": 0.05183199285075961, "grad_norm": 0.00036412960616871715, "learning_rate": 4.7408400357462024e-05, "loss": 0.018, "step": 1450 }, { "epoch": 0.05361930294906166, "grad_norm": 0.8087392449378967, "learning_rate": 4.731903485254692e-05, "loss": 0.0186, "step": 1500 }, { "epoch": 0.05540661304736372, "grad_norm": 0.4537216126918793, "learning_rate": 4.7229669347631816e-05, "loss": 0.0668, "step": 1550 }, { "epoch": 0.05719392314566577, "grad_norm": 0.0001737412385409698, "learning_rate": 4.714030384271671e-05, "loss": 0.0726, "step": 1600 }, { "epoch": 0.058981233243967826, "grad_norm": 0.08539757132530212, "learning_rate": 4.705093833780161e-05, "loss": 0.0042, "step": 1650 }, { "epoch": 0.06076854334226988, "grad_norm": 0.00496064405888319, "learning_rate": 4.6961572832886506e-05, "loss": 0.0476, "step": 1700 }, { "epoch": 0.06255585344057193, "grad_norm": 0.042360056191682816, "learning_rate": 4.687220732797141e-05, "loss": 0.0829, "step": 1750 }, { "epoch": 0.064343163538874, "grad_norm": 3.7637750210706145e-05, "learning_rate": 4.67828418230563e-05, "loss": 0.0436, "step": 1800 }, { "epoch": 0.06613047363717604, "grad_norm": 0.23298493027687073, "learning_rate": 4.66934763181412e-05, "loss": 0.0378, "step": 1850 }, { "epoch": 0.0679177837354781, "grad_norm": 0.0012479070574045181, "learning_rate": 4.66041108132261e-05, "loss": 0.0557, "step": 1900 }, { "epoch": 0.06970509383378017, "grad_norm": 0.006392527837306261, "learning_rate": 4.6514745308310994e-05, "loss": 0.0256, "step": 1950 }, { "epoch": 0.07149240393208221, "grad_norm": 0.007692915853112936, "learning_rate": 4.642537980339589e-05, "loss": 0.0492, "step": 2000 }, { "epoch": 0.07327971403038427, "grad_norm": 0.0017933073686435819, "learning_rate": 4.633601429848079e-05, "loss": 0.1599, "step": 2050 }, { "epoch": 0.07506702412868632, "grad_norm": 0.18109554052352905, "learning_rate": 4.624664879356568e-05, "loss": 0.1715, "step": 2100 }, { "epoch": 0.07685433422698838, "grad_norm": 0.16229896247386932, "learning_rate": 4.615728328865058e-05, "loss": 0.0534, "step": 2150 }, { "epoch": 0.07864164432529044, "grad_norm": 0.0001964129478437826, "learning_rate": 4.606791778373548e-05, "loss": 0.0263, "step": 2200 }, { "epoch": 0.08042895442359249, "grad_norm": 1.6310471296310425, "learning_rate": 4.597855227882037e-05, "loss": 0.0085, "step": 2250 }, { "epoch": 0.08221626452189455, "grad_norm": 4.2389536247355863e-05, "learning_rate": 4.5889186773905275e-05, "loss": 0.0656, "step": 2300 }, { "epoch": 0.0840035746201966, "grad_norm": 3.217921257019043, "learning_rate": 4.579982126899017e-05, "loss": 0.1063, "step": 2350 }, { "epoch": 0.08579088471849866, "grad_norm": 9.60240650177002, "learning_rate": 4.5710455764075075e-05, "loss": 0.0213, "step": 2400 }, { "epoch": 0.08757819481680071, "grad_norm": 0.22894810140132904, "learning_rate": 4.5621090259159964e-05, "loss": 0.0452, "step": 2450 }, { "epoch": 0.08936550491510277, "grad_norm": 19.863664627075195, "learning_rate": 4.553172475424487e-05, "loss": 0.0724, "step": 2500 }, { "epoch": 0.09115281501340483, "grad_norm": 0.0016126203117892146, "learning_rate": 4.5442359249329764e-05, "loss": 0.0507, "step": 2550 }, { "epoch": 0.09294012511170688, "grad_norm": 1.5360804796218872, "learning_rate": 4.535299374441466e-05, "loss": 0.028, "step": 2600 }, { "epoch": 0.09472743521000894, "grad_norm": 0.0007274563540704548, "learning_rate": 4.5263628239499557e-05, "loss": 0.0357, "step": 2650 }, { "epoch": 0.09651474530831099, "grad_norm": 6.482748985290527, "learning_rate": 4.517426273458445e-05, "loss": 0.0288, "step": 2700 }, { "epoch": 0.09830205540661305, "grad_norm": 0.005625843536108732, "learning_rate": 4.508489722966935e-05, "loss": 0.1256, "step": 2750 }, { "epoch": 0.10008936550491511, "grad_norm": 2.7138848304748535, "learning_rate": 4.4995531724754246e-05, "loss": 0.0259, "step": 2800 }, { "epoch": 0.10187667560321716, "grad_norm": 0.0008782164077274501, "learning_rate": 4.490616621983915e-05, "loss": 0.0787, "step": 2850 }, { "epoch": 0.10366398570151922, "grad_norm": 0.005403811577707529, "learning_rate": 4.481680071492404e-05, "loss": 0.0882, "step": 2900 }, { "epoch": 0.10545129579982127, "grad_norm": 0.011978592723608017, "learning_rate": 4.472743521000894e-05, "loss": 0.1246, "step": 2950 }, { "epoch": 0.10723860589812333, "grad_norm": 0.046049825847148895, "learning_rate": 4.463806970509384e-05, "loss": 0.0274, "step": 3000 }, { "epoch": 0.10902591599642537, "grad_norm": 0.017301073297858238, "learning_rate": 4.4548704200178734e-05, "loss": 0.0375, "step": 3050 }, { "epoch": 0.11081322609472744, "grad_norm": 0.00014423375250771642, "learning_rate": 4.445933869526363e-05, "loss": 0.0854, "step": 3100 }, { "epoch": 0.1126005361930295, "grad_norm": 3.674675463116728e-05, "learning_rate": 4.436997319034853e-05, "loss": 0.0106, "step": 3150 }, { "epoch": 0.11438784629133154, "grad_norm": 0.006831511855125427, "learning_rate": 4.428060768543342e-05, "loss": 0.0253, "step": 3200 }, { "epoch": 0.1161751563896336, "grad_norm": 0.005952741950750351, "learning_rate": 4.419124218051832e-05, "loss": 0.0196, "step": 3250 }, { "epoch": 0.11796246648793565, "grad_norm": 0.07499190419912338, "learning_rate": 4.410187667560322e-05, "loss": 0.0228, "step": 3300 }, { "epoch": 0.11974977658623771, "grad_norm": 7.443950744345784e-05, "learning_rate": 4.401251117068811e-05, "loss": 0.0018, "step": 3350 }, { "epoch": 0.12153708668453976, "grad_norm": 0.0009010107023641467, "learning_rate": 4.3923145665773015e-05, "loss": 0.0082, "step": 3400 }, { "epoch": 0.12332439678284182, "grad_norm": 0.022291332483291626, "learning_rate": 4.383378016085791e-05, "loss": 0.0415, "step": 3450 }, { "epoch": 0.12511170688114387, "grad_norm": 0.00010954050230793655, "learning_rate": 4.374441465594281e-05, "loss": 0.0191, "step": 3500 }, { "epoch": 0.12689901697944594, "grad_norm": 0.13758663833141327, "learning_rate": 4.3655049151027704e-05, "loss": 0.0157, "step": 3550 }, { "epoch": 0.128686327077748, "grad_norm": 0.002115897834300995, "learning_rate": 4.35656836461126e-05, "loss": 0.0152, "step": 3600 }, { "epoch": 0.13047363717605004, "grad_norm": 0.0017060822574421763, "learning_rate": 4.34763181411975e-05, "loss": 0.0008, "step": 3650 }, { "epoch": 0.1322609472743521, "grad_norm": 0.000241179543081671, "learning_rate": 4.3386952636282394e-05, "loss": 0.0612, "step": 3700 }, { "epoch": 0.13404825737265416, "grad_norm": 0.00036373737384565175, "learning_rate": 4.32975871313673e-05, "loss": 0.0551, "step": 3750 }, { "epoch": 0.1358355674709562, "grad_norm": 98.16144561767578, "learning_rate": 4.3208221626452186e-05, "loss": 0.0853, "step": 3800 }, { "epoch": 0.13762287756925826, "grad_norm": 0.00014869504957459867, "learning_rate": 4.311885612153709e-05, "loss": 0.0196, "step": 3850 }, { "epoch": 0.13941018766756033, "grad_norm": 0.035751309245824814, "learning_rate": 4.3029490616621986e-05, "loss": 0.0133, "step": 3900 }, { "epoch": 0.14119749776586238, "grad_norm": 0.1625383347272873, "learning_rate": 4.294012511170689e-05, "loss": 0.025, "step": 3950 }, { "epoch": 0.14298480786416443, "grad_norm": 33.116329193115234, "learning_rate": 4.285075960679178e-05, "loss": 0.0495, "step": 4000 }, { "epoch": 0.1447721179624665, "grad_norm": 0.6426512002944946, "learning_rate": 4.276139410187668e-05, "loss": 0.0476, "step": 4050 }, { "epoch": 0.14655942806076855, "grad_norm": 0.00031943226349540055, "learning_rate": 4.267202859696158e-05, "loss": 0.0191, "step": 4100 }, { "epoch": 0.1483467381590706, "grad_norm": 0.003275329479947686, "learning_rate": 4.258266309204647e-05, "loss": 0.0744, "step": 4150 }, { "epoch": 0.15013404825737264, "grad_norm": 1.0780760049819946, "learning_rate": 4.249329758713137e-05, "loss": 0.0119, "step": 4200 }, { "epoch": 0.15192135835567472, "grad_norm": 3.967891216278076, "learning_rate": 4.240393208221627e-05, "loss": 0.0824, "step": 4250 }, { "epoch": 0.15370866845397677, "grad_norm": 0.006215296685695648, "learning_rate": 4.231456657730116e-05, "loss": 0.059, "step": 4300 }, { "epoch": 0.1554959785522788, "grad_norm": 0.007833936251699924, "learning_rate": 4.222520107238606e-05, "loss": 0.1297, "step": 4350 }, { "epoch": 0.1572832886505809, "grad_norm": 0.0012796287192031741, "learning_rate": 4.213583556747096e-05, "loss": 0.0084, "step": 4400 }, { "epoch": 0.15907059874888294, "grad_norm": 0.00018260259821545333, "learning_rate": 4.204647006255585e-05, "loss": 0.0378, "step": 4450 }, { "epoch": 0.16085790884718498, "grad_norm": 0.0003368351317476481, "learning_rate": 4.1957104557640756e-05, "loss": 0.0048, "step": 4500 }, { "epoch": 0.16264521894548703, "grad_norm": 0.02242155559360981, "learning_rate": 4.186773905272565e-05, "loss": 0.0239, "step": 4550 }, { "epoch": 0.1644325290437891, "grad_norm": 2.047290802001953, "learning_rate": 4.177837354781055e-05, "loss": 0.0289, "step": 4600 }, { "epoch": 0.16621983914209115, "grad_norm": 0.001486947643570602, "learning_rate": 4.1689008042895445e-05, "loss": 0.0498, "step": 4650 }, { "epoch": 0.1680071492403932, "grad_norm": 0.0025149271823465824, "learning_rate": 4.159964253798034e-05, "loss": 0.0245, "step": 4700 }, { "epoch": 0.16979445933869527, "grad_norm": 0.0002470668696332723, "learning_rate": 4.151027703306524e-05, "loss": 0.0291, "step": 4750 }, { "epoch": 0.17158176943699732, "grad_norm": 0.0009205661481246352, "learning_rate": 4.1420911528150134e-05, "loss": 0.041, "step": 4800 }, { "epoch": 0.17336907953529937, "grad_norm": 0.029675321653485298, "learning_rate": 4.133154602323504e-05, "loss": 0.0241, "step": 4850 }, { "epoch": 0.17515638963360142, "grad_norm": 19.254920959472656, "learning_rate": 4.1242180518319926e-05, "loss": 0.02, "step": 4900 }, { "epoch": 0.1769436997319035, "grad_norm": 0.00010259875853080302, "learning_rate": 4.115281501340483e-05, "loss": 0.0452, "step": 4950 }, { "epoch": 0.17873100983020554, "grad_norm": 0.05092354863882065, "learning_rate": 4.1063449508489726e-05, "loss": 0.0068, "step": 5000 }, { "epoch": 0.1805183199285076, "grad_norm": 0.34378549456596375, "learning_rate": 4.097408400357462e-05, "loss": 0.0309, "step": 5050 }, { "epoch": 0.18230563002680966, "grad_norm": 0.0017653489485383034, "learning_rate": 4.088471849865952e-05, "loss": 0.0147, "step": 5100 }, { "epoch": 0.1840929401251117, "grad_norm": 0.01869867742061615, "learning_rate": 4.0795352993744415e-05, "loss": 0.0269, "step": 5150 }, { "epoch": 0.18588025022341376, "grad_norm": 0.0027590838726609945, "learning_rate": 4.070598748882931e-05, "loss": 0.0598, "step": 5200 }, { "epoch": 0.1876675603217158, "grad_norm": 0.11687185615301132, "learning_rate": 4.061662198391421e-05, "loss": 0.004, "step": 5250 }, { "epoch": 0.18945487042001788, "grad_norm": 0.05549556389451027, "learning_rate": 4.052725647899911e-05, "loss": 0.0229, "step": 5300 }, { "epoch": 0.19124218051831993, "grad_norm": 6.82716490700841e-05, "learning_rate": 4.0437890974084e-05, "loss": 0.0004, "step": 5350 }, { "epoch": 0.19302949061662197, "grad_norm": 3.671266794204712, "learning_rate": 4.0348525469168903e-05, "loss": 0.0297, "step": 5400 }, { "epoch": 0.19481680071492405, "grad_norm": 3.9267666339874268, "learning_rate": 4.02591599642538e-05, "loss": 0.0628, "step": 5450 }, { "epoch": 0.1966041108132261, "grad_norm": 0.3209056258201599, "learning_rate": 4.01697944593387e-05, "loss": 0.0531, "step": 5500 }, { "epoch": 0.19839142091152814, "grad_norm": 0.005282373167574406, "learning_rate": 4.008042895442359e-05, "loss": 0.0993, "step": 5550 }, { "epoch": 0.20017873100983022, "grad_norm": 0.0042650592513382435, "learning_rate": 3.9991063449508496e-05, "loss": 0.0542, "step": 5600 }, { "epoch": 0.20196604110813227, "grad_norm": 0.00035535218194127083, "learning_rate": 3.990169794459339e-05, "loss": 0.0064, "step": 5650 }, { "epoch": 0.2037533512064343, "grad_norm": 0.0763150155544281, "learning_rate": 3.981233243967828e-05, "loss": 0.0489, "step": 5700 }, { "epoch": 0.20554066130473636, "grad_norm": 0.09344036132097244, "learning_rate": 3.9722966934763185e-05, "loss": 0.0017, "step": 5750 }, { "epoch": 0.20732797140303844, "grad_norm": 0.011462843976914883, "learning_rate": 3.963360142984808e-05, "loss": 0.0149, "step": 5800 }, { "epoch": 0.20911528150134048, "grad_norm": 4.07599909522105e-05, "learning_rate": 3.954423592493298e-05, "loss": 0.0492, "step": 5850 }, { "epoch": 0.21090259159964253, "grad_norm": 3.350656032562256, "learning_rate": 3.9454870420017874e-05, "loss": 0.0381, "step": 5900 }, { "epoch": 0.2126899016979446, "grad_norm": 0.015672476962208748, "learning_rate": 3.936550491510278e-05, "loss": 0.1483, "step": 5950 }, { "epoch": 0.21447721179624665, "grad_norm": 0.0017092832131311297, "learning_rate": 3.9276139410187666e-05, "loss": 0.058, "step": 6000 }, { "epoch": 0.2162645218945487, "grad_norm": 8.521609561285004e-05, "learning_rate": 3.918677390527257e-05, "loss": 0.057, "step": 6050 }, { "epoch": 0.21805183199285075, "grad_norm": 0.004186888691037893, "learning_rate": 3.9097408400357466e-05, "loss": 0.0415, "step": 6100 }, { "epoch": 0.21983914209115282, "grad_norm": 0.0035520398523658514, "learning_rate": 3.900804289544236e-05, "loss": 0.0132, "step": 6150 }, { "epoch": 0.22162645218945487, "grad_norm": 0.0003482762258499861, "learning_rate": 3.891867739052726e-05, "loss": 0.048, "step": 6200 }, { "epoch": 0.22341376228775692, "grad_norm": 0.00028643032419495285, "learning_rate": 3.8829311885612155e-05, "loss": 0.0255, "step": 6250 }, { "epoch": 0.225201072386059, "grad_norm": 1.1610924048000015e-05, "learning_rate": 3.873994638069705e-05, "loss": 0.085, "step": 6300 }, { "epoch": 0.22698838248436104, "grad_norm": 8.35804094094783e-05, "learning_rate": 3.865058087578195e-05, "loss": 0.061, "step": 6350 }, { "epoch": 0.2287756925826631, "grad_norm": 0.0002916629600804299, "learning_rate": 3.856121537086685e-05, "loss": 0.0743, "step": 6400 }, { "epoch": 0.23056300268096513, "grad_norm": 1.7632960081100464, "learning_rate": 3.847184986595174e-05, "loss": 0.0814, "step": 6450 }, { "epoch": 0.2323503127792672, "grad_norm": 0.005348087754100561, "learning_rate": 3.8382484361036644e-05, "loss": 0.0667, "step": 6500 }, { "epoch": 0.23413762287756926, "grad_norm": 0.006519485265016556, "learning_rate": 3.829311885612154e-05, "loss": 0.0322, "step": 6550 }, { "epoch": 0.2359249329758713, "grad_norm": 0.013020873069763184, "learning_rate": 3.8203753351206436e-05, "loss": 0.0216, "step": 6600 }, { "epoch": 0.23771224307417338, "grad_norm": 0.011472865007817745, "learning_rate": 3.811438784629133e-05, "loss": 0.0045, "step": 6650 }, { "epoch": 0.23949955317247543, "grad_norm": 0.0032377566676586866, "learning_rate": 3.802502234137623e-05, "loss": 0.0086, "step": 6700 }, { "epoch": 0.24128686327077747, "grad_norm": 0.00016540393698960543, "learning_rate": 3.7935656836461125e-05, "loss": 0.0189, "step": 6750 }, { "epoch": 0.24307417336907952, "grad_norm": 0.28715601563453674, "learning_rate": 3.784629133154602e-05, "loss": 0.0456, "step": 6800 }, { "epoch": 0.2448614834673816, "grad_norm": 3.217639923095703, "learning_rate": 3.7756925826630925e-05, "loss": 0.0761, "step": 6850 }, { "epoch": 0.24664879356568364, "grad_norm": 0.8029662370681763, "learning_rate": 3.7667560321715814e-05, "loss": 0.0235, "step": 6900 }, { "epoch": 0.2484361036639857, "grad_norm": 0.0005404868279583752, "learning_rate": 3.757819481680072e-05, "loss": 0.033, "step": 6950 }, { "epoch": 0.25022341376228774, "grad_norm": 0.002369440160691738, "learning_rate": 3.7488829311885614e-05, "loss": 0.057, "step": 7000 }, { "epoch": 0.2520107238605898, "grad_norm": 2.6418510969961062e-05, "learning_rate": 3.739946380697052e-05, "loss": 0.0178, "step": 7050 }, { "epoch": 0.2537980339588919, "grad_norm": 7.692318916320801, "learning_rate": 3.7310098302055407e-05, "loss": 0.1066, "step": 7100 }, { "epoch": 0.2555853440571939, "grad_norm": 0.024961167946457863, "learning_rate": 3.722073279714031e-05, "loss": 0.0554, "step": 7150 }, { "epoch": 0.257372654155496, "grad_norm": 0.012517307884991169, "learning_rate": 3.7131367292225206e-05, "loss": 0.0008, "step": 7200 }, { "epoch": 0.25915996425379806, "grad_norm": 5.774623787146993e-05, "learning_rate": 3.7042001787310096e-05, "loss": 0.0183, "step": 7250 }, { "epoch": 0.2609472743521001, "grad_norm": 0.00010851398837985471, "learning_rate": 3.6952636282395e-05, "loss": 0.0462, "step": 7300 }, { "epoch": 0.26273458445040215, "grad_norm": 0.0003813335788436234, "learning_rate": 3.6863270777479895e-05, "loss": 0.0156, "step": 7350 }, { "epoch": 0.2645218945487042, "grad_norm": 0.00062342151068151, "learning_rate": 3.677390527256479e-05, "loss": 0.0891, "step": 7400 }, { "epoch": 0.26630920464700625, "grad_norm": 0.0014715425204485655, "learning_rate": 3.668453976764969e-05, "loss": 0.017, "step": 7450 }, { "epoch": 0.2680965147453083, "grad_norm": 0.00022484315559267998, "learning_rate": 3.659517426273459e-05, "loss": 0.0192, "step": 7500 }, { "epoch": 0.26988382484361034, "grad_norm": 0.02760065533220768, "learning_rate": 3.650580875781948e-05, "loss": 0.0281, "step": 7550 }, { "epoch": 0.2716711349419124, "grad_norm": 0.0006579934270121157, "learning_rate": 3.6416443252904384e-05, "loss": 0.0759, "step": 7600 }, { "epoch": 0.2734584450402145, "grad_norm": 1.039661169052124, "learning_rate": 3.632707774798928e-05, "loss": 0.0024, "step": 7650 }, { "epoch": 0.2752457551385165, "grad_norm": 0.00035826186649501324, "learning_rate": 3.6237712243074176e-05, "loss": 0.0111, "step": 7700 }, { "epoch": 0.2770330652368186, "grad_norm": 0.0013664854923263192, "learning_rate": 3.614834673815907e-05, "loss": 0.0297, "step": 7750 }, { "epoch": 0.27882037533512066, "grad_norm": 0.025907866656780243, "learning_rate": 3.605898123324397e-05, "loss": 0.0017, "step": 7800 }, { "epoch": 0.2806076854334227, "grad_norm": 1.7114611182478257e-05, "learning_rate": 3.5969615728328865e-05, "loss": 0.0211, "step": 7850 }, { "epoch": 0.28239499553172476, "grad_norm": 9.907771891448647e-05, "learning_rate": 3.588025022341376e-05, "loss": 0.0087, "step": 7900 }, { "epoch": 0.28418230563002683, "grad_norm": 2.670928370207548e-05, "learning_rate": 3.5790884718498665e-05, "loss": 0.0231, "step": 7950 }, { "epoch": 0.28596961572832885, "grad_norm": 0.0034552349243313074, "learning_rate": 3.5701519213583554e-05, "loss": 0.0128, "step": 8000 }, { "epoch": 0.2877569258266309, "grad_norm": 0.00045294914161786437, "learning_rate": 3.561215370866846e-05, "loss": 0.0045, "step": 8050 }, { "epoch": 0.289544235924933, "grad_norm": 0.0008908663294278085, "learning_rate": 3.5522788203753354e-05, "loss": 0.0123, "step": 8100 }, { "epoch": 0.291331546023235, "grad_norm": 0.0011008529691025615, "learning_rate": 3.543342269883825e-05, "loss": 0.0231, "step": 8150 }, { "epoch": 0.2931188561215371, "grad_norm": 0.014470423571765423, "learning_rate": 3.534405719392315e-05, "loss": 0.0353, "step": 8200 }, { "epoch": 0.2949061662198391, "grad_norm": 0.0011291592381894588, "learning_rate": 3.525469168900804e-05, "loss": 0.0236, "step": 8250 }, { "epoch": 0.2966934763181412, "grad_norm": 0.0008120378479361534, "learning_rate": 3.516532618409294e-05, "loss": 0.0638, "step": 8300 }, { "epoch": 0.29848078641644327, "grad_norm": 0.0017303444910794497, "learning_rate": 3.5075960679177836e-05, "loss": 0.0557, "step": 8350 }, { "epoch": 0.3002680965147453, "grad_norm": 0.026920760050415993, "learning_rate": 3.498659517426274e-05, "loss": 0.0242, "step": 8400 }, { "epoch": 0.30205540661304736, "grad_norm": 5.600203514099121, "learning_rate": 3.489722966934763e-05, "loss": 0.0211, "step": 8450 }, { "epoch": 0.30384271671134944, "grad_norm": 0.00029427200206555426, "learning_rate": 3.480786416443253e-05, "loss": 0.0007, "step": 8500 }, { "epoch": 0.30563002680965146, "grad_norm": 3.258937431382947e-05, "learning_rate": 3.471849865951743e-05, "loss": 0.0045, "step": 8550 }, { "epoch": 0.30741733690795353, "grad_norm": 0.0003963368944823742, "learning_rate": 3.462913315460233e-05, "loss": 0.0003, "step": 8600 }, { "epoch": 0.3092046470062556, "grad_norm": 0.00033222747151739895, "learning_rate": 3.453976764968722e-05, "loss": 0.0143, "step": 8650 }, { "epoch": 0.3109919571045576, "grad_norm": 0.0008735691662877798, "learning_rate": 3.4450402144772124e-05, "loss": 0.0777, "step": 8700 }, { "epoch": 0.3127792672028597, "grad_norm": 32.758811950683594, "learning_rate": 3.436103663985702e-05, "loss": 0.0444, "step": 8750 }, { "epoch": 0.3145665773011618, "grad_norm": 0.0014150891220197082, "learning_rate": 3.427167113494191e-05, "loss": 0.0312, "step": 8800 }, { "epoch": 0.3163538873994638, "grad_norm": 0.004327408503741026, "learning_rate": 3.418230563002681e-05, "loss": 0.0075, "step": 8850 }, { "epoch": 0.31814119749776587, "grad_norm": 0.027708660811185837, "learning_rate": 3.409294012511171e-05, "loss": 0.0012, "step": 8900 }, { "epoch": 0.3199285075960679, "grad_norm": 0.000973561720456928, "learning_rate": 3.4003574620196606e-05, "loss": 0.0161, "step": 8950 }, { "epoch": 0.32171581769436997, "grad_norm": 0.0005400975351221859, "learning_rate": 3.39142091152815e-05, "loss": 0.0094, "step": 9000 }, { "epoch": 0.32350312779267204, "grad_norm": 0.0005731348064728081, "learning_rate": 3.3824843610366405e-05, "loss": 0.0011, "step": 9050 }, { "epoch": 0.32529043789097406, "grad_norm": 0.0007925840909592807, "learning_rate": 3.3735478105451295e-05, "loss": 0.082, "step": 9100 }, { "epoch": 0.32707774798927614, "grad_norm": 18.95382308959961, "learning_rate": 3.36461126005362e-05, "loss": 0.0408, "step": 9150 }, { "epoch": 0.3288650580875782, "grad_norm": 0.0005464926944114268, "learning_rate": 3.3556747095621094e-05, "loss": 0.0406, "step": 9200 }, { "epoch": 0.33065236818588023, "grad_norm": 0.00010149468289455399, "learning_rate": 3.346738159070599e-05, "loss": 0.0253, "step": 9250 }, { "epoch": 0.3324396782841823, "grad_norm": 0.043079521507024765, "learning_rate": 3.337801608579089e-05, "loss": 0.0192, "step": 9300 }, { "epoch": 0.3342269883824844, "grad_norm": 3.083252886426635e-05, "learning_rate": 3.328865058087578e-05, "loss": 0.0064, "step": 9350 }, { "epoch": 0.3360142984807864, "grad_norm": 0.00011231891403440386, "learning_rate": 3.319928507596068e-05, "loss": 0.035, "step": 9400 }, { "epoch": 0.3378016085790885, "grad_norm": 0.09854020923376083, "learning_rate": 3.3109919571045576e-05, "loss": 0.0797, "step": 9450 }, { "epoch": 0.33958891867739055, "grad_norm": 0.00023640043218620121, "learning_rate": 3.302055406613048e-05, "loss": 0.0273, "step": 9500 }, { "epoch": 0.34137622877569257, "grad_norm": 0.00113078230060637, "learning_rate": 3.293118856121537e-05, "loss": 0.0031, "step": 9550 }, { "epoch": 0.34316353887399464, "grad_norm": 2.54555082321167, "learning_rate": 3.284182305630027e-05, "loss": 0.001, "step": 9600 }, { "epoch": 0.3449508489722967, "grad_norm": 0.00046690518502146006, "learning_rate": 3.275245755138517e-05, "loss": 0.0021, "step": 9650 }, { "epoch": 0.34673815907059874, "grad_norm": 0.0019335439428687096, "learning_rate": 3.2663092046470064e-05, "loss": 0.0044, "step": 9700 }, { "epoch": 0.3485254691689008, "grad_norm": 4.571199315250851e-05, "learning_rate": 3.257372654155496e-05, "loss": 0.0116, "step": 9750 }, { "epoch": 0.35031277926720283, "grad_norm": 0.0006864121532998979, "learning_rate": 3.248436103663986e-05, "loss": 0.0011, "step": 9800 }, { "epoch": 0.3521000893655049, "grad_norm": 0.0008427898283116519, "learning_rate": 3.2394995531724753e-05, "loss": 0.0079, "step": 9850 }, { "epoch": 0.353887399463807, "grad_norm": 0.00018227777036372572, "learning_rate": 3.230563002680965e-05, "loss": 0.0261, "step": 9900 }, { "epoch": 0.355674709562109, "grad_norm": 5.8319881645729765e-05, "learning_rate": 3.221626452189455e-05, "loss": 0.0244, "step": 9950 }, { "epoch": 0.3574620196604111, "grad_norm": 1.659814552112948e-05, "learning_rate": 3.212689901697944e-05, "loss": 0.0351, "step": 10000 }, { "epoch": 0.35924932975871315, "grad_norm": 0.4823216199874878, "learning_rate": 3.2037533512064346e-05, "loss": 0.0418, "step": 10050 }, { "epoch": 0.3610366398570152, "grad_norm": 0.8462158441543579, "learning_rate": 3.194816800714924e-05, "loss": 0.0142, "step": 10100 }, { "epoch": 0.36282394995531725, "grad_norm": 2.9845266908523627e-05, "learning_rate": 3.1858802502234145e-05, "loss": 0.046, "step": 10150 }, { "epoch": 0.3646112600536193, "grad_norm": 0.001422148197889328, "learning_rate": 3.1769436997319035e-05, "loss": 0.0096, "step": 10200 }, { "epoch": 0.36639857015192134, "grad_norm": 5.117091131978668e-05, "learning_rate": 3.168007149240393e-05, "loss": 0.0385, "step": 10250 }, { "epoch": 0.3681858802502234, "grad_norm": 0.0027402061969041824, "learning_rate": 3.1590705987488834e-05, "loss": 0.0262, "step": 10300 }, { "epoch": 0.3699731903485255, "grad_norm": 0.030330533161759377, "learning_rate": 3.1501340482573724e-05, "loss": 0.0468, "step": 10350 }, { "epoch": 0.3717605004468275, "grad_norm": 0.09292767196893692, "learning_rate": 3.141197497765863e-05, "loss": 0.0444, "step": 10400 }, { "epoch": 0.3735478105451296, "grad_norm": 0.00016525393584743142, "learning_rate": 3.132260947274352e-05, "loss": 0.0336, "step": 10450 }, { "epoch": 0.3753351206434316, "grad_norm": 0.00034486394724808633, "learning_rate": 3.123324396782842e-05, "loss": 0.0341, "step": 10500 }, { "epoch": 0.3771224307417337, "grad_norm": 0.008806136436760426, "learning_rate": 3.1143878462913316e-05, "loss": 0.0343, "step": 10550 }, { "epoch": 0.37890974084003576, "grad_norm": 2.650175338203553e-05, "learning_rate": 3.105451295799822e-05, "loss": 0.0007, "step": 10600 }, { "epoch": 0.3806970509383378, "grad_norm": 0.16857172548770905, "learning_rate": 3.096514745308311e-05, "loss": 0.0099, "step": 10650 }, { "epoch": 0.38248436103663985, "grad_norm": 0.009882774204015732, "learning_rate": 3.087578194816801e-05, "loss": 0.0459, "step": 10700 }, { "epoch": 0.38427167113494193, "grad_norm": 0.00250116060487926, "learning_rate": 3.078641644325291e-05, "loss": 0.0005, "step": 10750 }, { "epoch": 0.38605898123324395, "grad_norm": 7.572743925265968e-05, "learning_rate": 3.0697050938337804e-05, "loss": 0.0194, "step": 10800 }, { "epoch": 0.387846291331546, "grad_norm": 0.00021719296637456864, "learning_rate": 3.06076854334227e-05, "loss": 0.0813, "step": 10850 }, { "epoch": 0.3896336014298481, "grad_norm": 0.0006488068611361086, "learning_rate": 3.05183199285076e-05, "loss": 0.0073, "step": 10900 }, { "epoch": 0.3914209115281501, "grad_norm": 0.0017973796930164099, "learning_rate": 3.0428954423592494e-05, "loss": 0.0069, "step": 10950 }, { "epoch": 0.3932082216264522, "grad_norm": 6.656202458543703e-05, "learning_rate": 3.0339588918677393e-05, "loss": 0.0623, "step": 11000 }, { "epoch": 0.39499553172475427, "grad_norm": 0.006565978284925222, "learning_rate": 3.025022341376229e-05, "loss": 0.0177, "step": 11050 }, { "epoch": 0.3967828418230563, "grad_norm": 0.008500502444803715, "learning_rate": 3.0160857908847186e-05, "loss": 0.0406, "step": 11100 }, { "epoch": 0.39857015192135836, "grad_norm": 0.014150319620966911, "learning_rate": 3.0071492403932082e-05, "loss": 0.009, "step": 11150 }, { "epoch": 0.40035746201966044, "grad_norm": 0.0028486077208071947, "learning_rate": 2.9982126899016982e-05, "loss": 0.0467, "step": 11200 }, { "epoch": 0.40214477211796246, "grad_norm": 6.694492185488343e-05, "learning_rate": 2.9892761394101875e-05, "loss": 0.0038, "step": 11250 }, { "epoch": 0.40393208221626453, "grad_norm": 33.74071502685547, "learning_rate": 2.9803395889186775e-05, "loss": 0.0492, "step": 11300 }, { "epoch": 0.40571939231456655, "grad_norm": 2.357473385927733e-05, "learning_rate": 2.9714030384271675e-05, "loss": 0.0268, "step": 11350 }, { "epoch": 0.4075067024128686, "grad_norm": 0.011629421263933182, "learning_rate": 2.9624664879356567e-05, "loss": 0.0528, "step": 11400 }, { "epoch": 0.4092940125111707, "grad_norm": 0.002155030844733119, "learning_rate": 2.9535299374441467e-05, "loss": 0.0195, "step": 11450 }, { "epoch": 0.4110813226094727, "grad_norm": 0.0004045389359816909, "learning_rate": 2.9445933869526367e-05, "loss": 0.014, "step": 11500 }, { "epoch": 0.4128686327077748, "grad_norm": 8.811664883978665e-06, "learning_rate": 2.935656836461126e-05, "loss": 0.0871, "step": 11550 }, { "epoch": 0.41465594280607687, "grad_norm": 2.1142328478163108e-05, "learning_rate": 2.9267202859696156e-05, "loss": 0.0002, "step": 11600 }, { "epoch": 0.4164432529043789, "grad_norm": 0.21826158463954926, "learning_rate": 2.9177837354781056e-05, "loss": 0.0498, "step": 11650 }, { "epoch": 0.41823056300268097, "grad_norm": 0.03575194999575615, "learning_rate": 2.9088471849865956e-05, "loss": 0.0321, "step": 11700 }, { "epoch": 0.42001787310098304, "grad_norm": 0.00031422683969140053, "learning_rate": 2.899910634495085e-05, "loss": 0.0326, "step": 11750 }, { "epoch": 0.42180518319928506, "grad_norm": 0.001856229966506362, "learning_rate": 2.890974084003575e-05, "loss": 0.0002, "step": 11800 }, { "epoch": 0.42359249329758714, "grad_norm": 0.0010607796721160412, "learning_rate": 2.8820375335120648e-05, "loss": 0.081, "step": 11850 }, { "epoch": 0.4253798033958892, "grad_norm": 0.00241463934071362, "learning_rate": 2.873100983020554e-05, "loss": 0.0194, "step": 11900 }, { "epoch": 0.42716711349419123, "grad_norm": 0.00031845405464991927, "learning_rate": 2.864164432529044e-05, "loss": 0.0012, "step": 11950 }, { "epoch": 0.4289544235924933, "grad_norm": 0.0038351304829120636, "learning_rate": 2.8552278820375337e-05, "loss": 0.0779, "step": 12000 }, { "epoch": 0.4307417336907953, "grad_norm": 0.030567605048418045, "learning_rate": 2.8462913315460234e-05, "loss": 0.0444, "step": 12050 }, { "epoch": 0.4325290437890974, "grad_norm": 19.167110443115234, "learning_rate": 2.837354781054513e-05, "loss": 0.0333, "step": 12100 }, { "epoch": 0.4343163538873995, "grad_norm": 0.06379386782646179, "learning_rate": 2.828418230563003e-05, "loss": 0.0565, "step": 12150 }, { "epoch": 0.4361036639857015, "grad_norm": 0.014705734327435493, "learning_rate": 2.8194816800714923e-05, "loss": 0.0458, "step": 12200 }, { "epoch": 0.43789097408400357, "grad_norm": 0.0011007515713572502, "learning_rate": 2.8105451295799822e-05, "loss": 0.0057, "step": 12250 }, { "epoch": 0.43967828418230565, "grad_norm": 0.11051446944475174, "learning_rate": 2.8016085790884722e-05, "loss": 0.0098, "step": 12300 }, { "epoch": 0.44146559428060766, "grad_norm": 0.0006821333081461489, "learning_rate": 2.7926720285969615e-05, "loss": 0.0159, "step": 12350 }, { "epoch": 0.44325290437890974, "grad_norm": 1.3895423762733117e-05, "learning_rate": 2.7837354781054515e-05, "loss": 0.0047, "step": 12400 }, { "epoch": 0.4450402144772118, "grad_norm": 9.707266872283071e-05, "learning_rate": 2.7747989276139415e-05, "loss": 0.0091, "step": 12450 }, { "epoch": 0.44682752457551383, "grad_norm": 0.0005143424496054649, "learning_rate": 2.7658623771224308e-05, "loss": 0.0317, "step": 12500 }, { "epoch": 0.4486148346738159, "grad_norm": 0.00015378202078863978, "learning_rate": 2.7569258266309207e-05, "loss": 0.0172, "step": 12550 }, { "epoch": 0.450402144772118, "grad_norm": 3.0984447221271694e-05, "learning_rate": 2.7479892761394104e-05, "loss": 0.0203, "step": 12600 }, { "epoch": 0.45218945487042, "grad_norm": 0.004224927164614201, "learning_rate": 2.7390527256478997e-05, "loss": 0.0312, "step": 12650 }, { "epoch": 0.4539767649687221, "grad_norm": 0.0011511313496157527, "learning_rate": 2.7301161751563896e-05, "loss": 0.0354, "step": 12700 }, { "epoch": 0.45576407506702415, "grad_norm": 0.00011386737605789676, "learning_rate": 2.7211796246648796e-05, "loss": 0.0003, "step": 12750 }, { "epoch": 0.4575513851653262, "grad_norm": 0.013496979139745235, "learning_rate": 2.712243074173369e-05, "loss": 0.0011, "step": 12800 }, { "epoch": 0.45933869526362825, "grad_norm": 0.0001817968295654282, "learning_rate": 2.703306523681859e-05, "loss": 0.0007, "step": 12850 }, { "epoch": 0.46112600536193027, "grad_norm": 0.0013657300733029842, "learning_rate": 2.694369973190349e-05, "loss": 0.0019, "step": 12900 }, { "epoch": 0.46291331546023234, "grad_norm": 0.00011192862439202145, "learning_rate": 2.685433422698838e-05, "loss": 0.0011, "step": 12950 }, { "epoch": 0.4647006255585344, "grad_norm": 2.3545103073120117, "learning_rate": 2.676496872207328e-05, "loss": 0.0057, "step": 13000 }, { "epoch": 0.46648793565683644, "grad_norm": 0.011025574989616871, "learning_rate": 2.667560321715818e-05, "loss": 0.0143, "step": 13050 }, { "epoch": 0.4682752457551385, "grad_norm": 0.0017905368003994226, "learning_rate": 2.6586237712243074e-05, "loss": 0.001, "step": 13100 }, { "epoch": 0.4700625558534406, "grad_norm": 2.035365343093872, "learning_rate": 2.649687220732797e-05, "loss": 0.0611, "step": 13150 }, { "epoch": 0.4718498659517426, "grad_norm": 0.2169104963541031, "learning_rate": 2.640750670241287e-05, "loss": 0.0021, "step": 13200 }, { "epoch": 0.4736371760500447, "grad_norm": 6.568299431819469e-05, "learning_rate": 2.631814119749777e-05, "loss": 0.0064, "step": 13250 }, { "epoch": 0.47542448614834676, "grad_norm": 8.24251492304029e-06, "learning_rate": 2.6228775692582663e-05, "loss": 0.0155, "step": 13300 }, { "epoch": 0.4772117962466488, "grad_norm": 0.002219315618276596, "learning_rate": 2.6139410187667563e-05, "loss": 0.006, "step": 13350 }, { "epoch": 0.47899910634495085, "grad_norm": 0.00830235704779625, "learning_rate": 2.6050044682752462e-05, "loss": 0.0841, "step": 13400 }, { "epoch": 0.48078641644325293, "grad_norm": 1.912525294756051e-05, "learning_rate": 2.5960679177837355e-05, "loss": 0.0392, "step": 13450 }, { "epoch": 0.48257372654155495, "grad_norm": 0.0037906430661678314, "learning_rate": 2.5871313672922255e-05, "loss": 0.003, "step": 13500 }, { "epoch": 0.484361036639857, "grad_norm": 2.397968819423113e-05, "learning_rate": 2.578194816800715e-05, "loss": 0.0068, "step": 13550 }, { "epoch": 0.48614834673815904, "grad_norm": 0.0001541871315566823, "learning_rate": 2.5692582663092048e-05, "loss": 0.0384, "step": 13600 }, { "epoch": 0.4879356568364611, "grad_norm": 0.018565978854894638, "learning_rate": 2.5603217158176944e-05, "loss": 0.0114, "step": 13650 }, { "epoch": 0.4897229669347632, "grad_norm": 1.3260701894760132, "learning_rate": 2.5513851653261844e-05, "loss": 0.124, "step": 13700 }, { "epoch": 0.4915102770330652, "grad_norm": 0.08146532624959946, "learning_rate": 2.5424486148346737e-05, "loss": 0.0011, "step": 13750 }, { "epoch": 0.4932975871313673, "grad_norm": 0.0004648593894671649, "learning_rate": 2.5335120643431636e-05, "loss": 0.0228, "step": 13800 }, { "epoch": 0.49508489722966936, "grad_norm": 2.6997038730769418e-05, "learning_rate": 2.5245755138516536e-05, "loss": 0.0276, "step": 13850 }, { "epoch": 0.4968722073279714, "grad_norm": 0.00011526174785103649, "learning_rate": 2.515638963360143e-05, "loss": 0.0034, "step": 13900 }, { "epoch": 0.49865951742627346, "grad_norm": 0.00029504502890631557, "learning_rate": 2.506702412868633e-05, "loss": 0.014, "step": 13950 }, { "epoch": 0.5004468275245755, "grad_norm": 0.0016186943976208568, "learning_rate": 2.4977658623771225e-05, "loss": 0.0019, "step": 14000 }, { "epoch": 0.5022341376228776, "grad_norm": 0.0009279532241635025, "learning_rate": 2.4888293118856125e-05, "loss": 0.0454, "step": 14050 }, { "epoch": 0.5040214477211796, "grad_norm": 0.0004666298336815089, "learning_rate": 2.479892761394102e-05, "loss": 0.0033, "step": 14100 }, { "epoch": 0.5058087578194816, "grad_norm": 0.13235312700271606, "learning_rate": 2.4709562109025918e-05, "loss": 0.0069, "step": 14150 }, { "epoch": 0.5075960679177838, "grad_norm": 0.001180554274469614, "learning_rate": 2.4620196604110814e-05, "loss": 0.0002, "step": 14200 }, { "epoch": 0.5093833780160858, "grad_norm": 5.6808294175425544e-05, "learning_rate": 2.453083109919571e-05, "loss": 0.0019, "step": 14250 }, { "epoch": 0.5111706881143878, "grad_norm": 0.0004835101717617363, "learning_rate": 2.4441465594280607e-05, "loss": 0.0016, "step": 14300 }, { "epoch": 0.51295799821269, "grad_norm": 0.060612790286540985, "learning_rate": 2.4352100089365507e-05, "loss": 0.0202, "step": 14350 }, { "epoch": 0.514745308310992, "grad_norm": 0.012337922118604183, "learning_rate": 2.4262734584450403e-05, "loss": 0.001, "step": 14400 }, { "epoch": 0.516532618409294, "grad_norm": 0.0009960078168660402, "learning_rate": 2.4173369079535303e-05, "loss": 0.0027, "step": 14450 }, { "epoch": 0.5183199285075961, "grad_norm": 0.009885729290544987, "learning_rate": 2.40840035746202e-05, "loss": 0.0705, "step": 14500 }, { "epoch": 0.5201072386058981, "grad_norm": 0.014263806864619255, "learning_rate": 2.3994638069705095e-05, "loss": 0.0014, "step": 14550 }, { "epoch": 0.5218945487042002, "grad_norm": 0.0004742332093883306, "learning_rate": 2.390527256478999e-05, "loss": 0.014, "step": 14600 }, { "epoch": 0.5236818588025023, "grad_norm": 0.000258121348451823, "learning_rate": 2.3815907059874888e-05, "loss": 0.0046, "step": 14650 }, { "epoch": 0.5254691689008043, "grad_norm": 0.0008493911009281874, "learning_rate": 2.3726541554959784e-05, "loss": 0.0187, "step": 14700 }, { "epoch": 0.5272564789991063, "grad_norm": 0.0010454511502757668, "learning_rate": 2.3637176050044684e-05, "loss": 0.0232, "step": 14750 }, { "epoch": 0.5290437890974083, "grad_norm": 0.0003980633628088981, "learning_rate": 2.354781054512958e-05, "loss": 0.034, "step": 14800 }, { "epoch": 0.5308310991957105, "grad_norm": 0.017374487593770027, "learning_rate": 2.3458445040214477e-05, "loss": 0.0054, "step": 14850 }, { "epoch": 0.5326184092940125, "grad_norm": 0.001966976560652256, "learning_rate": 2.3369079535299377e-05, "loss": 0.1053, "step": 14900 }, { "epoch": 0.5344057193923145, "grad_norm": 0.0650547593832016, "learning_rate": 2.3279714030384273e-05, "loss": 0.0333, "step": 14950 }, { "epoch": 0.5361930294906166, "grad_norm": 1.191975206893403e-05, "learning_rate": 2.319034852546917e-05, "loss": 0.001, "step": 15000 }, { "epoch": 0.5379803395889187, "grad_norm": 0.01607808656990528, "learning_rate": 2.310098302055407e-05, "loss": 0.0206, "step": 15050 }, { "epoch": 0.5397676496872207, "grad_norm": 0.026940301060676575, "learning_rate": 2.3011617515638965e-05, "loss": 0.0149, "step": 15100 }, { "epoch": 0.5415549597855228, "grad_norm": 4.193421045783907e-05, "learning_rate": 2.2922252010723862e-05, "loss": 0.0005, "step": 15150 }, { "epoch": 0.5433422698838248, "grad_norm": 0.001876006368547678, "learning_rate": 2.2832886505808758e-05, "loss": 0.0106, "step": 15200 }, { "epoch": 0.5451295799821269, "grad_norm": 0.0055622560903429985, "learning_rate": 2.2743521000893654e-05, "loss": 0.0092, "step": 15250 }, { "epoch": 0.546916890080429, "grad_norm": 3.99901473429054e-05, "learning_rate": 2.2654155495978554e-05, "loss": 0.0057, "step": 15300 }, { "epoch": 0.548704200178731, "grad_norm": 1.553464426251594e-05, "learning_rate": 2.256478999106345e-05, "loss": 0.0026, "step": 15350 }, { "epoch": 0.550491510277033, "grad_norm": 4.381528854370117, "learning_rate": 2.2475424486148347e-05, "loss": 0.0166, "step": 15400 }, { "epoch": 0.5522788203753352, "grad_norm": 0.006728729698807001, "learning_rate": 2.2386058981233247e-05, "loss": 0.0524, "step": 15450 }, { "epoch": 0.5540661304736372, "grad_norm": 0.0004343650653026998, "learning_rate": 2.2296693476318143e-05, "loss": 0.0252, "step": 15500 }, { "epoch": 0.5558534405719392, "grad_norm": 0.0003802645078394562, "learning_rate": 2.220732797140304e-05, "loss": 0.0312, "step": 15550 }, { "epoch": 0.5576407506702413, "grad_norm": 0.7540137767791748, "learning_rate": 2.211796246648794e-05, "loss": 0.0012, "step": 15600 }, { "epoch": 0.5594280607685433, "grad_norm": 6.478348950622603e-05, "learning_rate": 2.2028596961572835e-05, "loss": 0.0084, "step": 15650 }, { "epoch": 0.5612153708668454, "grad_norm": 2.4346041679382324, "learning_rate": 2.193923145665773e-05, "loss": 0.0193, "step": 15700 }, { "epoch": 0.5630026809651475, "grad_norm": 0.02704077772796154, "learning_rate": 2.1849865951742628e-05, "loss": 0.0304, "step": 15750 }, { "epoch": 0.5647899910634495, "grad_norm": 0.06117241084575653, "learning_rate": 2.1760500446827525e-05, "loss": 0.0011, "step": 15800 }, { "epoch": 0.5665773011617515, "grad_norm": 5.2681320084957406e-05, "learning_rate": 2.167113494191242e-05, "loss": 0.0141, "step": 15850 }, { "epoch": 0.5683646112600537, "grad_norm": 5.393799438024871e-05, "learning_rate": 2.158176943699732e-05, "loss": 0.007, "step": 15900 }, { "epoch": 0.5701519213583557, "grad_norm": 0.007310281973332167, "learning_rate": 2.1492403932082217e-05, "loss": 0.0117, "step": 15950 }, { "epoch": 0.5719392314566577, "grad_norm": 0.0003883357858285308, "learning_rate": 2.1403038427167117e-05, "loss": 0.0032, "step": 16000 }, { "epoch": 0.5737265415549598, "grad_norm": 0.00011131053179269657, "learning_rate": 2.1313672922252013e-05, "loss": 0.0349, "step": 16050 }, { "epoch": 0.5755138516532619, "grad_norm": 0.3866327404975891, "learning_rate": 2.122430741733691e-05, "loss": 0.0018, "step": 16100 }, { "epoch": 0.5773011617515639, "grad_norm": 0.022152794525027275, "learning_rate": 2.1134941912421806e-05, "loss": 0.0433, "step": 16150 }, { "epoch": 0.579088471849866, "grad_norm": 0.0004128245636820793, "learning_rate": 2.1045576407506702e-05, "loss": 0.0019, "step": 16200 }, { "epoch": 0.580875781948168, "grad_norm": 0.00041269129724241793, "learning_rate": 2.09562109025916e-05, "loss": 0.0005, "step": 16250 }, { "epoch": 0.58266309204647, "grad_norm": 0.005800791084766388, "learning_rate": 2.0866845397676498e-05, "loss": 0.0025, "step": 16300 }, { "epoch": 0.5844504021447721, "grad_norm": 7.66227749409154e-05, "learning_rate": 2.0777479892761395e-05, "loss": 0.0244, "step": 16350 }, { "epoch": 0.5862377122430742, "grad_norm": 0.00028066098457202315, "learning_rate": 2.068811438784629e-05, "loss": 0.0502, "step": 16400 }, { "epoch": 0.5880250223413762, "grad_norm": 0.010334786958992481, "learning_rate": 2.059874888293119e-05, "loss": 0.0271, "step": 16450 }, { "epoch": 0.5898123324396782, "grad_norm": 0.06601165235042572, "learning_rate": 2.0509383378016087e-05, "loss": 0.0215, "step": 16500 }, { "epoch": 0.5915996425379804, "grad_norm": 0.007633717264980078, "learning_rate": 2.0420017873100983e-05, "loss": 0.0012, "step": 16550 }, { "epoch": 0.5933869526362824, "grad_norm": 0.007739843335002661, "learning_rate": 2.0330652368185883e-05, "loss": 0.0185, "step": 16600 }, { "epoch": 0.5951742627345844, "grad_norm": 5.8659989008447155e-06, "learning_rate": 2.024128686327078e-05, "loss": 0.0081, "step": 16650 }, { "epoch": 0.5969615728328865, "grad_norm": 0.001345645054243505, "learning_rate": 2.0151921358355676e-05, "loss": 0.0017, "step": 16700 }, { "epoch": 0.5987488829311886, "grad_norm": 0.012781122699379921, "learning_rate": 2.0062555853440572e-05, "loss": 0.0023, "step": 16750 }, { "epoch": 0.6005361930294906, "grad_norm": 8.988818990474101e-06, "learning_rate": 1.997319034852547e-05, "loss": 0.015, "step": 16800 }, { "epoch": 0.6023235031277927, "grad_norm": 0.007313130889087915, "learning_rate": 1.9883824843610368e-05, "loss": 0.0406, "step": 16850 }, { "epoch": 0.6041108132260947, "grad_norm": 0.003129061311483383, "learning_rate": 1.9794459338695265e-05, "loss": 0.0118, "step": 16900 }, { "epoch": 0.6058981233243967, "grad_norm": 0.007756507955491543, "learning_rate": 1.970509383378016e-05, "loss": 0.0734, "step": 16950 }, { "epoch": 0.6076854334226989, "grad_norm": 0.0032883670646697283, "learning_rate": 1.961572832886506e-05, "loss": 0.0191, "step": 17000 }, { "epoch": 0.6094727435210009, "grad_norm": 0.000193546453374438, "learning_rate": 1.9526362823949957e-05, "loss": 0.0007, "step": 17050 }, { "epoch": 0.6112600536193029, "grad_norm": 1.1219638586044312, "learning_rate": 1.9436997319034853e-05, "loss": 0.0201, "step": 17100 }, { "epoch": 0.613047363717605, "grad_norm": 0.007620728109031916, "learning_rate": 1.9347631814119753e-05, "loss": 0.0053, "step": 17150 }, { "epoch": 0.6148346738159071, "grad_norm": 1.7706630387692712e-05, "learning_rate": 1.9258266309204646e-05, "loss": 0.0293, "step": 17200 }, { "epoch": 0.6166219839142091, "grad_norm": 0.011270659975707531, "learning_rate": 1.9168900804289542e-05, "loss": 0.0481, "step": 17250 }, { "epoch": 0.6184092940125112, "grad_norm": 0.001038248767144978, "learning_rate": 1.9079535299374442e-05, "loss": 0.0096, "step": 17300 }, { "epoch": 0.6201966041108132, "grad_norm": 7.489488052669913e-05, "learning_rate": 1.899016979445934e-05, "loss": 0.0288, "step": 17350 }, { "epoch": 0.6219839142091153, "grad_norm": 0.0007915996829979122, "learning_rate": 1.8900804289544235e-05, "loss": 0.0059, "step": 17400 }, { "epoch": 0.6237712243074174, "grad_norm": 5.489266914082691e-05, "learning_rate": 1.8811438784629135e-05, "loss": 0.0166, "step": 17450 }, { "epoch": 0.6255585344057194, "grad_norm": 1.8211525457445532e-05, "learning_rate": 1.872207327971403e-05, "loss": 0.0526, "step": 17500 }, { "epoch": 0.6273458445040214, "grad_norm": 0.0003753642668016255, "learning_rate": 1.863270777479893e-05, "loss": 0.0092, "step": 17550 }, { "epoch": 0.6291331546023236, "grad_norm": 0.00022277800599113107, "learning_rate": 1.8543342269883827e-05, "loss": 0.0042, "step": 17600 }, { "epoch": 0.6309204647006256, "grad_norm": 0.007644360419362783, "learning_rate": 1.8453976764968723e-05, "loss": 0.0039, "step": 17650 }, { "epoch": 0.6327077747989276, "grad_norm": 0.00518200546503067, "learning_rate": 1.836461126005362e-05, "loss": 0.0002, "step": 17700 }, { "epoch": 0.6344950848972297, "grad_norm": 0.01000925898551941, "learning_rate": 1.8275245755138516e-05, "loss": 0.0018, "step": 17750 }, { "epoch": 0.6362823949955317, "grad_norm": 0.3702799379825592, "learning_rate": 1.8185880250223413e-05, "loss": 0.0236, "step": 17800 }, { "epoch": 0.6380697050938338, "grad_norm": 0.0007268276531249285, "learning_rate": 1.8096514745308312e-05, "loss": 0.0814, "step": 17850 }, { "epoch": 0.6398570151921358, "grad_norm": 2.1542404283536598e-05, "learning_rate": 1.800714924039321e-05, "loss": 0.0187, "step": 17900 }, { "epoch": 0.6416443252904379, "grad_norm": 0.0001183008062071167, "learning_rate": 1.7917783735478105e-05, "loss": 0.0142, "step": 17950 }, { "epoch": 0.6434316353887399, "grad_norm": 0.0016776048578321934, "learning_rate": 1.7828418230563005e-05, "loss": 0.0918, "step": 18000 }, { "epoch": 0.645218945487042, "grad_norm": 0.0006614942103624344, "learning_rate": 1.77390527256479e-05, "loss": 0.0069, "step": 18050 }, { "epoch": 0.6470062555853441, "grad_norm": 0.00020736618898808956, "learning_rate": 1.7649687220732797e-05, "loss": 0.0008, "step": 18100 }, { "epoch": 0.6487935656836461, "grad_norm": 0.0003252045717090368, "learning_rate": 1.7560321715817697e-05, "loss": 0.0285, "step": 18150 }, { "epoch": 0.6505808757819481, "grad_norm": 0.4281199872493744, "learning_rate": 1.7470956210902594e-05, "loss": 0.0135, "step": 18200 }, { "epoch": 0.6523681858802503, "grad_norm": 0.00044872984290122986, "learning_rate": 1.738159070598749e-05, "loss": 0.0025, "step": 18250 }, { "epoch": 0.6541554959785523, "grad_norm": 0.0005819129291921854, "learning_rate": 1.7292225201072386e-05, "loss": 0.0519, "step": 18300 }, { "epoch": 0.6559428060768543, "grad_norm": 0.00571415014564991, "learning_rate": 1.7202859696157283e-05, "loss": 0.0259, "step": 18350 }, { "epoch": 0.6577301161751564, "grad_norm": 1.6940903151407838e-05, "learning_rate": 1.7113494191242182e-05, "loss": 0.0085, "step": 18400 }, { "epoch": 0.6595174262734584, "grad_norm": 1.9965871615568176e-05, "learning_rate": 1.702412868632708e-05, "loss": 0.0006, "step": 18450 }, { "epoch": 0.6613047363717605, "grad_norm": 0.008692293427884579, "learning_rate": 1.6934763181411975e-05, "loss": 0.0021, "step": 18500 }, { "epoch": 0.6630920464700626, "grad_norm": 0.02017422392964363, "learning_rate": 1.6845397676496875e-05, "loss": 0.0197, "step": 18550 }, { "epoch": 0.6648793565683646, "grad_norm": 0.00021708336134906858, "learning_rate": 1.675603217158177e-05, "loss": 0.0003, "step": 18600 }, { "epoch": 0.6666666666666666, "grad_norm": 0.00038431567372754216, "learning_rate": 1.6666666666666667e-05, "loss": 0.0162, "step": 18650 }, { "epoch": 0.6684539767649688, "grad_norm": 2.9442389859468676e-05, "learning_rate": 1.6577301161751567e-05, "loss": 0.0611, "step": 18700 }, { "epoch": 0.6702412868632708, "grad_norm": 0.00041814154246822, "learning_rate": 1.648793565683646e-05, "loss": 0.0006, "step": 18750 }, { "epoch": 0.6720285969615728, "grad_norm": 5.710513505619019e-05, "learning_rate": 1.6398570151921357e-05, "loss": 0.0264, "step": 18800 }, { "epoch": 0.6738159070598749, "grad_norm": 0.00713470159098506, "learning_rate": 1.6309204647006256e-05, "loss": 0.0336, "step": 18850 }, { "epoch": 0.675603217158177, "grad_norm": 0.0040970491245388985, "learning_rate": 1.6219839142091153e-05, "loss": 0.0127, "step": 18900 }, { "epoch": 0.677390527256479, "grad_norm": 0.0007213613134808838, "learning_rate": 1.613047363717605e-05, "loss": 0.0205, "step": 18950 }, { "epoch": 0.6791778373547811, "grad_norm": 0.0005440552486106753, "learning_rate": 1.604110813226095e-05, "loss": 0.0008, "step": 19000 }, { "epoch": 0.6809651474530831, "grad_norm": 3.6667373933596537e-05, "learning_rate": 1.5951742627345845e-05, "loss": 0.0124, "step": 19050 }, { "epoch": 0.6827524575513851, "grad_norm": 0.005729563534259796, "learning_rate": 1.5862377122430745e-05, "loss": 0.0323, "step": 19100 }, { "epoch": 0.6845397676496873, "grad_norm": 0.00031316946842707694, "learning_rate": 1.577301161751564e-05, "loss": 0.0107, "step": 19150 }, { "epoch": 0.6863270777479893, "grad_norm": 0.00034565600799396634, "learning_rate": 1.5683646112600538e-05, "loss": 0.0098, "step": 19200 }, { "epoch": 0.6881143878462913, "grad_norm": 0.007016350980848074, "learning_rate": 1.5594280607685434e-05, "loss": 0.0357, "step": 19250 }, { "epoch": 0.6899016979445934, "grad_norm": 0.0009075969574041665, "learning_rate": 1.550491510277033e-05, "loss": 0.0069, "step": 19300 }, { "epoch": 0.6916890080428955, "grad_norm": 0.001256062532775104, "learning_rate": 1.5415549597855227e-05, "loss": 0.0155, "step": 19350 }, { "epoch": 0.6934763181411975, "grad_norm": 0.3069056272506714, "learning_rate": 1.5326184092940126e-05, "loss": 0.0457, "step": 19400 }, { "epoch": 0.6952636282394995, "grad_norm": 1.3622390724776778e-05, "learning_rate": 1.5236818588025023e-05, "loss": 0.0186, "step": 19450 }, { "epoch": 0.6970509383378016, "grad_norm": 0.00046906014904379845, "learning_rate": 1.5147453083109919e-05, "loss": 0.0282, "step": 19500 }, { "epoch": 0.6988382484361036, "grad_norm": 0.0385785773396492, "learning_rate": 1.5058087578194819e-05, "loss": 0.004, "step": 19550 }, { "epoch": 0.7006255585344057, "grad_norm": 0.0003727893636096269, "learning_rate": 1.4968722073279715e-05, "loss": 0.0012, "step": 19600 }, { "epoch": 0.7024128686327078, "grad_norm": 0.00044182888814248145, "learning_rate": 1.4879356568364611e-05, "loss": 0.0028, "step": 19650 }, { "epoch": 0.7042001787310098, "grad_norm": 0.9426965117454529, "learning_rate": 1.478999106344951e-05, "loss": 0.0028, "step": 19700 }, { "epoch": 0.7059874888293118, "grad_norm": 4.756861744681373e-05, "learning_rate": 1.4700625558534406e-05, "loss": 0.0006, "step": 19750 }, { "epoch": 0.707774798927614, "grad_norm": 0.0004083750245627016, "learning_rate": 1.4611260053619302e-05, "loss": 0.0007, "step": 19800 }, { "epoch": 0.709562109025916, "grad_norm": 2.2984992028796114e-05, "learning_rate": 1.4521894548704202e-05, "loss": 0.0084, "step": 19850 }, { "epoch": 0.711349419124218, "grad_norm": 0.01631794683635235, "learning_rate": 1.4432529043789098e-05, "loss": 0.0053, "step": 19900 }, { "epoch": 0.7131367292225201, "grad_norm": 0.00013000282342545688, "learning_rate": 1.4343163538873996e-05, "loss": 0.0006, "step": 19950 }, { "epoch": 0.7149240393208222, "grad_norm": 3.866414772346616e-05, "learning_rate": 1.4253798033958893e-05, "loss": 0.0266, "step": 20000 }, { "epoch": 0.7167113494191242, "grad_norm": 0.0024818661622703075, "learning_rate": 1.4164432529043789e-05, "loss": 0.001, "step": 20050 }, { "epoch": 0.7184986595174263, "grad_norm": 0.0025555104948580265, "learning_rate": 1.4075067024128689e-05, "loss": 0.0018, "step": 20100 }, { "epoch": 0.7202859696157283, "grad_norm": 0.11053712666034698, "learning_rate": 1.3985701519213583e-05, "loss": 0.0026, "step": 20150 }, { "epoch": 0.7220732797140303, "grad_norm": 0.8080301284790039, "learning_rate": 1.389633601429848e-05, "loss": 0.0032, "step": 20200 }, { "epoch": 0.7238605898123325, "grad_norm": 0.05114757642149925, "learning_rate": 1.380697050938338e-05, "loss": 0.0003, "step": 20250 }, { "epoch": 0.7256478999106345, "grad_norm": 9.302706166636199e-05, "learning_rate": 1.3717605004468276e-05, "loss": 0.0438, "step": 20300 }, { "epoch": 0.7274352100089365, "grad_norm": 0.25427255034446716, "learning_rate": 1.3628239499553172e-05, "loss": 0.0009, "step": 20350 }, { "epoch": 0.7292225201072386, "grad_norm": 0.010419056750833988, "learning_rate": 1.353887399463807e-05, "loss": 0.0042, "step": 20400 }, { "epoch": 0.7310098302055407, "grad_norm": 0.0004255430249031633, "learning_rate": 1.3449508489722967e-05, "loss": 0.0079, "step": 20450 }, { "epoch": 0.7327971403038427, "grad_norm": 0.0002166570775443688, "learning_rate": 1.3360142984807863e-05, "loss": 0.0234, "step": 20500 }, { "epoch": 0.7345844504021448, "grad_norm": 0.00016183908155653626, "learning_rate": 1.3270777479892763e-05, "loss": 0.0091, "step": 20550 }, { "epoch": 0.7363717605004468, "grad_norm": 0.0008138740668073297, "learning_rate": 1.3181411974977659e-05, "loss": 0.0127, "step": 20600 }, { "epoch": 0.7381590705987489, "grad_norm": 0.0011914832284674048, "learning_rate": 1.3092046470062557e-05, "loss": 0.0003, "step": 20650 }, { "epoch": 0.739946380697051, "grad_norm": 4.5319968194235116e-05, "learning_rate": 1.3002680965147454e-05, "loss": 0.0056, "step": 20700 }, { "epoch": 0.741733690795353, "grad_norm": 6.168178515508771e-05, "learning_rate": 1.291331546023235e-05, "loss": 0.0919, "step": 20750 }, { "epoch": 0.743521000893655, "grad_norm": 5.0896484026452526e-05, "learning_rate": 1.282394995531725e-05, "loss": 0.1042, "step": 20800 }, { "epoch": 0.7453083109919572, "grad_norm": 0.0003189992276020348, "learning_rate": 1.2734584450402146e-05, "loss": 0.0004, "step": 20850 }, { "epoch": 0.7470956210902592, "grad_norm": 0.004061105195432901, "learning_rate": 1.2645218945487042e-05, "loss": 0.0141, "step": 20900 }, { "epoch": 0.7488829311885612, "grad_norm": 0.011431465856730938, "learning_rate": 1.255585344057194e-05, "loss": 0.0034, "step": 20950 }, { "epoch": 0.7506702412868632, "grad_norm": 0.00021052254305686802, "learning_rate": 1.2466487935656837e-05, "loss": 0.0001, "step": 21000 }, { "epoch": 0.7524575513851653, "grad_norm": 9.472965029999614e-05, "learning_rate": 1.2377122430741735e-05, "loss": 0.0035, "step": 21050 }, { "epoch": 0.7542448614834674, "grad_norm": 0.3230762183666229, "learning_rate": 1.2287756925826631e-05, "loss": 0.0012, "step": 21100 }, { "epoch": 0.7560321715817694, "grad_norm": 3.0450801205006428e-05, "learning_rate": 1.219839142091153e-05, "loss": 0.0003, "step": 21150 }, { "epoch": 0.7578194816800715, "grad_norm": 0.0009597355965524912, "learning_rate": 1.2109025915996426e-05, "loss": 0.0114, "step": 21200 }, { "epoch": 0.7596067917783735, "grad_norm": 0.0003465202753432095, "learning_rate": 1.2019660411081324e-05, "loss": 0.0219, "step": 21250 }, { "epoch": 0.7613941018766756, "grad_norm": 3.194337841705419e-05, "learning_rate": 1.193029490616622e-05, "loss": 0.0616, "step": 21300 }, { "epoch": 0.7631814119749777, "grad_norm": 0.00045533524826169014, "learning_rate": 1.1840929401251118e-05, "loss": 0.0119, "step": 21350 }, { "epoch": 0.7649687220732797, "grad_norm": 0.00026991829508915544, "learning_rate": 1.1751563896336016e-05, "loss": 0.0071, "step": 21400 }, { "epoch": 0.7667560321715817, "grad_norm": 0.009153268299996853, "learning_rate": 1.166219839142091e-05, "loss": 0.0154, "step": 21450 }, { "epoch": 0.7685433422698839, "grad_norm": 0.0002896921359933913, "learning_rate": 1.1572832886505809e-05, "loss": 0.0364, "step": 21500 }, { "epoch": 0.7703306523681859, "grad_norm": 11.182766914367676, "learning_rate": 1.1483467381590707e-05, "loss": 0.0349, "step": 21550 }, { "epoch": 0.7721179624664879, "grad_norm": 0.0034106161911040545, "learning_rate": 1.1394101876675605e-05, "loss": 0.0078, "step": 21600 }, { "epoch": 0.77390527256479, "grad_norm": 0.00010709511843742803, "learning_rate": 1.1304736371760501e-05, "loss": 0.0043, "step": 21650 }, { "epoch": 0.775692582663092, "grad_norm": 1.2288366633583792e-05, "learning_rate": 1.1215370866845398e-05, "loss": 0.0197, "step": 21700 }, { "epoch": 0.7774798927613941, "grad_norm": 0.0007678976398892701, "learning_rate": 1.1126005361930296e-05, "loss": 0.0052, "step": 21750 }, { "epoch": 0.7792672028596962, "grad_norm": 0.002781949006021023, "learning_rate": 1.1036639857015192e-05, "loss": 0.0008, "step": 21800 }, { "epoch": 0.7810545129579982, "grad_norm": 1.678530861681793e-05, "learning_rate": 1.094727435210009e-05, "loss": 0.0249, "step": 21850 }, { "epoch": 0.7828418230563002, "grad_norm": 0.0001823195198085159, "learning_rate": 1.0857908847184988e-05, "loss": 0.0011, "step": 21900 }, { "epoch": 0.7846291331546024, "grad_norm": 0.0021642933133989573, "learning_rate": 1.0768543342269884e-05, "loss": 0.0083, "step": 21950 }, { "epoch": 0.7864164432529044, "grad_norm": 0.011181133799254894, "learning_rate": 1.067917783735478e-05, "loss": 0.0001, "step": 22000 }, { "epoch": 0.7882037533512064, "grad_norm": 0.0007323863101191819, "learning_rate": 1.0589812332439679e-05, "loss": 0.0035, "step": 22050 }, { "epoch": 0.7899910634495085, "grad_norm": 5.042120392317884e-06, "learning_rate": 1.0500446827524577e-05, "loss": 0.0204, "step": 22100 }, { "epoch": 0.7917783735478106, "grad_norm": 2.0223915271344595e-05, "learning_rate": 1.0411081322609473e-05, "loss": 0.095, "step": 22150 }, { "epoch": 0.7935656836461126, "grad_norm": 1.75688346644165e-05, "learning_rate": 1.0321715817694371e-05, "loss": 0.0009, "step": 22200 }, { "epoch": 0.7953529937444147, "grad_norm": 20.508712768554688, "learning_rate": 1.0232350312779268e-05, "loss": 0.0277, "step": 22250 }, { "epoch": 0.7971403038427167, "grad_norm": 0.00028106712852604687, "learning_rate": 1.0142984807864164e-05, "loss": 0.0013, "step": 22300 }, { "epoch": 0.7989276139410187, "grad_norm": 0.22000660002231598, "learning_rate": 1.0053619302949062e-05, "loss": 0.0057, "step": 22350 }, { "epoch": 0.8007149240393209, "grad_norm": 8.083471038844436e-05, "learning_rate": 9.96425379803396e-06, "loss": 0.0031, "step": 22400 }, { "epoch": 0.8025022341376229, "grad_norm": 0.055748261511325836, "learning_rate": 9.874888293118856e-06, "loss": 0.0004, "step": 22450 }, { "epoch": 0.8042895442359249, "grad_norm": 6.724525519530289e-06, "learning_rate": 9.785522788203753e-06, "loss": 0.0063, "step": 22500 }, { "epoch": 0.8060768543342269, "grad_norm": 0.004222277086228132, "learning_rate": 9.69615728328865e-06, "loss": 0.0161, "step": 22550 }, { "epoch": 0.8078641644325291, "grad_norm": 6.366088200593367e-05, "learning_rate": 9.606791778373549e-06, "loss": 0.0118, "step": 22600 }, { "epoch": 0.8096514745308311, "grad_norm": 0.00010139405640074983, "learning_rate": 9.517426273458445e-06, "loss": 0.0003, "step": 22650 }, { "epoch": 0.8114387846291331, "grad_norm": 0.0005521093844436109, "learning_rate": 9.428060768543343e-06, "loss": 0.0185, "step": 22700 }, { "epoch": 0.8132260947274352, "grad_norm": 0.00011474806524347514, "learning_rate": 9.33869526362824e-06, "loss": 0.0012, "step": 22750 }, { "epoch": 0.8150134048257373, "grad_norm": 0.0009986285585910082, "learning_rate": 9.249329758713138e-06, "loss": 0.0186, "step": 22800 }, { "epoch": 0.8168007149240393, "grad_norm": 2.7077521735918708e-05, "learning_rate": 9.159964253798034e-06, "loss": 0.0019, "step": 22850 }, { "epoch": 0.8185880250223414, "grad_norm": 0.017945552244782448, "learning_rate": 9.070598748882932e-06, "loss": 0.0005, "step": 22900 }, { "epoch": 0.8203753351206434, "grad_norm": 0.009763717651367188, "learning_rate": 8.98123324396783e-06, "loss": 0.0019, "step": 22950 }, { "epoch": 0.8221626452189454, "grad_norm": 0.0006894482066854835, "learning_rate": 8.891867739052725e-06, "loss": 0.0014, "step": 23000 }, { "epoch": 0.8239499553172476, "grad_norm": 4.397491647978313e-05, "learning_rate": 8.802502234137623e-06, "loss": 0.0321, "step": 23050 }, { "epoch": 0.8257372654155496, "grad_norm": 0.01145015936344862, "learning_rate": 8.71313672922252e-06, "loss": 0.0868, "step": 23100 }, { "epoch": 0.8275245755138516, "grad_norm": 7.181058026617393e-05, "learning_rate": 8.623771224307419e-06, "loss": 0.0035, "step": 23150 }, { "epoch": 0.8293118856121537, "grad_norm": 0.0008214627159759402, "learning_rate": 8.534405719392315e-06, "loss": 0.0055, "step": 23200 }, { "epoch": 0.8310991957104558, "grad_norm": 15.694032669067383, "learning_rate": 8.445040214477212e-06, "loss": 0.0244, "step": 23250 }, { "epoch": 0.8328865058087578, "grad_norm": 6.911622222105507e-06, "learning_rate": 8.35567470956211e-06, "loss": 0.002, "step": 23300 }, { "epoch": 0.8346738159070599, "grad_norm": 7.621223630849272e-05, "learning_rate": 8.266309204647006e-06, "loss": 0.0006, "step": 23350 }, { "epoch": 0.8364611260053619, "grad_norm": 0.20208632946014404, "learning_rate": 8.176943699731904e-06, "loss": 0.0039, "step": 23400 }, { "epoch": 0.838248436103664, "grad_norm": 0.00231548142619431, "learning_rate": 8.087578194816802e-06, "loss": 0.0033, "step": 23450 }, { "epoch": 0.8400357462019661, "grad_norm": 2.214570668002125e-05, "learning_rate": 7.998212689901698e-06, "loss": 0.0468, "step": 23500 }, { "epoch": 0.8418230563002681, "grad_norm": 1.7991300410358235e-05, "learning_rate": 7.908847184986595e-06, "loss": 0.0116, "step": 23550 }, { "epoch": 0.8436103663985701, "grad_norm": 0.7063977122306824, "learning_rate": 7.819481680071493e-06, "loss": 0.0023, "step": 23600 }, { "epoch": 0.8453976764968723, "grad_norm": 0.0005979365087114275, "learning_rate": 7.730116175156391e-06, "loss": 0.0001, "step": 23650 }, { "epoch": 0.8471849865951743, "grad_norm": 9.088612387131434e-06, "learning_rate": 7.640750670241287e-06, "loss": 0.0064, "step": 23700 }, { "epoch": 0.8489722966934763, "grad_norm": 0.006399601697921753, "learning_rate": 7.5513851653261844e-06, "loss": 0.0017, "step": 23750 }, { "epoch": 0.8507596067917784, "grad_norm": 0.00021108676446601748, "learning_rate": 7.462019660411082e-06, "loss": 0.022, "step": 23800 }, { "epoch": 0.8525469168900804, "grad_norm": 0.9258911609649658, "learning_rate": 7.372654155495978e-06, "loss": 0.0041, "step": 23850 }, { "epoch": 0.8543342269883825, "grad_norm": 8.936112863011658e-05, "learning_rate": 7.283288650580876e-06, "loss": 0.0001, "step": 23900 }, { "epoch": 0.8561215370866846, "grad_norm": 1.7370651960372925, "learning_rate": 7.193923145665773e-06, "loss": 0.0027, "step": 23950 }, { "epoch": 0.8579088471849866, "grad_norm": 0.0029333913698792458, "learning_rate": 7.104557640750671e-06, "loss": 0.0357, "step": 24000 }, { "epoch": 0.8596961572832886, "grad_norm": 0.01217588223516941, "learning_rate": 7.015192135835568e-06, "loss": 0.0135, "step": 24050 }, { "epoch": 0.8614834673815907, "grad_norm": 0.008493458852171898, "learning_rate": 6.925826630920465e-06, "loss": 0.0001, "step": 24100 }, { "epoch": 0.8632707774798928, "grad_norm": 0.0002583667519502342, "learning_rate": 6.836461126005363e-06, "loss": 0.0182, "step": 24150 }, { "epoch": 0.8650580875781948, "grad_norm": 3.6770063161384314e-05, "learning_rate": 6.747095621090259e-06, "loss": 0.0002, "step": 24200 }, { "epoch": 0.8668453976764968, "grad_norm": 0.0009357187082059681, "learning_rate": 6.6577301161751565e-06, "loss": 0.0484, "step": 24250 }, { "epoch": 0.868632707774799, "grad_norm": 0.0004051830619573593, "learning_rate": 6.5683646112600545e-06, "loss": 0.1176, "step": 24300 }, { "epoch": 0.870420017873101, "grad_norm": 4.589395393850282e-05, "learning_rate": 6.478999106344952e-06, "loss": 0.0087, "step": 24350 }, { "epoch": 0.872207327971403, "grad_norm": 5.9133399190614e-05, "learning_rate": 6.389633601429848e-06, "loss": 0.0355, "step": 24400 }, { "epoch": 0.8739946380697051, "grad_norm": 0.00040170710417442024, "learning_rate": 6.300268096514745e-06, "loss": 0.0013, "step": 24450 }, { "epoch": 0.8757819481680071, "grad_norm": 4.549537607090315e-06, "learning_rate": 6.2109025915996425e-06, "loss": 0.0119, "step": 24500 }, { "epoch": 0.8775692582663092, "grad_norm": 0.00019794557010754943, "learning_rate": 6.1215370866845405e-06, "loss": 0.0199, "step": 24550 }, { "epoch": 0.8793565683646113, "grad_norm": 0.0002817718777805567, "learning_rate": 6.032171581769437e-06, "loss": 0.0013, "step": 24600 }, { "epoch": 0.8811438784629133, "grad_norm": 2.2904992874828167e-05, "learning_rate": 5.942806076854335e-06, "loss": 0.0096, "step": 24650 }, { "epoch": 0.8829311885612153, "grad_norm": 0.1459171175956726, "learning_rate": 5.853440571939232e-06, "loss": 0.0003, "step": 24700 }, { "epoch": 0.8847184986595175, "grad_norm": 0.0002032791671808809, "learning_rate": 5.7640750670241285e-06, "loss": 0.0002, "step": 24750 }, { "epoch": 0.8865058087578195, "grad_norm": 0.0002909420873038471, "learning_rate": 5.6747095621090265e-06, "loss": 0.0009, "step": 24800 }, { "epoch": 0.8882931188561215, "grad_norm": 0.0027191690169274807, "learning_rate": 5.585344057193923e-06, "loss": 0.0017, "step": 24850 }, { "epoch": 0.8900804289544236, "grad_norm": 0.0009960250463336706, "learning_rate": 5.495978552278821e-06, "loss": 0.0001, "step": 24900 }, { "epoch": 0.8918677390527256, "grad_norm": 9.508804942015558e-05, "learning_rate": 5.406613047363718e-06, "loss": 0.0097, "step": 24950 }, { "epoch": 0.8936550491510277, "grad_norm": 0.0010277976980432868, "learning_rate": 5.317247542448615e-06, "loss": 0.0035, "step": 25000 }, { "epoch": 0.8954423592493298, "grad_norm": 2.4904526071622968e-05, "learning_rate": 5.2278820375335125e-06, "loss": 0.0005, "step": 25050 }, { "epoch": 0.8972296693476318, "grad_norm": 0.0015225968090817332, "learning_rate": 5.138516532618409e-06, "loss": 0.0152, "step": 25100 }, { "epoch": 0.8990169794459338, "grad_norm": 8.117647666949779e-05, "learning_rate": 5.049151027703307e-06, "loss": 0.0002, "step": 25150 }, { "epoch": 0.900804289544236, "grad_norm": 0.005292691756039858, "learning_rate": 4.959785522788204e-06, "loss": 0.0152, "step": 25200 }, { "epoch": 0.902591599642538, "grad_norm": 5.475033321999945e-05, "learning_rate": 4.870420017873101e-06, "loss": 0.0402, "step": 25250 }, { "epoch": 0.90437890974084, "grad_norm": 0.00039539620047435164, "learning_rate": 4.7810545129579985e-06, "loss": 0.0478, "step": 25300 }, { "epoch": 0.9061662198391421, "grad_norm": 0.2715374827384949, "learning_rate": 4.691689008042896e-06, "loss": 0.0082, "step": 25350 }, { "epoch": 0.9079535299374442, "grad_norm": 0.003237704513594508, "learning_rate": 4.602323503127793e-06, "loss": 0.0011, "step": 25400 }, { "epoch": 0.9097408400357462, "grad_norm": 1.7789971025194973e-05, "learning_rate": 4.51295799821269e-06, "loss": 0.0413, "step": 25450 }, { "epoch": 0.9115281501340483, "grad_norm": 9.451475489186123e-05, "learning_rate": 4.423592493297587e-06, "loss": 0.0049, "step": 25500 }, { "epoch": 0.9133154602323503, "grad_norm": 1.6161772012710571, "learning_rate": 4.3342269883824845e-06, "loss": 0.0199, "step": 25550 }, { "epoch": 0.9151027703306523, "grad_norm": 0.0006955278222449124, "learning_rate": 4.244861483467382e-06, "loss": 0.002, "step": 25600 }, { "epoch": 0.9168900804289544, "grad_norm": 0.00014403380919247866, "learning_rate": 4.155495978552279e-06, "loss": 0.0153, "step": 25650 }, { "epoch": 0.9186773905272565, "grad_norm": 3.223052408429794e-05, "learning_rate": 4.066130473637176e-06, "loss": 0.0011, "step": 25700 }, { "epoch": 0.9204647006255585, "grad_norm": 0.00047142524272203445, "learning_rate": 3.976764968722073e-06, "loss": 0.0007, "step": 25750 }, { "epoch": 0.9222520107238605, "grad_norm": 0.010394671000540257, "learning_rate": 3.8873994638069705e-06, "loss": 0.0242, "step": 25800 }, { "epoch": 0.9240393208221627, "grad_norm": 0.0024043857119977474, "learning_rate": 3.798033958891868e-06, "loss": 0.0184, "step": 25850 }, { "epoch": 0.9258266309204647, "grad_norm": 0.00401880219578743, "learning_rate": 3.708668453976765e-06, "loss": 0.0003, "step": 25900 }, { "epoch": 0.9276139410187667, "grad_norm": 2.407459578535054e-05, "learning_rate": 3.6193029490616625e-06, "loss": 0.001, "step": 25950 }, { "epoch": 0.9294012511170688, "grad_norm": 0.00016128833522088826, "learning_rate": 3.5299374441465597e-06, "loss": 0.0031, "step": 26000 }, { "epoch": 0.9311885612153709, "grad_norm": 7.472659490304068e-05, "learning_rate": 3.4405719392314565e-06, "loss": 0.0006, "step": 26050 }, { "epoch": 0.9329758713136729, "grad_norm": 3.585060767363757e-05, "learning_rate": 3.351206434316354e-06, "loss": 0.0066, "step": 26100 }, { "epoch": 0.934763181411975, "grad_norm": 7.751063822070137e-05, "learning_rate": 3.2618409294012513e-06, "loss": 0.0287, "step": 26150 }, { "epoch": 0.936550491510277, "grad_norm": 0.0006985155632719398, "learning_rate": 3.172475424486149e-06, "loss": 0.0106, "step": 26200 }, { "epoch": 0.938337801608579, "grad_norm": 0.00020505161955952644, "learning_rate": 3.0831099195710457e-06, "loss": 0.009, "step": 26250 }, { "epoch": 0.9401251117068812, "grad_norm": 3.1767311156727374e-05, "learning_rate": 2.993744414655943e-06, "loss": 0.0044, "step": 26300 }, { "epoch": 0.9419124218051832, "grad_norm": 0.019595185294747353, "learning_rate": 2.90437890974084e-06, "loss": 0.0067, "step": 26350 }, { "epoch": 0.9436997319034852, "grad_norm": 0.00015039286518003792, "learning_rate": 2.8150134048257373e-06, "loss": 0.0002, "step": 26400 }, { "epoch": 0.9454870420017873, "grad_norm": 0.00041445993701927364, "learning_rate": 2.7256478999106345e-06, "loss": 0.0024, "step": 26450 }, { "epoch": 0.9472743521000894, "grad_norm": 7.043827736197272e-06, "learning_rate": 2.6362823949955317e-06, "loss": 0.0058, "step": 26500 }, { "epoch": 0.9490616621983914, "grad_norm": 0.0001892998261610046, "learning_rate": 2.546916890080429e-06, "loss": 0.026, "step": 26550 }, { "epoch": 0.9508489722966935, "grad_norm": 0.0011999151902273297, "learning_rate": 2.457551385165326e-06, "loss": 0.0033, "step": 26600 }, { "epoch": 0.9526362823949955, "grad_norm": 7.256813842104748e-05, "learning_rate": 2.3681858802502238e-06, "loss": 0.0092, "step": 26650 }, { "epoch": 0.9544235924932976, "grad_norm": 0.05680149793624878, "learning_rate": 2.278820375335121e-06, "loss": 0.0003, "step": 26700 }, { "epoch": 0.9562109025915997, "grad_norm": 4.024420559289865e-05, "learning_rate": 2.1894548704200177e-06, "loss": 0.0338, "step": 26750 }, { "epoch": 0.9579982126899017, "grad_norm": 0.0002506797027308494, "learning_rate": 2.100089365504915e-06, "loss": 0.0034, "step": 26800 }, { "epoch": 0.9597855227882037, "grad_norm": 2.6599795818328857, "learning_rate": 2.0107238605898126e-06, "loss": 0.0036, "step": 26850 }, { "epoch": 0.9615728328865059, "grad_norm": 8.052168413996696e-05, "learning_rate": 1.9213583556747098e-06, "loss": 0.0323, "step": 26900 }, { "epoch": 0.9633601429848079, "grad_norm": 3.59859004674945e-05, "learning_rate": 1.831992850759607e-06, "loss": 0.0001, "step": 26950 }, { "epoch": 0.9651474530831099, "grad_norm": 0.028114568442106247, "learning_rate": 1.7426273458445042e-06, "loss": 0.0002, "step": 27000 }, { "epoch": 0.966934763181412, "grad_norm": 0.0005747485556639731, "learning_rate": 1.6532618409294012e-06, "loss": 0.0028, "step": 27050 }, { "epoch": 0.968722073279714, "grad_norm": 0.20126579701900482, "learning_rate": 1.5638963360142984e-06, "loss": 0.0036, "step": 27100 }, { "epoch": 0.9705093833780161, "grad_norm": 9.089746163226664e-05, "learning_rate": 1.4745308310991958e-06, "loss": 0.011, "step": 27150 }, { "epoch": 0.9722966934763181, "grad_norm": 0.0010499503696337342, "learning_rate": 1.385165326184093e-06, "loss": 0.0, "step": 27200 }, { "epoch": 0.9740840035746202, "grad_norm": 0.00011471787729533389, "learning_rate": 1.2957998212689904e-06, "loss": 0.0004, "step": 27250 }, { "epoch": 0.9758713136729222, "grad_norm": 0.00010378209117334336, "learning_rate": 1.2064343163538874e-06, "loss": 0.005, "step": 27300 }, { "epoch": 0.9776586237712243, "grad_norm": 0.00020865820988547057, "learning_rate": 1.1170688114387846e-06, "loss": 0.0007, "step": 27350 }, { "epoch": 0.9794459338695264, "grad_norm": 0.009171546436846256, "learning_rate": 1.027703306523682e-06, "loss": 0.0573, "step": 27400 }, { "epoch": 0.9812332439678284, "grad_norm": 0.00013850234972778708, "learning_rate": 9.383378016085791e-07, "loss": 0.0385, "step": 27450 }, { "epoch": 0.9830205540661304, "grad_norm": 8.932514174375683e-05, "learning_rate": 8.489722966934764e-07, "loss": 0.0014, "step": 27500 }, { "epoch": 0.9848078641644326, "grad_norm": 0.001844526268541813, "learning_rate": 7.596067917783736e-07, "loss": 0.0, "step": 27550 }, { "epoch": 0.9865951742627346, "grad_norm": 5.615004062652588, "learning_rate": 6.702412868632708e-07, "loss": 0.0036, "step": 27600 }, { "epoch": 0.9883824843610366, "grad_norm": 0.006281781475991011, "learning_rate": 5.80875781948168e-07, "loss": 0.0499, "step": 27650 }, { "epoch": 0.9901697944593387, "grad_norm": 0.00031924585346132517, "learning_rate": 4.915102770330653e-07, "loss": 0.0093, "step": 27700 }, { "epoch": 0.9919571045576407, "grad_norm": 4.5880449761170894e-05, "learning_rate": 4.021447721179625e-07, "loss": 0.0, "step": 27750 }, { "epoch": 0.9937444146559428, "grad_norm": 1.7767933968571015e-05, "learning_rate": 3.1277926720285975e-07, "loss": 0.0079, "step": 27800 }, { "epoch": 0.9955317247542449, "grad_norm": 0.033302705734968185, "learning_rate": 2.2341376228775692e-07, "loss": 0.0005, "step": 27850 }, { "epoch": 0.9973190348525469, "grad_norm": 0.0001265561586478725, "learning_rate": 1.3404825737265417e-07, "loss": 0.0115, "step": 27900 }, { "epoch": 0.9991063449508489, "grad_norm": 0.0001866740349214524, "learning_rate": 4.4682752457551387e-08, "loss": 0.0018, "step": 27950 } ], "logging_steps": 50, "max_steps": 27975, "num_input_tokens_seen": 0, "num_train_epochs": 1, "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": 0.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }